{"Affiliation":[{"label":"Affiliation","value":"Land and Food Systems, Faculty of","attrs":{"lang":"en","ns":"http:\/\/vivoweb.org\/ontology\/core#departmentOrSchool","classmap":"vivo:EducationalProcess","property":"vivo:departmentOrSchool"},"iri":"http:\/\/vivoweb.org\/ontology\/core#departmentOrSchool","explain":"VIVO-ISF Ontology V1.6 Property; The department or school name within institution; Not intended to be an institution name."}],"AggregatedSourceRepository":[{"label":"Aggregated Source Repository","value":"DSpace","attrs":{"lang":"en","ns":"http:\/\/www.europeana.eu\/schemas\/edm\/dataProvider","classmap":"ore:Aggregation","property":"edm:dataProvider"},"iri":"http:\/\/www.europeana.eu\/schemas\/edm\/dataProvider","explain":"A Europeana Data Model Property; The name or identifier of the organization who contributes data indirectly to an aggregation service (e.g. Europeana)"}],"Campus":[{"label":"Campus","value":"UBCV","attrs":{"lang":"en","ns":"https:\/\/open.library.ubc.ca\/terms#degreeCampus","classmap":"oc:ThesisDescription","property":"oc:degreeCampus"},"iri":"https:\/\/open.library.ubc.ca\/terms#degreeCampus","explain":"UBC Open Collections Metadata Components; Local Field; Identifies the name of the campus from which the graduate completed their degree."}],"Creator":[{"label":"Creator","value":"Mandal, Ronit","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/creator","classmap":"dpla:SourceResource","property":"dcterms:creator"},"iri":"http:\/\/purl.org\/dc\/terms\/creator","explain":"A Dublin Core Terms Property; An entity primarily responsible for making the resource.; Examples of a Contributor include a person, an organization, or a service."}],"DateAvailable":[{"label":"Date Available","value":"2022-07-15T22:37:42Z","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/issued","classmap":"edm:WebResource","property":"dcterms:issued"},"iri":"http:\/\/purl.org\/dc\/terms\/issued","explain":"A Dublin Core Terms Property; Date of formal issuance (e.g., publication) of the resource."}],"DateIssued":[{"label":"Date Issued","value":"2022","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/issued","classmap":"oc:SourceResource","property":"dcterms:issued"},"iri":"http:\/\/purl.org\/dc\/terms\/issued","explain":"A Dublin Core Terms Property; Date of formal issuance (e.g., publication) of the resource."}],"Degree":[{"label":"Degree (Theses)","value":"Doctor of Philosophy - PhD","attrs":{"lang":"en","ns":"http:\/\/vivoweb.org\/ontology\/core#relatedDegree","classmap":"vivo:ThesisDegree","property":"vivo:relatedDegree"},"iri":"http:\/\/vivoweb.org\/ontology\/core#relatedDegree","explain":"VIVO-ISF Ontology V1.6 Property; The thesis degree; Extended Property specified by UBC, as per https:\/\/wiki.duraspace.org\/display\/VIVO\/Ontology+Editor%27s+Guide"}],"DegreeGrantor":[{"label":"Degree Grantor","value":"University of British Columbia","attrs":{"lang":"en","ns":"https:\/\/open.library.ubc.ca\/terms#degreeGrantor","classmap":"oc:ThesisDescription","property":"oc:degreeGrantor"},"iri":"https:\/\/open.library.ubc.ca\/terms#degreeGrantor","explain":"UBC Open Collections Metadata Components; Local Field; Indicates the institution where thesis was granted."}],"Description":[{"label":"Description","value":"Pulsed UV light (PL) is a non-thermal food preservation method that employs high-intensity short-duration pulses of polychromatic light (200-1200 nm) to inactivate microorganisms while having minimal effects on nutrients and sensory acceptability. PL is an excellent decontamination technology for solid foods. However, for liquids, parameters like clarity, turbidity, UV transmissivity govern the PL efficacy. Therefore, my work is focused on developing an understanding of PL processing using modeling and computational simulation of different liquids. The PL system used consists of two different quartz reactors (annular [AT] and coiled tube [CT]) with a cylindrical lamp at the axis. The 3-D light energy distribution was modeled using water, model liquid foods (water + red\/green dye), and skim milk around the lamp to predict lamp energy at any point in space around the lamp. Liquids were treated in the reactors at various flow rates (14-75 L\/h) and pulse frequency (1-5 Hz) after inoculation with Escherichia coli ATCC 29055, Listeria innocua ATCC 33090 and Clostridium sporogenes ATCC 7955. A batch collimation experiment was done to ascertain UV dosage for microbial inactivation in the liquids. The inactivation was in the order: water > water + red dye > water + green dye > milk. Then, milk samples (0%, 1%, 2% and 3.25% fat) were treated under the same conditions and analyzed for microbial inactivation, pH, colour, vitamin B2 and C content, lipid, and protein oxidation. Up to 4 logs reduction of microorganisms were obtained for milk. The pH of the treated samples was similar to the control. The colour parameter b* decreased as treatment intensified, while vitamin B2 and C decreased significantly. There was significant lipid oxidation and protein oxidation. Red grape and watermelon juice were treated under the same conditions and analyzed for microbial inactivation, pH, colour, total phenolics, antioxidant capacity, lycopene, and anthocyanin contents. While there was complete inactivation (>7 logs) of microbes in juices, there were minimal effects on colour and pH; but the phenolics, antioxidants, and anthocyanins were affected greatly. The results of this study will inform the food industry on how to use this technology effectively.","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/description","classmap":"dpla:SourceResource","property":"dcterms:description"},"iri":"http:\/\/purl.org\/dc\/terms\/description","explain":"A Dublin Core Terms Property; An account of the resource.; Description may include but is not limited to: an abstract, a table of contents, a graphical representation, or a free-text account of the resource."}],"DigitalResourceOriginalRecord":[{"label":"Digital Resource Original Record","value":"https:\/\/circle.library.ubc.ca\/rest\/handle\/2429\/82156?expand=metadata","attrs":{"lang":"en","ns":"http:\/\/www.europeana.eu\/schemas\/edm\/aggregatedCHO","classmap":"ore:Aggregation","property":"edm:aggregatedCHO"},"iri":"http:\/\/www.europeana.eu\/schemas\/edm\/aggregatedCHO","explain":"A Europeana Data Model Property; The identifier of the source object, e.g. the Mona Lisa itself. This could be a full linked open date URI or an internal identifier"}],"FullText":[{"label":"Full Text","value":"DESIGN, DEVELOPMENT, AND VALIDATION OF CONTINUOUS-FLOW PULSED UV LIGHT SYSTEMS FOR LIQUID FOOD PROCESSING by  Ronit Mandal  B.Tech., West Bengal University of Animal and Fishery Sciences, 2016 M.Tech., Indian Institute of Technology Kharagpur, 2018  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Food Science)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)   July 2022   \u00a9 Ronit Mandal, 2022       ii  The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the dissertation entitled: Design, development, and validation of continuous-flow pulsed UV light systems for liquid food processing  submitted by Ronit Mandal in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Food Science  Examining Committee: Dr. Anubhav Pratap-Singh, Assistant Professor, Food Science, UBC Supervisor  Dr. Derek Dee, Assistant Professor, Food Science, UBC Supervisory Committee Member  Dr. Fariborz Taghipour, Professor, Chemical and Biological Engineering, UBC Supervisory Committee Member Dr. Siyun Wang, Associate Professor, Food Science, UBC University Examiner Dr. Orlando Rojas, Professor, Chemical and Biological Engineering, UBC University Examiner   iii  Abstract Pulsed UV light (PL) is a non-thermal food preservation method that employs high-intensity short-duration pulses of polychromatic light (200-1200 nm) to inactivate microorganisms while having minimal effects on nutrients and sensory acceptability. PL is an excellent decontamination technology for solid foods. However, for liquids, parameters like clarity, turbidity, UV transmissivity govern the PL efficacy. Therefore, my work is focused on developing an understanding of PL processing using modeling and computational simulation of different liquids. The PL system used consists of two different quartz reactors (annular [AT] and coiled tube [CT]) with a cylindrical lamp at the axis. The 3-D light energy distribution was modeled using water, model liquid foods (water + red\/green dye), and skim milk around the lamp to predict lamp energy at any point in space around the lamp. Liquids were treated in the reactors at various flow rates (14-75 L\/h) and pulse frequency (1-5 Hz) after inoculation with Escherichia coli ATCC 29055, Listeria innocua ATCC 33090 and Clostridium sporogenes ATCC 7955. A batch collimation experiment was done to ascertain UV dosage for microbial inactivation in the liquids. The inactivation was in the order: water > water + red dye > water + green dye > milk. Then, milk samples (0%, 1%, 2% and 3.25% fat) were treated under the same conditions and analyzed for microbial inactivation, pH, colour, vitamin B2 and C content, lipid, and protein oxidation. Up to 4 logs reduction of microorganisms were obtained for milk. The pH of the treated samples was similar to the control. The colour parameter b* decreased as treatment intensified, while vitamin B2 and C decreased significantly. There was significant lipid oxidation and protein oxidation. Red grape and watermelon juice were treated under the same conditions and analyzed for microbial inactivation, pH, colour, total phenolics, antioxidant capacity, lycopene, and anthocyanin contents. While there was complete inactivation (>7 logs) of microbes in juices, there were minimal effects on colour and pH; but the phenolics, antioxidants, and anthocyanins were affected greatly. The results of this study will inform the food industry on how to use this technology effectively.   iv  Lay Summary  The modern consumers are more aware about food and nutritional science than before and demand safe as well as fresh-tasting, nutritious food. Due to these trends, the food industry is shifting towards processes which can render food safe and keep them fresh-like with minimal processing. Pulsed UV light is one such process which use light pulse application for processing of foods. However, to adopt such technology especially for liquid foods like juices and milk, proper studies involving process characterization, and validation is necessary. Therefore, in this research project, we carried out characterization of the process for treatment of milk, grape and watermelon juices using computational fluid dynamics, actinometry and biodosimetry. The liquid foods were inoculated with three test microorganisms and then treated by pulsed UV light to ascertain the level of microbial inactivation after processing. Also, changes in nutritional and physico-chemical properties of treated foods were studied. This project will help in developing an alternate technology for processing liquid foods. v  Preface The thesis is based on the work carried out at the University of British Columbia. The design of the experiments conducted was based on the research proposals submitted to the Supervisory Committee. I and Dr. Pratap-Singh designed the experimental procedures that were conducted. I carried out the CFD simulations which form a part of chapters 3 and 4. I performed other experimental works based on the published research works according to the defined methodologies in proposals with the help of fellow lab mates. These works have been included as a part of chapters 4, 5, 6 and 7. Chapter 2 is based on our review paper which was part of our collaboration in pulsed light research in our group. This was published as Mandal, R., Mohammadi, X., Wiktor, A., Singh, A., & Pratap Singh, A. (2020). Applications of pulsed light decontamination technology in food processing: An overview. Applied Sciences, 10(10), 3606. I was responsible for drafting the manuscript by conducting the literature search, data curation. Xanyar Mohammadi helped in drafting and reviewing the manuscript. Dr. Artur Wiktor gave his valuable suggestions. Dr. Anubhav Pratap-Singh planned the paper structure and reviewed and supervised the team. A version of chapter 3 has been published as Mandal, R., & Pratap-Singh, A. (2021). Characterization of continuous-flow pulsed UV light reactors for processing of liquid foods in annular tube and coiled tube configurations using actinometry and computational fluid dynamics. Journal of Food Engineering, 304, 110590. I carried out the experiments, carried out CFD simulations and drafted the original manuscript. Dr. Pratap-Singh reviewed the manuscript, gave his suggestions for improvements, and submitted the paper as a corresponding author.  Currently, chapters 4, 5, 6 and 7 are currently being drafted for submission in peer-reviewed journals.  Some of the research work carried out as a part of the thesis were presented as oral and poster presentations in various conferences as follows: vi  1) Mandal R, Pratap-Singh A. \u201cNumerical modeling and simulation of a continuous pulsed UV light food processing system\u201d (2021). Conference proceedings. IFT\/EFFoST Nonthermal processing workshop, Monterrey, Mexico, November 3-6, 2019. Oral 2) Mandal R, Pratap-Singh A. \u201cImpact of Pulsed UV light treatment on microbiological, nutritional and quality parameters of milk\u201d (2021). Dairy Science and Technology Symposium, Delivering with dairy: from primary production to primary purpose, Arhus University, Denmark, June 21-25, 2021. (Online) Oral 3) Mandal R, Pratap-Singh A. \u201c Pulsed UV Light Treatment of Milk: Influence on Microbiological and Quality Parameters\u201d (2021). IFT Annual event, July 18-21, 2021. (Online) Oral & Poster 4) Mandal R, Pratap-Singh A \u201cEffects of pulsed UV light processing on microbiological, nutritional, and quality characteristics of bovine milk\u201d (2021) 35th EFFoST International Conference, Lausanne, Switzerland, November 1-4, 2021. Poster The figures (especially graphs, bar diagrams and schematics) used in this thesis have been generated by me using Microsoft Excel and PowerPoint. CFD simulations images have been exported from CFD software (ANSYS) and slightly differ from the versions published online. Other figures are drawn by me based on previously published papers. Data used in Tables have been curated from published works or based on experimental calculations.   vii  Table of Contents  Abstract\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026.iii Lay Summary\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026....iv Preface\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026.v Table of Contents\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026.vii List of Tables\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026..x List of Figures\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026..xii List of Abbreviations\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026..xxi Acknowledgements\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026...xxiii Dedication\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026...xxv Chapter 1: Introduction ................................................................................................................... 1 1.1 Brief concepts .............................................................................................................. 1 1.2 Thesis background ....................................................................................................... 2 1.3 Objectives, hypothesis, and research design ................................................................ 2 Chapter 2: Literature review ........................................................................................................... 4 2.1 Pulsed UV light technology ......................................................................................... 4 2.2 Effect of PL on food quality characteristics and nutritional value ............................ 35 2.3 PL processing of liquid foods .................................................................................... 41 2.4 Challenges with PL processing of liquid foods ......................................................... 47 Chapter 3: Design & Characterization of PL reactors in annular tube and coiled tube configurations using chemical actinometry and computational fluid dynamics ........................... 50 3.1 Summary .................................................................................................................... 50 viii  3.2 Introduction ................................................................................................................ 50 3.3 Materials and methods ............................................................................................... 51 3.4 Results and discussion ............................................................................................... 63 3.5 Conclusion ................................................................................................................. 83 Chapter 4: Microbial validation and UV dose determination for continuous-flow pulsed UV light reactors .......................................................................................................................................... 84 4.1 Summary .................................................................................................................... 84 4.2 Introduction ................................................................................................................ 84 4.3 Materials and methods ............................................................................................... 86 4.4 Results and discussion ............................................................................................... 93 4.5 Conclusion ............................................................................................................... 111 Chapter 5: Inactivation of microorganisms in milk, grape, and watermelon juices ................... 112 5.1 Summary .................................................................................................................. 112 5.2 Introduction .............................................................................................................. 112 5.3 Materials and methods ............................................................................................. 114 5.4 Results and discussion ............................................................................................. 122 5.5 Conclusion ............................................................................................................... 143 Chapter 6: Effect of pulsed UV light treatment of physico-chemical and nutritional parameters of milk ............................................................................................................................................. 144 6.1 Summary .................................................................................................................. 144 6.2 Introduction .............................................................................................................. 144 6.3 Materials and methods ............................................................................................. 145 6.4 Results and discussion ............................................................................................. 151 ix  6.5 Conclusion ............................................................................................................... 186 Chapter 7: Effect of pulsed UV light treatment of physico-chemical and nutritional parameters of red grape and watermelon juice .................................................................................................. 188 7.1 Summary .................................................................................................................. 188 7.2 Introduction .............................................................................................................. 188 7.3 Materials and methods ............................................................................................. 189 7.4 Results and discussion ............................................................................................. 194 7.5 Conclusion ............................................................................................................... 216 Chapter 8: Conclusion, significance, limitations, and future directions ..................................... 217 8.1 Modeling and simulation of PL processing ............................................................. 217 8.2 Validation using challenge microorganisms in model and real liquid foods ........... 218 8.3 PL-treated milk\u2014changes in nutritional and quality parameters ............................ 219 8.4 PL-treated juices\u2014changes in nutritional and quality parameters .......................... 220 8.5 Significance .............................................................................................................. 220 8.6 Limitations ............................................................................................................... 221 8.7 Future research directions ........................................................................................ 221 Bibliography\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026...223 Appendices\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026..243 Appendix A1: Supplemental information for Chapter 3\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026..243 Appendix A2: Supplemental information for Chapter 6\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026..246 Appendix A3: Supplemental information for Chapter 7\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026\u2026..249  x  List of Tables Table 2.1 Previously published research on PL processing ............................................................ 5 Table 2.2 PL systems for food decontamination .......................................................................... 12 Table 2.3 The effects of PL treatments on microbial inactivation in vitro based on several published data ............................................................................................................................... 22 Table 2.4 The effects of PL on some liquid food products based on several published research 28 Table 2.5 Continuous PL systems for processing of liquid food products ................................... 42  Table 3.1 Calculated properties of the liquid products ................................................................. 53  Table 3.2 Properties defined or the domains in the CFD code ..................................................... 61  Table 3.3 Values of constants for Eq. 3.21 for air and the model liquid foods ............................ 66 Table 3.4 Residence time distribution (RTD) parameters for AT and CT reactors ...................... 69  Table 3.5 Simulated results for the flow velocity, pressure drop and residence time .................. 76  Table 3.6 Computed values of total fluence rates based on volume-averaged irradiance for different liquids at various flow rates and 3 Hz frequency ........................................................... 80 Table 4.1 Calculated total delivered fluence (Fo,CFD) [J\/cm2] values for Annular (AT) reactor .. 93 Table 4.2 Calculated total delivered fluence (Fo,CFD) [J\/cm2] values for coiled (CT) reactor ...... 94 Table 4.3 Values of constant obtained by fitting the survivor curves for water, water+red dye, water+green dye, and skim milk in AT reactor .......................................................................... 101 Table 4.4 Values of constant obtained by fitting the survivor curves for water, water+red dye, water+green dye, and skim milk in CT reactor........................................................................... 105 Table 4.5 Calculated D-values of microbial strains in water, water+red dye, water+green dye, and skim milk .............................................................................................................................. 110 Table 4.6 Reduction equivalent fluence (REF) values for different treatment conditions ......... 111 Table 5.1 Calculated thermophysical and optical properties of milk (various fat%), red grape, and watermelon juice ......................................................................................................................... 117 Table 5.2 Experimental design of the PL treatment of liquid foods ........................................... 119 Table 5.3 Values of constant obtained by fitting the survivor curves for milk, grape, and watermelon juice in AT reactor .................................................................................................. 131 Table 5.4 Values of constant obtained by fitting the survivor curves for milk, grape, and watermelon juice in CT reactor ................................................................................................... 136 xi  Table 6.1 Experimental design of the PL treatment of milk ....................................................... 147 Table 6.2 Effect of different treatments on pH of milk samples ................................................ 152 Table 6.3 Effect of different treatments on L* of milk samples ................................................. 155 Table 6.4 Effect of different treatments on a* of milk samples .................................................. 158 Table 6.5 Effect of different treatments on b* value of milk samples ........................................ 161 Table 6.6 Effect of different treatments on total colour difference (\u2206E) of milk samples ......... 164 Table 7.1 Experimental design of the PL treatment of red grape and watermelon juice ............ 190 Table 7.2 Effect of PL treatments on pH and colour of watermelon juice ................................. 195 Table 7.3 Effect of PL treatments on pH and colour of red grape juice ..................................... 196     xii  List of Figures Figure 2.1 Increase in the number of publications in the field of pulsed UV light ........................ 5 Figure 2.2 Fluence concept. Total cumulative radiant energy incident on the small sphere from all directions per unit cross section area of the sphere .................................................................... 8 Figure 2.3 Interaction of Light with product. When light falls on a body, part of it reflects, part of it gets absorbed and part of it transmitted into the inner layers .................................................... 10 Figure 2.4 PL generation system. PL generation system consists of an electrical power source from where electric energy enters the capacitors, which accumulate the energy and get discharged into high-intensity light pulses ................................................................................... 15 Figure 2.5 DNA bases photoproduct formation due to UV application ....................................... 19 Figure 2.6 Penetration depth of light in food products ................................................................. 32 Figure 2.7 Annular thin-film PL reactor ....................................................................................... 45 Figure 2.8 Taylor-Couette reactor................................................................................................. 46 Figure 2.9 Coiled tube or Dean flow reactor ................................................................................ 46 Figure 3.1 PL reactor configurations. Schematic representation of (a) AT and (b) CT Reactor .. 54 Figure 3.2 Light intensity measurement through liquid solutions ................................................ 56 Figure 3.3 Meshed geometry of (a) AT reactor and (b) CT reactor ............................................. 60 Figure 3.4 Scatter-plots of light energy distribution in (a) x-z plane, (b) y-z plane ..................... 64 Figure 3.5 Light energy distribution contours in (a) x-z plane, (b) y-z plane ............................... 65 Figure 3.6 Absorbed UV-C energy dose. Graphical representation of absorbed UV-C dose by actinometry for (a) AT reactor, (b) CT reactor. Numbers (Annular_# or Coiled tube_#)  in the figure legends denote pulse frequency [Hz] ................................................................................. 71 Figure 3.7 Velocity distribution. Velocity distribution for (a) AT reactor (0.162 m\/s) shown by velocity vectors; (b) CT reactor (0.28 m\/s) velocity contours in YZ plane at x=0; XY plane at z =0, with inlet and outlet profiles ................................................................................................... 74 Figure 3.8 Irradiance distribution in air with minimum, maximum and volume-averaged irradiance. Contours of irradiance profile in XZ plane at y=0 and YZ plane at x=0 .................... 77 Figure 3.9 Irradiance distribution with minimum, maximum and volume-averaged irradiance. Irradiance contours for different liquids in XZ plane for (a) AT reactor at y=0 and YZ plane at x= 0, -30.5, -15, 15, and 30.5 cm. (b) CT reactors in XZ plane at y=0 ........................................ 78 xiii  Figure 3.10 PL process Uniformity. Graphical representations of (a) Velocity profile for AT reactor; (b) Irradiance profile for AT reactor ................................................................................ 81 Figure 3.11 PL process Uniformity. Graphical representations of (a) Velocity profile for CT reactor; (b) Irradiance profile for CT reactor ................................................................................ 81 Figure 4.1 Particle position at time \ud835\udc95\ud835\udfcf (point 1) and point 2 at time \ud835\udc95\ud835\udfd0 ........................................ 88 Figure 4.2 Particle x, y, z positions at time t in the reactor .......................................................... 88 Figure 4.3 Collimator tube experimental setup............................................................................. 90 Figure 4.4 Inactivation curves of microorganisms in different liquids in annular (AT) reactor (a) Water, (b) water+red dye, (c) water+green dye, and (d) skim milk ............................................. 95 Figure 4.5 Inactivation curves of microorganisms in different liquids in coiled (CT) reactor (a) Water, (b) water+red dye, (c) water+green dye, and (d) skim milk ............................................. 97 Figure 4.6 Fitting of survivors curves of microorganisms into log-linear with tail model in annular (AT) reactor for (a) water, (b) water+red dye, (c) water+green dye, and (d) skim milk . 99 Figure 4.7 Fitting of survivors curves of microorganisms into log-linear with tail model in coiled (CT) reactor for (a) water, (b) water+red dye, (c) water+green dye, and (d) skim milk ............ 100 Figure 4.8 Dose-response curves for different microorganisms in water, water+red dye, water+green dye, and skim milk ................................................................................................. 109 Figure 5.1 Inactivation curve for microorganisms in milk of different fat % in annular (AT) reactor. (a) E. coli, (b) L. innocua and (c) C. sporogenes ........................................................... 123 Figure 5.2 Inactivation curve for microorganisms in milk of different fat % in coiled (CT) reactor. (a) E. coli, (b) L. innocua and (c) C. sporogenes ........................................................... 124 Figure 5.3 Inactivation curve for E. coli, L. innocua and C. sporogenes in fruit juices. (a) red grape juice (annular), (b) red grape juice (coiled), (c) watermelon juice (annular) and (d) watermelon juice (coiled) ........................................................................................................... 126 Figure 5.4 Fitting of survivors curves of microorganisms into log-linear with tail model in annular (AT) reactor for (a) 3.25%, (b) 2%, (c) 1%, (d) 0%, (e) red grape juice, and (f) watermelon juice ......................................................................................................................... 129 Figure 5.5 Fitting of survivors curves of microorganisms into log-linear with tail model in coiled (CT) reactor for (a) 3.25%, (b) 2%, (c) 1%, (d) 0%, (e) red grape juice, and (f) watermelon juice..................................................................................................................................................... 129 xiv  Figure 6.1 Effect of PL treatment on vitamin C content of milk samples of different fat% in AT reactor. (a) 3.25%, (b) 2%, (c) 1%, and (d) 0%. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols ....................................................................................................................................... 166 Figure 6.2 Effect of PL treatment on vitamin C content of milk samples of different fat% in CT reactor. (a) 3.25%, (b) 2%, (c) 1%, and (d) 0%. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols ....................................................................................................................................... 168 Figure 6.3 Effect of PL treatment on vitamin B2 content of milk samples of different fat% in AT reactor. (a) 3.25%, (b) 2%, (c) 1%, and (d) 0%. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols ....................................................................................................................................... 171 Figure 6.4 Effect of PL treatment on vitamin B2 content of milk samples of different fat% in CT reactor. (a) 3.25%, (b) 2%, (c) 1%, and (d) 0%. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols ....................................................................................................................................... 173 Figure 6.5 Effect of PL treatment on lipid oxidation content of milk samples of different fat% in AT reactor. (a) 3.25%, (b) 2%, (c) 1%. A * symbol with the bar beneath denotes significant xv  difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols .... 176 Figure 6.6 Effect of PL treatment on lipid oxidation content of milk samples of different fat% in CT reactor. (a) 3.25%, (b) 2%, (c) 1%. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols .... 177 Figure 6.7 Effect of PL treatment on protein oxidation content of milk samples of different fat% in AT reactor. (a) 3.25%, (b) 2%, (c) 1%, and (d) 0%. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols ....................................................................................................................................... 179 Figure 6.8 Effect of PL treatment on protein oxidation content of milk samples of different fat% in CT reactor. (a) 3.25%, (b) 2%, (c) 1%, and (d) 0%. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols ....................................................................................................................................... 181 Figure 6.9 Rotated biplot for the principal components for various analyses carried out for milk (3.25%) in (a) AT and (b) CT reactor ......................................................................................... 184 Figure 7.1 Changes in vitamin C content in PL-treated watermelon juice in annular (AT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols xvi  +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols ................................................. 201 Figure 7.2 Changes in vitamin C content in PL-treated watermelon juice in coiled (CT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols ................................................. 201 Figure 7.3 Changes in trans-Resveratrol content in PL-treated red grape juice in annular (AT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols ............................................................................................ 202 Figure 7.4 Changes in trans-Resveratrol content in PL-treated red grape juice in coiled (CT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols ............................................................................................ 203 Figure 7.5 Changes in total phenolics content in PL-treated red grape juice in annular (AT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols ................................................. 204 Figure 7.6 Changes in total phenolics content in PL-treated red grape juice in coiled (CT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols ................................................. 204 xvii  Figure 7.7 Changes in total phenolics content in PL-treated watermelon juice in annular (AT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; Symbols +, ^, &, and # indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols ............................................................................................ 205 Figure 7.8 Changes in total phenolics content in PL-treated watermelon juice in coiled (CT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; Symbols +, ^, &, and # indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols ............................................................................................ 205 Figure 7.9 Changes in antioxidant activity of PL-treated red grape juice in annular (AT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols ................................................. 206 Figure 7.10 Changes in antioxidant activity of PL-treated red grape juice in coiled (CT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols ................................................. 207 Figure 7.11 Changes in antioxidant activity of PL-treated watermelon juice in annular (AT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, and # indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols ................................................. 207 Figure 7.12 Changes in antioxidant activity of PL-treated watermelon juice in coiled (CT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three xviii  experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, and # indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols ................................................. 208 Figure 7.13 Changes in total anthocyanin content in PL-treated red grape juice in annular (AT) reactor. A * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols .... 209 Figure 7.14 Changes in total anthocyanin content in PL-treated red grape juice in coiled (CT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, and # indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols ................................................. 209 Figure 7.15 Changes in lycopene content in PL-treated watermelon juice in annular (AT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, and # indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols ................................................. 210 Figure 7.16 Changes in lycopene content in PL-treated watermelon juice in coiled (CT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, and # indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols ................................................. 211 Figure 7.17 Rotated biplot for the principal components for various analyses carried out for grape juice in (a) AT and (b) CT reactor .................................................................................... 212 Figure 7.18 Rotated biplot for the principal components for various analyses carried out for watermelon juice in (a) AT and (b) CT reactor .......................................................................... 214 xix  Figure A1-1 PL equipment setup: Annular and coiled tube reactors, PL lamp and controller... 243 Figure A1-2: PL flashlamp emission spectrum (200-670 nm) as provided by Solaris Disinfection Inc.  ............................................................................................................................................. 243 Figure A1-3 Standard curve for conductivity vs concentration of NaCl used in residence time distribution studies ...................................................................................................................... 244 Figure A1-4 Predicted energy as per Eq. 3.21 versus experimental energy values .................... 244 Figure A1-5 Residence time distribution (curve) showing NaCl concentration vs time for real and ideal reactors. For ideal reactor, the spread of the curve is narrower compared to the real reactor ......................................................................................................................................... 245 Figure A2-1 Standard curve for vitamin C standard solutions used for quantification of vitamin C in milk samples ........................................................................................................................... 246 Figure A2-2 Standard curve for vitamin B2 standard solutions used for quantification of vitamin B2 in milk samples...................................................................................................................... 246 Figure A2-3 Chromatogram for determination of Vitamin C content in vitamin C standard solution. Peak is shown at retention time of around 2 min ......................................................... 247 Figure A2-4 Chromatogram for determination of Vitamin C content in milk samples. Peak is shown at retention time of around 2 min .................................................................................... 247 Figure A2-5 Chromatogram for determination of Vitamin B2 content in vitamin B2 standard solution. Peak is shown at retention time of around 5.7 min ...................................................... 247 Figure A2-6 Chromatogram for determination of Vitamin B2 content in milk samples. Peak is shown at retention time of around 5.7 min ................................................................................. 248 Figure A3-1 Standard curve of gallic acid solutions for estimation of total phenolic content in juice samples ............................................................................................................................... 249 Figure A3-2 Standard curve for trans-Resveratrol standard solutions used for quantification of trans-Resveratrol in grape juice samples .................................................................................... 249 Figure A3-3 Chromatogram for determination of Vitamin C content in watermelon juice samples. Peak is shown at retention time of around 2 min ......................................................... 250 Figure A3-4 Chromatogram for determination of trans-Resveratrol content in trans-Resveratrol standard solution. Peak is shown at retention time of around 17 min ........................................ 250 xx  Figure A3-5 Chromatogram for determination of trans-Resveratrol content in grape juice sample. Peak is shown at retention time of around 17 min ...................................................................... 250 xxi  List of Abbreviations .csv Comma separated value file 3-D Three dimensional AC Alternating current  ALA 5-aminolevulinic acid ANOVA Analysis of variance AT Annular tube reactor ATCC American type culture collection BHI Brain heart infusion BHT Butylated hydroxy toluene BSA Bovine serum albumin BSPL Broad spectrum pulsed light C-3-GE Cyanidin-3-glucoside equivalent CE Catechin equivalents CFD Computational fluid dynamics CFU Colony forming units CIE International Commission on Illumination Corp. Corporation CP Cold plasma CPU Central processing unit CT Coiled tube reactor DAD Diode array detector DF Dilution factor DNA Deoxyribonucleic acid DNPH 2,4-dintrophenyl hydrazine DO Discrete ordinance FC Folin-Ciocalteau FDA United States Food and Drug Administration GAE Gallic acid equivalents H202 Hydrogen peroxide HAV Hepatitis A virus HIPL High-intensity pulsed light HPLC High performance liquid chromatography HPP High-pressure processing HTST High temperature short time I3- Triiodide ions ILP Intense light pulses IR Infrared radiation K3Fe(C2O4)3 Potassium ferrioxalate KI\/IO3- Potassium iodide-iodate  xxii  LB Lysogeny broth LSD Least significant difference MDA Malondialdehyde MNV Murine norovirus MRSA Methylene resistant Staphylococcus aureus MTCC Microbial Type Culture Collection and Gene Bank MW Microwave NCTC National collection of type cultures NDe Dean number NRe Reynolds number oBx Degree brix oC Degree Celsius OD Optical density OMF Oscillating magnetic field OSI Oxygen stability index PCA Principal component analysis PEF Pulsed electric field PES Polyether sulfone pH Potential of hydrogen PL Pulsed UV light PWL Pulsed white light RAM Random access memory RCM Reinforced Clostridial medium REF Reduction equivalent fluence  RF Radio frequency RMSE Root mean squared error RTD Residence time distribution SEM Scanning electron microorganisms TEAC Trolox equivalent antioxidant capacity TPC Total phenolic content TS% Total solid % TT 6,4 PP thymine-thymine 6,4 photoproduct TT CPD thymine-thymine cyclobutene pyrimidine dimers UHT Ultra high temperature US Ultrasonication UV Ultraviolet UV-C Ultraviolet-C region UV-LED Ultraviolet-Light emitting diode Vis-NIR Visible-near infrared  \u0394E Total colour difference \u0394P Pressure drop xxiii  Acknowledgements While this part comes sequentially at the beginning in the Table of Contents, I kept procrastinating writing it until the very end of this thesis. It is because I truly want to acknowledge the support of each and every one who has played their part in this wonderful three odd years. The time period of my PhD research went so fast! First and foremost, I want to acknowledge the constant support of my PhD supervisor Anubhav Pratap-Singh. I was fortunate that we are from same alma mater. Thank you for the chance you gave me in your laboratory by bestowing your belief in me from day one. I also want to thank you for allowing me to TA and take tutorials in the Food Processing course that helped me gain some experience in giving lectures. Starting from help related with my research, to any other issues and well-being, you were always available. Your advice helped me a lot to grow as a researcher. Secondly, I would like to thank my supervisory committee members Dr. Fariborz Taghipour, Dr. Derek Dee for their continuous guidance, tips, inputs and help with understanding and developing my research  into a well-rounded thesis. I would like to thank  Dr. Xiaonan Lu (though he only initially served as a member and left as he had to move to McGill university). I would also like to thank Dr. Yada, Dr. David Kitts, Dr. Christine Scaman, Dr. Simone Castellarin. I was fortunate to have interacted with you all. I would take this opportunity to thank Peter Hoffman, who has always ensured that the laboratories and the equipment in the department are functional. His training and technical support over the course of my research has played an important role in ensuring timely completion of my research work. I would also like to thank Dr. Joana Pico Carbajo for helping me with analytical chemistry. I always enjoyed having a talk with you. I would also like to thank Dr. Artur Wiktor from Warsaw University of Life Sciences, Poland, with whom I had the opportunity to work initially on my research project. I really enjoyed hiking with you and glad to catch up with you recently in Switzerland at the 35th EFFoST international conference.  I would like to acknowledge the support of Solaris Disinfection Inc. team, especially Val Ramanand, Robby Dhillon, Keval, and Arjun. Thanks for providing funding for the project and most importantly the lamps and pulsed UV light unit. Without this nothing would have been xxiv  possible. I really appreciate your technical support over the years. I would like to thank the funding agencies NSERC and also the support of departmental for supporting me financially. I want to extend my thanks to families of Mary and David Macaree, Shuryo Nakai, Eunice Li-Chan, and also Kin\u2019s Farm Market for their financial awards and support extended to me.   I also must specially thank my lab mates Dr. Anika Singh, Dr. Farahnaz Fathardoobady, Dr. Alberto Baldelli, Yigong Guo, Mithun Dey, Xanyar Mohammadi, Maryam Shojaei, Rocio, Loraine, Hannah, Andrew, Madeleine, Winnie, Yuan, Sofia. Thank you, Yigong for helping me with the HPLC and chemical analysis. Thank you, Alejandro for your straight-forward and easy-going attitude. Thank you, Philip for always making the us laugh with your jokes and puns, and especially our four-cool guys montages. Thanks Madeleine and Winnie for helping me with experiments and chemical analyses. I would also like to extend my gratitude to my roommates during this period, Mrinmoy and Santanil. We had so much fun together and I enjoyed our trips to Sechelt, Yosemite, Victoria, and Tofino. I also want to thank my other roommates especially Jerry and Deb. I also thank Mehtab, who helped me initially when I first moved to Vancouver. I met many other people whom I would like to name and thank, but that would be a lengthy list. Thanks everyone who have been there for this exciting journey. Special thanks are owed to my parents, Mr. Himangshu Mandal and Mrs. Shanti Mandal who have made sacrifices and supported me throughout my years of education, both morally and financially. Love and care for my sweet sister Rittika, with whom I share a special bond and have fun together. Lastly and most importantly, I want to thank my lovely wife Payal Karmakar from the bottom of my heart. We met, fell in love and now we have tied the wedding knot. Thanks for being with me and tolerating me since the last ten years. Thanks for making me a better person and teaching me so many things. I am also grateful that we shared our journey since our undergraduate degree and as PhD students. Through the lows of the seminar presentation, meeting deadlines, submission, to the highs of travelling around India, and Switzerland, thank you for being there. Without your constant support and sacrifice, my PhD would have been impossible.   xxv  Dedication   To the endless pursuit of unravelling the mysteries of sciences                      1  Chapter 1: Introduction 1.1 Brief concepts  Currently, there is a growing concern for nutrition and food safety. With the advent of technology, and the urge of seamlessly researching anything on the internet, the consumers of the 21st century are more aware and demanding than ever before. They want safe, sound, and healthy foods of good quality, having the adequate supply of required nutrients, but devoid of any preservatives [1].  Conventionally, the food industry has been using thermal processing for extension of shelf-life of foods since time immemorial for processing of foods. The main objectives of thermal food processing are to reduce microbial load in food, inactivate food quality-degrading enzymes, product transformation (protein denaturation, starch gelatinization) [1]. Thermal processing formally started with the works of Nicholas Appert on canning in 1809, and the works of Dr. Louis Pasteur (1822-1895) on microbial basis of food spoilage, followed by works of scientists on laying down the principles of thermal processes (development of ultra-high temperature (UHT) and high temperature-short time (HTST) treatment) during 1950s. However, the deleterious effect of heat on the nutritional attributes (destruction of vitamins) and sensory qualities (colour changes and  off-flavour development) of food is now also a common knowledge [2, 3]. The food industry is experiencing a paradigm shift to minimal and non-thermal technologies. This has prompted food processing researchers to develop various technologies like ohmic heating, radiofrequency (RF), microwave (MW), and infrared (IR) heating, high-pressure processing (HPP), ultrasound (US), pulsed electric field (PEF), pulsed light (PL), Cold plasma (CP), oscillating magnetic field (OMF) etc. as alternates to the conventional thermal processes. These have a potential to gradually replace the thermal processing practiced by the food industry, especially because these are carried out at ambient and sub-lethal temperatures and in this way do not have antagonistic impacts unlike thermal processing technologies.  PL is a light-based non-thermal technology used for decontamination of food products. The technology is based on using broad spectrum light ranging from germicidal ultraviolet (UV) region to infrared region targeted for inactivation of microorganisms in various solid and liquid foods. While solid foods are suitable for PL treatment because of presence of microorganisms on their surface, liquid foods pose a challenge due to limited penetration of light in the liquids. Therefore, 2  to extend the PL process for liquid food processing, a thorough investigation is necessary. For the current thesis research work, PL processing is studied, characterized, and validated by developing continuous processing systems for liquid food processing. 1.2 Thesis background In the framework of the current study, an equipment was designed and tested for lamp energy distribution, reactor hydrodynamic performance, actinometry and computational fluid dynamics. Based on the findings, the reactors were validated for model (coloured liquids, water, skim milk) and real liquid foods (milk of different fat% and fruit juices) by microbial challenge studies. These liquid foods were also assessed for any changes in their nutritional and quality characteristics. 1.3 Objectives, hypothesis, and research design The overall objective of this doctoral degree\u2019s dissertation was to design, develop,  characterize and validate continuous-flow pulsed UV light reactors for processing of liquid foods viz. milk, red grape, and watermelon juice. 1.3.1 Hypothesis The central hypothesis of my doctoral research was that despite the limited penetration of light in liquid foods, a thin-profile treatment close to that of the penetration depth of light in liquid foods will enable reaching 5-log reductions in liquid foods. The designing and characterization of thin-film reactors in annular and coiled configuration based on residence time distribution studies, UV dose determination, computational fluid dynamics, followed by microbial validation can help in development of  PL technology for processing of milk, red grape, and watermelon juice. Due to its non-thermal nature, quality characteristics, nutritional value of these liquid foods can be retained.  The hypothesis was tested on milk and fruit juices by developing a continuous processing system. 1.3.2 Specific objectives  The hypothesis was tested by the following specific objectives: 3  Objective 1: Design & Characterization of PL reactors in annular tube and coiled tube configurations using hydrodynamics, chemical actinometry and computational fluid dynamics studies Objective 2: Microbial validation and UV dose determination for continuous-flow pulsed UV light reactors Objective 3: Inactivation of microorganisms in milk, grape, and watermelon juices Objective 4: Effect of pulsed UV light treatment of physico-chemical and nutritional parameters of milk Objective 5: Effect of pulsed UV light treatment of physico-chemical and nutritional parameters of red grape and watermelon juice  4  Chapter 2: Literature review 2.1 Pulsed UV light technology  Pulsed UV light (PL) technology has emerged in the recent decades as a novel and non-thermal technology for food processing. PL treatment generally consist of applying a series of high-intensity, short bursts of light pulses to inactivate microorganisms. PL is synonymous with the terms like intense light pulse (ILP), high-intensity broad-spectrum pulsed light (BSPL), high-intensity pulsed UV light (HIPL) or pulsed white light (PWL) [4,5]. PL has a great potential to enhance the keeping quality of food products [6,7] by destroying microorganisms such as Listeria monocytogenes within a few seconds [1]. PL was successful in reducing the total microbial load in vitro when inoculated on surface of agar plates [1]. So, it can be claimed to have potential to preserve food materials.  In addition, PL technology can be effectively used to sterilize packaging materials [8], and equipment surfaces [9,10]. However, PL technology has its limitations to be used for food applications due to the opacity of food products and their non-uniform surfaces, possible temperature rise leading to deterioration in organoleptic qualities. In 1996, the U.S. Food and Drug Administration (FDA) approved the PL technology applications \u2018for production, processing and handling of foods\u2019 and decontaminating food contact surfaces [11]. They recommended the use of PL treatment using xenon lamp having surface emission of wavelengths (\u03bb) between 200-1100 nm, with the cumulative treatment of not more than 12 J\/cm2 [4]. Since then, a great increase in the research and development in the field of PL technology in food industry has taken place. As a result, more than 250 articles including reviews, patents, chapters have been published till now.  Figure 2.1 shows a steady rise in publications after the year 2000. Eventually, numerous PL systems and configurations have been manufactured. PL technology can be thought of as a successful intervention of processing fresh-cut fruits and vegetables, equipment-contact surfaces, clear liquid foods in continuous flow-through modules etc. to decontaminate them. Efforts are on to extend it to turbid and opaque liquid foods, seeds, powders, ready-to-eat meals, and other thermally-labile food commodities. With the growing demand for safe and nutrient rich foods, thorough research should make the PL process ready to be adopted by the food industry sooner or later. Thus, the main objective of this chapter is to review the recent developments in PL processes and applications that have been in practice. It is apparent 5  that there have been many reviews earlier about PL technology and they have been considered for developing and formulating this literature review. The Table 2.1 discusses the particulars of some of those reviews conducted earlier. Thus, after reading all those reviews we tried to incorporate the latest trends in the PL processing and develop this literature review, which formed the bedrock of my doctoral dissertation. Processing systems that have been developed earlier are summarized. The detailed inactivation mechanisms of microorganisms by using the PL technology are also discussed within the sections. Further, the different PL applications in the food industry, effects on food quality, various challenges, are also highlighted.   Figure 2.1: Increase in the number of publications in the field of pulsed UV light Table 2.1: Previously published research on PL processing 6  Paper Title, Authors, and Year of Publication Key Contents Pulsed light technology;  Barbosa-Canovas et al. [12] \u2022 Systems developed for PL processing,  \u2022 Possible bacterial inactivation mechanisms, \u2022 Critical factors affecting the PL process efficacy, \u2022 Results from industrial sources, \u2022 Research needs Pulsed light for food decontamination: a review; G\u00f3mez-L\u00f3pez et al. [13] \u2022 Terminologies related to PL,  \u2022 Microbial inactivation mechanisms, \u2022 Factors affecting the inactivation, \u2022 Decontamination of food, \u2022 Effects on nutritional and toxicological aspects Pulsed-light system as a novel food decontamination technology: a review;  Elmnasser et al. [14] \u2022 Principles of PL,  \u2022 Microbial inactivation in vitro, \u2022 Main limitations, \u2022 Decontamination of several food products Pulsed Light Treatments for Food Preservation. A Review; Oms-Oliu et al. [15] \u2022 Principles of PL, \u2022 Microbial inactivation in vitro, \u2022 Microbial inactivation in several foods, \u2022 Critical process factors, \u2022 Microbial inactivation on packaging materials and food contact surfaces, \u2022 Limitations The Sensory Quality of Meat, Game, Poultry, Seafood and Meat Products as Affected by Intense Light Pulses: A Systematic Review;  Tomasevic and Rajkovic [16] \u2022 Effect of PL on the sensory quality of 16 varieties of meat, meat products, game, poultry, seafood, \u2022 Effect on colour and odour of meat Pulsed Light Treatment of Different Food Types with a Special Focus on Meat: A Critical Review; Heinrich et al. [17] \u2022 Principles of PL processing, \u2022 Efficacy of processing unpackaged and packaged meat products, \u2022 Decontamination of meat cutting, packaging surfaces, \u2022 Industrial implementation needs Pulsed light processing of foods for microbial safety; Bhavya and Hebbar [18] \u2022 Principles of PL processing, PL processing systems, \u2022 Mode of action on microorganisms, \u2022 PL treatment on several food matrices, \u2022 Combination of PL, with other processing technologies 7  2.1.1 Overview of PL technology PL technology involves the application of intense light in the form of short pulses on a target of interest (food or food contact surface or illumination through a fluid food matrix) to destroy bacteria, yeasts, molds, and viruses present therein. The PL is nothing but ordinary white light like sunlight, except having a very high intensity, and applied over a very short period of time. Sunlight is an electromagnetic radiation, which is all around us. PL has an electromagnetic spectrum similar to that of sunlight, ranging from Ultraviolet (UV) to near infrared (IR) wavelengths [12]. This range of electromagnetic radiation includes ultraviolet (UV) rays (\u03bb = 200\u2013400 nm; UV is again divided into UV-C: 200\u2013280 nm, UV-B: 280\u2013315 nm, UV-A: 315\u2013400 nm), visible light (\u03bb = 400\u2013700 nm) and infrared (IR) rays (\u03bb = 700\u20131100 nm) [4,14]. However, the spectral output of PL being rich in UV regions unlike solar spectrum which is rich in visible regions [19] (pp. 236). The light pulses generation is carried out by excitation of inert gases, like xenon in flashlamps, and collision of gaseous molecules, due to electrical pulse application. The light energy is then released in the form of short-duration light bursts in highly concentrated manner (lasting for a few hundred microseconds, usually 1 \u03bcs to 0.1 s) with high peak power [18,20]. The light pulses produced are short-lived, but are of high intensity, summing up to around twenty thousand times the sunlight intensity at the sea level [1,6,14].  Broad spectrum pulsed light inactivates microorganisms by a combination of photo-chemical, photo-thermal, and photo-physical effects [21], which are discussed later. Mainly the lethal effect is owed to the photo-chemical effect where light photons of the UV-C wavelengths, particularly 253.7 nm, are absorbed by the microbial DNA. Thus, it shows the antimicrobial effect alike continuous UV processing [22].  2.1.2 Terms related to PL technology Similar to continuous UV processing, the rules and terms associated with photophysics and photochemistry laws have been applied for PL treatment characterization since its inception. Light based processes are mainly quantified in terms of the cumulative energy dose imparted during a treatment. Cumulative energy dosage is important so that the results of PL can be uniformly reported. Typically, the term Fluence is used to quantify the radiant energy for a treatment. Fluence (\ud835\udc39\ud835\udc5c) is the cumulative radiant energy comprised of all wavelengths coming from all direction 8  passing through an imaginary infinitesimal sphere with cross section \ud835\udc51\ud835\udc34 per unit area \ud835\udc51\ud835\udc34 (unit: J\/cm2) (Figure 2.2).   Figure 2.2: Fluence concept. Total cumulative radiant energy incident on the small sphere from all directions per unit cross section area of the sphere The PL inactivation efficacy is dependent upon the delivered fluence [4]. Parameters affecting the fluence (at a particular distance from lamp) are the pulse width (\ud835\udf0f), treatment time (\ud835\udc61), number of pulses (\ud835\udc5b), pulse repetition rate or frequency (\ud835\udc53) [23]. The fluence is mathematically calculated as energy for each pulse multiplied by the total number of pulses times pulse frequency as expressed by the equation 2.1 [24]. \ud835\udc39\ud835\udc5c = \ud835\udc52 \u00d7 \ud835\udc61 \u00d7 \ud835\udc53,                                                (2.1) Where, \ud835\udc39\ud835\udc5c = fluence [J\/cm2], \ud835\udc52 = energy of a pulse [J\/cm2 per pulse], \ud835\udc61 = treatment time [s] and \ud835\udc53 = pulse frequency [Hz]. Other terms related to PL like fluence rate [W\/cm2], spectral irradiance [W\/(cm2-nm)], radiance exposure [J\/cm2] etc. have been described by G\u00f3mez-L\u00f3pez and Bolton [23]. The researchers tried to introduce the fundamentals of photochemistry and biodosimetry as applicable to continuous UV light wave processing to PL tests.  Terms (described by G\u00f3mez-L\u00f3pez and Bolton [23]) Radiant exposure (\ud835\udc3b) \u2013 It is the total energy \u201cof all wavelengths incident from all upward directions on a small element of surface containing the point under consideration divided by the area of the element.\u201d [J\/cm2] Fluence rate (\ud835\udc38\ud835\udc5c) \u2013 Denotes the rate of applied fluence [W\/cm2] Irradiance (\ud835\udc38) \u2013 Denotes the rate of radiance exposure [W\/cm2] 9  Spectroradiometer devices are used to measure the irradiance and radiant exposure; however, it is difficult to measure treatment fluence. Other terms like  photon fluence (\ud835\udc39\ud835\udc5d,\ud835\udc5c) [Einsteinb\/cm2]a , spectral photon irradiance (\ud835\udc38\ud835\udc5d,\ud835\udf06) [Einstein\/s-cm2-nm], incident spectral photon irradiance (\ud835\udc38\ud835\udc5d,\ud835\udf06\ud835\udf0a ) [Einstein\/s-cm2-nm] have been defined by G\u00f3mez-L\u00f3pez and Bolton [23]. But these terms require collimatorc PL devices for quantification and these are not available yet. 2.1.3 Pulsed and continuous UV light Light energy as continuous UV has been used since a long time for disinfection purpose for decontamination in hospitals, laboratories etc. However, light pulse application is a slightly different approach. During PL application, the electrical energy is concentrated over a long period of time and discharged in a very short time in the form of light pulses. So, the short-lived high-intensity light flashes have apparently more efficacy than continuous UV light application and peak power provided by the pulses are more than what continuous wave of equivalent total energy provides. It may be due to the fact that, if the total radiant energy is equal for both, the shorter pulse duration leads to higher impulsive power. This is similar to the concept of \u2018impulsive force\u2019, wherein an impact force (analogous to using a hammer) produces greater impact if the time of impact is short. Pulsed UV systems can dissipate ~35 MW of impulsive power for a fluence level of 3 J\/cm2 [25]. On the contrary, the continuous wave UV system dissipates power in the range of 0.1 to 1 kW [26]. PL treatment is proven to be more efficient in reducing pathogens like Escherichia coli O157:H7 and L. monocytogenes than continuous UV light irradiation [17]. In a recent study done by Zou et al. [27], the results show E. coli inactivation by using pulsed UV-LED irradiation is more effective than continues UV. By applying the same dose, the inactivation rate of E. coli significantly rose as the duty cycle reduced from 100% to 5% [27]. In addition to a higher decontamination power, light pulses have a much higher penetrating ability through materials than continuous UV light [28,29]. Though it is quite debatable whether pulsed light has more penetration than continuous UV light. However, it can be supported by the conjecture that penetration of light waves depends on wavelength\u2014penetration depth of light in a substance increases as the wavelength of light decreases. As the PL includes wavelength from 200 nm-1100 nm, the lower wavelength of around 200 nm having higher penetration than continuous wave UV (considering it has wavelength around 253.7 nm). aSquared bracket [ ] denotes the unit of the term being defined; bOne Einstein = one mole of photon bCollimator system is a device consisting of a long tube placed between lamp and sample which produces parallel rays of light 10  There is some also heating attributed to the use of PL due to IR regions, which is beneficial as it may add to the decontamination process; however, it has created problems for PL application on fresh produces on the basis on degradation of sensory qualities [30]. The pulse frequency to some extent controls the heating; with lower frequencies leading to less heating as there is enough time for the generated heat to dissipate. Continuous UV light processing does not contribute to product heating as the UV light wavelengths lack IR region, which are associated with heating. 2.1.4 Interaction of light and matter Mathematically, the interaction of PL with matter can be studied by Lambert-Beer\u2019s law. When a light radiation of energy intensity \ud835\udc3c\ud835\udc5c falls on a product superficially, it gets transmitted into its depth and then gets absorbed by the layers of product [4] as in Figure 2.3.   Figure 2.3: Interaction of Light with product. When light falls on a body, part of it reflects, part of it gets absorbed and part of it transmitted into the inner layers The light fraction transmitted light having energy \ud835\udc3c(\ud835\udc65) is transmitted as a function of distance \u2018\ud835\udc65\u2019 under the product surface, the wavelength of light, and the material extinction (absorption) coefficient as per Lambert-Beer\u2019s law (Equation 2.2): Most food materials show a decay in light intensity while penetrating into their bulk ([4], which appears in the form of heat, leading to temperature rise, given by Equation (2.3). The extinction coefficient, \ud835\udefc in Equation (2) is related to the transparency or opacity of the material. Such temperature rise creates a temperature gradient (\ud835\udc51\ud835\udc47\ud835\udc51\ud835\udc51) (equation 2.4) between the surface and the internal layers of the product setting up conductive heat transfer (\ud835\udc58=thermal conductivity of material, [W\/m-K]) within the material.  \ud835\udc3c(\ud835\udc65) = \ud835\udc3c\ud835\udc5c \ud835\udc52(\u2212[\ud835\udefc]\ud835\udc65),                                           (2.2) 11  \u2206\ud835\udc47 =\ud835\udc3c\ud835\udc51\ud835\udf0c\ud835\udc36\ud835\udc5d\ud835\udc34\ud835\udc51 ,                                                   (2.3) where, \ud835\udefc = extinction coefficient; \ud835\udf0c = density of body [kg\/m3], \ud835\udc36\ud835\udc5d = specific heat of the body [J\/kg-K] and \ud835\udc34 = surface area of body [m2]. \ud835\udc3c\ud835\udc51 = absorbed energy (from the transmitted light) conducted up to a depth d, given by: \ud835\udc3c\ud835\udc51 = \ud835\udc3c(\ud835\udc65)(\ud835\udc3c\ud835\udc5c \ud835\udc52(\u2212[\ud835\udefc]\ud835\udc65)),                                  (2.4) \ud835\udc51\ud835\udc47\ud835\udc51\ud835\udc51=\ud835\udc3c\ud835\udc51\ud835\udc58\ud835\udc34 ,                                                    (2.5) This dissipated heat may be utilized for the decontamination of the food products [4]. However, it must be emphasized that this heating is not the primary mechanism of PL action on microbes (discussed in Section 2.1.7). Although, this heating may be utilized to enhance decontamination potential, and to potentially reach pasteurization temperatures in selected cases. Like, in a PL treatment at 59.4 J\/cm2, raspberries reached 75 \u00b0C from initial temperature of 30 \u00b0C, which reduced Salmonella by 3.4 logs and E. coli O157:H7 by around 3 logs [31]. On higher fluence level (72 J\/cm2) temperature raised up to 80 \u00b0C. Due to short duration of light pulses a steady-state thermal conduction is never achieved. Higher pulse frequencies, however, can lead to greater product heating. Nevertheless, only a thin surface layer gets rapidly heated sparing the inner layers of material unlike high overall temperature achievement as in continuous systems [4]. Thus, PL treatment can be claimed to be a \u2018non-thermal\u2019 processing method by virtue of the fact that the inner layers are not prone to any temperature rise ensuring destruction of pathogenic and spoilage causing microorganisms as well as retention of nutrient and bio-active components. 2.1.5 Merits of PL technology Overall, PL seems very effective for treating surfaces of food products, packaging materials as its action on a thin surface layer is sufficient enough to destroy superficial vegetative cells [21]. PL is thus a rapid process of decontamination of food and packaging surfaces. PL processing offers several benefits over conventional processing like bulk of the product remains around the ambient temperatures due to shorter energy delivery time leading to no damage of inner parts of products. The PL systems are advantageous as flashlamps used to generate light pulses in PL systems do not require mercury unlike continuous UV systems [17,24]. Apart from that, the PL systems also impart lesser risk from pathogens, improved quality of the product   than   thermal processing, better economics while food processing due to low operational costs, greater flexibility and leaves no chemical\/biochemical residues [17]. PL technology also stands out well as compared to other 12  non-thermal technologies like HPP and PEF. In case of HPP, food is needed to be usually packaged before treatment. Also, mild HPP treatment seems unable to inactivate spores, rather the process may promote spore germination [2]. On the other hand, PL effectively inactivates spores [18, 26] and decontaminate food and packaging materials. Food contact surface decontamination can also be achieved using PL [9,10]. PEF processing, although being an excellent food preservation technique, has some limitations like lower processing capacities (small treatment chambers), erosion of electrodes, electrochemical reactions, arcing [7]. These challenges could be mitigated to a great extent using PL processing systems. 2.1.6 Pulsed UV light equipment Initially, PL treatments were used for aesthetic applications like cosmetic, skin, and hair treatment areas and later on after approval of FDA in 1996, there has been tremendous development in the field of PL based disinfection has further been spread to food industries, pharmaceuticals, and water treatment plants. During the 1970s, Hiramoto [32] patented a pulsed UV light system for microbial inactivation [4]. Later on, works were done by Dunn et al. [33] who patented systems for sterilization of food and packaging materials. Since then, a major organization pioneering the PL based research in the world for food and drinking water decontamination was the Purepulse Technologies Inc., San Diego, California, subsidiary of Xenon Corporation [17]. They acquired Hiramoto\u2019s patent and did further research to commercialize the PL equipment for food applications under the name PureBright\u2122 system [4]. Subsequently, after several advancements in this field, the major companies manufacturing PL based decontamination systems are SteriBeam Systems from Germany, Wek-tec Systems from Germany, Xenon Corp. from USA, and Claranor from France. The Table 2.2 summarizes the available PL systems for food decontamination.  Table 2.2: PL systems for food decontamination 13  PL System   Characteristics References Pure Bright\u2122 System for sterilizing liquid products (fruit juices) \u2022 Annular PL processing chamber with PL lamp inside and housed inside a highly reflective material \u2022 Tube-quartz made; arrangement-spherical, spiral, etc. \u2022 Electrodes- metallic (tungsten) electrodes \u2022 Flashlamps-filled with inert gases like xenon, krypton, or a noble gases mixture at various pressures System for sterilization of a flexible film for aseptic packaging Flexible film drawn through rollers into an absorption enhancement solution and then making them into a cylindrical tube (with product filler and lamp inside) like structure, sealing the longitudinal and bottom seals. The lamp disinfects the interior. Then product fills them and film is drawn forward. Then another seal is done transversally to totally seal the first package and create bottom seal of next package. System for sterilizing preformed containers Packaging containers moved in line under the flashlamps and then pre-sterilized product fills them sequentially. [33] Wek-tec\u00ae  The system had a stainless steel enclosure,  Test area dimensions: 16 cm wide \u00d7 12 cm deep \u00d7 15 cm high. Flashlamps \u2013 linear; xenon filled Fluence- 0.1 J\/cm2 to 8 J\/cm2 Pulse duration - 200 \u00b5s [34] Sterilization system for inner surfaces of preformed containers The containers for packaging milk, juices placed on a conveyer belt, which move under the lamps that flash them. Then, the container progresses forward. [35] Bench top PL system Lamps-lamp house is at the top center and filled with Xenon gas. [36] 14  A typical PL system works by generation of light pulses in the pulse forming device. Pulses of light are produced by multiplying electrical power many folds by concentrating electrical energy in a capacitor over less than a second and releasing this energy in a relatively shorter duration (millionths or 10-6 of a second) to a flashlamps; the flashlamps, filled with inert gases give off light in the form of flashes [12]. The pulses per second or pulse frequency varies mostly as 1-3 Hz [24]. The pulsed light systems basically have of three main units [14] as given in schematic diagram (Figure 2.4): The control module and light source connected by a control cable, to modulate the electric current to get specific pulse repetition rate, pulse width and peak power. OneShot EN2\/2143\u20131 unit 3-fluidized bed PL system mounted with adjustable air nozzles to blow compressed air for fluidization of the food powders. Water circulation in surrounding quartz jacket around the flash lamp to counter overheating and a reflecting cylinder. [37] RDT350 model, La Calhene, USA Treatment chamber-250 cm diameter  Lamps no.-8 xenon lamps arranged on the periphery of the chamber Distance of samples from the xenon lamps was 13.5 cm  Lamps Fluence-1.5 J\/cm2  Pulse duration-300 \u00b5s  Wavelength-between 200\u20131200 nm [14] SteriPulse\u00ae XL-3000 system Fluence: 1.27 J\/cm2 per pulse Pulse duration-360 \u00b5s; 3 pulses\/s Broad spectrum UV with high germicidal power  Stainless steel sterilization chamber, removable lamp housing  [38] Claranor, France PL system with multiple xenon lamps [39] The LytBot, Solaris Disinfection Inc., Canada PL system for Hospital Disinfection   [37] Table 2.2: Continued 15  Power unit: comprising electrical power supplier where high-voltage DC power is obtained from low-voltage AC power. Pulse configuration device: comprising of high-voltage capacitors joined in parallel that concentrate energy from the power supplying unit in the charge cycle and release that during the discharge cycle, generating high electrical current. It is also connected to special high-power handling switches that perform on\/off cycles of very short time converting the continuous low-electrical power into a pulsed high-electrical power. Lamp unit: comprising specially designed batteries of flashlamps containing inert gases that are excited due to pulsed high-electrical power from the pulsing device. While transitioning to lower energy or ground states, they give off energy in the form of high-intensity pulses. The high-intensity PL is delivered to the products by various lamp configurations.   Figure 2.4: PL generation system. PL generation system consists of an electrical power source from where electric energy enters the capacitors, which accumulate the energy and get discharged into high-intensity light pulses 2.1.7 Fundamentals of microbial decontamination by PL PL leads to inactivation of microbes based on various extensive physico-chemical processes. The microbial destruction is generally attributed to mechanisms like photo-chemical, photo-thermal, and photo-physical effects. These mechanisms are responsible for inactivation of microorganisms 16  independently or in combination with each other, depending on treatment conditions and target organisms [13]. Although photo-thermal and photo-physical effects might have an influence on inactivation, most works indicate the photo-chemical effect of light to be the main factor that drives the destruction of microorganisms.   Other than the physico-chemical mechanisms, the spectral range and microbial morphology also govern the microbial destruction by PL to an extent. The effect of spectral range of light pulses has also been extensively studied. The spectral range found to be most efficient in microbial destruction was UV range [17], with UV-C component (200-280 nm) having the major role and a theoretical inactivation peak at 253.7 nm [13]. Wang et al. [41], however, reported that maximum reduction of E. coli took place at 270 nm, in the UV-C range. Ramos-Villarroel et al. [42] also obtained greater inactivation in fresh-cut avocadoes by using UV-C component than Vis-NIR spectra PL. However, the IR and visible regions of light radiation were not effective in PL based inactivation process [17]. The IR and visible regions only had the antibacterial effect at higher powers [14].  Microbial morphology does not show consistency with respect to PL treatment. Rowan et al. [43] concluded the susceptibility of microorganisms to PL in the manner: Gram-negative bacteria > Gram-positive bacteria > fungal spores. In another study, Anderson et al. [44] observed similar trend in microbial sensitivity. Contrarily, G\u00f3mez-L\u00f3pez et al. [45], however, did not report similar pattern of susceptibility in their studies. Spores have lower susceptibility than bacteria vegetative cells which was stated to be due to inherent DNA repair mechanisms of cell [44]. However, this tendency of repairing of DNA is adversely affected by PL application as compared to continuous UV application which is why PL is more effective [46]. Pigments in microorganisms can also be a determinant factor for PL efficacy. Anderson et al. [44] inferred that spore colour may have effect on action of PL. They showed that dark coloured spores of the mold Aspergillus niger were resistant to PL than pink coloured Fusrium culmorum, due to more absorption of UV-C light.  To answer some of the fundamental questions of microbial inactivation by PL, an understanding of behavior of light and its associated laws is necessary.  17  2.1.7.1 Quantum mechanical background of light The light was considered to behave like a wave, as per the studies and experiments conducted during the early 19th century. However, due to the works of Max Planck in late 19th century and followed by Albert Einstein in early 20th century, it became evident that light also behaves like particles, or quanta. Planck explained his blackbody radiation by assuming that light is quantized and consists of discrete stream of particles (called photons) with an energy proportional to the frequency of the light radiation. Einstein was also able to explain the photoelectric effect (for which he was awarded the Nobel prize in Physics in 1921). These ideas generated the popular Planck law of radiation as in equations 2.6 and 2.7. \ud835\udc62 = \u210e\ud835\udf08 = \u210e\ud835\udc50 \ud835\udf06\u2044                                                            (2.6) \ud835\udc48 = \ud835\udc41\ud835\udc34\u210e\ud835\udf08 = \ud835\udc41\ud835\udc34\u210e\ud835\udc50\ud835\udf06\u2044                                                       (2.7)  Where, \ud835\udc62 = energy of  a photon, [J]; \ud835\udf08 = frequency of radiation, [Hz]; \ud835\udc50 = speed of light, [2.998 \u00d7 108 m\/s]; \ud835\udf06 = wavelength, [m]; \u210e = Planck\u2019s constant, that fit the empirical model that Planck developed for his work on blackbody radiation, [6.626 \u00d7 10-34 J-s]; \ud835\udc41\ud835\udc34 = Avogadro\u2019s constant, [6.022 \u00d7 1023\/mol]; \ud835\udc48 = energy of one mole of photons or Einstein. 2.1.7.2 Laws of photochemistry  Reactions mediated by light are governed by the Laws of photochemistry [23]. These are stated as follows: First law of Photochemistry: Light has to be \u201cabsorbed by a molecule to produce any photochemical change in the molecule\u201d Second law of Photochemistry: A photo-chemical change in a molecule occurs when that molecule absorbs a single photon of light. The photo-chemical yield of a reaction is derived from the number of photons absorbed (Stark-Einstein law). 2.1.7.3 Photo-chemical effect Researchers firmly claim the chemical effect of PL to be the most important factor for microbial inactivation. UV absorption by DNA is the major factor of germicidal effect of PL [13]. UV light shows antibacterial action by virtue of the formation of pyrimidine dimers, mostly by forming thymine photoproducts (thymine-thymine cyclobutene pyrimidine dimers, TT CPD and thymine-thymine 6,4 photoproduct, TT 6,4 PP, figure 2.5) and [1,13]. The UV-C light, particularly 253.7 nm corresponds to the threshold, which lead to dimer formation in microbial DNA. As per the 18  Laws of photochemistry, photons of 253.7 nm wavelength or below can be absorbed by the \ud835\udc36(5) =\ud835\udc36(6) double bonds in thymine, which lead to thymine dimerization [19]. A single photon at 253.7 nm (frequency of 1.182 \u00d7 1015 Hz; energy of 4.88 eV or 7.83 \u00d7 10-19 J) is able to dissociate the \ud835\udc36(5) = \ud835\udc36(6) double bond of a thymine molecule in the DNA. Wavelengths in visible region (1.77-3.10 eV) and IR region (0.12-1.77 eV) have insufficient energies to be able to affect the thymine \ud835\udc36(5) = \ud835\udc36(6) bonds.  These thymine dimers hinder the bacterial DNA replication process, thus resulting in the inactivation via mutagenic action. Other than DNA dimerization, G\u00f3mez-L\u00f3pez et al. [13] also reported that UV-C treatment of spores forms \u2018\u2018spore photoproduct\u2019\u2019 like 5-thyminyl-5,6-dihydrothymine and cyclo-butane pyrimidine dimers. UV light also gets absorbed by conjugated systems like carbon-carbon double bonds in protein molecules, thus causing structural alterations [1,17]. wavelengths >295 nm have been demonstrated to have capacity to destroy viruses. In another study, wavelength >400 nm has shown to destroy virus due to the rupture of phage capsid [46]. Photo-sensitization is another facet of the photo-chemical effect. It involves the administration of certain photoactive compound (called \u2018photo-sensitizer\u2019) that gets accumulated in the target microorganism cells and followed by light flashes [47]. 19   Figure 2.5: DNA bases photoproduct formation due to UV application The interaction of these compounds with light, in excess of oxygen leads to the destruction of cells. Important advances have been made in photo-sensitized antimicrobial chemotherapy, like disinfection of the blood and blood products. Luk\u0161iene et al. [48] carried out 5-aminolevulinic acid (ALA)\u2013based Bacillus cereus photo-sensitization in vitro and on the surface of packaging material. Due to the consequences of the photo-sensitization, several \u2018cytotoxic reactions\u2019 were reported to occur in the cells. Disruption of cellular membrane, enzymes inactivation and damage of DNA took place as a result [47]. Photosensitization based microbial inactivation by PL could be a new area of application of light-based technologies. 2.1.7.4 Photo-thermal effect Photo-thermal effects of light is shown by wavelengths corresponding to IR regions. Owing to their lower energy than UV region in the PL lamp spectrum, they are only absorbed by molecules and compounds in products and they impart kinetic energies to the molecules which produce heat. Thermal effects of PL may add to inactivation of microorganisms if final temperature reaches pasteurization temperature. Although higher temperature by intense PL treatment might add to 20  inactivation it has been shown that the food sensory qualities are adversely affected. It has been previously emphasized in several studies that light pulses heat the surface layers of foods. The temperature of surface layer increases which can destroy superficial pathogenic organisms. The heat can also penetrate inside of the food by conduction, but as mentioned, due to a short process the heat gets dissipated before it reaches the bulk. Nevertheless, a study by Hiramoto [32] showed that light pulses may be absorbed into A. niger cells, heating the molds instantaneously, which might lead to cell death. Dunn et al. [33] also reported the possibility of light pulses heating the food surfaces to destroy inhabiting bacteria. Changes in the structure and physiology of the cells might be the reason of cell death, what could be called photo-thermal stress. Wekhof et al. [34] showed that regional overheating of bacteria due to PL exposure may cause damage and rupture leading to \u2018explosion\u2019 of cellular contents. They showed scanning electron-microscope (SEM) images of PL-treated A. niger spores, visualizing the release of overheated contents of spores. This happens be due to the differential heating rate of cellular components leading to localized expansion and bursting of cell. Nicorescu et al. [49] demonstrated that in PL treatment of Bacillus subtilis, the cell wall got disrupted due to that photo-thermal stress. Xu and Wu [50] showed that on PL treatment of E. coli, there were structural changes, possibly due to \u2018overheating, intercellular water vaporization, and subsequent membrane disruption\u2019. Though, the studies showed the heat generation due to light pulses should only be a lesser contributing factor towards microbial inactivation, as compared to photo-chemical effect. 2.1.7.5 Photo-physical effect Photo-physical effect of PL has also been identified to be a contributory factor for microbial inactivation. The tendency of high energy pulses to disrupt the cell structure was demonstrated by Krishnamurthy et al. [51]. They showed that Staphylococcus aureus cells suffered cell wall damage, shrinkage of plasma membrane, leakage of cell organelles. In another study, Takeshita et al. [52] observed changes in cell shape and cell membrane, elusion of intracellular components and proteins in S. cerevisiae due to PL. Ramos-Villarroel et al. [53] also observed cytoplasmic damages to L. innocua and E. coli after PL treatment. Macias-Rodriguez et al. [54] carried out PL treatment of eggs to study photo-physical changes to E. coli cells and observed structural damage in the form of central depression in cells. These all studies show a contributory effect of photo-physical effect on microbial inactivation exclusive to PL. The structure damaging tendency of PL 21  is higher than that of continuous UV [46]. It might be due to application of high energy \u2018pulses\u2019 as compared to a continuous nature of the latter. 2.1.8 Role of pulsed light technology in food safety The PL technology has been shown to effectively inactivate several microorganisms like bacteria, yeasts, molds, and even viruses. The efficacy of its decontaminating several food substances has been well demonstrated and documented [15]. Several studies have been done in the past decades with respect to application of PL in ensuring microbiological safety [1,4,43,44,55]. Some results from these and other studies in vitro and decontamination of several liquid food products are shown in Table 2.3 and 2.4, respectively.   22  Table 2.3: The effects of PL treatments on microbial inactivation in vitro based on several published data Microorganism Media Pulse Energy [J\/cm2] Treatment Time [s]a Log10 Reductions Reference B. subtilis Tryptic soy broth 4 2 \u03bcs 10 [33] 1 4 \u03bcs 10 E. coli Tryptic soy agar 4 1 \u03bcs 10 1.5 2 \u03bcs 10 A. niger Potato dextrose agar 4 4 \u03bcs 10 12 1 \u03bcs 10 S. aureus Tryptic soy agar 0.75 2 \u03bcs 10 0.2 4 \u03bcs 8 Saccharomyces cerevisiae Tryptic soy agar 0.4 4 \u03bcs 10 E. coli O157:H7 Tryptone Soya-Yeast Extract Agar 3 1\u2013512 \u03bcs 6.82 [25] L. monocytogenes 4b 6.25 Pseudomonas aeruginosa Tryptone Soya-yeast Extract Agar 3 20 \u03bcs 5.8 [43] B. cereus 4.9 E. coli O157:H7 6.2 L. monocytogenes 4.4 S. enteritidis 5.6 S. cerevisiae 4.9 S. aureus 5.1 23  B. cereus, E. coli, and S. enteritidis Tryptones soy-yeast extract broth 3 85 \u03bcs About 8 [44] A. niger, Fusarium culmorum Malt extract agar 4.5 A. niger Buffer solution 1 1000 \u03bcs 4.8 [34] Cryptosporidium parvum oocysts Solution containing approximately 25 \u00d7 106 Cryptosporidium oocysts 0.11 - 3\u20135 [56] 0.22 - B. subtilis  Sterile deionized water  1\u20134 \u03bcs 2\u20135 [38] S. cerevisiae Sterile potassium phosphate buffer suspensions 0.7 1200 \u03bcs 6 [52] Botrytis cinerea Rose Bengal agar - 250 s 3  [57] Monilinia fructigena Malt extract agar - 250 s 4 S. aureus Baird-Parker agar plates 5.6 5400 \u03bcs 7.5 [28] 5.6 5400 \u03bcs 8.5 Table 2.3: Continued 24  Alicyclobacillus acidoterrestris Orange serum agar 7 1500 \u03bcs >5.2 [45] B. circulans Nutrient agar >4.1 Brochotrix thermosphacta Nutrient agar 3.1 Leuconostoc mesenteroides de Man, Rogosa and Sharpe agar 4.0 Photobacterium phosphoreum Nutrient agar >4.4 P. fluorescens Nutrient agar 4.2 Shewanella putrefaciens Nutrient agar 3.9 Clostridium perfringens Nutrient agar >2.9 B. cereus Nutrient agar 3.0 E. coli O157:H7 Nutrient agar 4.7 L. monocytogenes Nutrient agar 2.8 S. typhimurium Nutrient agar 3.2 Shigella flexnii Nutrient agar 3.8 S. aureus Nutrient agar 5.1 Yersina enterocolitica Nutrient agar 3.9 A. flavus Potato dextrose agar 2.2 Table 2.3: Continued 25  Candida lambica Yeast glucose choramphenicol 2.8 Rhodotorula mucilaginosa Yeast glucose choramphenicol >2.8 B. cinerea Potato dextrose agar 1.2 Alternaria alternate, A. niger, B. cinerea, F. oxysporum, F. roseum, M. fructicola, Penicillium expansum, P. digitatum, Phytophthora citrophthora and Rhizopus stolonifer Solid culture media Up to 0.2 - Controlled completely [58] L. innocua  Clear liquid broth 12 - >6 [59] S. aureus, E. coli NCTC 9001, Methicillin-resistant Phosphate-buffered saline suspension 630 30 min 5 for S. aureus and MRSA strains [60] Table 2.3: Continued 26  S. aureus (MRSA) LMG 15,975 and MRSA 16a Negligible for E. coli NCTC 9001 S. enterica serovar Enteritidis Noble agar 0.7 - 6.7 [61] E. coli ATCC 25,922 and E. coli O157:H7 Butterfield\u2019s phosphate buffer 13.1 4 s >8.5 [62] Tryptic soy broth 4 s (12 pulses @ 3 pulses\/s) 3.5 Murine norovirus (MNV-1) and hepatitis A virus (HAV) Viral suspensions  2 s 5 [63] L. innocua NCTC 11,288 Maximum recovery diluent  28 8 s 5.13 [64] E. coli K12 DSM 160 4.67 L. monocytogenes Tryptone soy agar plate 0.00175 180 s 6 [65] B. subtilis strain ATCC 6633 Cell suspension (OD580 nm - 0.8) 0.6 - 8.7 [49] B. cereus Luria-Bertani Agar 1.95 200 s About 7  [66] L. monocytogenes About 7 Table 2.3: Continued 27  B. subtilis Nutrient agar for vegetative cells Tryptic soy broth for spores (cell density 109 cells\/mL in both) 12 - About 4 of sporses >8 for vegetative cells [67] Geobacillus stearothermophillus Nutrient agar for both vegetative cells and spores - About 0.5 for spores About 1 for vegetative cells P. aeruginosa Plate count agar 1.13 30 s 3.63 [68] About 11 300 s About 6.5 L. monocytogenes Scott A Solid tryptic soy agar Petri plates - 20 s 5 [1] 0.1% peptone water solution in whirl pak bags - 93 \u00b1 5 s 1 Bacillus subtilis spores Clear liquid medium 0.017  - 1 [69] Murine norovirus (MNV-1) and  Phosphate buffered saline 4.94 - >5.8 [70] Table 2.3: Continued 28  Tulane virus (TV) - >6.0 E. coli K-12 MG1655 Lysogeny broth (LB) 76 J\/cm2 wavelength of 190 nm - 11 [71] 95 J\/cm2 wavelength of 240 nm 6 a Time mentioned in s unless specified Table 2.4: The effects of PL on some liquid food products based on several published research.  Food Products Pulse Energy [J\/cm2] Treatment Time [s]a Log10 Reductions Reference Milk 25.1 114 s >2.0 of Serratia marcescens [72] Milk 1.27 - 0.55\u20137.26 of S. aureus [51] Apple juice 8.8 3 s 7.29 of E. coli ATCC 2592 [62] Apple juice 4 - 4 of E.Coli [73] 2.98 of L. innocua Orange juice - 2.90 of E. Coli 0.93 of L. innocua Apple juice 28 8 s >4.7 of E. coli 1.93 of L. innocua [64] Orange juice About 1 for both L. innocua and E. coli Table 2.3: Continued 29  Milk 1.06 of E. coli  0.51\u20130.84 of L. innocua Apple juice 5.1 1.52 3.9 of E. coli  [74] Orange juice 5.1 2.81 s 2.42 of E. coli [75] Apple juice 5.1 300 s 4.9 of E. coli [76] Milk (9.8% total solids) 8.4 - 2.5 of E. coli ATCC 25922 [77] Concentrated milk (45% total solids) 8.4 < 1 of E. coli ATCC 25922 Skim milk 14.9 3.4 of E. coli ATCC 25922 Milk (2% fat) 14.9 >2.5 of E. coli ATCC 25922 Whole milk 14.9 >2.5 of E. coli ATCC 25922 Orange Juice 71.6 60 s Up to 1 of L. innocua, Up to 2 of E. coli  Up to 1.5 of S. Enteritidis Up to 0.5 of S. cerevisiae [78] Strawberry juice Up to 0.2 of L. innocua, Up to 0.3 of S. cerevisiae Apple juice Up to 4.5 of L. innocua  Up to 2 of E. coli  Up to 4 of S. Enteritidis Up to 6 of S. cerevisiae Table 2.4: Continued 30  Melon juice Up to 6 of L. innocua and E. coli Up to 5 of S. Enteritidis Up to 7 of S. cerevisiae Blueberry wine - 60 s 9.6 \u00b1 0. 9 of S. cerevisiae [79] White grape wine - 40 s 5.9 \u00b1 0.5 of S. cerevisiae Raw milk 26.25 - Up to 3.2 of total microbial count [80] Whey 1.1 - 0.5 of L. innocua [81] Orange Juice 71.6 60 s 0.3\u20130.8 of L. innocua, E. coli and S. Enteritidis Up to 0.5 of S. cerevisiae [82] Strawberry juice 0.3\u20130.8 of L. innocua, E. coli and S. Enteritidis Apple juice 1.6 of L. innocua 2.1 of E. coli 2.4 of S. Enteritidis Up to 1 of S. cerevisiae Goat milk 10 - 6 of E. coli [83] Coconut water 19.2 - 5.2 of E. coli MTCC 433 [84] Coconut water 95.2 15 s 4 of E. coli MTCC 433 [85] Orange juice 4.5 of E. coli MTCC 433 Pineapple juice 5.33 of E. coli MTCC 433 Pineapple juice 1479 120 s 5 of aerobic mesophiles and the yeast and mold [86] Mixed fruit beverage 2400 100 s [87] Table 2.4: Continued 31  5000 167 s Aerobic mesophilic and yeasts and mold reduced below detection limit aTime mentioned in s unless specified32  2.1.9 Factors PL decontamination of solid and liquid foods PL activity is different for solids and liquid food products. Solid foods absorb the PL and there is an exponential decay of light intensity as light goes on being absorbed as a function of the depth of food and its absorption coefficient. The extent to which PL penetrates the food products is termed its penetration depth, which has been defined as the distance where the light fluence rate is reduced to 37 % (1\/\ud835\udc52; \ud835\udc52=2.7128) of its initial value [19]. Some other convention states penetration depth where light intensity is reduced to 10% of surface incident intensity. This is depicted in Figure 2.6.  Figure 2.6: Penetration depth of light in food products The penetration depth is defined so because reduction of light intensity to such lower percentage is unable to carry out any inactivation beyond that depth. Heinrich et al. [17] reported that lower absorption coefficient, and higher transmission coefficients increase the penetration of PL. Uesugi and Moraru [88] demonstrated a penetration depth of 2.3 mm in Vienna sausages by the PL for a fluence of 9.4 J\/cm2 at 50.8 mm sample distance from lamp and achieving a 1.39 log reduction in L. innocua. Vimont et al. [89] reported penetration of PL up to 10 mm depth in whey protein gels. Sauer and Moraru [62] obtained penetration depth of 41.7 and 15.9 mm for apple juice and cider respectively at fluences of 8.8 and 11.7 J\/cm2 and 50.8 mm sample distance from lamp. They also reported greater reductions of E. coli of more than 7 and 5.5 for apple juice and cider respectively. In another study, coconut water of depth 5 mm was successfully treated by PL at fluence level of 19.2 J\/cm2 and lamp distance of 50 mm. Considering almost same light intensity at the surface of any substrate, this indicates that PL has greater penetration in liquid foods than solid foods, with a steeper decay in light intensity in the solid foods. 33  2.1.9.1 Solid food products Since in case of solid foods products, the light does not penetrate well, PL application is mostly limited to surface inactivation and these are superficially decontaminated [13]. For, its superficial inactivation characteristics, PL would be effective for totally smooth products, where microorganisms would be totally exposed to light flashes. However, surfaces of solid foods are especially not smooth. Minute surface cracks and crevices could possibly shelter microorganisms, where light may not reach. Therefore, the entire superficial area of the foods has to be flashed so as to achieve the total decontamination of its surface despite surface unevenness, which makes the process very complex [21]. Several research have been done on surface-treatment of fruits for microbial reductions. The possibility of PL to decontaminate the surfaces of fruits like strawberries and raspberries was demonstrated by Bialka and Demirci [31], with insignificant damage of fruits being reported. They treated raspberry surfaces using fluence of 72 J\/cm2 and obtained E. coli 0157:H7 and Salmonella reduction by 3.9 and 3.4 logs, respectively. Similarly, in strawberries they obtained E. coli 0157:H7 and Salmonella reduction by 3.3 and 4.3 logs respectively at 64.8 J\/cm2. However, very high fluence values were used. Marine and poultry products have also been treated efficiently by PL. Successful decontamination of surfaces was demonstrated for shrimp and fish [90]. Listeria on shrimp surface were reduced by 1-3 logs using 4-8 flashes of 1-2 J\/cm2 PL. Whereas fish treated with 3 PL flashes at intensity of 10 J\/cm2 gave reduction of about 2 logs of surface psychrotrophic bacteria. Pa\u0161kevi\u010di\u016bte et al. [91] treated chicken surface with high-power PL of 1000 pulses and dose 5.4 J\/cm2 and reduced S. Typhimurium and L. monocytogenes by 2\u20132.4 logs and total aerobic mesophiles by 2 logs. Eggshells have also shown to be successfully decontaminated using light pulses. Exposure to PL of fluence levels of 2.1 J\/cm2 led to inactivation of Salmonella cells (5 log reductions per eggshell) on the egg surface [92]. Hierro et al. [61] performed PL treatment of shell eggs using 12 J\/cm2 fluence.  They obtained about 2.49 log reduction of S. Enteritidis on eggshell. The discrepancy between the results of the latter in spite of using higher fluence could be due to employing of different cell recovery methods after treatment. Lasagabster et al. [92] used a shell rinse method for recovery of cells from shell surface, while Hierro et al. [61] used shell crush method, indicating the former an inefficient cell recovery method [93]. Nevertheless, solid foods ranging from fruits, vegetables, poultry to marine products have 34  been treated successfully by PL. Though, surface irregularities remain a challenge for the solid food decontamination. 2.1.9.2 Liquid food products Liquid samples treatment with PL is more challenging because microorganisms dwell in the whole volume of liquids. So, PL efficacy is particularly based on several factors like sample distance from lamp, exposure, turbidity, optical and physico-chemical properties of sample, and sample depth [19,83]. Higher exposure time of liquid foods leads to higher inactivation. Also, the sample distance from lamp determines treatment efficiency; the lower the distance the better. Optical properties like transparency, and turbidity affect the PL efficacy. Turbid liquid foods might hinder process efficiency because of light scattering. On the other hand, transparent liquids like clear juices are treated by PL efficiently [17]. Huffman et al. [94] treated water using PL at 0.25 J\/cm2 and obtained reductions in bacteria Klebsiella terrigena of more than 7.4 logs, and in viruses (Poliovirus and Rotavirus) and parasite (Cryptosporidium parvum) of above 4 logs. Milk was shown to be effectively treated by the exposure of 56 s to PL at a dose of 25.1 J\/cm2 [72]. S. aureus was reduced by 7.26 logs at 1.27 J\/cm2 when subjecting milk to PL treatment [51]. This shows that the higher fluence values required for opaque food like milk as compared to water. Food composition also affects the PL processing. Miller et al. [77] carried out PL treatment of milk (concentrated milk, whole milk, skim milk) and showed that PL efficiency decreased as the content of milk fat and total solids increased. They obtained highest inactivation for skim milk and concluded that the presence of milk fat globules scattered light. Other liquid foods like Apple juice and orange juice (inoculated with E coli DH5-\u03b1 and L. innocua 11288) were subjected to continuous PL system with fluence of 4 J\/cm2 [73]. They the inactivation levels of 4.00 and 2.90 log reductions respectively in apple and orange juices for Escherichia coli. Similarly, 2.98 and 0.93 logs for L. innocua were obtained for apple and orange juices, respectively. Higher turbidity and cloudiness in orange juice lowered the microbial inactivation by PL. Hillegas and Demirci [26] obtained 0.97 log reductions of C. sporogenes spores inoculated in 2 mm depth of honey after 135 pulses intense PL treatment of 5.6 J\/cm2 per pulse. They concluded that microbial inactivation increased by varying the process parameters (lamp distance, number of pulses, sample depth), but complete inactivation was not achieved due to poor penetration of pulsed UV system in honey. Thus, depth of liquid foods too affects the PL treatment. As light energy decreases with depth, thin 35  film or profiles are suitable for PL processing. The physico-chemical properties like viscosity and density affect the flow properties of liquid foods and thus have to be considered while designing a continuous PL treatment system for them. Nevertheless, most of the studies show that PL is equally efficient and has great potential for processing of several liquid foods. With the decontamination of clear liquids by PL successful, more research has to be extended towards turbid and opaque liquid foods.  2.2 Effect of PL on food quality characteristics and nutritional value One of the main limitations of conventional processes like thermal processing is the impact on food quality aspects including the taste, flavour, colour, nutritional and bio-active components. Being a non-thermal process, PL processing is expected to have no harmful impacts on the food quality while making it decontaminated. The potentiality of PL applications for food processing is thus affected by how it affects their colour, taste, nutrients, and other properties of food products. Various research works have been done to understand the effects on PL on food quality parameters. The effects of PL processing on various food quality attributes are discussed below.  2.2.1 Effects on organoleptic properties (colour, texture, and flavour) of food products Fruits and vegetables contain several phytochemicals and pigments which impart colour and other sensory characteristics to them. It is very important to have a knowledge of the effects of PL on these product characteristics. In earlier studies conducted by Dunn et al. [33], PL was claimed to maintain sensory attributes intact without any significant changes. Many studies have emerged thereafter which have shown both similar and opposite results of effect of PL on sensory properties of products like fruits and vegetables. Bialka and Demirci [31,95] treated blueberries, raspberries and strawberries using pulsed UV light and observed insignificant colour parameter and sensory changes, as well as no damages on the treated samples as compared to untreated ones. However, Bialka and Demirci [31] reported high inactivation at high fluence level of 32.4 J\/cm2 at 3 cm sample distance from lamp but carried out colour and sensory analysis at lower level of 22.6 and 11.3 J\/cm2 at 8 cm distance.  In another study, colour and firmness of strawberries remained acceptable after PL treatments [66]. Anugu [79] treated blueberry wine by PL and did not observe any significant effects on hue and lightness values of blueberry wine.  However, many studies have shown significant undesirable changes by PL treatment. Fine and Geravis [37] observed that undesirable colour changes occurred on PL treatment of black pepper 36  and wheat flour, which was apparent well before achieving required microbial inactivation. G\u00f3mez-L\u00f3pez et al. [96] reported discoloration in PL-treated iceberg lettuce and plastic like off-odour in shredded white cabbage immediately after treatment of 2700 pulses of 7 J intensity at 12.8 cm distance. This indicates a severe PL treatment might lead to significant changes to food quality. Slight colour changes were also observed on PL-treated apple juice [76]. Similarly, Ramos-Villarroel et al. [42,53,97] showed that UV-C component of PL significantly hampered the firmness and colour in fresh cut avocado, watermelons and mushrooms as UV-C wavelengths corresponding to higher energy have tendency to bring about undesirable sensory changes. There was a significant change in lightness value of the fresh-cut avocado and post-storage browning [42] after PL treatment with the main role of UV-C region. There was significant decrease in colour parameters and firmness of water-melons by PL application contributed significantly by UV-C. Colour changes were attributed to decrease in carotenoid concentration [92] which might be due to slight heating of samples and exposure to UV-C. Similar decrease in lightness and firmness values of mushrooms were observed [53] after PL application. Abida et al. [20] also reported such effects of PL treatment on mushroom texture due to thermal damage by photo-thermal effect. Luk\u0161iene et al. [98] affirmed that PL treatment affected the tactile properties of fruit and vegetables like texture and firmness, which was shown by Ramos-Villarroel et al. [42,53,97]. Slight decrease in the firmness of PL-treated tomatoes was observed [99]. In PL-treated mango pulp, there was 40 % decrease in fruit firmness post-storage [100]. There was also significant improvement in colour in PL-treated mango peel and pulp after storage for 7 days. Chroma values increased up to 140 % in the peel and 130 % in the pulp post-storage which was attributed to increments in their carotenoid concentration due to chlorophyll degradation. However, this contradicts the study by Ramos-Villarroel et al. [97] and effects depend on product matrix; like in watermelon the PL decreases carotenoids, wherein in mango leads to chlorophyll degradation. While severe treatment conditions are detrimental to product sensory quality, optimized treatment conditions at particular light intensity, number of pulses and lamp distance will minimize these changes. Lipids in milk products might be quite susceptible to off-flavour generation by auto-oxidation phenomenon, due to light. Slight aroma changes were observed in PL-treated goat milk [83] possibly due to photo-chemical changes in milk components and lipid oxidation. Krishnamurthy et al. [51] showed that adequately designed pulsed UV treatment process of milk does not induce 37  lipid oxidation. Thus, properly designed fluid milk product PL treatment system are needed based on optimized flow rate, fluence and residence time to minimize the effect on their colour and sensory properties. Dunn [33] showed that no changes in colour and taste occurred by PL in dry cottage cheese curd after severe treatment of 32 J\/cm2. Contrastingly, however, a smaller dose of 9.22 J\/cm2 adversely affected the sensory quality cheese [101]. Colour changes were significant for cheese treated by PL at 53.4 J\/cm2 for 40 s at 5 cm distance [102]. The colour values a* and b* were significantly different from untreated ones, indicating changes in yellowness and redness.  The effects of PL application on sensory qualities of meat, sea-food, fish, and other processed foods have been extensively studied. Dunn et al. [103] showed the sensory attributes of the PL-treated fish remained acceptable and close to the control samples. They concluded that fish was acceptable even after 15 days of refrigerated storage. However, pulses of 2.5-5 J\/cm2 of UV light did not change colour in packaged catfish fillets [104]. In another study, salmon fillets, on PL treatment for 30 s at 3 cm distance and for 45 s at 5 cm distance showed excessive heating which resulted in colour changes visually [105]. Pa\u0161kevi\u010diute & Luk\u0161iene [106] after PL treatment of 5.4 J\/cm2 did not observe any changes in colour and flavour changes in chicken breast meat. Keklik et al. [107] observed significant quality alterations in PL-treated chicken frankfurters. The fluence level up to 67 J\/cm2 for 60 s treatment induced darker, greenish, and yellowish tint in the treated frankfurters. This could be due to application of higher fluence. The sensory qualities of PL-treated bologna slices were observed to be adversely affected [108]. Researchers noticed change in redness value and sensory attributes for dose above 2.1 J\/cm2. Wambura and Verghese [109] studied the impact of pulsed UV light on quality attributes of ham. They observed darkening of ham after processing and storage up to 7 days with increasing L* value, slight increase in yellowness due to increase in b* and decreasing redness (due to decrease in a*) value possibly because of accumulation of metmyoglobin on surface. They also observed decrease in firmness of meat post-processing. Pork and salmon, after PL treatment showed significant colour and odour changes [110] after PL treatment. Tomasevic and Rajkovic [16] found that PL treatment degraded the organoleptic properties of cooked meat, whereas it did affect the sensory attributes of dry cured meats and fermented sausages. The higher fluence dosage (17 J\/cm2) significantly changed the colour parameter values of meat products. No significant differences in sensory and colour quality were found for sea foods. Sulphur off-odour was detected in fresh egg pasta after PL treatment of 38  1.75 J\/cm2; however, no significant colour change was observed [111]. As a matter of fact, fat content remains a decisive factor in PL treatment of meats with higher fat content leading to significant sensory changes. Fat however tends to increase microbial resistance to PL and thus fat rich meat products are needed to be treatment under severe conditions jeopardizing their sensory characteristics. 2.2.2 Effects on physico-chemical properties of foods The effects of PL processing on various physico-chemical properties of food products like viscosity, pH, functional properties have also been studied by several researchers. Shuwaish et al. [104] treated catfish fillets by PL of 2.5-5 J\/cm2 fluence and observed no effects on the shear forces. Marquenie et al. [112] did not observe any effect of PL on firmness of strawberries either singly (160 s treatment) or in combination with UV-C (254 nm, 0.1 J\/cm2 and 120 s PL). Krishnamurthy et al. [51] observed increase in milk temperature up to 38 \u00b0C by continuous-flow PL treatment of milk, which led to fouling as well as alterations in milk quality. Kasahara et al. [83] noticed a slight decrease in viscosity, pH, and density of milk after PL application. Palgan et al. [64] did not observe any significant change in \u00b0Brix and pH of apple juice, after PL treatment. Eggs treated with PL showed no drastic effect on the quality of egg albumin and on the sensory and functional properties [92]. The firmness and pH of the fresh-cut avocado after PL application however decreased; also, textural properties of fresh-cut watermelon and mushrooms [42,53,97] were affected. PL also showed no impact on the soluble solids (in terms of \u00b0Brix), pH, and non-enzymatic browning index, except for slight effect on the colour of apple juice [76]. Slight decrease in pH, \u00b0Brix and rise in titratable acidity (in terms of g of anhydrous citric acid\/ 100 g tomato) took place in tomatoes as a result storage of 20 days after PL treatment of 8 J\/cm2 [113]. Slight changes in cell wall integrity and weight decrease were observed in tomatoes after PL treatment of 8 J\/cm2 [99]. The physico-chemical properties are said to be affected in some studies and not in others. However, no conclusive results could be drawn, and it could be said that the effect of PL might be driven by the type of product matrices. 2.2.3 Effects on nutrients and bio-active components As already emphasized, PL processing tends to interact with the superficial layers of the food products which are in the vicinity of the light source. So, it does not appear to hamper their inner layers; hence, nutrients and bio-active components present within the inner layers remain intact as 39  untreated samples. Research has been done in plenty to study the effect on PL processing on food nutrients and bio-active components. On carrying out a nutritional evaluation of PL-treated frankfurters, Dunn et al. [6] observed no differences in proteins, vitamins like riboflavin and vitamin C, nitrosamine, benzopyrene content compared to untreated samples, while a strong loss of riboflavin was reported in foods due to heat, light, and oxygen. Contrastingly, riboflavin content in products like chicken, beef and fish were did not decrease using pulsed UV treatment [114]. However, riboflavin and vitamin E contents were reduced only to 95 % by 4 light pulses and to 85 % by 8 light pulses of their initial value, respectively. Wambura and Verghese [109] demonstrated pronounced oxidation in the treated ham samples in terms of the oxidative stability index (OSI) in hours, which showed a decreasing trend as storage period progressed. Abida et al. [20] reported PL treatment of mushroom reduced some phenolic compounds and vitamin C content.  Lipid and protein oxidation is induced in many food products post-PL processing and remains a challenge [91,102,109,110]. Pa\u0161kevi\u010diute et al. [91] observed PL induced lipid oxidation on chicken surfaces. The intensity of lipid oxidation in control and treated chicken samples showed a difference of 0.16 mg malondialdehyde (MDA) \/ kg of chicken meat. Can et al. [102], in their PL treatment study of cheese at 5 cm distance for 40 s, found that levels of lipid peroxidation as MDA concentration was found to be about 30.7 and 34.8 \u03bcg MDA\/g cheese in packaged condition and unpackaged conditions respectively, with unpackaged cheese undergoing higher oxidation due to exposure to oxygen. MDA levels in PL-treated salmon and pork increased by 39.3% and 25.5 %, respectively, after PL processing [110]. After PL treatment of milk, Elmnasser et al. [115] did not notice any lipid oxidation and changes in amino acid sequence and composition of proteins. Kasahara et al. [83] however talked about the possibility PL induced photo-chemical changes in milk lipids, proteins, and oxidation of vitamins. Fernandez et al. [116] studied the protein oxidation of cheese as affected by PL, taking bovine serum albumin (BSA) as a model protein. They measured oxidation in terms of carbonyl content increase (nmol carbonyls \/ mg BSA), obtaining an increase in carbonyls above 10 nmol \/ mg BSA for fluence of 11.4 J\/cm2. Thus, lipid and protein oxidation tend to remain a bottleneck in PL processing of fat and protein rich products. Fruits like grapes and berries are rich source of phytonutrients and polyphenolic compounds. The retention of these compounds is necessary from the nutritional point of view. Impact of PL 40  processing on these compounds has been a topic of continuous research. There was a decrease in total phenol and antioxidant activity (in Trolox equivalents) and impact on sensory attributes by 8 s (fluence of 1.17 J\/cm2 per pulse) PL treatment of apple juice [64]. However, light-based technologies, like continuous UV-C processing has been shown to increase the concentration of phytochemicals in fruits. This could be due to the activation of plant defense response to severe treatments, which express the formation of the phytochemicals like phenolic compounds. In a study, level of resveratrol in grapes increased by more than 10-fold [117]. Given the similarities with UV, PL is claimed to have the similar effect of increasing phytochemicals [13]. Nevertheless, a study by Mu\u00f1oz et al. [75] showed no effect on the antioxidant capacity at fluence level of 5.1 J\/cm2 and assisted thermosonication for orange juice. Pulsed UV exposure of blueberries enhanced their antioxidant capacity, phytochemicals, and enzyme activity [79]. Pulsed UV treatment of blueberry wine enhanced the total anthocyanins [as mg of Cyanidin-3-glucoside equivalents (C-3-GE)\/L of wine] and total flavonoids [as mg of Catechin equivalents (CE)\/L of wine] but did not significantly reduce total phenolics [as mg of Gallic acid equivalents (GAE)\/L of wine] and antioxidant activity [as mmol Trolox equivalent\/L of wine] [79]. There was increase in bio-active components like \u03b2-carotene, \u03b1-carotene, and total lycopene contents with some lycopene isomerization in PL-treated tomatoes [118]. Green tomatoes showed increase in phenolic compounds, antioxidant activity, lycopene content, and total carotenoids after treatment [119]. Ag\u00fcero et al. [120] observed little increase in polyphenolic content and antioxidant activity after PL treatment of spinach. There was an increase in antioxidant activity in mango pulp [100]. The researchers found an increase in phenylalanine ammonia lyase enzyme content indicating increase in phenolic content. There was also little reduction in Vitamin C content in the mango pulp on storage after PL exposure. Braga et al. [121] shows that using pulsed UV as a pre-treatment for drying mangoes can reduce vitamins loss as compared to untreated dried mangoes. They found that the level of Vitamin C and carotenoids in untreated dried mangoes were 10 % and 40 % lower than mangoes treated with fluences between 3.6-10.8 J\/cm2. In addition, in samples subjected to fluences between 3.6 and 7.2 J\/cm2 the level of vitamins B1, B2 and B5 risen by 10 to 25%. However, Vitamin B6 was decreased by 40 to 50 % [121]. In a study, our research group showed the effect of PL on gallic acid solution (model solution of phenolic compounds in fruits and vegetables) [122]. It was observed that gallic acid solution turned brownish due to photo-41  degradation and a critical fluence of 3.82 J\/cm2 at 2 cm distance, was identified below which the gallic acid solution showed little\/no degradation. As bio-active components being important from nutritional point of view, studies on effect of PL on these components seem important. 2.3 PL processing of liquid foods PL processing of many liquid food products have been done as a strategy for their decontamination. The liquid food products mainly treated by PL have been fruit juices, milk, coconut water, alcoholic products like beer, wine etc. Batch PL systems are offered by PL manufacturing companies that comprise of a cylindrical PL lamp housed within a lamp housing. On the other hand, commercial continuous PL systems are scarce currently. Therefore, researchers have designed continuous PL systems by suitable modifications for their PL equipment [51,73].  2.3.1 Batch or static mode Batch mode of PL systems comprise of a PL lamp used for treatment of liquid foods placed at certain distance from lamp and the liquid thickness is varied. For experimental design, researchers have considered the parameters such as sample distance from lamp, sample volume, or sample depth, pulse frequency, operating voltage, etc. Effect of sample distance and pulse frequency are encompassed by the treatment fluence. On the other hand, sample volume and depth affect the uniformity of treatment due to light attenuation. Sample voltage affects the lamp emission spectrum by displacing the spectra to lower wavelengths at higher voltage [123]. Choi et al. [124] studied the effects on PL treatment on L. monocytogenes in a batch PL system. The system consisted of a PL lamp and sample dish (with product 2 mm thick) was 60 mm away from it. The operating parameter included pulse width of 1.5 \u03bcs, 0-600 s, and 10 Hz frequency at an operating voltage of up to 15 kV. Up to 5 log reduction in infant beverage was obtained.  Hwang et al. [125] treated liquid foods (apple, grape, plum, orange juice) by intense PL treatment in batch mode. The system comprised of  a PL lamp with the sample underneath at 3.5 cm distance from lamp. Up to 7 log reduction of Pseudomonas aeruginosa after PL treatment of 24.35 J\/cm2, for apple and plum juice. For grape juice 1.9-log reduction was reported after total fluence of 29.21 J\/cm2. Chakraborty et al. [87] treated a mixed fruit beverage formulated with apple ber juice, carambola juice and black table grape juice using PL system in batch mode. For treatments, 50 mL of 42  beverage sample was kept at 2.5 cm distance from lamp and the liquid thickness was 5 mm. Application of fluence up to 5000 J\/cm2 led to inactivation of natural microbiota with an enhancement of levels of antioxidants, vitamin C, and phenolic compounds.   Different level of fluences from 24-5000 J\/cm2 have been used for processing of liquid foods in batch mode. While processing in batch system is easier to control and replicate, it is necessary that the treatment protocols are reported correctly. The measurement of fluence and its reporting is very important. Since batch mode systems consist of a sample dish which has a finite size (diameter), and the radiometer measures the light energy falling only on a small surface, the measurement of energy values from radiometer needs to be measured correctly. Also, sample thickness and radial light attenuation need to be considered. Stirring the liquid sample in petri dishes during PL treatment can also help to make the PL treatment in batch mode uniform to some extent and will also offer higher microbial inactivation. 2.3.2 Continuous-flow reactors Although commercial flow-through PL reactors are lacking, one of the first continuous PL reactor was patented in 1990s [33]. The continuous system comprised of an annular quartz PL chamber with the lamp placed within the chamber. The system was designed for processing of liquid foods like juices, milk. While there are many key manufacturers of PL equipment\/lamps\/systems as discussed earlier, an important challenge lie for them is to design a continuous PL treatment system. The main design consideration for such continuous reactors is that all the liquid volume should get exposed to light while being pumped through the reactor inlet and outlet. Such challenge has been taken by numerous researchers who have worked earlier on continuous processing of liquid foods like fruit juices and beverages, milk, coconut water, carbonated beverages, tea, etc. Table 2.5 lists the continuous PL systems that have been used earlier by researchers. Table 2.5: Continuous PL systems for processing of liquid food products Product Treatment parameters Log reductions References Milk Milk was passed though quartz tube placed on a v-groove reflector with adjustable height.  7.23 log reduction of S. aureus at 8 cm distance, 20 mL\/min flow rate and 1 pass Krishnamurthy et al. [51] 43  Voltage = 3.8 kV Pulse frequency = 3 Hz Pulse energy = 1.28 J\/cm2 Sample distance = 5-11 cm Flow rate = 20-40 mL\/min Passes = 1-3 Apple juice Juices were passed though quartz tube placed 1.9 below PL quartz window Voltage = 3.8 kV Pulse frequency = 3 Hz Pulse energy = 1.21 J\/cm2 Flow rate = 38.4 mL\/min 4.00 and 2.98 for E. coli and L. innocua, respectively Pataro et al. [73] Orange juice 2.90 and 0.93 for E. coli and L. innocua, respectively Water Dynamic flow-through system from Claranor, France Voltage = 1-3 kV Total fluence = 10 J\/cm2 Flow rate = 1-5 L\/min Up to 6 log reduction for L. innocua Artiguez et al. [126] Whey Dynamic flow-through system from Claranor, France Voltage = 1-3 kV Total fluence = 10 J\/cm2 Flow rate = 1-5 L\/min Up to 6 log reduction for L. innocua Artiguez et al. [81] Apple juice Juice was passed though quartz tube placed 10 below PL quartz window Voltage = 3.8 kV Pulse frequency = 3 Hz Up to 4.2 log reduction for E .coli Ferrario and Guerrero [127] 44  Pulse energy = 1.21 J\/cm2 Total fluence = 0.73 J\/cm2 Flow rate = 155 mL\/min Reynolds no. = 220 Orange juice Product was passed though quartz tube placed 20 below PL quartz window Voltage = 0.5-1.5 kV Pulse frequency = 3 Hz Total fluence = 8.1-756 J\/cm2 Flow rate = 100 mL\/min Up to 4.0 log reduction for E .coli Preetha et al. [85] Pineapple juice Up to 4.5 log reduction for E .coli Coconut water Up to 5.33 log reduction for E .coli Grape juice Juice was passed through a spiral tube (tube diameter = 1mm; coil diameter = 50 mm; length = 220 mm) placed at 5 cm below the lamp Energy per pulse = 0.13-0.66 J\/cm2 Total fluence = 6.6-66 J\/cm2 Flow rate = 30-60 mL\/min Up to 2.6 log reduction of E. coli for flow rate 60 mL\/min at 0.66 J\/cm2\/pulse Wang et al. [128] For continuous treatments, the researchers have studied the effects of flow rates, number of passes, voltage, and tube distance form lamp. Increasing the number of passes increase the residence time of liquid food inside the reactor, which offer higher log reduction, but can also lead to undesirable sensory and nutritional changes in the product. Flow rates used previously by the researchers were very less (up to 155 mL\/min or about 9 L\/h) compared to conventional pasteurization or sterilization systems. The tube distance was also varied and distances up to 20 cm have been used. Since the light energy decays with distance, the tubes should be placed in the vicinity of the lamp. Table 2.5: Continued 45  Therefore, a proper design of PL reactor geometry would enable exposure of all the liquid volumes uniformly, leading to a uniform treatment. Lastly, the treatment regime (turbulent or laminar), Reynolds number has been overlooked mostly previously. A proper characterization of the PL reactors based on light energy delivery, and hydrodynamics studies need to be carried out. Liquid treatment using continuous UV-C reactors has been explored in detail by many researchers. Therefore, different geometries of reactors have been used and studied previously. 2.3.2.1 Annular thin-film reactors To overcome the limited penetration depth of light on liquid foods, a thin film of liquid food is treated in an annular reactor system placed around the lamp axis. The system comprises of two thin stationary cylindrical (quartz) tubes placed concentrically so that a thin film (1 mm) of liquid can be passed in between (Figure 2.7). These are operated at laminar condition with parabolic flow profile and turbulent flow profiles as well [129]. Such systems have been widely explored for disinfection of liquid foods like apple juice, mango nectar [130,131].  Figure 2.7: Annular thin-film PL reactor Such reactors in flow-through mode have not been explored yet in PL processing. Wiktor et al. [122] and Mandal et al. [132] have used annular PL reactors in batch mode for processing of gallic acid solution and black tea infusions, respectively.    2.3.2.2 Taylor-Couette reactors Even though a thin film is used in annular tubes, since the tubes are stationary, the treatment is less efficient. Therefore, sometimes the inner cylinder is rotated to offer additional turbulence to the liquid and hence adequate mixing. The Taylor-Couette system comprises of an inner cylinder (rotor) which rotates and outer cylinder (stator) which is stationary (Figure 2.8). The lamp is placed around the outer stator. Due to the rotation of the rotor, the liquid film is subjected to vortices and an effective mixing under plug flow condition is achieved. Such systems have been widely studied 46  in UV-C processing [133]. However, PL systems have not been developed based on such geometry yet.   Figure 2.8: Taylor-Couette reactor 2.3.2.3 Coiled tube or Dean flow reactors Such reactors consist of  a coiled treatment chamber\/tube which is wound helically around the lamp (Figure 2.9). Due to rotation of the liquid inside the tube, centrifugal force acts on the liquid, which leads to formation of eddies or vortices which are superimposed on the primary forward flow of the liquid in the tube. These vortices are named after Dean, which described them [134]. Due to these secondary vortices, the radial mixing is effective even under laminar flow conditions [129]. These reactors offer homogenous treatment even for low UV transmissivity opaque liquid foods [129]. Such reactor systems have been studied by Wang et al. [128] for grape juice treatment. We also are using coiled tube reactors for treatment of liquids foods and published a paper [135] that forms the basis of chapter 3.  Figure 2.9: Coiled tube or Dean flow reactor 47  2.4 Challenges with PL processing of liquid foods The characterization of PL treatments demand that the terms are properly defined. One of the bottlenecks towards adoption of the technology is that the processing parameters have not been described properly and fundamental rules defining PL technology have not been considered [23]. This creates hurdles when comparing and replicating the data for PL experiments carried out earlier. It is apparent that the parameters describing the PL process (pulse width, energy per pulse, voltage) are inherent to PL system used by different researchers. So, the same pulse width settings and pulse frequency may lead to different results in different systems. Also, reactor configuration determines the performance and efficacy of treatment. Lamp emittance and volumetric dose distribution also determine the performance. In case of liquid food matrices, additional parameters (absorption coefficients, UV transmissivity, clarity, flow properties and reactor hydrodynamics) determine reactor performance [17,24,62,126]. Design and scale-up of an effective PL reactor needs considerations all the important process and product parameters and validation if the delivered dose inactivates microbial population by required logarithmic cycles. 2.4.1 UV dose determination Proper calculation of UV dose that is delivered during any PL treatment is an effective factor that governs the application, reproduction and scale up of the treatments. For batch systems, it is easier comparatively to determine the UV dose due to lack of complex geometries. The radiometer can be employed for measurement of energy intensity, fluence rates at a surface in bench top batch systems. For continuous flow-through systems, lack of direct energy measurement systems offers problems for dose quantification. Radiometers cannot measure energy inside the reactor directly due to space constraints of reactor, which are usually thin-film. For determination of UV dose in flow-through systems considerations such as flow geometry, flow regime, mixing conditions, liquid food properties (opaque\/ clear liquids, absorption coefficients of liquids) have to be made [129]. Total UV dose delivery inside the reactor as well as three-dimensional intensity distribution have to be ascertained. Some of the methods that have been developed for determination UV dose in flow-through reactors are discussed in the following subsections. 2.4.1.1 Chemical actinometry method Actinometer is a chemical substance, which undergoes photo-chemical reaction, when exposed to UV\/Visible light [136]. The extent of reaction taking place can be quantified in terms of the 48  quantum yield of the reaction [129,137]. The actinometer under light exposure changes colour whose absorbance can be measured and the delivered energy is calculated mathematically. Chemical actinometers can be calibrated against a radiometer. The most widely used chemical for actinometry are potassium ferrioxalate (K3Fe(C2O4)3) and potassium iodide-iodate (KI\/IO3-) solutions. In potassium ferrioxalate system, Iron in 3+ oxidation state is photo-reduced to 2+ state, to form ferrioxalate complex [129]. Absorbance measurement at 510 nm is carried out to measure Fe(I1)-I.lo-phenanthroline complex. For iodide-iodate actinometer, the iodide ions yield triiodide ions (I3-) which can be measured spectrophotometrically at 254 nm to calculate the UV dose. Chemical actinometry have been widely used for UV-C processing. However, suitable actinometry study for PL systems are few. Bohrerova [136] developed a protocol for measurement of actinometer UV dose for polychromatic light sources as PL. One limitation of actinometers is that they need to be dissolved in pure water. However, their dissolution in liquid foods for dose estimation seems impractical since they can react with the chemical compounds like phenolic compounds, phytochemicals, flavonoids in foods, lead to improper results.  2.4.1.2 Biodosimetry method Biodosimetry method has widely been used for estimation of UV dose for flow-through reactors. It is a biological method for estimation of UV dose. For the assay, one or more target microorganisms are inoculated in the liquid food product. Then, the product is exposed to UV light on a collimated beam apparatus (for which UV dose estimation is easier) to develop a dose-response curve for each microorganisms in the liquid food [129,133]. Using regression techniques, a regression equation is developed to model the dose-response curves. Thereafter, the same microorganisms in the liquid foods are treated in continuous reactors. The microbial inactivation level achieved in continuous systems are then compared with the regression equations previously developed to ascertain the UV dose delivered  in the continuous systems. These dose is termed as the reduction equivalent dose or fluence. Though, there might be differences between the dose calculated by collimated systems compared to the dose for UV reactors, due to differences in the volumetric difference of the systems [129]. 2.4.1.3 Computational fluid dynamics Multiphysics modeling and simulation tools are important for UV dose distribution and process validation. Computation fluid dynamics (CFD) is a numerical solution method which takes into 49  consideration the principles of fluid flow, transport phenomena to predict the flow of fluid within a geometry. The experimental determination of fluid flow profile is very difficult due to the physical constraint of the reactors. CFD can help in flow visualization by solving the governing equations of fluid flow like the equation of continuity, momentum transfer and so on. For CFD simulation, the reactor geometry (as per the dimensions) is developed into a model using drawing tools in the CFD software. Thereafter, the geometry is divided into smaller sections by meshing operation. For solving the governing equations in the geometry to get the flow distribution, the fluid and material properties (density, specific heat, absorption coefficient etc.) are set based on calculation of data, and boundary conditions on the geometry are defined based on the experimental design. Distribution of parameters like velocity, pressure can be visualized in the form of contour plots, streamlines, vectors in the reactor geometry.  Continuous UV processing has been extensively studied and simulated using CFD for mathematical modeling, UV dose distribution determination [129,138]. The application of CFD in UV processing has been in the design of UV reactor considering the delivery of adequate UV dose for a particular flow rate and residence time. Additionally, CFD can also be used to detect the areas in the reactor which receives very high or less UV light. Particle tracking in the simulated flow field can be carried out to determine the cumulative energy received by the particle. Total UV dose is thus determined using CFD for a particular liquid food, reactor geometry, flow rate. PL technology resembles the UV treatment to a great extent. A better understanding process of PL interaction with foods and design of continuous PL systems could be achieved by the CFD modeling and simulation. However, till date CFD simulation of PL processing are scarce. We have carried out CFD simulations of flow-through PL reactors to visualize the light energy distribution, velocity profile, and process uniformity [135] that are mentioned in detail in chapter 3. 50  Chapter 3: Design & Characterization of PL reactors in annular tube and coiled tube configurations using chemical actinometry and computational fluid dynamics  3.1 Summary Pulsed UV light (PL) processing is a growing non-thermal method of treatment of solid and liquid foods. Proper design of liquid food requires understanding of hydrodynamics, and PL dose distribution within the PL chamber. In this study, three dimensional light energy distribution around a PL lamp was modeled based on three-parameter Gaussian model in air, model liquid foods and skim milk. Two different types of reactors, annular tube, and coiled tube are characterized based on hydrodynamics, potassium iodide-iodate based actinometry and computational fluid dynamics. Hydrodynamics result revealed a better performance of coiled tube having narrower residence time distribution, secondary flow vortices and more turbulence in the reactor which induce radial mixing as opposed to axial mixing in annular tube. Actinometry results showed higher PL dose (36.36-1148.63 J\/L) is absorbed in the coiled tube geometry than annular (22.22-322.22 J\/L). Numerical simulations also showed a higher process uniformity in the coiled tube PL reactors than annular tube. 3.2 Introduction  PL processing has been used in treatment of several liquid foods in batch and flow-through modes [51,73,81,127]. Unlike UV processing, the way PL treatments have been characterized has varied. This creates hurdles when comparing and replicating the earlier PL experiments as the parameters describing the PL process (pulse width, energy per pulse, voltage) are inherent to PL system used by different laboratories. Reactor configuration also determines the performance and efficacy of treatment along with lamp emittance and volumetric dose distribution. Liquid food parameters like absorption coefficients, UV transmission, clarity, and flow properties also determine reactor performance [17,62,126]. Proper design of PL reactors for liquid food products requires that all the volume elements of the fluid encounter uniform treatment while they flow through the reactor.  Thus, design and scale-up of an effective PL reactor is possible after taking into consideration all important process and product parameters, and by validating the design properly to ensure that the delivered dose inactivates microbial population satisfactorily. However, important considerations 51  of light energy distribution, reactor hydrodynamics and absorbed UV-C dose by the liquid in the PL reactors are lacking in previous studies. Computational fluid dynamics (CFD) based numerical simulations are important method for studying light-based systems, wherein they enable understanding of treatment systems considering all the underlying physics within the treatment. CFD also offers a non-expensive solution to test different configurations and measurement of properties which are otherwise difficult to measure due to physical constraints of the systems. Unlike the conventional UV-C treatment for which the CFD analysis has been well studied [138-141], PL systems have not been analyzed numerically yet to the best of our knowledge, which could be beneficial for better understanding.  Therefore, the main objective of this study is to characterize the PL reactors in two different configurations based on the modeling of spatial distribution of light energy, hydrodynamic performance of the reactors and actinometric dosage calculations. Finally, CFD analysis has been carried out to understand the transport process phenomena occurring inside PL reactors. 3.3 Materials and methods Liquid products of different clarity were used: deionized water, skim milk, red and green coloured model liquid foods (to simulate real foods like juices). Pasteurized Skim milk (Dairyland, Saputo Dairy products, BC, Canada) was purchased from the market. Red and green liquid solutions were obtained by dissolving 0.1 mL of food colour\/dye (McCormick Inc., London, ON, Canada) in 100 mL of deionized water. Physico-chemical properties were evaluated for the coloured model liquid solutions (Water+Red\/Green dye), milk and water. Total solids (TS %) was obtained for milk by gravimetric method [142]. Colour parameters (L*, a* and b*) were obtained by using colorimeter (HunterLab, model LabScanTM XE Plus, Hunter Associates Laboratory, Reston, VA, USA). Density (\ud835\udf0c) of all liquids were measured by method of weight of known volume. Viscosity (\ud835\udf07) was measured for the liquids using a rheometer (Model MCR 302, Anton Paar GmbH, Graz, Austria) under 0.1-100 \/s shear rate. Optical properties (Napierian absorption coefficient, \ud835\udefc\ud835\udc52,254, [\/m], UV transmission %, \ud835\udc47\ud835\udc48\ud835\udc49, [%], and penetration depth, \ud835\udeff37%, [m]) were calculated from the absorbance value at 254 nm (\ud835\udc34254) (measured on a cuvette of 1 cm path length with UV-Vis spectrophotometer (Shimadzu 1800, Tokyo, Japan)) using the Eq. 3.1-3.4. All the measured properties are given in the Table 3.1. 52  \ud835\udc3c = \ud835\udc3c\ud835\udc5c\ud835\udc52\u2212\ud835\udc34254 = \ud835\udc3c\ud835\udc5c\ud835\udc52\u2212\ud835\udefc\ud835\udc52,254\ud835\udc51                                                   (3.1) \ud835\udefc\ud835\udc52,254 = \ud835\udc34254\/\ud835\udc51                                                           (3.2) \ud835\udc47\ud835\udc48\ud835\udc49 % =\ud835\udc3c\ud835\udc3c\ud835\udc42\u00d7 100 = \ud835\udc52\u2212\ud835\udc34254 \u00d7 100                                              (3.3) \ud835\udeff37% =0.9943\ud835\udefc\ud835\udc52,254                                                              (3.4) where, \ud835\udc3c\ud835\udc5c = Incident radiation intensity, [W\/cm2]; \ud835\udc3c = Transmitted radiation intensity, [W\/cm2]; \ud835\udc51 = Path length in cuvette , [1 cm]; \ud835\udeff37% = Penetration depth where \ud835\udc3c\ud835\udc5c decreases to 37% of its value, [mm]. 53  Table 3.1: Calculated properties of the liquid products All data are expressed as mean \u00b1 standard deviation \u03c1 = Density; \u03bc = Viscosity; \ud835\udefc\ud835\udc52,254 = Napierian absorption coefficient; \u03b437% = penetration depth Product \ud835\udf36\ud835\udc86,\ud835\udfd0\ud835\udfd3\ud835\udfd2 [\/m] \ud835\udc7b\ud835\udc7c\ud835\udc7d [%] \u03b437% [cm] \u03c1 [kg\/m3] \u03bc [Pa-s] L* a* b* TS % Water 6.467 \u00b1 2.325 93.76 \u00b1 2.167 16.61 \u00b1 5.209 1002.7 \u00b1 3.449 0.869 \u00b1 0.046 _ _ _ _ Water+Red dye 24.07 \u00b1 0.115 78.61 \u00b1 0.091 4.132 \u00b1 0.020 1004.1 \u00b1 3.523 0.789 \u00b1 0.073 9.52 \u00b1 0.19 -3.87 \u00b1 0.14 9.97 \u00b1 0.66 _ Water+Green dye 37.60 \u00b1 0.608 68.66 \u00b1 0.418 2.645 \u00b1 0.043 1003 \u00b1 2.000 0.844 \u00b1 0.060 4.61 \u00b1 0.29 7.22 \u00b1 0.22 0.03 \u00b1 0.14 _ Skim Milk 244.3 \u00b1 12.27 8.737 \u00b1 1.082 0.408 \u00b1 0.021 1023.3 \u00b1 5.774 2.037 \u00b1 0.593 72.04 \u00b1 0.04 -5.68 \u00b1 0.01 -1.31 \u00b1 0.01 9.31 \u00b1 0.11 54  3.3.1 Pulsed UV light system Pulsed UV light system consisted of a cylindrical xenon flashlamp emitting high-intensity light pulses in UV-Visible-IR region (~200-1100nm), provided by Solaris Disinfection Inc. (Mississauga, ON, Canada). The lamp dimensions were\u2014length, \ud835\udc3f = 62 cm; diameter, \ud835\udc51\ud835\udc3f = 9 mm. The lamp emitted 30 J of energy per pulse. Considering a pulse width of 100 ms (since light fluctuation due to pulsation was hardly noticeable beyond 9 pulses\/s), the lamp sleeves emitted energy intensity per pulse of 17122.12 mW\/cm2.  Treatment of liquids was characterized for the PL reactors designed in two different configurations. Annular tube reactor (AT) configuration: The AT reactor was designed at Faculty of Land and Food Systems, University of British Columbia (Vancouver, BC, Canada) with the assistance of Solaris Disinfection Inc. The reactor consisted of a cylindrical annular treatment chamber with the lamp at the axis (Figure 3.1a) made from fused quartz glass of high clarity. The dimensions of the chamber were: length, \ud835\udc3f\ud835\udc34\ud835\udc47 = 62 cm; outer diameter, \ud835\udc37\ud835\udc5c= 4.5 cm. The liquid product was pumped into the system using an inlet section (9 mm internal dimeter) at one. The thickness (\ud835\udf06) of the liquid in the chamber was 1 mm (based on fabricator\u2019s specification) imparting a thin profile. The outlet section was present at the other longitudinal end of the reactor.  (a)  (b) Figure 3.1:  PL reactor configurations. Schematic representation of (a) AT and (b) CT Reactor 55  Coiled tube reactor (CT) configuration: The CT reactor was designed with the help of Solaris Disinfection Inc. and fabricated by Sandfire Scientific Ltd. (Gibsons, BC, Canada). The reactor consisted of a quartz glass coiled chamber with the lamp at the helical axis (Figure 3.2b). The mean helical or coil diameter (\ud835\udc51\ud835\udc50\ud835\udc5c\ud835\udc56\ud835\udc59) was 3.8 cm. The tube inner diameter (\ud835\udc51\ud835\udc61\ud835\udc62\ud835\udc4f\ud835\udc52) was 9 mm with a pitch (\ud835\udc5d) of 1.2 cm and 49 turns. The thickness of the liquid layer around the lamp was 9 mm (same as the inner diameter of tube). Both the systems included a peristaltic pump (Masterflex L\/S model 7554-90, Cole-Parmer Instruments, IL, United States) with a variable speed drive. The PL lamps were controlled by a digital controller by varying frequency and time. PL setup is given in Figure A1-1. Lamp emission spectrum is given in Figure A1-2. 3.3.2 Fluence mapping 3.3.2.1 Measurements in air The spatial distribution of lamp energy per pulse in air was mapped by measuring light energy around the lamp as per Hsu and Moraru [143], using a pyroelectric head sensor (PE80BF-DIF-C, Ophir-Spiricon LLC, UT, United states), placed under the lamp, with Nova II display (Ophir-Spiricon LLC, UT, United states) to record the energy. The circular sensor included a 1 cm2 circular aperture on it. Energy per pulse was recorded at various distances: along the x-axis (along the lamp length) every 2 cm; along y-axis (across the lamp at x =0 cm) every 0.5 cm; for 1-7 cm distances along the z-axis (vertically from the lamp). Measurements were done in triplicate. The sensor was rested for 30 s between subsequent readings to avoid saturation.   3.3.2.2 Measurements in liquid products Distribution of lamp energy within the liquids was measured to mimic the condition of real treatment within the reactors. For CT and AT reactors, the liquid layer around the lamp started radially at distances of 1.3 cm and 2 cm, respectively. Thus, petri dishes with the test liquids were kept under the lamp such that the liquid surface was at a mean distance 2.6 cm from lamp sleeve. The petri dish was centered over the sensor aperture to measure light energy through the liquid layer. Thickness of the liquid in petri dish was varied from 0.25-1 cm by adding more liquid and manipulating the dish distances from lamp. The arrangement is shown diagrammatically in Figure 3.2. For the mentioned thicknesses of liquids, the light energy was measured along the x-axis (every 2 cm) in triplicates.  56    Fig. 3.2: Light intensity measurement through liquid solutions 3.3.2.3 Modeling spatial distribution of light energy Preliminary investigation of the light energy distribution showed that the energy peaks at the lamp center and becomes lowest at either ends. Since the distribution was symmetric about the lamp center, a three-parameter Gaussian model for symmetrical curves as in Eq. 3.5 was used [143].  \ud835\udc3c(\ud835\udc65) = \ud835\udc34\ud835\udc52(\u22120.5(\ud835\udc65\u2212\ud835\udc4f\ud835\udc65)2\ud835\udc50\ud835\udc652 )                                                       (3.5) where, \ud835\udc3c(\ud835\udc65)=Light energy per pulse at x distance along x-axis, [mJ\/cm2]; \ud835\udc34= Amplitude or peak energy, [mJ\/cm2]; \ud835\udc4f\ud835\udc65= Constant; \ud835\udc50\ud835\udc65= \ud835\udc52\u22122 width along x-axis. For generating the three-dimensional energy distribution for air and model liquid foods, the energy data was analyzed using non-linear regression and were fitted to the Gaussian model to obtain \ud835\udc4f\ud835\udc65 and \ud835\udc50\ud835\udc65. Similar equation was used to model the energy distribution across the y-axis. The mean value (of triplicates) of energy data was fitted to the exponential model along the z-axis (Eq. 3.6).  \ud835\udc3c(\ud835\udc67) = \ud835\udc34\ud835\udc52\u2212\ud835\udc35\ud835\udc67                                                           (3.6) where, \ud835\udc3c(\ud835\udc67)= Light energy at distance z cm, [mJ\/cm2]; \ud835\udc35= Constant.  The experimental data along z-axis was fitted to the Eq. 3.6 and the constants, \ud835\udc34 and \ud835\udc35, were determined.  3.3.3 Hydrodynamic characterization 3.3.3.1 Calculation of flow properties The flow and hydrodynamic characterization of PL reactors was carried out for deionized water. Flow properties were calculated based on the flow conditions in the reactors. Deionized water was pumped at five different flow rates (14-75 L\/h). The flow rate (?\u0307?, [m3\/s]) of water was calculated 57  by measuring volume of water collected in a beaker after specific time. Average flow velocity (?\u0305?, [m\/s]) was calculated by Eq. 3.7.   ?\u0305? =?\u0307?\ud835\udf0b\ud835\udc51\u210e2\/4                                                                (3.7) where, \ud835\udc51\u210e= Hydraulic diameter, [mm],  calculated by dividing the cross-section area of reactor channel by wetted perimeter. The \ud835\udc51\u210e for AT and CT reactors were 2 mm and 9 mm, respectively. The flow conditions within the reactors for each flow operations were characterized by the Reynolds number (\ud835\udc41\ud835\udc45\ud835\udc52) (Eq. 3.8). \ud835\udc41\ud835\udc45\ud835\udc52 =\ud835\udf0c\ud835\udc51\u210e?\u0305?\ud835\udf07                                                              (3.8) The flow condition was characterized as laminar for \ud835\udc41\ud835\udc45\ud835\udc52< 2300 and transition zone above \ud835\udc41\ud835\udc45\ud835\udc52 of 104 with the transition zone lying in-between: 2300 < \ud835\udc41\ud835\udc45\ud835\udc52 < 104 [144]. In addition to \ud835\udc41\ud835\udc45\ud835\udc52, another dimensionless parameter Dean number (\ud835\udc41\ud835\udc37\ud835\udc52), accounting for the secondary flow (vortices) in the liquid over the primary forward flow in CT reactor (Eq. 3.9). \ud835\udc41\ud835\udc37\ud835\udc52 = \ud835\udc41\ud835\udc45\ud835\udc52\u221a\ud835\udc51\ud835\udc61\ud835\udc62\ud835\udc4f\ud835\udc52\ud835\udc51\ud835\udc50\ud835\udc5c\ud835\udc56\ud835\udc59                                                           (3.9) Dean flow occurs in the turbulent regime when \ud835\udc41\ud835\udc37\ud835\udc52 is greater than 400 [145]. 3.3.3.2 Residence time distribution The residence time distribution (RTD) was evaluated for the reactors by tracer-response method adopted by USEPA [146]. Water was pumped through the reactors at different flow rates. A tracer of sodium chloride, NaCl (10 mg\/mL) solution  mixed with green coloured dye was injected for 1 s. Time was recorded when the tracer solution enters inlet section and samples were drawn continuously after dye appears at outlet. The drawn samples were measured for electrical conductivity using conductivity meter (CDM 210, Radiometer Analytical SAS, Lyon, France). Conductivity values were converted into NaCl concentration using a standard curve (0-10 mg\/mL of NaCl) (shown in Figure A1-3). The tracer concentration versus time curve was analyzed to calculate RTD parameters. The experiments were done in triplicates. Time at which tracer first appears (\ud835\udc61\ud835\udc53) and when peak concentration appears (\ud835\udc61\ud835\udc5d) were calculated. Timepoints when 10, 50 and 90 % of tracer passes, denoted by \ud835\udc6110, \ud835\udc6150 and \ud835\udc6190 were also calculated. Theoretical residence time (\ud835\udc61\ud835\udc5f\ud835\udc52\ud835\udc60,\ud835\udc61\u210e) was calculated for all conditions using the Eq. 3.10. 58  \ud835\udc61\ud835\udc5f\ud835\udc52\ud835\udc60,\ud835\udc61\u210e =\ud835\udc49\ud835\udc61\u210e?\u0307?                                                               (3.10) where, \ud835\udc49\ud835\udc61\u210e= Theoretical volume of the reactor, [m3].  Mean residence time, \ud835\udc61\ud835\udc5a, was calculated based on the centroid of the distribution, by calculating the first moment (Eq. 3.11), approximated by Eq. 3.12. \ud835\udc61\ud835\udc5a =\u222b \ud835\udc36\ud835\udc61\ud835\udc51\ud835\udc61\u221e0\u222b \ud835\udc36\ud835\udc51\ud835\udc61\u221e0                                                            (3.11) where, \ud835\udc36 = Concentration [mg\/mL]; \ud835\udc61 = time [s].  \ud835\udc61\ud835\udc5a =\u2211\ud835\udc36\ud835\udc56\ud835\udc61\ud835\udc56\u2211\ud835\udc36\ud835\udc56                                                              (3.12) The dimensionless number, dispersion number (\ud835\udc41\ud835\udc37\ud835\udc56), showing the axial dispersion in the reactors was calculated using the variance \ud835\udf0e2 (or spread) of the distribution (Eq. 3.13 and 3.14). \ud835\udf0e2 =\u2211\ud835\udc36\ud835\udc56\ud835\udc61\ud835\udc562\u2211\ud835\udc36\ud835\udc56\u2212 \ud835\udc61\ud835\udc5a2                                                           (3.13) \ud835\udc41\ud835\udc37\ud835\udc56 =\ud835\udf0e22\ud835\udc61\ud835\udc5a2                                                                 (3.14) Using above values, \ud835\udc61\ud835\udc53\ud835\udc61\ud835\udc5f\ud835\udc52\ud835\udc60,\ud835\udc61\u210e , \ud835\udc61\ud835\udc5d\ud835\udc61\ud835\udc5f\ud835\udc52\ud835\udc60,\ud835\udc61\u210e, Morrill dispersion index ( \ud835\udc6190\ud835\udc6110), \ud835\udc61\ud835\udc5a\ud835\udc61\ud835\udc5f\ud835\udc52\ud835\udc60,\ud835\udc61\u210e, and \ud835\udc6150\ud835\udc61\ud835\udc5a were calculated for the reactors. 3.3.4 Potassium iodide-iodate based actinometry The absorbed UV dose at 254 nm (having peak germicidal effect) by a liquid in the reactor was calculated by potassium iodide-iodate (KI-KIO3) based actinometry. The actinometry method was used based on Rahn et al. [137], with suitable modifications by Bohrerova et al. [136] for PL treatment. The actinometry solution (0.6 M KI and 0.1 M KIO3 [VWR Chemicals LLC., OH, United States] in 0.01 M sodium borate (Fisher Scientific, IL, United States) buffer; pH 9.25) absorbs UV-C light and forms triiodide (I3-) ions as per (Eq. 3.15). 8\ud835\udc3e\ud835\udc3c + \ud835\udc3e\ud835\udc3c\ud835\udc423 + 3\ud835\udc3b2\ud835\udc42     \u210e\ud835\udf08    \u2192    3\ud835\udc3c3\u2212 + 6\ud835\udc42\ud835\udc3b\u2212 + 9\ud835\udc3e+                               (3.15) The actinometry solution was pumped through the reactors at the same flow rates. The PL lamp was operated at 1, 3, and 5 Hz pulse frequency while the solution was flowing in the reactor. The selected pulse frequency values were based on literature [51]. Initial and final temperatures (\ud835\udc47\ud835\udc56 and \ud835\udc47\ud835\udc53) of the solution were recorded using a resistance thermometer (BIOS Medical, Newmarket, ON, Canada). After treatments, the absorbance of the solution was measured at 352 nm (\ud835\udc34352). The 59  absorbance of untreated solution was measured at 300 nm (\ud835\udc34300). Experiments were done in triplicates. The UV-C dose at 254 nm (\ud835\udc37254) was calculated using the Eq. 3.16 [147]. \ud835\udc37254 =\ud835\udc34352\u00d7\ud835\udc43254\ud835\udc5d\ud835\udc50\u00d7\ud835\udf19\u00d7\ud835\udf00352                                                   (3.16) where, \ud835\udc43254= Number of Joules per Einstein of 254 nm photons (4.751\u00d7105 J\/Einstein for PL processing [136]); \ud835\udc5d\ud835\udc50= Path length of cuvette, [1 cm]; \ud835\udf00352= Molar absorption coefficient of triiodide at 352 nm, [2.76\u00d7104 L\/(mol-cm)]; \ud835\udf19 = Quantum yield (Effects per photon, [Mol\/Einstein]) calculated as per Eq. 3.17 [136]. \ud835\udf19 = 0.47 \u00d7 [1 + 0.23 \u00d7 (\ud835\udc50\ud835\udc56 \u2212 0.577)] \u00d7 [1 + 0.02 \u00d7 (\ud835\udc47\ud835\udc56 \u2212 20.7)]                 (3.17) The Eq. 3.17 applied for temperature and concentration correction to quantum yield. \ud835\udc50\ud835\udc56 is the concentration of iodide (\ud835\udc34300\/1.061). 3.3.5 Numerical simulation The numerical simulation of fluid flow and light radiation distribution inside the reactor domains was carried out using CFD to gain an insight on the zones of minimum and maximum flow velocities and light energy values. 3.3.5.1 Three-dimensional geometry development and meshing The CFD software Ansys Fluent 2019 R2 (Ansys Inc., PA, United States) was used for simulation. The computer system was Intel Quad Core i5-3470S CPU with 16 GB RAM. The three-dimensional (3-D) geometry of the reactors was developed using Ansys DesignModeler. The AT and CT reactors\u2019 liquid domain volume were developed and modeled as fluid type geometry. The solid walls of the reactors (quartz) were modeled as solid type. The lamp was modeled as a hollow cylinder whose external walls emit light. Air domain was defined in-between external lamp walls and reactor walls. The 3-D geometry was then meshed to discretized into smaller finite control volumes. Different zones in the geometry were meshed into different shaped elements (tetrahedrons, quadrilaterals etc.). The mesh was made finer near the lamp walls and reactor walls. The AT and CT reactors\u2019 geometries were meshed into 425, 824 and 3, 810, 810 number of elements, respectively (Figure 3.3).  60    (a)  (b) Fig. 3.3. Meshed geometry of (a) AT reactor and (b) CT reactor 3.3.5.2 Model initial and boundary conditions The meshed geometry was imported into Ansys FLUENT tool. Fully developed flow conditions were assumed within the reactor. The turbulent conditions within the reactor were accounted for by the \u03ba-\u03b5 turbulence model in realizable mode. Enhanced wall treatment condition was chosen to resolve complex flow patterns near the reactor walls. Lamp outer wall and reactor walls were 61  defined as quartz glass; liquid domains were set as deionized water, red and green model liquid foods, and skim milk. The properties of the domains (obtained experimentally and from literature) are given in Table 3.2. Table 3.2. Properties defined or the domains in the CFD code Domain \ud835\udf36\ud835\udc86,\ud835\udfd0\ud835\udfd3\ud835\udfd2 [\/m] \ud835\udf48\ud835\udc86,\ud835\udfd0\ud835\udfd3\ud835\udfd2 [\/m] Refractive index \u03c1 [kg m3] \u03bc [cP] \ud835\udc8c [Wm-K] \ud835\udc84\ud835\udc91 [J\/kg-K] Air 0 0 1c 1.184g 0.0184h 0.026g 1007g Quartz 1 0.001 1.459d 2200d - 1.38d 670j Water 6.47 0.005a 1.333e 1003 0.869 0.6i 4186k Water+Red dye 24.07 0.005a 1.333e 1004 0.789 0.6i 4186k Water+Green dye 37.6 0.005a 1.333e 1003 0.844 0.6i 4186k Skim milk 244.3 20000b 1.340f 1023 2.037 0.54f 3800f \u03c1 = density; \ud835\udefc\ud835\udc52,254 = Napierian absorption coefficient; \ud835\udf0e\ud835\udc52,254 = scattering coefficient aJonasz and Fournier [148] bAernouts et al. [149] cRefractive index database [150] dProperties of fused silica [151] eIndex of refraction [152] f Walstra et al. [153] g Viscosity (\u03bc) of air, dynamic and kinematic [154] hAir - Dynamic and Kinematic Viscosity [155] iThermal conductivity, (k) [156]  jProperties of fused quartz [157] kSpecific heat, (cp) [158]  Flow velocity field was characterized assuming the liquids as isothermal (with initial temperature at 298.15 K), Newtonian, incompressible, inert, and constant thermophysical and optical properties were considered for the liquids. The inlet section was set as velocity inlet varying from 0.0309-0.162 m\/s for AT reactor and from 0.0625-0.3271 m\/s in the CT reactor based on the calculated average velocities (?\u0305?). Turbulent intensity was set at 5%. Outlet section was set as set as pressure outlet boundary condition. Fluid flow and hydrodynamics calculations were performed 62  numerically by solving the Reynolds-Averaged Navier-Stokes equations: continuity equation (Eq. 3.18) and momentum conservation equation (Eq. 3.19). \ud835\udf15\ud835\udf0c\ud835\udf15\ud835\udc61+ \u2207\ud835\udf0c?\u0305? = 0                                                           (3.18) \ud835\udf0c\ud835\udf15?\u0305?\ud835\udf15\ud835\udc61+ \ud835\udf0c(?\u0305?. \u2207)?\u0305? = \u2212\u2207\ud835\udc5d + \ud835\udf0c\ud835\udc54 + \ud835\udf07\u22072?\u0305?                                       (3.19) where, \u2207\ud835\udc5d= Pressure gradient, [Pa]; \ud835\udc54= Gravitational acceleration, [m\/s2]. The analytical solution of the above nonlinear partial differential equations for the used complex reactors geometries does not exist. Thus, the equations were solved numerically by finite volume method for each smaller control volume (elements). The residence time of particles in the fluid was obtained by Lagrangian particle track scheme in FLUENT by discrete phase modeling. Inert particles of size 1.0\u00d710-6 m were injected from the inlet section surface with flow rate scaled for inlet surface area. Particle movement was set according to Discrete random walk model. The inlet and outlet surfaces were set as \u2018escape\u2019 for particles, while the reactor wall was set as \u2018reflective\u2019 boundary.  The irradiance [W\/m2] distribution around the PL lamp was assumed be unaffected by the water flow field inside the reactor. The numerical modeling of the irradiance distribution around the lamp and within the reactors was carried out by solving the radiation transfer equation (Eq. 3.20) numerically using  the discrete ordinate (DO) method, which has been widely used for the modeling of the UV processing [140]. The equation describes the travel of photons from a light source and their losses of energy due to absorption and scattering effects and its solution gives the value of light intensity at any point around the lamp and within the reactor volume.  \ud835\udc51\ud835\udc3c\ud835\udf06,\u03a9\ud835\udc51\ud835\udc60+ (\ud835\udefc\ud835\udc52,\ud835\udf06 + \ud835\udf0e\ud835\udf06)\ud835\udc3c\ud835\udf06,\u03a9 =\ud835\udf0e\ud835\udf064\ud835\udf0b\u222b \ud835\udc5d(\u03a9\u2032 \u2192 \u03a9)\ud835\udc3c\ud835\udf06,\u03a9\ud835\udc51\u03a9\u2032.\u03a9\u2032=4\ud835\udf0b                         (3.20) where, \ud835\udc3c\ud835\udf06,\u03a9 = Light intensity of wavelength, \ud835\udf06, along \u03a9 direction; \ud835\udf0e\ud835\udf06 = Scattering coefficient of medium, [\/m]; \ud835\udc5d(\u03a9\u2032 \u2192 \u03a9) = Phase function describing the directional distribution of radiation.  The Eq. 3.20 was solved over several discretized solid angles with certain directional vectors with angular discretization at 5\u00d75 divisions and 3\u00d73 pixelations. Wavelength was set based on PL from 200-1100 nm. The lamp external wall was set as semi-transparent fully diffuse wall having multiple point sources emitting light in all direction . The irradiance value was 10000 W\/m2 (lamp 63  surface irradiance value, assumed based on the radiometer readings) for a 100 ms pulse (considering peak single pulse energy at 100 mJ\/cm2). External radiation temperature was set as 1 K to minimize at interference of outside radiation.  3.3.5.3 Solutions The numerical solution was carried out by Coupled method, with all the parameters, pressure, momentum, turbulent kinetic energy, eddy dissipation rate, and energy set at second order upwind [138]. The iterative solution procedure was selected until the residuals converged below 10-6.  3.3.6 Statistical analysis The data were expressed in terms of means \u00b1 standard deviation. The models for the energy distribution were determined for goodness of fit by standard error of mean, adjusted R2, and correlation coefficient using MS Excel (Microsoft Corp., WA, United States) and SigmaPlot 14.0 (Systat Software Inc., CA, United States). The graphs for energy distribution along x- and y- axes were plotted using MS Excel and MATLAB code (Mathworks, MA, United States). 3.4 Results and discussion 3.4.1 Fluence mapping The spatial distribution of light energy was measured by considering the lamp centre as the origin (0, 0, 0). The light energy was not uniform around the lamp and decreased with distance from lamp due to the attenuation of light energy by absorption, scattering and reflection. At the lamp center, the energy decreased exponentially with the distance from lamp, which is in accordance with the Beer-Lambert\u2019s law. Similar findings have been reported by Hsu and Moraru [143]. Light energy along the x-axis was symmetric, being highest at the lamp center, with very low energy at either ends of the lamp. The light energy values at the lamp ends were 5.13 and 23.6 % of the value at center at 1 and 7 cm distances from lamp, respectively. The values in-between [-24, 24] were considered for modeling of the distribution of light energy around the lamp as there was sharp decline along the ends (with decline of >50% of peak energy beyond these distances). The light energy values along the x-axis and y-axis for air for distances along z-axis is given in the scatterplots in Figure 3.4 and contours were generated using MATLAB codes as in Figure 3.5.  64   (a)  (b) Figure 3.4: Scatter-plots of light energy distribution in (a) x-z plane, (b) y-z plane 0102030405060708090-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30Fluence [mJ\/cm2]x-axis [cm]z=7 cm z=6 cm z=5 cm z=4 cm z=3 cm z=2 cm z=1 cm0102030405060708090-4 -3.5 -3 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5 3 3.5 4Fluence [mJ\/cm2]y-axis [cm]z=7 cm z=6 cm z=5 cm z=4 cm z=3 cm z= 2 cm z=1 cm65   Figure 3.5: Light energy distribution contours in (a) x-z plane, (b) y-z plane While for the x-axis, the curve was flatter, along the y-axis, the distribution was narrower as light energy decays rapidly about the origin in y-axis [139,159]. Fitting the experimental data to the exponential model yielded the constants, \ud835\udc34 and \ud835\udc35, with goodness of fit based on R2>0.92. Similarly, the constants for the Gaussian model were calculated. The predictive equations showed slight under-prediction for the light energy closer distances as well as farther distances. The Gaussian model predicted well at the intermediary distances for both the x- and y-axes. The energy distribution around lamp seems consistent with Hsu and Moraru [143].  The attenuation of light in the liquid solutions is an important consideration when designing PL processes to factor in how the light reaches the microorganisms. In the experiment, the radiometer was centered below the liquid layer in a petri dish. Ideally the radiometer sensor should be directly in contact with the liquid in the dish which was not possible. Thus, the petri dish adds to the attenuation of energy, leading to under-prediction. The light energy decreased with the increasing liquid depth. It was assumed that the light attenuates uniformly throughout a small section of the liquid column [143]. However, light scattering (for turbid liquid foods, microbial cells) offers additional challenge to this simplistic assumption.  The Gaussian model was used to fit the value of light energy for the model liquid foods. The light attenuation in the liquids was observed to be greater than the air as observed by the higher 66  exponential constant values for the liquids. This is expected due to the greater absorption coefficients of liquids than air. The overall equation for predicting the fluence distribution of light energy at any point (x, y, z) around the lamp in air, as well as the liquids can be obtained by combining equations for all the three axes (Eq. 3.21). \ud835\udc3c(\ud835\udc65, \ud835\udc66, \ud835\udc67) = \ud835\udc34\ud835\udc52\u2212(0.5(\ud835\udc65\u2212\ud835\udc4f\ud835\udc65)2\ud835\udc50\ud835\udc652 +0.5(\ud835\udc66\u2212\ud835\udc4f\ud835\udc66)2\ud835\udc50\ud835\udc662 +\ud835\udc35\ud835\udc67)                                         (3.21) where, (\ud835\udc65, \ud835\udc66, \ud835\udc67) = position vector of the point from the origin (0, 0, 0). The values of the constants in the Eq. 3.21 are given in Table 3.3. Figure A1-4 shows the plot for comparison of predicted values of energy versus the experimental values. Table 3.3: Values of constants for Eq. 3.21 for air and the model liquid foods Medium \ud835\udc68 \ud835\udc83\ud835\udc99 \ud835\udc84\ud835\udc99 \ud835\udc83\ud835\udc9a \ud835\udc84\ud835\udc9a \ud835\udc6a R2 Adjusted R2 Standard error of mean Air 89.689 0 43.174 0 2.845 0.235 0.925 0.924 4.59 Water 75.326 0 36.311 2.098 2.776 0.256 0.819 0.817 1.171 Water+Red dye 98.643 0 36.22 1.309 2.523 0.4152 0.897 0.896 1.059 Water+Green dye 80.802 0 35.18 1.395 2.803 0.3479 0.864 0.862 1.144 Skim milk 79.71 0 35.133 2.072 4.369 0.3471 0.864 0.862 1.144 The light energy values at the ends of lamp to the center of lamp was found to be constant for water, water+red dye and water+green dye liquids (~16, 12.5, 12 %, respectively) with the depth of liquids. For skim milk, the light energy at the lamp ends was 4.7 and 13.8 % of the value at the lamp center for depths of 1 and 0.25 cm, respectively. This reflects greater attenuation of light in the skim milk, which might be due to the scattering properties of the milk proteins [160]. This means that the PL processes for clear liquid foods like juices is favoured unlike opaque foods like milk. 3.4.2 Hydrodynamic characterization  Hydrodynamic performance of the AT and CT reactors was studied using deionized water. With pump speed control knob, ?\u0307? increased proportionally, with an increase in ?\u0305? and decrease in \ud835\udc61\ud835\udc5f\ud835\udc52\ud835\udc60,\ud835\udc61\u210e. The ?\u0305? values ranged from 0.0309-0.1620 m\/s and 0.0625-0.3271 m\/s for AT and CT reactor, 67  respectively. The  ?\u0305? values were significantly higher (p < 0.05) for CT reactor than AT reactor. Reynolds numbers (\ud835\udc41\ud835\udc45\ud835\udc52) were also calculated for the various flow conditions. The flow regime was laminar (\ud835\udc41\ud835\udc45\ud835\udc52< 2300) for all flow conditions in AT reactor; \ud835\udc41\ud835\udc45\ud835\udc52 were significantly lower (p < 0.05) than the CT reactor. The flow regime was transitional (\ud835\udc41\ud835\udc45\ud835\udc52 > 2300) for higher flow rates in CT reactor. Dean number (\ud835\udc41\ud835\udc37\ud835\udc52) which characterized the secondary flow vortices with the primary forward flow in the CT reactor due to coiled flow channel, was calculated. The secondary flow or vortices was first studied by Dean [134], arise due to centrifugal forces exerted on the fluid volume while rotation. The vortices are highly promoted when \ud835\udc51\ud835\udc61\ud835\udc62\ud835\udc4f\ud835\udc52\ud835\udc51\ud835\udc50\ud835\udc5c\ud835\udc56\ud835\udc59\u2208 (0.03, 0.10), which generate liquid mixing [161,162]. The \ud835\udc51\ud835\udc61\ud835\udc62\ud835\udc4f\ud835\udc52\ud835\udc51\ud835\udc50\ud835\udc5c\ud835\udc56\ud835\udc59 value for the CT reactor was 0.237, which is higher than the upper limit of desired \ud835\udc51\ud835\udc61\ud835\udc62\ud835\udc4f\ud835\udc52\ud835\udc51\ud835\udc50\ud835\udc5c\ud835\udc56\ud835\udc59 values. However, due to higher \ud835\udc41\ud835\udc45\ud835\udc52 and turbulent conditions, effective radial mixing and uniform treatment conditions exist in the CT reactor as compared to AT reactor. The values of ?\u0307?, \ud835\udc61\ud835\udc5f\ud835\udc52\ud835\udc60,\ud835\udc61\u210e, \ud835\udc41\ud835\udc45\ud835\udc52, and \ud835\udc41\ud835\udc37\ud835\udc52 for both the reactors are given in Table 3.4. For both the reactors, under all flow conditions, the RTD curve showed a steep spike of tracer concentration with a finite area. Ideally, the spike area should be of zero width mimicking plug flow conditions A typical RTD curve is shown in Figure A1-5. The parameter \ud835\udc61\ud835\udc53\ud835\udc61\ud835\udc5f\ud835\udc52\ud835\udc60,\ud835\udc61\u210e indicates the degree of \u201cmost severe short-circuiting\u201d which means the axial mixing of tracer. The value is ~1.0 for plug flow conditions while 0 for mixing [146]. In case of AT reactors, the values were << 1.0, showing excessive short-circuiting except at flow rate of 75 L\/h. For the CT reactor, the values approached 1.0 with increased \ud835\udc41\ud835\udc45\ud835\udc52, mimicking ideal reactor. Similar observations were made by M\u00fcller et al. [163], who obtained higher \ud835\udc61\ud835\udc53\ud835\udc61\ud835\udc5f\ud835\udc52\ud835\udc60,\ud835\udc61\u210e  for CT reactors than AT. \ud835\udc61\ud835\udc5d\ud835\udc61\ud835\udc5f\ud835\udc52\ud835\udc60,\ud835\udc61\u210e, indicating the effective volume of the reactor, should approach 1.0 for a good design [146]. The values approached 1.0 for both reactors, except it increased for AT reactor for high \ud835\udc41\ud835\udc45\ud835\udc52. Morrill dispersion index ( \ud835\udc6190\ud835\udc6110) reflects the width of the RTD curve and should be <2.0 for effective design [146]. For the AT reactor, the values were >2.0, while CT reactor showed values <2.0. The parameter \ud835\udc61\ud835\udc5a\ud835\udc61\ud835\udc5f\ud835\udc52\ud835\udc60,\ud835\udc61\u210e shows the usefulness of the whole reactor volume [146, 163]. Value of ~1.0 shows the effective volume of reactor is closer to the actual volume. For AT reactor, the values were >1.0 which is in 68  accordance with the results obtained by Koutchma and Parisi [164]. However, contrastingly Gayan et al. [165] showed effective use of such reactor volume by obtaining values of 0.944. The CT reactor showed effective use of reactor volume as reflected by values of ~1.0. The ratio \ud835\udc6150\ud835\udc61\ud835\udc5a shows the skewness of the RTD curve, where value <1.0 gives left skewness, and hence ineffective design of reactor. AT reactor showed value <1.0, where CT reactor showed values ~1.0. The dispersion was characterized by dimensionless number, \ud835\udc41\ud835\udc37\ud835\udc56, where AT reactor showed high dispersion (\ud835\udc41\ud835\udc37\ud835\udc56 > 0.1), while CT reactor showed moderate dispersion (\ud835\udc41\ud835\udc37\ud835\udc56 \u2208 [0.01, 0.10]). Overall, the CT reactor showed better hydrodynamic performance than AT reactor, as the latter showed more axial mixing than CT reactors.  It highlights the fact that CT geometries can help in obtaining better plug flow characteristics than AT reactor in case of UV processing.69  Table 3.4: Residence time distribution (RTD) parameters for AT and CT reactors Type of reactor ?\u0307? [L\/h] \ud835\udc75\ud835\udc79\ud835\udc86 \ud835\udc75\ud835\udc6b\ud835\udc86 \ud835\udc95\ud835\udc93\ud835\udc86\ud835\udc94,\ud835\udc95\ud835\udc89 [s] \ud835\udc95\ud835\udc8e [s] \ud835\udc95\ud835\udc87\ud835\udc95\ud835\udc93\ud835\udc86\ud835\udc94,\ud835\udc95\ud835\udc89 \ud835\udc95\ud835\udc91\ud835\udc95\ud835\udc93\ud835\udc86\ud835\udc94,\ud835\udc95\ud835\udc89 \ud835\udc95\ud835\udfd7\ud835\udfce\ud835\udc95\ud835\udfcf\ud835\udfce \ud835\udc95\ud835\udc8e\ud835\udc95\ud835\udc93\ud835\udc86\ud835\udc94,\ud835\udc95\ud835\udc89 \ud835\udc95\ud835\udfd3\ud835\udfce\ud835\udc95\ud835\udc8e \ud835\udc75\ud835\udc6b\ud835\udc8a AT 14.32 71.06 \u00b1 1.715 - 16.35 \u00b1 0.399 26.41 \u00b1 3.474 0.530 \u00b1 0.289 1.121 \u00b1 0.255 2.855 \u00b1 0.469 1.615 \u00b1 0.213 0.756 \u00b1 0.077 0.105 \u00b1 0.031 32.48 161.2 \u00b1 2.723 - 7.206 \u00b1 0.123 10.19 \u00b1 1.347 0.601 \u00b1 0.080 1.110 \u00b1 0.139 2.500 \u00b1 0.289 1.414 \u00b1 0.187 0.993 \u00b1 0.131 0.180 \u00b1 0.052 49.63 246.3 \u00b1 6.661 - 4.717 \u00b1 0.129 7.637 \u00b1 1.277 0.353 \u00b1 0.122 1.060 \u00b1 0.212 3.306 \u00b1 0.337 1.619 \u00b1 0.271 0.786 \u00b1 0.025 0.260 \u00b1 0.128 64.13 318.3 \u00b1 6.108 - 3.650 \u00b1 0.069 7.322 \u00b1 0.629 0.457 \u00b1 0.158 1.553 \u00b1 0.158 2.861 \u00b1 0.428 2.006 \u00b1 0.172 0.871 \u00b1 0.123 0.117 \u00b1 0.022 74.90 371.8 \u00b1 4.833 - 3.124 \u00b1 0.041 8.361 \u00b1 0.582 1.067 \u00b1 0.185 2.240 \u00b1 0.320 1.802 \u00b1 0.216 2.676 \u00b1 0.186 0.997 \u00b1 0.0426 0.056 \u00b1 0.011 CT 14.32 562.7 \u00b1 13.58 273.8 \u00b1 6.608 83.01 \u00b1 2.027 81.53 \u00b1 8.376 0.932 \u00b1 0.026 1.137 \u00b1 0.036 1.235 \u00b1 0.018 0.982 \u00b1 0.102 1.143 \u00b1 0.094 0.057 \u00b1 0.027 32.48 1276 \u00b1 21.56 621.1 \u00b1 10.49 36.59 \u00b1 0.624 39.62 \u00b1 2.745 0.948 \u00b1 0.032 1.139 \u00b1 0.018 1.233 \u00b1 0.023 1.084 \u00b1 0.088 1.089 \u00b1 0.073 0.038 \u00b1 0.018 49.63 1951 \u00b1 52.74 949.3 \u00b1 25.67 23.95 \u00b1 0.655 26.701 \u00b1 5.081 0.947 \u00b1 0.030 1.156 \u00b1 0.074 1.323 \u00b1 0.062 1.114 \u00b1 0.204 1.128 \u00b1 0.230 0.049 \u00b1 0.058 70  64.13 2520 \u00b1 48.36 1227 \u00b1 23.53 18.53 \u00b1 0.355 22.33 \u00b1 0.358 0.917 \u00b1 0.036 1.098 \u00b1 0.037 1.295 \u00b1 0.0862 1.205 \u00b1 0.034 0.986 \u00b1 0.016 0.014 \u00b1 0.003 74.90 2944 \u00b1 38.27 1433 \u00b1 18.62 15.86 \u00b1 0.206 17.91 \u00b1 4.272 0.988 \u00b1 0.048 1.177 \u00b1 0.039 1.343 \u00b1 0.075 1.129 \u00b1 0.270 1.210 \u00b1 0.354 0.064 \u00b1 0.085 ?\u0307? = volumetric flow rate; \ud835\udc41\ud835\udc45\ud835\udc52 = Reynolds number; \ud835\udc41\ud835\udc37\ud835\udc52 = Dean number; \ud835\udc41\ud835\udc37\ud835\udc56 = Dispersion number; \ud835\udc61\ud835\udc53 = time at which tracer first appears; \ud835\udc61\ud835\udc5d = time when peak concentration appears;  \ud835\udc6110, \ud835\udc6150 and \ud835\udc6190 = timepoints when 10, 50 and 90 % of tracer passes;  \ud835\udc61\ud835\udc5f\ud835\udc52\ud835\udc60,\ud835\udc61\u210e = theoretical residence time; \ud835\udc61\ud835\udc5a = mean residence time  71  3.4.3 Actinometric dose calculation The UV-C dose (\ud835\udc37254) absorbed by a liquid flowing through the reactor was obtained by actinometry. The actinometric solution absorbs light photons while flowing in the reactors when the PL system was running at pulse frequencies of 1, 3 and 5 pulses\/s [Hz]. The absorbed \ud835\udc37254 dose for AT and CT reactor ranged from 36.36-1149 J\/L and 22.22-322.2 J\/L, respectively. The values of the \ud835\udc37254 for AT and CT reactors are shown graphically in Fig. 3.6(a) and 3.6(b), respectively. As the pulse frequency increased, the absorbed \ud835\udc37254 dose increased for the same flow rates, due to absorption of more photons. However, the absorbed dose decreased with increasing flow rates. This was indicated by higher correlation coefficients (AT reactor: r = 0.756; p < 0.05 and CT reactor: r = 0.756; p < 0.05). With increased flow rate and \ud835\udc41\ud835\udc45\ud835\udc52, the residence time decreased, leading to lesser absorption of \ud835\udc37254 dose. The decrease in UV-C dose with flow rate have been reported earlier [147,163]. These results necessitate the selection of flow rates of liquids and pulse frequency which impart better germicidal effect.  (a) 05010015020025030035014.32 32.48 49.63 64.13 74.90Energy dose [J\/L]Flow rate [L\/h]Annular tube_1 Annular tube_3 Annular tube_572   (b) Figure 3.6. Absorbed UV-C energy dose. Graphical representation of absorbed UV-C dose by actinometry for (a) AT reactor, (b) CT reactor. Numbers (Annular_# or Coiled tube_#)  in the figure legends denote pulse frequency [Hz] 3.4.4 Numerical simulation 3.4.4.1 Flow velocity profile Flow field and hydrodynamics in both the reactors were analyzed by CFD. The flow conditions were varied in numerical scheme based on the experimental conditions (with only water) used. Since the results for the velocity in case of other liquids (water+red\/green dye, and skim milk; data not shown) did not vary much as compared to water, only results for the water are discussed here. Numerically calculated radial velocity distribution in the reactors was analyzed. Ideally, the radial velocity profiles in the straight tube or AT reactors are parabolic with transition to a flatter profile with increased turbulence [138]. This is characterized by ?\u0305?\/\ud835\udc63\ud835\udc5a\ud835\udc4e\ud835\udc65 (ratio of average velocity in the reactor to maximum velocity) of ~0.5. For all flow conditions in the AT reactor, the ?\u0305?\/\ud835\udc63\ud835\udc5a\ud835\udc4e\ud835\udc65 ratio was ~0.5. This agrees with Sozzi and Taghipour [`38]. In case of the CT reactor, the resultant  ?\u0305?\/\ud835\udc63\ud835\udc5a\ud835\udc4e\ud835\udc65 values were >0.5 (ranging from 0.689-0.752), which show greater turbulence and radial liquid mixing. The pressure drops in liquid (\u2206\ud835\udc43) while flowing through the reactors was also found numerically for all the flow conditions. The \u2206\ud835\udc43 values for the CT reactor were significantly higher (p < 0.05) than AT reactor, which was also reported by Dean [134]. The values of \u2206\ud835\udc43 also increased 020040060080010001200140014.32 32.48 49.63 64.13 74.90Energy dose [J\/L]Flow rate [L\/h]Coiled tube_1 Coiled tube_3 Coiled tube_573  with flow velocities (for AT: r = 0.995; CT: r = 0.993). The values for ?\u0305?\/\ud835\udc63\ud835\udc5a\ud835\udc4e\ud835\udc65 and \u2206\ud835\udc43 are given in Table 1.4. The simulated velocity profiles in the reactors are given in Fig. 3.7.  The results for residence time for particles simulated by CFD showed good comparison with RTD results obtained experimentally, especially for CT reactor. There are some discrepancies between the results for residence times calculated by CFD compared to experimental \ud835\udc61\ud835\udc5f\ud835\udc52\ud835\udc60,\ud835\udc61\u210e. This is due to the assumption that the geometries modeled in software were simplistic version of the ones used experimentally. The results have been summarized in Table 3.5.74    (a) 75   (b) Figure 3.7: Velocity distribution. Velocity distribution for (a) AT reactor (0.162 m\/s) shown by velocity vectors; (b) CT reactor (0.28 m\/s) velocity contours in YZ plane at x=0; XY plane at z =0, with inlet and outlet profiles. 76  Table 3.5: Simulated results for the flow velocity, pressure drop and residence time ?\u0305? = average velocity; \ud835\udc63\ud835\udc5a\ud835\udc4e\ud835\udc65 = maximum velocity; \u2206\ud835\udc43 = Pressure drop; \ud835\udc61\ud835\udc5a = mean residence time Type of reactor ?\u0305? [m\/s] Simulated maximum velocity (\ud835\udc97\ud835\udc8e\ud835\udc82\ud835\udc99) ?\u0305?\/\ud835\udc97\ud835\udc8e\ud835\udc82\ud835\udc99 \u2206\ud835\udc77 [Pa] Experimental \ud835\udc95\ud835\udc8e [s] Simulation Minimum residence time (\ud835\udc95\ud835\udc8e\ud835\udc8a\ud835\udc8f, [s]) Maximum residence time (\ud835\udc95\ud835\udc8e\ud835\udc82\ud835\udc99, [s]) AT 0.0309 0.0691 0.4472 68.40 26.41 \u00b1 3.474 3.000 6.154 0.070 0.1593 0.4394 185.01 10.19 \u00b1 1.347 0.000 15.790 0.107 0.2404 0.4451 326.22 7.640 \u00b1 1.277 3.000 6.234 0.138 0.3138 0.4398 471.36 7.320 \u00b1 0.629 2.000 4.346 0.162 0.3717 0.4358 591.74 8.360 \u00b1 0.582 15.90 52.90 CT 0.0625 0.0912 0.6855 297.90 81.53 \u00b1 8.376 63.90 76.64 0.1418 0.1861 0.7620 890.29 39.62 \u00b1 2.745 28.20 36.35 0.2167 0.2772 0.7818 1631.30 26.70 \u00b1 5.081 18.43 24.10 0.2800 0.3548 0.7892 2406.55 22.33 \u00b1 0.358 11.95 16.50 0.3271 0.4128 0.7923 3082.44 17.91 \u00b1 4.272 5.000 12.93 77  3.4.4.2 Irradiance profile Single pulse irradiance was modeled for air and the reactors through which flow of different liquids were simulated. Velocity profile has no effect on the irradiance distribution within the reactor. Though, the velocity associated with different particles within the flow field affect the fluence rate and hence UV-C dose during the flow. Figure 3.8 shows the contours of irradiance profile around the lamp. It can be seen that the diffused lamp settings approximate the light energy distribution as obtained earlier experimentally (Fig. 3.5). Figure 3.9 (a) and 3.9 (b) shows the contours of irradiance for the AT and CT configurations, respectively along with minimum irradiance (\ud835\udc3c\ud835\udc5a\ud835\udc56\ud835\udc5b), maximum irradiance (\ud835\udc3c\ud835\udc5a\ud835\udc4e\ud835\udc65) and volume-averaged irradiance (\ud835\udc3c\ud835\udc63\ud835\udc5c\ud835\udc59\u0305\u0305 \u0305\u0305\u0305). The irradiance distribution for AT reactor agrees with results of Casado et al. [140]. However, the irradiance distribution for CT reactor could not be compared as no previous CFD analysis has been carried out for CT reactors. The values reveal an over-estimation of irradiance values for all the conditions for CT reactor. This could be due to complex design of the reactor, whereby, the tubular coils reflect and refract light within the reactor. The values for fluence rates computed for different liquid products at different flow rates for 3 Hz PL frequency is given in Table 3.6.  Figure 3.8: Irradiance distribution in air with minimum, maximum and volume-averaged irradiance. Contours of irradiance profile in XZ plane at y=0 and YZ plane at x=078   (a) 79   (b) Figure 3.9:  Irradiance distribution with minimum, maximum and volume-averaged irradiance. Irradiance contours for different liquids in XZ plane for (a) AT reactor at y=0 and YZ plane at x= 0, -30.5, -15, 15, and 30.5 cm. (b) CT reactors in XZ plane at y=0 80  Table 3.6. Computed values of total fluence rates based on volume-averaged irradiance for different liquids at various flow rates and 3 Hz frequency Type of reactor Medium Total fluence rates [W\/cm2] Flow rates [L\/h] 14.3 32.5 49.6 64.1 74.9 AT Water 16.13 7.107 4.652 3.599 3.081 Water+Red dye 15.72 6.927 4.534 3.508 3.003 Water+Green dye 15.42 6.796 4.448 3.442 2.946 Skim milk 1.522 0.671 0.439 0.339. 0.291 CT Water 80.83 35.63 23.32 18.04 15.45 Water+Red dye 72.06 31.76 20.79 16.09 13.77 Water+Green dye 53.78 23.70 15.51 12.00 10.28 Skim milk 1.263 0.557 0.364 0.282 0.241 The computed value of irradiance at the lamp sleeve i.e., 17079.66 W\/m2 matches the value found experimentally (for a 62 cm long and 9 mm wide lamp, emitting 30 J\/pulse at 100 ms pulse width, the incident irradiance is 17122.12 W\/m2). 3.4.4.3 Process uniformity For PL reactor design for liquid food treatment, it is necessary that all the liquid elements receive appropriate and uniform UV-C dose. The velocities of particles determine the flow field and the resultant RTD within the reactor. Faster particles receive lesser UV-C dose than slower moving particles. Calculation based on average UV-C dose and velocities within the reactors might not give total picture of microbial inactivation by PL. Distribution of UV-C dose and velocities is thus necessary. In case of AT reactor, the velocity profile (obtained numerically) within the annulus is almost flatter (Fig. 3.10a). Also, the irradiance profile does not vary to a great extent, except for skim milk (Fig. 3.10b). This will ensure that the fluid elements receive uniform treatment within the AT reactor. However, the AT reactor led to enhanced axial mixing, which is unfavorable and lead to under-treatment during PL processing.  81   (a)  (b) Figure 3.10: PL process Uniformity. Graphical representations of (a) Velocity profile for AT reactor; (b) Irradiance profile for AT reactor  (a) 00.020.040.060.080.10.120.140.160.180.20 0.0002 0.0004 0.0006 0.0008 0.001 0.0012Velocity [m\/s]Distance [m]14.2 L\/h 32.48 L\/h 49.63 L\/h 64.13 L\/h 74.9 L\/h00.20.40.60.811.20 0.0002 0.0004 0.0006 0.0008 0.001 0.0012I\/I maxDistance (m)Water Skim milk Red Liquid Green Liquid00.050.10.150.20.250.30.350.40.450 0.002 0.004 0.006 0.008 0.01Velocity [m\/s]Distance [m]14.32 L\/h 32.48 L\/h 49.63 L\/h 64.13 L\/h 74.9 L\/h82   (b) Figure 3.11: PL process Uniformity. Graphical representations of (a) Velocity profile for CT reactor; (b) Irradiance profile for CT reactor In case of the CT reactor, the velocity profile across the tube diameter is non-parabolic with faster fluid elements residing along the outer walls of the tube (Fig. 3.11a). Whereas the light energy and hence irradiance decrease exponentially along the tube width (Fig. 3.11b). There was a good agreement between experimental and simulated values (water: R2 = 0.89 and standard error = 0.073; Water+Red dye: R2 = 0.94 and standard error = 0.066; Water+Green dye: R2 = 0.88 and standard error = 0.085; skim milk: R2 = 0.98 and standard error = 0.060). The irradiance decreased as the clarity and absorbance coefficient (\ud835\udefc\ud835\udc52,254) of the liquids increased (Water > Water+Red dye > Water+Green dye >Skim milk). This is expected as the \ud835\udefc\ud835\udc52,254 of the medium increases, the light transmission decreases. Water having the least \ud835\udefc\ud835\udc52,254 (6.467 \/m) resulted the maximum light transmission through it as compared to the skim milk (\ud835\udefc\ud835\udc52,254=244.3 \/m) [133]. This again indicates that the clear liquid foods tend to be better suited to PL and similar technologies. Also, it seems that the faster particles away from the lamp may not receive required UV-C dose and escape untreated, which could be an additional hurdle in case of opaque liquids like milk. But due to coiled geometry, the generation of Dean vortices will lead to proper radial mixing and hence more uniform treatment. Overall, better uniformity and higher UV-C dose absorption also confirm the 0 0.002 0.004 0.006 0.008 0.0105001000150020002500300035004000450000.20.40.60.811.2Experimental irradiance [W\/m2]I\/I maxDistance [m]Water Skim Milk Red liquidGreen liquid Water_exp Red Liquid_expGreen Liquid_exp Skim milk_exp83  results of the CFD that CT reactors performs better than AT reactors, which need to be further validated using microbial challenge studies, that form the topic of subsequent sections.  3.5 Conclusion In this study, the light energy distribution of PL lamp in air and in different liquids were modeled by three-parameter Gaussian model, whereby it was found that light energy decreased as the liquid clarity decreased. Hydrodynamic performances of the coiled and annular geometry PL reactors were compared and evaluated by hydrodynamic and actinometric studies. The coiled geometry offered better hydrodynamic performance by virtue of narrower RTD and approximation of plug flow. Additionally, the Dean vortices induced radial mixing in the reactor. Higher UV-C dose absorption in coiled reactor was obtained. CFD analysis also helped in visualizing the flow field and irradiance profiles within the reactors. CFD results also showed the decrease in irradiance values with clarity of liquid. While clear liquid foods like juices will be more suitable to PL processing, the opaque liquids like milk could also be treated using coiled reactors due to better mixing and higher UV-C dose absorption. However, these results need to be validated with microbial challenge studies which will help in design and scale up of a continuous PL processing system for various clear and opaque\/turbid liquid foods.               84  Chapter 4: Microbial validation and UV dose determination for continuous-flow pulsed UV light reactors 4.1 Summary The light energy distribution in liquids varies as a function of clarity and optical properties of the liquids. This means that the susceptibility of microorganisms to UV light in different liquids will also vary according to the optical properties of the liquids. In this chapter, the delivered fluences for different process parameters (flow rate and pulse frequency) for the model liquids (water, water+red dye, water+green dye, skim milk) for AT and CT reactor were calculated based on the particle tracking scheme of the CFD package. Then for validation, the model liquids were inoculated with test microorganisms (E. coli, L. innocua, C. sporogenes) and PL-treated (flow rate: 14.3-74.9 L\/h; Frequency: 1-5 Hz) in AT and CT reactor, and microbial inactivation kinetics was modeled. Collimated tube experiment was also carried out to ascertain the UV dose (D-value) of microorganisms in different liquids. It was observed that the inactivation of microorganisms depended on the optical properties of the liquid and varied as: water > water+red dye >  water+green dye  > skim milk. Also, greater inactivation was achieved for CT reactor than AT. The microbial inactivation kinetics was best fitted to log-linear with tail order. The inactivation levels of the microorganisms in different liquids also followed the same order (water > water+red dye >  water+green dye > skim milk). For the microorganisms susceptibility to PL, inactivation followed the order E. coli > L. innocua > C. sporogenes.  4.2 Introduction An appropriate measurement of PL fluence during PL experiments involving a continuous-flow reactors are challenging and as such processes like chemical actinometry, biodosimetry have been employed. This is because it is difficult to measure energy using the radiometers inside the reactors due to volume constraints of the reactor chambers. CFD is one of the most important tool that used for visualization of flow and performance modeling of reactors. CFD has also been used to calculate imparted UV dose and dose distribution in continuous UV-C processing [138]. The UV dose and dose distribution is calculated employing CFD using Lagrangian method, whereby finite number of particles are tracked in the flow field in a reactor by balancing the forces (buoyancy, gravity, hydrodynamic) on the particles. Based on the position of the particles, the absorbed dose is integrated along the path of the particle using the CFD code to predict the cumulative fluence 85  delivered. The absorbed dose at each point along the reactor path is calculated from the product of fluence rate at that point and the time period the particle was exposed to that fluence rate. CFD particle tracking analysis of PL process is lacking. Carrying out CFD particle tracking to ascertain the delivered fluence in continuous reactors will be beneficial.  The light energy values and irradiances in liquids has been observed to vary according to the absorption coefficient, UV transmissivity and penetration depths of light in liquid [135]. Clear liquids tend to be decontaminated easily compared to the semi-transparent and opaque foods. Inactivation rates of microorganisms in liquids subjected to PL treatment thus will depend on the liquid properties. This will further affect the inactivation kinetics of microorganisms in liquids during PL treatment. Therefore, a deviation from conventional log-linear kinetics has been observed during PL treatment [85]. Biodosimetry has been used for estimation of the UV dose of test microorganisms using a batch collimated tube setup, whereby a sample inoculated with the test microorganisms is subjected to different fluences to calculate the susceptibility of the microorganisms to light. D-value which represents the amount of energy required to bring 90% reduction in microorganisms population, is calculated using collimated tube experiments since, the light rays are mostly perpendicular to the test sample surface and the light energy\/fluence measurements can be accurately carried [23]. The results of the biodosimetry could also be extended to determine the reduction equivalent fluence value for the reactors [166]. While biodosimetry has been widely used in UV-C processing, till now, no such application exists for PL processing. The objective of this research is to calculate the delivered fluence in different model liquids under different PL processing conditions (by varying flow rate and frequency) using CFD. Microbial validation after inoculation of model liquids with challenge microorganisms and PL treatment for the same conditions has been carried out to see the effect on microbial inactivation in the liquids. Microbial inactivation kinetics is also modeled using kinetic models. Lastly, biodosimetry using the challenge microorganisms is also caried out to ascertain the D-values of microorganisms in different liquids.   86  4.3 Materials and methods 4.3.1 Materials Liquid products of different clarity were used: deionized water, skim milk, red and green coloured model liquid foods (to simulate real foods like juices). Skim milk (Dairyland, Saputo Dairy products, BC, Canada) was purchased from the market. Autoclaved ultrapure water (18 M\u03a9) was used for dilution and making reagents for analyses. Red and green liquid solutions were obtained by dissolving 0.1 mL of red and green food dye (McCormick Inc., London, ON, Canada) in 100 mL of autoclaved ultrapure water. Physico-chemical properties were evaluated for the coloured model liquid solutions (Water+Red\/Green dye), milk and water. Total solids (TS %) was obtained for milk by gravimetric method [142]. Colour parameters (L*, a* and b*) were obtained by using colorimeter (HunterLab, model LabScanTM XE Plus, Hunter Associates Laboratory, Reston, VA, USA). Density (\ud835\udf0c) of all liquids were measured by method of weight of known volume. Viscosity (\ud835\udf07) was measured for the liquids using a rheometer (Model MCR 302, Anton Paar GmbH, Graz, Austria) under 0.1-100 \/s shear rate. Optical properties (Napierian absorption coefficient, \ud835\udefc\ud835\udc52,254, [\/m], UV transmission %, \ud835\udc47\ud835\udc48\ud835\udc49, [%], and penetration depth, \ud835\udeff37%, [m]) were calculated from the absorbance value at 254 nm (\ud835\udc34254) (measured on a cuvette of 1 cm path length with UV-Vis spectrophotometer (Shimadzu 1800, Tokyo, Japan)) using the Eq. 3.1-3.4 (chapter 3). All the measured properties are given in the Table 3.1 (chapter 3). 4.3.2 Microbial strains Escherichia coli K-12 ATCC 29055, Listeria innocua ATCC 33090, and Clostridium sporogenes ATCC 7955 were used as challenge microorganisms for microbial inactivation studies in water, model liquids and skim milk using PL. The growth and preparation of strains are mentioned below. All the microbial strains were procured from  American Type Culture Collection (ATCC, Manassas, VA, United States) as lyophilized pellets. E. coli ATCC 29055: E. coli ATCC 29055 is a vegetative bacteria which has been used in PL studies as a non-pathogenic surrogate for E. coli O157:H7 in previous PL processing studies [167,168]. The lyophilized pellet was aseptically suspended in 15 mL aliquot of autoclaved nutrient broth (Sigma Aldrich, St. Louis, MO, United States), which was then mixed back to the nutrient broth bottle. The culture was grown for 24 h at 37 oC in an incubator. The broth was then centrifuged at 5000g for 15 min at 4 oC and supernatant was discarded. The pellet was resuspended 87  in autoclaved double distilled water. The previous washing and centrifugation steps were repeated twice before final resuspension of recovered cell pellet into autoclaved double distilled water. The final stock solution contained approximately 108 CFU\/mL. This solution was used as inoculum for PL treatments. The stock culture was stored at -80 oC added with 30% (v\/v) glycerol solution for long-term storage.  L. innocua ATCC 33090: L. innocua 33090 is a vegetative bacteria, which was used as a surrogate for L. monocytogenes, by virtue of their phenotypic similarity [82]. L. monocytogenes has been observed to be one of the most resistant bacteria to PL and L. innocua is proposed to be applicable for validation studies [92,96]. The lyophilized pellet was aseptically suspended in 15 mL aliquot of autoclaved brain heart infusion (BHI) broth (BD Bacto, Sparks, MD, United States), which was then mixed back to the BHI broth bottle. The medium was incubated at 37 oC for 24 h in an incubator. The cells were recovered from the broth by centrifugation at 4000g for 15 min at 4 oC, and by discarding the supernatant. The pellet was resuspended in autoclaved double distilled water aseptically. The previous washing and centrifugation steps were repeated twice before final resuspension of recovered cell pellet into autoclaved double distilled water. The bacterial suspension contained around 107-108 CFU\/mL which was used as inoculum for PL treatments. The stock culture was stored at -80 oC added with 30% (v\/v) glycerol solution for long-term storage. C. sporogenes ATCC 7955: C. sporogenes ATCC 7955 is an anaerobic, mesophilic spore forming bacteria. The strain has been used for thermal processing studies as a surrogate bacteria for C. botulinum [168]. The lyophilized pellet was aseptically suspended in 15 mL aliquot of autoclaved reinforced Clostridial medium (RCM) (Thermo Scientific, United Kingdom), which was then mixed back to the RCM bottle. The medium was incubated at 37 oC for 5 days under anaerobic conditions using an anaerobic jar. The cells were recovered from the RCM by centrifugation at 4000g for 15 min at 4 oC, and by discarding the supernatant. The pellet was resuspended in autoclaved double distilled water aseptically. The previous washing and centrifugation steps were repeated twice before final resuspension of recovered cell pellet into autoclaved double distilled water. The spore bacterial suspension contained around 107-108 CFU\/mL which was used as 88  inoculum for PL treatments. The stock culture was stored at -80 oC added with 30% (v\/v) glycerol solution for long-term storage. 4.3.3 Calculation of delivered fluence 4.3.3.1 Particle tracking by computational fluid dynamics The particle tracking was carried out using CFD in FLUENT (Ansys) as described in the chapter 3 (section 3.3.5.2) by Lagrangian particle track scheme by discrete phase modeling. Inert particles of size 1.0\u00d710-6 m were injected from the inlet section surface with flow rate scaled for inlet surface area. Particle movement was set according to Discrete random walk model. The inlet and outlet surfaces were set as \u2018escape\u2019 for particles, while the reactor wall was set as \u2018reflective\u2019 boundary. The inlet velocity inlet was set from 0.0309-0.162 m\/s for AT reactor and from 0.0625-0.3271 m\/s in the CT reactor based on the calculated average velocities (?\u0305?). After solving the CFD code, the particle tracking data was exported into an Excel comma separated value (.csv) file. The parameters like x, y, z positions and components of velocity (\ud835\udc63\ud835\udc65, \ud835\udc63\ud835\udc66, \ud835\udc63\ud835\udc67) were obtained for each time points as shown in the figure 4.1.  Figure 4.1: Particle position at time \ud835\udc95\ud835\udfcf (point 1) and point 2 at time \ud835\udc95\ud835\udfd0   Figure 4.2: Particle x, y, z positions at time t in the reactor 89  4.3.3.2 Fluence calculation The equation 3.21 (chapter 3) which predicts the fluence at any position in space around the lamp was coupled with the particle position-time data for calculation of total delivered fluence (\ud835\udc39\ud835\udc5c,\ud835\udc36\ud835\udc39\ud835\udc37).  Total fluence (\ud835\udc39\ud835\udc5c,\ud835\udc36\ud835\udc39\ud835\udc37) delivered for each test liquid was calculated by integrating by summation (equation 4.1) the fluence at each position in the reactors adjusting for the lamp ON-OFF time based on 1, 3 and 5 Hz pulse frequencies.  \ud835\udc39 = \u2211 \ud835\udc3c\ud835\udc61\ud835\udc56(\ud835\udc65\ud835\udc56, \ud835\udc66\ud835\udc56, \ud835\udc67\ud835\udc56)\ud835\udc5b1 = \ud835\udc3c\ud835\udc611(\ud835\udc651, \ud835\udc661, \ud835\udc671)+\ud835\udc3c\ud835\udc612(\ud835\udc652, \ud835\udc662, \ud835\udc672)+ \ud835\udc3c\ud835\udc613(\ud835\udc653, \ud835\udc663, \ud835\udc673) +\u2026\u2026+\ud835\udc3c\ud835\udc61\ud835\udc5b(\ud835\udc65\ud835\udc5b, \ud835\udc66\ud835\udc5b, \ud835\udc67\ud835\udc5b) (4.1) Based on the values of coefficients \ud835\udc4f\ud835\udc65, \ud835\udc4f\ud835\udc66, \ud835\udc50\ud835\udc65, \ud835\udc50\ud835\udc66, \ud835\udc34, \ud835\udc35, and \ud835\udc36 as in Table 2.3, \ud835\udc39\ud835\udc5c,\ud835\udc36\ud835\udc39\ud835\udc37 was calculated for the test liquids (water, water+red dye, water+green dye, skim milk), velocities (0.0309-0.162 m\/s for AT reactor and from 0.0625-0.3271 m\/s for CT reactor), and pulse frequencies (1, 3 and 5 Hz). 4.3.4 Collimator tube experiment Collimator system is a device which when placed between a lamp source and target surface, produces parallel beam of light that is used for dose-response analysis in light-based technologies [23]. The experimental conditions and delivered light energy can be controlled to a great extent which are suitable to develop dose-response curve and calculate the susceptibilities of different microorganisms in terms of traditional D-value like thermal processes.  4.3.4.1 Experimental setup and treatments The experimental setup for the collimator system was fabricated by using a narrow tube (diameter, \ud835\udc51\ud835\udc50\ud835\udc61 = 5 cm, length, \ud835\udc59\ud835\udc50\ud835\udc61 = 5 cm), placed (centrally along the axis of PL lamp) under the lamp and masking the lamp footprint, to block all the light from elsewhere. The system allowed light rays to pass only through the tube opening on the target (microbial suspension in a Petri dish (diameter 5 cm) placed directly under the tube). The experimental setup is shown in figure 4.3. 90   Figure 4.3: Collimator tube experimental setup 4.3.4.2 UV dose estimation  For experimentation, 10 mL of test liquids (water, water+red dye, water+green dye, skim milk) with 0.1 mL of different microbial suspensions were taken in sterile petri dishes and flashed with PL for 0, 5, 15, 25, 35, 45, 55, 65 and 75 s at 3 Hz frequency. The liquids in petri dishes were constantly stirred by magnetic stirring to assist proper mixing. The energy\/pulse was measured to be 0-2.3625 J\/cm2 using a pyroelectric radiometer (PE80BF-DIF-C, Ophir-Spiricon LLC, UT, United states). Microbial enumeration was carried out following the protocol in section 3.3.5. Thereafter a dose-response curve (Fluence-inactivation curve) was developed for all liquids. The curves were analyzed by linear regression technique to ascertain the D-value, which is the dose required for reduction of microorganisms by one log.  4.3.5 Pulsed UV light treatment in flow-through reactors Each microbial suspensions were added at the level of 1 mL per 100 mL of sample. The sample containing the inoculum  was thoroughly swirled. The samples were kept at rest to acclimatize the microorganisms and then, they were subjected to PL treatments. PL treatment system consisted of a cylindrical xenon flashlamp emitting high-intensity light pulses in UV-Visible-IR region (~200-1100 nm), provided by Solaris Disinfection Inc. (Mississauga, ON, Canada). The lamp dimensions were: length, \ud835\udc3f = 62 cm; diameter, \ud835\udc51\ud835\udc3f = 9 mm. The lamp emitted 30 J of energy per pulse. Considering a pulse width of 100 ms (since light fluctuation due to pulsation was hardly noticeable beyond 9 pulses\/s), the lamp sleeves emitted energy intensity per pulse of 17122.12 mW\/cm2. Treatments were carried out in both the type of reactor configurations, viz. AT and CT reactors. Liquids\u2014Water, water+red dye, water+green dye and skim milk samples in volumes 100 mL and 300 mL for AT and CT reactor were transferred into a stainless steel container with lid. Liquids 91  were pumped through the reactors using a peristaltic pump (Masterflex L\/S model 7554-90, Cole-Parmer Instruments, IL, United States) with a variable speed drive. Flow rates of  14.32, 32.48, 49.63, 64.13 and 74.90 L\/h were used and when the liquid reached the inlet, the PL system was flashed (pulse frequency used were 1,3 and 5 Hz) until all the liquid was collected from outlet in a bottled covered with aluminum foil. After each run, the reactor was disinfected with 3% (v\/v) H2O2 and finally rinsed with autoclaved double distilled water. There was a total of 15 PL treatments and a control. The energy per pulse at distances 1.5 and 2 cm (for CT and AT rectors, respectively) from lamp were measured along the length of the lamp to calculate average energy per pulse, using a pyroelectric head sensor (PE80BF-DIF-C, Ophir-Spiricon LLC, UT, United states), with Nova II display (Ophir-Spiricon LLC, UT, United states) placed under the lamp. All the treatments were carried out in triplicates. After treatment, the liquid samples were kept refrigerated (4 oC) until analyses.  After the PL treatments, the PL-treated samples and control samples were serially diluted with autoclaved ultapure water. The dilutions were spread plated on nutrient agar (Sigma Aldrich, St. Louis, MO, United States) for E. coli ATCC 29055, BHI agar (BD Bacto, Sparks, MD, United States) for L. innocua ATCC 33090 and tryptic soy agar (BD Difco\u2122, Sparks, MD, United States) for C. sporogenes ATCC 7955. The plates were incubated at conditions specific for each bacteria: 37 oC for 24 h for E. coli and L. innocua; 37 oC for 48 h for C. sporogenes. 4.3.5.1 Inactivation kinetics modeling The microbial counts for all the microbial strains in tested liquid foods for both types of reactors were transformed into logarithmic values and fitted into kinetic models using GInaFiT add-in [170] (version 1.6) for Microsoft Excel (Microsoft Corporation, Redmond, WA, USA).  Log-linear inactivation model [171]: This model is a conventional model being used to fit first order reaction kinetics and is widely applicable in thermal processing. The model assumes that the cells have similar resistance to applied stress (PL treatment) and no. of survivors decrease over time exponentially. When expressed in log values and plotted against treatment fluence, a straight line is obtained with a negative slope. The model is given in equation 4.2. \ud835\udc3f\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc61 = \ud835\udc3f\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c \u2212\ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65\ud835\udc39\ud835\udc5c\ud835\udc59\ud835\udc5b10                                               (4.2) 92  Where, \ud835\udc41\ud835\udc5c, \ud835\udc41\ud835\udc61 = Initial viable count and number of survivors after PL treatment with specific fluence, \ud835\udc39\ud835\udc5c, [CFU\/mL]; \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 = Rate constant of microbial inactivation, [cm2\/J]. Weibull inactivation model [172]: The Weibull inactivation model is a non-linear kinetic model and assumes that the different fractions of microorganisms in the population show different resistance to PL treatment. Overall survival curve of the population is described by a cumulative exponential distribution (equation 4.3).  \ud835\udc3f\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc61 = \ud835\udc3f\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c \u2212 (\ud835\udc39\ud835\udc5c\ud835\udeff)\ud835\udc5d                                               (4.3) Where, \ud835\udc41\ud835\udc5c, \ud835\udc41\ud835\udc61 = Initial viable count and number of survivors after PL treatment with specific fluence, \ud835\udc39\ud835\udc5c, [CFU\/mL]; \ud835\udeff = Fluence required for first decimal reduction (\ud835\udc41\ud835\udc5c \u2192\ud835\udc41\ud835\udc5c10\u2044 ), [J\/cm2]; \ud835\udc5d = shape parameter (when \ud835\udc5d<1, curve is concave upward; when \ud835\udc5d>1, curve is concave downward). Log-linear with tail [173]: The Log-linear with tail model is composed of two parts; one, where the inactivation is linear and second, where the inactivation is very low, leaving residual microorganisms unaffected with further applied fluence. The model is given by equation 4.4. \ud835\udc3f\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc61 = \ud835\udc3f\ud835\udc5c\ud835\udc5410[(10\ud835\udc3f\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c \u2212 10\ud835\udc3f\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60) \u00d7 \ud835\udc52\u2212\ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65\ud835\udc39\ud835\udc5c + 10\ud835\udc3f\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60]            (4.4) Where, \ud835\udc41\ud835\udc5c, \ud835\udc41\ud835\udc61 = Initial viable count and number of survivors after PL treatment with specific fluence, \ud835\udc39\ud835\udc5c, [CFU\/mL]; \ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60 = residual microbial population, [CFU\/mL]; \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 = Rate constant of microbial inactivation, [cm2\/J]. The goodness of fit was tested by using coefficient of determination (R2) and root mean squared error (RMSE). 4.3.5.2 Reduction equivalent fluence (REF) estimation The dose-response curves obtained from collimated tube experiment  were transformed in such a way that x-axis was set as log reductions, \ud835\udc3f\ud835\udc5c\ud835\udc5410 (\ud835\udc41\ud835\udc5c\ud835\udc41\ud835\udc61\u2044 ) and y-axis was set as fluence (\ud835\udc39\ud835\udc5c). The data was fitted into a polynomial model (\ud835\udc66 = \ud835\udc4e\ud835\udc652 + \ud835\udc4f\ud835\udc65 + \ud835\udc50) as per equation 4.5 to calculate \ud835\udc39\ud835\udc5c,\ud835\udc45\ud835\udc38\ud835\udc39 [166].  93  \ud835\udc39\ud835\udc5c,\ud835\udc45\ud835\udc38\ud835\udc39 = \ud835\udc4e [\ud835\udc3f\ud835\udc5c\ud835\udc5410 (\ud835\udc41\ud835\udc5c\ud835\udc41\ud835\udc61\u2044 )]2+ \ud835\udc4f [\ud835\udc3f\ud835\udc5c\ud835\udc5410 (\ud835\udc41\ud835\udc5c\ud835\udc41\ud835\udc61\u2044 )] + \ud835\udc50                              (4.5) 4.4 Results and discussion 4.4.1 Fluence calculation The calculated total delivered fluence (\ud835\udc39\ud835\udc5c,\ud835\udc36\ud835\udc39\ud835\udc37) from the CFD analyses are shown in Table 4.1 and 4.2. The fluence values were greatly affected by the type of liquid, pulse frequency and flow rate. For AT reactor, the fluences values were highest for water and the lowest for skim milk. However, the reverse trend was observed in case of CT reactor. This could be due to slight overestimation of fluence values for the milk samples using equation 3.21. With the increase of flow rate, due to decreasing residence time, the fluence values decreased for all liquids. Also, the fluence values increased with the pulse frequency, as PL fluence is directly proportional to the pulse frequency. The calculated values of fluence for different liquids were significantly higher (p < 0.05) for CT reactor than the AT reactor. This is expected since the liquids have higher residence time, hence are exposed with more UV light in CT reactor. Table 4.1: Calculated total delivered fluence (\ud835\udc6d\ud835\udc90,\ud835\udc6a\ud835\udc6d\ud835\udc6b) [J\/cm2] values for Annular (AT) reactor  Flow rate [L\/h] Water Water+red dye Water+green dye Skim milk Pulse frequency Pulse frequency Pulse frequency Pulse frequency 1 3 5 1 3 5 1 3 5 1 3 5 14.32 1.091 3.272 5.453 0.9347 2.8041 4.6735 0.895 2.684 4.473 0.893 2.679 4.464 32.48 0.339 1.017 1.695 0.3684 1.1053 1.8422 0.361 1.082 1.804 0.338 1.014 1.690 49.63 0.345 1.036 1.726 0.3291 0.9873 1.6455 0.309 0.927 1.544 0.296 0.887 1.478 64.13 0.278 0.833 1.389 0.2521 0.7564 1.2607 0.238 0.714 1.190 0.228 0.685 1.142 74.90 0.194 0.581 0.969 0.1740 0.5219 0.8698 0.165 0.494 0.823 0.156 0.468 0.780  94  Table 4.2: Calculated total delivered fluence (\ud835\udc6d\ud835\udc90,\ud835\udc6a\ud835\udc6d\ud835\udc6b) [J\/cm2] values for coiled (CT) reactor Flow rate [L\/h] Water Water+red dye Water+green dye Skim milk Pulse frequency Pulse frequency Pulse frequency Pulse frequency 1 3 5 1 3 5 1 3 5 1 3 5 14.32 3.411 10.232 17.053 5.166 15.499 25.832 4.228 12.683 21.139 4.495 13.484 22.474 32.48 1.060 3.179 5.299 1.605 4.816 8.027 1.314 3.941 6.568 1.397 4.190 6.983 49.63 1.079 3.236 5.393 1.634 4.901 8.169 1.337 4.011 6.685 1.421 4.264 7.107 64.13 0.869 2.607 4.345 1.316 3.949 6.582 1.077 3.232 5.386 1.145 3.436 5.727 74.90 0.606 1.819 3.032 0.919 2.756 4.593 0.752 2.255 3.759 0.799 2.398 3.996  4.4.2 Microbial inactivation by PL The Figures 4.4 and 4.5 show the inactivation curves for E. coli, L. innocua and C. sporogenes for different liquids (water, water+red dye, water+green dye and skim milk) treated in AT and CT reactors, respectively. The calculated fluences (section 4.4.1) were used to express the results for the inactivation curves. For all the liquids treated, the inactivation for each bacterial strain were significantly different (p < 0.05) from each other. Also, the inactivation was significantly different (p < 0.05) for a specific microbial strain, for all the treated liquids. Regardless of the treated liquid, the inactivation of E. coli was significantly higher than the other while C. sporogenes showed the least inactivation. For AT reactor, the maximum log reduction of E. coli were 5.736 \u00b1 0.000, 5.324 \u00b1 0.102, 5.136 \u00b1 0.072, 1.088 \u00b1 0.049 for water, water+red dye, water+green dye and skim milk, respectively. The values for C. sporogenes were 4.846 \u00b1 0.039, 4.709 \u00b1 0.026, 4.54 \u00b1 0.039, 0.818 \u00b1 0.016 for water, water+red dye, water+green dye and skim milk, respectively. Similar observations were made in the case of  CT reactor, where the E. coli was least resistant to PL and C. sporogenes was the most. The greater resistance of spore forming bacteria could be to the presence of components in the cell wall which were absent in the vegetative cell leading to lower susceptibility. For CT reactor, the maximum log reduction of E. coli were 7.816 \u00b1 0.174, 7.672 \u00b1 0.102, 7.576 \u00b1 0.199, 4.093 \u00b1 0.034 for water, water+red dye, water+green dye and skim milk, 95  respectively. The values for C. sporogenes were 7.453 \u00b1 0.088, 7.457 \u00b1 0.034, 7.034 \u00b1 0.053, 3.607 \u00b1 0.069 for water, water+red dye, water+green dye and skim milk, respectively.   (a)  (b) -6-5-4-3-2-100 1 2 3 4 5 6Log [N\/No]Fluence [J\/cm2]E. coli L. innocua C. sporogenes-6-5-4-3-2-100 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5Log [N\/No]Fluence [J\/cm2]E. coli L. innocua C. sporogenes96   (c)  (d) Figure 4.4: Inactivation curves of microorganisms in different liquids in annular (AT) reactor (a) Water, (b) water+red dye, (c) water+green dye, and (d) skim milk -6-5-4-3-2-100 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5Log [N\/No]Fluence [J\/cm2]E. coli L. innocua C. sporogenes-6-5-4-3-2-100 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5Log (N\/No)Fluence [J\/cm2]E. coli L. innocua C. sporogenes97   (a)  (b) -8-7-6-5-4-3-2-100 2 4 6 8 10 12 14 16 18Log [N\/No]Fluence [J\/cm2]E. coli L. innocua C. sporogenes-8-7-6-5-4-3-2-100 5 10 15 20 25Log [N\/No]Fluence [J\/cm2]E. coli L. innocua C. sporogenes98   (c)  (d) Figure 4.5: Inactivation curves of microorganisms in different liquids in coiled (CT) reactor (a) Water, (b) water+red dye, (c) water+green dye, and (d) skim milk The data from the inactivation of the microbial strains were transformed into survivor curve to model the inactivation based on conventional log-linear, Weibull and log-linear with tail model. For all the microbial strains, regardless of the type of reactor and treatment liquid, the survivor -8-7-6-5-4-3-2-100 5 10 15 20Log [N\/No]Fluence [J\/cm2]E. coli L. innocua C. sporogenes-8-7-6-5-4-3-2-100 5 10 15 20Log [N\/No]Fluence [J\/cm2]E. coli L. innocua C. sporogenes99  curve showed a steady decrease in bacterial load, followed by stabilization, where further inactivation was minimal. The survivor curves were fitted best by the log-linear with tail model, and the goodness of fit described by lowest RMSE error (0.048-0.596) and highest R2 values (0.925-0.988). The survivor curves fitted to log-linear with tail model are shown in Figures 4.6 and 4.7. The data of constants for different fitted models along with their RMSE error and R2 values are shown in Table 4.3 and 4.4.      (a)                                                                      (b)     (c)                                                                      (d) Figure 4.6: Fitting of survivors curves of microorganisms into log-linear with tail model in annular (AT) reactor for (a) water, (b) water+red dye, (c) water+green dye, and (d) skim milk    024680 1 2 3 4 5 6Log10NtFluence [J\/cm2]E. coli L. innocuaC. sporogenes C. sporogenes_L-L-TE. coli_L-L-T L. innocua_L-L-T024680 1 2 3 4 5Log10NtFluence [J\/cm2]E. coli L. innocuaC. sporogenes C. sporogenes_L-L-TE. coli_L-L-T L. innocua_L-L-T024680 1 2 3 4 5Log10NtFluence [J\/cm2]E. coli L. innocuaC. sporogenes C. sporogenes_L-L-TE. coli_L-L-T L. innocua_L-L-T00.511.522.533.544.555.566.577.580 1 2 3 4 5Log10NtFluence [J\/cm2]E. coli L. innocuaC. sporogenes C. sporogenes_L-L-TE. coli_L-L-T L. innocua_L-L-T100      (a)                                                                      (b)     (c)                                                                      (d) Figure 4.7: Fitting of survivors curves of microorganisms into log-linear with tail model in coiled (CT) reactor for (a) water, (b) water+red dye, (c) water+green dye, and (d) skim milk  024680 5 10 15 20Log10NtFluence [J\/cm2]E. coli L. innocuaC. sporogenes C. sporogenes_L-L-TE. coli_L-L-T L. innocua_L-L-T024680 5 10 15 20 25 30Log10NtFluence [J\/cm2]E. coli L. innocuaC. sporogenes C. sporogenes_L-L-TE. coli_L-L-T L. innocua_L-L-T024680 5 10 15 20 25Log10NtFluence [J\/cm2]E. coli L. innocuaC. sporogenes C. sporogenes_L-L-TE. coli_L-L-T L. innocua_L-L-T024680 5 10 15 20 25Log10NtFluence [J\/cm2]E. coli L. innocuaC. sporogenes C. sporogenes_L-L-TE. coli_L-L-T L. innocua_L-L-T101  Table 4.3: Values of constant obtained by fitting the survivor curves for water, water+red dye, water+green dye, and skim milk in AT reactor Product Parameters Model Log-linear R2 RMSE Weibull R2 RMSE Log-linear plus tail R2 RMSE E. coli Water \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 2.405 0.620 1.087 _ 0.861 0.658 7.647 0.987 0.198 \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 6.001 8.302 7.643 \ud835\udeff [J\/cm2] _ 0.034 _ \ud835\udc5d [-] _ 0.393 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 2.401  Water+red dye \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 2.477 0.619 0.967 _ 0.882 0.539 8.992 0.970 0.272 \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 5.911 8.129 7.573 \ud835\udeff [J\/cm2] _ 0.029 _ \ud835\udc5d [-] _ 0.370 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 3.075 Water+green dye \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 2.457 0.601 0. 953 _ 0.878 0.527 9.540 0.976 0.234 \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 5.863 8.073 7.548 102  \ud835\udeff [J\/cm2] _ 0.026 _ \ud835\udc5d [-] _ 0.358 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 3.205 Skim milk \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.51 0.838 0.106 _ 0.980 0.037 1.246 0.958 0.054 \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 7.27 7.56 7.408 \ud835\udeff [J\/cm2] _ 3.35 _ \ud835\udc5d [-] _ 0.46 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 6.488 L. innocua Water \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 2.369 0.707  0.881  _ 0.899  0.517  6.014 0.994  0.125  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 6.401 8.244 7.607 \ud835\udeff [J\/cm2] _ 0.069 _ \ud835\udc5d [-] _ 0.440 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 2.470 Water+red dye \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 2.552 0.656 0.919 _ 0.869 0.566 7.628 0.985 0.189 \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 6.319 8.205 7.736 \ud835\udeff [J\/cm2] _ 0.054 _ Table 4.3: Continued 103  \ud835\udc5d [-] _ 0.418 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 3.119 Water+green dye \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 2.601 0.656 0.897 _ 0.873 0.545 7.739 0.981 0.210 \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 6.327 8.194 7.695 \ud835\udeff [J\/cm2] _ 0.052 _ \ud835\udc5d [-] _ 0.414 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _  _ 3.193 Skim milk \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.570 0.746 0.158 _ 0.890 0.104 1.782 0.977 0.048 \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 7.477 7.767 7.716 \ud835\udeff [J\/cm2] _ 2.723 _ \ud835\udc5d [-] _ 0.496 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 6.676 C. sporogenes Water \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 2.201 0.741  0.751  _ 0.904  0.457  5.286 0.995  0.100  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 6.676 8.178 7.701 \ud835\udeff [J\/cm2] _ 0.116 _ Table 4.3: Continued 104  \ud835\udc5d [-] _ 0.472 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 2.963 Water+red dye \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 2.455 0.692 0.813 _ 0.888 0.492 6.580 0.986 0.174 \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 6.485 8.166 7.672 \ud835\udeff [J\/cm2] _ 0.074 _ \ud835\udc5d [-] _ 0.435 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 3.303 Water+green dye \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 2.539 0.689 0.813 _ 0.888 0.488 6.920 0.988 0.158 \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 6.513 8.179 7.708 \ud835\udeff [J\/cm2] _ 0.073 _ \ud835\udc5d [-] _ 0.436 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-]  _ 3.368 Skim milk \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.481 0.702 0.149 _ 0.868 0.099 1.790 0.978 0.403 \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 7.747 7.747 7.712 \ud835\udeff [J\/cm2] _ 3.635 _ \ud835\udc5d [-] _ 0.474 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 6.813  Table 4.3: Continued 105  Table 4.4: Values of constant obtained by fitting the survivor curves for water, water+red dye, water+green dye, and skim milk in CT reactor Product Parameters Model Log-linear R2 RMSE Weibull R2 RMSE Log-linear plus tail R2 RMSE E. coli Water \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.913 0.464  1.774  _ 0.810  1.055  4.440 0.977  0.371  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 4.580 8.608 7.665 \ud835\udeff [J\/cm2] _ 0.010 _ \ud835\udc5d [-] _ 0.312 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 0.762 Water+red dye \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.592 0.475  1.700  _ 0.829  0.972  2.706 0.971  0.397  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 4.647 8.617 7.514 \ud835\udeff [J\/cm2] _ 0.016 _ \ud835\udc5d [-] _ 0.309 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 0.811 Water+green dye \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.701 0.501  1.567  _ 0.846  0.871  3.198 0.956  0.463  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 4.800 8.538 7.493 \ud835\udeff [J\/cm2] _ 0.016 _ 106  \ud835\udc5d [-] _ 0.312 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 1.229 Skim milk \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.452     _     1.232 0.965  0.250  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 6.720 8.172 7.763 \ud835\udeff [J\/cm2] _ 0.499 _ \ud835\udc5d [-] _ 0.442 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 3.703 L. innocua Water \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 1.003 0.536  1.684  _ 0.836  1.001  3.904 0.985  0.299  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 5.148 8.828 7.867 \ud835\udeff [J\/cm2] _ 0.022 _ \ud835\udc5d [-] _ 0.351 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 0.690 Water+red dye \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.625 0.543  1.569  _ 0.848  0.904  2.504 0.963  0.449  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 5.184 8.743 7.770 \ud835\udeff [J\/cm2] _ 0.033 _ \ud835\udc5d [-] _ 0.342 _ Table 4.4: Continued 107  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 1.140 Water+green dye \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.742 0.537  1.541  _ 0.848  0.884  2.867 0.968  0.403  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 5.218 8.708 7.675 \ud835\udeff [J\/cm2] _ 0.029 _ \ud835\udc5d [-] _ 0.341 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 1.159 Skim milk \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.446     _     1.126 0.925  0.359  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 6.890 8.081 7.806 \ud835\udeff [J\/cm2] _ 0.787 _ \ud835\udc5d [-] _ 0.486 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 3.841 C. sporogenes Water \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 1.025 0.583  1.566  _ 0.840  0.969  3.593 0.983  0.318  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 5.756 8.976 8.210 \ud835\udeff [J\/cm2] _ 0.044 _ \ud835\udc5d [-] _ 0.386 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 1.146 Table 4.4: Continued 108  Water+red dye \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.670 0.589  1.531  _ 0.846  0.937  2.182 0.983  0.313  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 5.723 8.915 7.986 \ud835\udeff [J\/cm2] _ 0.067 _ \ud835\udc5d [-] _ 0.385 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 0.959 Water+green dye \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.759 0.569  1.478  _ 0.839  0.905  2.684 0.984  0.287  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 5.765 8.859 8.058 \ud835\udeff [J\/cm2] _ 0.056 _ \ud835\udc5d [-] _ 0.375 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 1.480 Skim milk \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.422     _     0.989 0.924  0.343  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 7.154 8.115 7.926 \ud835\udeff [J\/cm2] _ 1.184 _ \ud835\udc5d [-] _ 0.525 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 4.173  Table 4.4: Continued 109  4.4.3 Collimated tube experiment 4.4.3.1 UV dose estimation The UV dose was determined using biodosimetry method where, liquid products being tested were inoculated with different microorganisms and treated at different fluences to develop the dose-response curves. Figure 4.8 (a), (b), (c), and (d) show the dose-response curves for the microorganisms in water, water+red dye, water+green dye and skim milk. The calculated UV dose (in terms of D-value \u2013 the energy required for inactivation of 1 log of microbial population) (Table 4.5) were highest for skim milk and lowest for water, with the values for coloured liquids in-between water and milk. The higher the D-value, the more the time required for the microbial population to be inactivated. In skim milk, the D-value of the bacteria is highest compared to other liquids, since light is attenuated and scattering owing to the high absorption and scattering coefficients for milk. The results obtained hereby confirm the observations made previously by CFD in chapter 4, where water showed the highest irradiance values and milk showed the least irradiance values. Additionally, regardless of the liquids, the D-values increased from E. coli < L. innocua < C. sporgenes.      (a)                                                                                                    (b)  y = -4.2103x + 0.0142R\u00b2 = 0.9829y = -3.0588x - 0.3653R\u00b2 = 0.9694y = -2.7823x - 0.1144R\u00b2 = 0.9854-8-7-6-5-4-3-2-100 0.5 1 1.5Log N\/NoFluence [J\/cm2]WaterE. coli_W L. innocua_W C. sporogenes_Wy = -5.0286x - 0.1513R\u00b2 = 0.9486y = -3.007x - 0.2299R\u00b2 = 0.9335y = -2.7778x - 0.0944R\u00b2 = 0.9763-8-7-6-5-4-3-2-100 0.5 1 1.5Log N\/NoFluence [J\/cm2]Water +  Red dyeE. coli_W+R L. innocua_W+R C. sporogenes_W+R110       (c)                                                                                                    (d) Figure 4.8: Dose-response curves for different microorganisms in water, water+red dye, water+green dye, and skim milk  Table 4.5: Calculated D-values of microbial strains in water, water+red dye, water+green dye, and skim milk  Liquid product D-values [J\/cm2] E. coli L. innocua C. sporogenes Water 0.177 0.208 0.241 Water+red dye 0.199 0.256 0.326 Water+green dye 0.197 0.258 0.313 Skim milk 4.226 6.437 7.661  4.4.3.2 Reduction equivalent fluence (REF)  The REF values were calculated from the obtained regression equation (equation 4.5) based on the biodosimetry method for the maximum inactivation achieved under different conditions). The calculated values of REF and the corresponding fluence values from CFD are shown in Table 4.6. It is apparent that the fluence values were higher for the CT reactor than the AT reactor, owing to higher inactivation level obtained in case if the former. The REF values were highest in case of skim milk, which means that the required fluence to inactivate same number of microbial cells\/population were higher in skim milk compared to the other liquids.  y = -5.0697x - 0.0254R\u00b2 = 0.9162 y = -2.996x - 0.2285R\u00b2 = 0.9373y = -2.5677x - 0.1974R\u00b2 = 0.9717-8-7-6-5-4-3-2-100 0.5 1 1.5Log N\/NoFluence [J\/cm2]Water + Green dyeE. coli_W+G L. innocua_W+G C. sporogenes_W+Gy = -0.2288x - 0.0332R\u00b2 = 0.9684y = -0.1499x - 0.0351R\u00b2 = 0.9811y = -0.1259x - 0.0355R\u00b2 = 0.9755-0.8-0.7-0.6-0.5-0.4-0.3-0.2-0.100 0.5 1 1.5 2 2.5Log N\/NoFluence [J\/cm2]MilkE. coli_M L. innocua_M C. sporogenes_M111  Table 4.6: Reduction equivalent fluence (REF) values for different treatment conditions Liquid product E. coli L. innocua C. sporogenes Calculated fluence [J\/cm2] REF [J\/cm2] Calculated fluence [J\/cm2] REF [J\/cm2] Calculated fluence [J\/cm2] REF [J\/cm2] AT CT AT CT AT CT AT CT AT CT AT CT Water 5.453 17.05 1.067 1.133 5.453 17.05 1.846 3.303 5.453 17.05 1.629 2.329 Water+red dye 4.674 25.83 1.642 3.512 4.674 25.83 2.012 4.921 4.674 25.83 1.966 4.138 Water+green dye 4.473 21.14 1.220 2.379 4.473 21.14 2.168 4.291 4.473 21.14 1.728 4.016 Skim milk 4.464 22.47 7.336 71.02 4.464 22.47 10.250 119.5 4.464 22.47 9.379 121.8  4.5 Conclusion This chapter deals with calculation of fluence and UV dose of PL process. The delivered fluence in the model liquid foods for different processing parameters were calculated by coupling CFD particle tracking results with lamp energy prediction model. Inactivation of microorganisms inoculated in liquid foods followed the order E. coli > L. innocua > C. sporogenes. This highlights the resistance of spore forming Clostridium bacteria to PL compared to the vegetative bacteria strains. Also, the inactivation was dependent on the sample optical properties with water showing the highest inactivation and skim milk the least. Log-linear plus tail model fitted the inactivation data of microorganisms best, which highlights the fact that during PL treatment, fluences up to a certain value are useful for carrying out inactivation after which the inactivation rate slows, which may only contribute to product deterioration. Further fluence application only tends to affect the quality parameters of food. Biodosimetry using collimator tube experiment revealed that D-value of the microorganisms depended on the liquid optical properties. These results highlight importance of sample optical properties during PL treatment of liquid foods.    112  Chapter 5: Inactivation of microorganisms in milk, grape, and watermelon juices 5.1 Summary PL process was validated after inoculation of challenge microorganisms in model liquid foods. While the model shows that the inactivation depends on the optical properties and clarity of the liquid samples, it is important to extend the method to real liquid foods of different optical properties, composition and thermophysical properties. Based on this criteria, the chosen liquid foods were milk (0, 1, 2, 3.25% fat), red grape juice and watermelon juice. The liquid products were inoculated with the challenge microorganisms (E. coli, L. innocua, and C. sporogenes) and subjected to PL treatment as mentioned in previous sections. The inactivation of microorganisms followed the order: watermelon juice > red grape juice > milk, in both AT and CT reactors. The obtained log reduction in AT reactor was >5 logs for both the juice types while for milk up to 1 log reduction as achieved. In case of CT reactor, >7 logs reduction was achieved in juices, and reduction of >4 logs were achieved for milk. Fat% significantly (p < 0.05) affected the microbial inactivation due to its scattering properties\u2014skim milk showed the greatest inactivation while 3.25% milk showed the least inactivation. The inactivation kinetics were also modeled using three different kinetic models and log-linear with tail model fitted the inactivation data best. Inactivation of the test microorganisms in milk was compared against HTST treatment of milk, in which case >5 logs inactivation was achieved.  5.2 Introduction Liquid foods like milk and fruit juices are processed thermally using techniques like HTST and UHT processing. In these techniques the products are heated up to the desired heating temperature for a specific time (72 oC\/15 s or 145-150 oC\/fraction of a second) to carry out inactivation of microorganisms or pathogens related to the public health concern. However, due to thermal processing, undesirable sensory changes like cooked flavour development and nutritional losses take place in milk. Therefore, emerging technologies are being explored for pasteurization of these products with minimal sensory and nutritional changes. UV-C and PL have been used for processing of liquid foods like milk, and opaque and clear fruit juices previously. 113  As mentioned earlier, treatment of dairy matrices, especially fluid milk products like milk, whey etc. have been subjected to processing by PL earlier. Smith et al. [72] treated bovine milk using PL at a fluence of 25.1 J\/cm2 to obtain >2 log reduction of Serratia marcescens. Krishnamurthy et al. [51] passed milk through a quartz tube placed on a V-groove reflector with adjustable height at flow rates 20 mL\/min and energy \/pulse of 1.28 J\/cm2 to obtain 7.23 log reduction of S. aureus. Palgan et al. [64] subjected milk to 28 J\/cm2 fluence and observed reduction of 1.06 log of E. coli and 0.51\u20130.84 of L. innocua. Miller et al. [77] treated milk (9.8% total solids) and  concentrated milk (45% total solids) at 8.4 J\/cm2 to obtain 2.5 log and < 1 reduction of E. coli ATCC 25922. They also obtained log reduction of 3.4 for skim milk, > 2.5 for 2% fat and whole milk after 14.9 J\/cm2 fluence application. It is evident that the treatment of milk by PL is affected by the solids content in milk. Higher solids in milk lead to lower inactivation levels which is due to greater absorption and scattering of light by milk solids. Overall, it seems a potent technology for processing of milk. However, comparing of these data is a tedious task since the way fluence measurement was carried out varies. Moreover, there exists a lacking in use of continuous PL treatment systems of milk processing.  PL is also an excellent technology for decontamination of fruit juices. The clearer fruit juices with lower total soluble solids tend to be treated excellently using PL. Not much literature exists for treatment of watermelon and grape juice processing using PL. Xu et al. [174] processed red grape juice in a spiral tube using PL at energy per pulse of 0.66 J\/cm2, flow rate of 40 mL\/min and 80 pulses (total fluence 52.8 J\/cm2) to obtain 4.89 log reduction of E. coli. Wang et al. [128] treated grape juice by passing through a spiral tube at an energy per pulse of 0.66 J\/cm2 and total fluence up to 66 J\/cm2, flow rate of 60 mL\/min to obtain up to 2.6 log reduction of E. coli. This shows that the results varied even for the same microorganisms and food product. Also, the flow rates used in these studies were very less compared to the flow regimes used in our studies. It would be interesting to compare these results by carrying out juice processing in continuous PL processing systems.  Therefore, the main objective of this experiment to treat milk of various fat%, red grape juice and watermelon juice in continuous PL reactors (AT and CT) at various flow rates and frequencies to 114  ascertain the level of inactivation of challenge microorganisms in these products. Also, the inactivation kinetics will be modeled based on different inactivation kinetic models.   5.3 Materials and methods 5.3.1 Microbial strains E. coli K-12 (ATC 29055), L. innocua ATCC 33090, and C. sporogenes ATCC 7955 were used as challenge microorganisms for microbial inactivation studies in milk, red grape and watermelon juices using PL. The growth and preparation of strains are mentioned below. All the microbial strains were procured from  American Type Culture Collection (ATCC, Manassas, VA, United States) as lyophilized pellets. E. coli ATCC 29055: E. coli ATCC 29055 is a vegetative bacteria which has been used in PL studies as a non-pathogenic surrogate for E. coli O157:H7 in previous PL processing studies [167,168]. The lyophilized pellet was aseptically suspended in 15 mL aliquot of autoclaved nutrient broth (Sigma Aldrich, St. Louis, MO, United States), which was then mixed back to the nutrient broth bottle. The culture was grown for 24 h at 37 oC in an incubator. The broth was then centrifuged at 5000g for 15 min at 4 oC and supernatant was discarded. The pellet was resuspended in autoclaved double distilled water. The previous washing and centrifugation steps were repeated twice before final resuspension of recovered cell pellet into autoclaved double distilled water. The final stock solution contained approximately 109 CFU\/mL. This solution was used as inoculum for PL treatments. The stock culture was stored at -80 oC added with 30% (v\/v) glycerol solution for long-term storage.  L. innocua ATCC 33090: L. innocua 33090 is a vegetative bacteria, which was used as a surrogate for L. monocytogenes, by virtue of their phenotypic similarity [82]. L. monocytogenes has been observed to be one of the most resistant bacteria to PL and L. innocua is proposed to be applicable for validation studies [92,96]. The lyophilized pellet was aseptically suspended in 15 mL aliquot of autoclaved brain heart infusion (BHI) broth (BD Bacto, Sparks, MD, United States), which was then mixed back to the BHI broth bottle. The medium was incubated at 37 oC for 24 h in an incubator. The cells were recovered from the broth by centrifugation at 4000g for 15 min at 4 oC, and by discarding the supernatant. The pellet was resuspended in autoclaved double distilled water aseptically. The previous washing and centrifugation steps were repeated twice before final 115  resuspension of recovered cell pellet into autoclaved double distilled water. The bacterial suspension contained around 109 CFU\/mL which was used as inoculum for PL treatments. The stock culture was stored at -80 oC added with 30% (v\/v) glycerol solution for long-term storage. C. sporogenes ATCC 7955: C. sporogenes ATCC 7955 is an anaerobic, mesophilic spore forming bacteria. The strain has been used for thermal processing studies as a surrogate bacteria for C. botulinum [168]. The lyophilized pellet was aseptically suspended in 15 mL aliquot of autoclaved reinforced Clostridial medium (RCM) (Thermo Scientific, United Kingdom), which was then mixed back to the RCM bottle. The medium was incubated at 37 oC for 5 days under anaerobic conditions using an anaerobic jar. The cells were recovered from the RCM by centrifugation at 4000g for 15 min at 4 oC, and by discarding the supernatant. The pellet was resuspended in autoclaved double distilled water aseptically. The previous washing and centrifugation steps were repeated twice before final resuspension of recovered cell pellet into autoclaved double distilled water. The spore bacterial suspension contained around 108-109 CFU\/mL which was used as inoculum for PL treatments. The stock culture was stored at -80 oC added with 30% (v\/v) glycerol solution for long-term storage. 5.3.2 Liquid food samples The liquid foods considered for treatment with PL were selected based on their varying pH, thermophysical and optical properties. The liquid foods chosen were: Pasteurized Milk (0%, 1%, 2% and 3.25% milk fat; from Dairyland, Saputo Dairy products, BC, Canada), clear red grape (Welch Foods Inc., Concord, MA, United States) and watermelon juice (Simply Orange Juice Company, Apopka, FL, United States), which were purchased from local grocery stores. The milk samples contained 3.2% (w\/v) protein and 4.8% (w\/v) carbohydrate, (mainly lactose). The grape and watermelon juices contained 16.4% (w\/v) and 12% (w\/v) sugars, respectively. Autoclaved ultrapure water (18 M\u03a9) was used for dilution and making reagents for analyses. Physico-chemical properties were evaluated for milk and fruit juices. Total solids (TS %) was obtained for milk by gravimetric method [142]. Degree Brix (oBx) was measured for the grape and watermelon juice using a handheld refractometer (Sper Scientific, Scottsdale, AZ, United States). Colour parameters (L*, a* and b*) were obtained by using colorimeter (HunterLab, model LabScanTM XE Plus, Hunter Associates Laboratory, Reston, VA, USA). Density (\ud835\udf0c) of all liquids were measured by method of weight of known volume. Viscosity (\ud835\udf07) was measured for the liquids using a rheometer 116  (Model MCR 302, Anton Paar GmbH, Graz, Austria) under 0.1-100 \/s shear rate. Optical properties (Napierian absorption coefficient, \ud835\udefc\ud835\udc52,254, [\/m], UV transmission %, \ud835\udc47\ud835\udc48\ud835\udc49, [%], and penetration depth, \ud835\udeff37%, [m]) were calculated from the absorbance value at 254 nm (\ud835\udc34254) (measured on a cuvette of 1 cm path length with UV-Vis spectrophotometer (Shimadzu 1800, Tokyo, Japan)) using the Eq. 3.1-3.4 (chapter 3). All the measured properties are given in the Table 5.1.          117  Table 5.1: Calculated thermophysical and optical properties of milk (various fat%), red grape, and watermelon juice \ud835\udf0c = density;  \u03bc = viscosity; \ud835\udefc\ud835\udc52,254 = Napierian absorption coefficient; \u03b437% = penetration depth aTS% was measured for milk and oBx was measured for juices n = 3; samples expressed as mean \u00b1 standard deviation; different subscripts in each row for the same column shows significant differences (p < 0.05) between the samplesProduct \ud835\udf36\ud835\udc86,\ud835\udfd0\ud835\udfd3\ud835\udfd2 [\/m] \ud835\udc7b\ud835\udc7c\ud835\udc7d [%] \u03b437% [mm] \u03c1 [kg\/m3] \u03bc [Pa-s] L* a* b* TS % (or oBx)a Water 6.467 \u00b1 2.325e 93.76 \u00b1 2.167a 16.61 \u00b1 5.209a 1002.7 \u00b1 3.45g 0.869 \u00b1 0.046c _ _ _ _ Milk (0%) 244.3 \u00b1 12.27b 8.737 \u00b1 1.082d 4.080 \u00b1 0.21c 1023 \u00b1 5.774c 2.037 \u00b1 0.593b 72.04 \u00b1 0.04d -5.68 \u00b1 0.01f -1.31 \u00b1 0.01e 9.31 \u00b1 0.11d Milk (1%) 248.9 \u00b1 6.700b  8.310 \u00b1 0.548d 3.996 \u00b1 0.106c 1013 \u00b1 0.025d 1.947 \u00b1 0.533b 74.73 \u00b1 0.70c -3.61 \u00b1 0.05e 3.65 \u00b1 0.24d 10.95 \u00b1 0.00c Milk (2%) 261.4 \u00b1 3.138a 7.327 \u00b1 0.231e 3.804 \u00b1 0.046d 1009 \u00b1 0.013e 1.805 \u00b1 0.171b 78.83 \u00b1 0.72b -2.90 \u00b1 0.03d 5.42 \u00b1 0.09c 13.41 \u00b1 3.25b Milk (3.25%) 280.4 \u00b1 29.86a 6.220 \u00b1 1.680e 3.570 \u00b1 0.36d 1010 \u00b1 0.003f 2.467 \u00b1 0.79a 81.40 \u00b1 0.37a -2.30 \u00b1 0.02c 6.07 \u00b1 0.14b 12.11 \u00b1 0.00b Red grape juice 156.5 \u00b1 15.61c 21.09 \u00b1 3.430c 6.400 \u00b1 0.678b 1072 \u00b1 2.646a 2.937 \u00b1 0.00a 10.16 \u00b1 0.36f 20.1 \u00b1 0.62a 13.1 \u00b1 1.15a 16.00 \u00b1 0.20a Watermelon juice 110.6 \u00b1 7.43d 33.14 \u00b1 2.480b 9.015 \u00b1 0.613b 1049 \u00b1 2.517b 1.830 \u00b1 0.00b 42.20 \u00b1 0.04e 4.18 \u00b1 0.02b 3.89 \u00b1 0.02d 8.30 \u00b1 0.10e 118   5.3.3 Inoculation and pulsed UV light treatment Each microbial suspensions were added at the level of 1 mL per 100 mL of sample. The sample was containing the inoculum  was thoroughly swirled. The samples were kept at rest to acclimatize the microorganisms and then, they were subjected to PL treatments. PL treatment system consisted of a cylindrical xenon flashlamp emitting high-intensity light pulses in UV-Visible-IR region (~200-1100 nm), provided by Solaris Disinfection Inc. (Mississauga, ON, Canada). The lamp dimensions were: length, \ud835\udc3f = 62 cm; diameter, \ud835\udc51\ud835\udc3f = 9 mm. The lamp emitted 30 J of energy per pulse. Considering a pulse width of 100 ms (since light fluctuation due to pulsation was hardly noticeable beyond 9 pulses\/s), the lamp sleeves emitted energy intensity per pulse of 17122.12 mW\/cm2. Treatments were carried out in both the type of reactor configurations, viz. AT and CT reactors. Milk and juice samples in volumes 100 mL and 300 mL for AT and CT reactor were transferred into a stainless steel container with lid. Liquids were pumped through the reactors using a peristaltic pump (Masterflex L\/S model 7554-90, Cole-Parmer Instruments, IL, United States) with a variable speed drive. Flow rates of  14.32, 32.48, 49.63, 64.13 and 74.90 L\/h were used and when the liquid reached the inlet, the PL system was flashed (pulse frequency used were 1,3 and 5 Hz) until all the liquid was collected from outlet in a bottled lined with aluminum foil. After each run, the reactor was disinfected with 3% (v\/v) H2O2 and finally rinsed with autoclaved double distilled water. There was a total of 15 PL treatments (PL1-PL15) and control samples for all the milk fat%, fruit juices tested. The energy per pulse at distances 1.5 and 2 cm (for CT and AT rectors, respectively) from lamp were measured along the length of the lamp to calculate average energy per pulse, using a pyroelectric head sensor (PE80BF-DIF-C, Ophir-Spiricon LLC, UT, United states), with Nova II display (Ophir-Spiricon LLC, UT, United states) placed under the lamp. The average energies were 53 and 43 mJ\/cm2, at 1.5 and 2 cm, respectively. All the treatments were carried out in triplicates. After treatment, the liquid food samples were kept refrigerated (4 oC) until analyses. The experimental design is shown in the Table 5.2. High temperature short time (HTST) pasteurization (in water bath, maintained at 80 oC, milk heated 72 \u00b1 0.5 oC, and held for 15 s) of milk samples were also carried out to ascertain the effect of heating on microbial inactivation. 119  Table 5.2: Experimental design of the PL treatment of liquid foods Pulse frequency [Hz] Flow rate [L\/h] Treatment  Fluence [J\/cm2]a Reynolds number (\ud835\udc75\ud835\udc79\ud835\udc86) 0% 1% 2% 3.25% Re grape juice Watermelon juice AT CT AT CT AT CT AT CT AT CT AT CT AT CT 1 14.3 PL1 0.703 4.399 30.97 282.16 32.08 293.45 34.43 313.66 25.25 230.07 22.5 205.2 35.4 322.5 32.5 PL2 0.309 1.573 70.25 640.02 72.77 665.64 78.09 711.48 57.28 521.89 51.1 465.4 80.3 731.6 49.6 PL3 0.203 1.029 107.35 978.12 111.21 1017.26 119.34 1087.32 87.54 797.57 78.1 711.2 122.7 1118.1 64.1 PL4 0.157 0.797 138.71 1263.78 143.69 1314.35 154.19 1404.87 113.10 1030.50 100.9 918.9 158.6 1444.6 74.9 PL5 0.135 0.683 162.01 1476.11 167.83 1535.18 180.10 1640.91 132.17 1203.64 117.8 1073.3 185.2 1687.3 3 14.3 PL6 2.109 10.708 30.97 282.16 32.08 293.45 34.43 313.66 25.25 230.07 22.5 205.2 35.4 322.5 32.5 PL7 0.929 4.719 70.25 640.02 72.77 665.64 78.09 711.48 57.28 521.89 51.1 465.4 80.3 731.6 49.6 PL8 0.609 3.089 107.35 978.12 111.21 1017.26 119.34 1087.32 87.54 797.57 78.1 711.2 122.7 1118.1 64.1 PL9 0.471 2.390 138.71 1263.78 143.69 1314.35 154.19 1404.87 113.10 1030.50 100.9 918.9 158.6 1444.6 74.9 PL10 0.403 2.046 162.01 1476.11 167.83 1535.18 180.10 1640.91 132.17 1203.64 117.8 1073.3 185.2 1687.3 5 14.3 PL11 3.515 17.846 30.97 282.16 32.08 293.45 34.43 313.66 25.25 230.07 22.5 205.2 35.4 322.5 32.5 PL12 1.549 7.866 70.25 640.02 72.77 665.64 78.09 711.48 57.28 521.89 51.1 465.4 80.3 731.6 120  49.6 PL13 1.014 5.149 107.35 978.12 111.21 1017.26 119.34 1087.32 87.54 797.57 78.1 711.2 122.7 1118.1 64.1 PL14 0.785 3.984 138.71 1263.78 143.69 1314.35 154.19 1404.87 113.10 1030.50 100.9 918.9 158.6 1444.6 74.9 PL15 0.672 3.410 162.01 1476.11 167.83 1535.18 180.10 1640.91 132.17 1203.64 117.8 1073.3 185.2 1687.3 aTotal fluence (\ud835\udc39\ud835\udc5c) was calculated as Average energy per pulse \u00d7 reactor residence time (\ud835\udc61\ud835\udc5f\ud835\udc52\ud835\udc60,\ud835\udc61\u210e) 121  5.3.4 Analyses 5.3.4.1 Microbial enumeration After the PL treatments, the PL-treated samples and control samples were serially diluted with autoclaved double distilled water. The dilutions were spread plated on nutrient agar (Sigma Aldrich, St. Louis, MO, United States) for E. coli ATCC 29055, BHI agar (BD Bacto, Sparks, MD, United States) for L. innocua ATCC 33090 and tryptic soy agar (BD Difco\u2122, Sparks, MD, United States) for C. sporogenes ATCC 7955. The plates were incubated at conditions specific for each bacteria: 37 oC for 24 h for E. coli and L. innocua; 37 oC for 48 h for C. sporogenes. 5.3.4.2 Inactivation kinetics modeling The microbial counts for all the microbial strains in tested liquid foods for both types of reactors were transformed into logarithmic values and fitted into kinetic models using GInaFiT add-in [170] (version 1.6) for Microsoft Excel (Microsoft Corporation, Redmond, WA, USA).  Log-linear inactivation model [171]: This model is a conventional model being used to fit first order reaction kinetics and is widely applicable in thermal processing. The model assumes that the cells have similar resistance to applied stress (PL treatment) and no. of survivors decrease over time exponentially. When expressed in log values and plotted against treatment fluence, a straight line is obtained with a negative slope. The model is given in equation 5.1. \ud835\udc3f\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc61 = \ud835\udc3f\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c \u2212\ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65\ud835\udc39\ud835\udc5c\ud835\udc59\ud835\udc5b10                                               (5.1) Where, \ud835\udc41\ud835\udc5c, \ud835\udc41\ud835\udc61 = Initial viable count and number of survivors after PL treatment with specific fluence, \ud835\udc39\ud835\udc5c, [CFU\/mL]; \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 = Rate constant of microbial inactivation, [cm2\/J]. Weibull inactivation model [172]: The Weibull inactivation model is a non-linear kinetic model and assumes that the different fractions of microorganisms in the population show different resistance to PL treatment. Overall survival curve of the population is described by a cumulative exponential distribution (equation 5.2).  \ud835\udc3f\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc61 = \ud835\udc3f\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c \u2212 (\ud835\udc39\ud835\udc5c\ud835\udeff)\ud835\udc5d                                               (5.2) Where, \ud835\udc41\ud835\udc5c, \ud835\udc41\ud835\udc61 = Initial viable count and number of survivors after PL treatment with specific fluence, \ud835\udc39\ud835\udc5c, [CFU\/mL]; \ud835\udeff = Fluence required for first decimal reduction (\ud835\udc41\ud835\udc5c \u2192\ud835\udc41\ud835\udc5c10\u2044 ), [J\/cm2]; 122  \ud835\udc5d = shape parameter (when \ud835\udc5d<1, curve is concave upward; when \ud835\udc5d>1, curve is concave downward). Log-linear with tail [173]: The Log-linear with tail model is composed of two parts; one, where the inactivation is linear and second, where the inactivation is very low leaving residual microorganisms unaffected with applied fluence. The model is given by equation 5.3. \ud835\udc3f\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc61 = \ud835\udc3f\ud835\udc5c\ud835\udc5410[(10\ud835\udc3f\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c \u2212 10\ud835\udc3f\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60) \u00d7 \ud835\udc52\u2212\ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65\ud835\udc39\ud835\udc5c + 10\ud835\udc3f\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60]            (5.3) Where, \ud835\udc41\ud835\udc5c, \ud835\udc41\ud835\udc61 = Initial viable count and number of survivors after PL treatment with specific fluence, \ud835\udc39\ud835\udc5c, [CFU\/mL]; \ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60 = residual microbial population, [CFU\/mL]; \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 = Rate constant of microbial inactivation, [cm2\/J]. The goodness of fit was tested by using coefficient of determination (R2) and root mean squared error (RMSE). 5.3.5 Statistical analysis The results were expressed as mean \u00b1 standard deviation. The results were analyzed using two-way ANOVA and Fisher\u2019s least significant difference (LSD) was carried out as post-hoc test to determine if there was a significant difference (p < 0.05) between treatment means. The statistical analyses were carried out using SPSS software version 27 (IBM\u00ae Corp., Armonk, NY, USA) and Microsoft Excel (Microsoft Corporation, Redmond, WA, USA) with XLSTAT package. 5.4 Results and discussion 5.4.1 Microbial inactivation 5.4.1.1 Milk of various fat% The inactivation curves of different microorganisms (E. coli, L. innocua, and C. sporogenes) in milk of different fat % are given in Figure 5.1 and 5.2 for AT and CT reactors, respectively. The E. coli population was reduced by 1.088, 1.005, 0.950 and 0.894 logs in 0%, 1%, 2% and 3.25% fat milk in AT reactor. Similarly, 4.093, 3.991, 3.855 and 3.789 log reduction were observed in CT reactor. Similarly, L. innocua was reduced by 0.858-0.988 log in AT reactor and 3.569-3.752 logs in CT reactor. Also, C. sporogenes was reduced by 0.745-0.818 in AT reactor and by 3.384-3.607 logs in CT reactor. Regardless of type of strain, the inactivation was significantly higher (p < 0.05) in case milk of lower fat% compared to highest fat% milk. This behavior could be ascribed to the higher absorption coefficient and scattering effects of 3.25% milk than milk with lower fat% 123  (0%, 1% and 2%) [77]. Also, fat tends to enhance the resistance of microorganisms to applied stress. For HTST pasteurized milk samples, the inactivation achieved were >5 logs for all the samples. This is expected as the heat treatment tends to inactivate a greater number of bacteria than the PL treatment. Interestingly, the maximum residence times for PL in CT reactors was 85 s (leading to ~4 log reduction), compared to HTST samples where 15 s resulted in >5 log.  (a)  (b) -1.5-1.2-0.9-0.6-0.300 0.5 1 1.5 2 2.5 3 3.5 4Log [N\/No]Fluence [J\/cm2]E. coli0% 1% 2% 3.25%-1.5-1.2-0.9-0.6-0.300 0.5 1 1.5 2 2.5 3 3.5 4Log [N\/No]Fluence [J\/cm2]L. innocua0% 1% 2% 3.25%124   (c) Figure 5.1: Inactivation curve for microorganisms in milk of different fat % in annular (AT) reactor. (a) E. coli, (b) L. innocua and (c) C. sporogenes  (a) -1.5-1.2-0.9-0.6-0.300 0.5 1 1.5 2 2.5 3 3.5 4Log [N\/No]Fluence [J\/cm2]C. sporogenes0% 1% 2% 3.25%-5-4-3-2-100 2 4 6 8 10 12 14 16 18 20Log [N\/No]Fluence [J\/cm2]E. coli0% 1%2% 3.25%125    (b)  (c) Figure 5.2: Inactivation curve for microorganisms in milk of different fat % in coiled (CT) reactor. (a) E. coli, (b) L. innocua and (c) C. sporogenes  5.4.1.2 Red grape and watermelon juice The inactivation curves for microbial strains in grape and watermelon juice in AT and CT reactor are given in Figure 5.3. The level of inactivation were significantly higher (p < 0.05) for E. coli than L. innocua and C. sporogenes. The achieved inactivation in watermelon juice were -5-4-3-2-100 2 4 6 8 10 12 14 16 18 20Log [N\/No]Fluence [J\/cm2]L. innocua 0% 1% 2% 3.25%-5-4-3-2-100 2 4 6 8 10 12 14 16 18 20Log [N\/No]Fluence [J\/cm2]C. sporogenes0% 1% 2% 3.25%126  significantly higher (p < 0.05) in case of E. coli and L. innocua in AT reactor than grape juice, while this was not true for C. sporogenes in CT reactor. The higher level of inactivation in watermelon juice could be ascribed to lower absorption coefficient and oBrix of watermelon juice (absorption coefficient = 110.6 \/m; TS = 8.3% ) than grape juice (absorption coefficient =156.5\/m; TS = 16.0%). The level of inactivation in the juices were also significantly higher than milk, which could be ascribed to the milk fat showing greater absorption and scattering of light, and also poor penetration. This has been observed previously by Hwang et al. [125], who showed higher inactivation apple juice, grape juice compared to milk, for the same level of applied fluence.  (a) -6-5-4-3-2-100 0.5 1 1.5 2 2.5 3 3.5 4Log [N\/No]Fluence [J\/cm2]AT ReactorE. coli L. innocua C. sporogenes127   (b)  (c) -8-7-6-5-4-3-2-100 2 4 6 8 10 12 14 16 18 20Log [N\/No]Fluence [J\/cm2]CT Reactor E. coli L. innocua C. sporogenes-6-5-4-3-2-100 0.5 1 1.5 2 2.5 3 3.5 4Log [N\/No]Fluence [J\/cm2]AT Reactor E. coli L. innocua C. sporogenes128   (d) Figure 5.3: Inactivation curve for E. coli, L. innocua and C. sporogenes in fruit juices. (a) red grape juice (annular), (b) red grape juice (coiled), (c) watermelon juice (annular) and (d) watermelon juice (coiled)  5.4.1.3 Microbial inactivation kinetics The data from the inactivation of the microbial strains were transformed into survivor curve to model the inactivation based on conventional log-linear, Weibull and log-linear with tail model. For all the microbial strains, regardless of the type of reactor and tested liquid, the survivor curve showed a steady decrease in bacterial load, followed by stabilization, where further inactivation was minimal. The survivor curves were fitted best by the log-linear with tail model, and the goodness of fit described by lowest RMSE error (0.045-0.497) and highest R2 values (0.939-0.990). The survivor curves fitted to log-linear with tail model are shown in Figures 5.4 and 5.5. The data of constants for different fitted models along with their RMSE error and R2 values are shown in Table 5.2 and 5.3.  -8-7-6-5-4-3-2-100 2 4 6 8 10 12 14 16 18 20Log [N\/No]Fluence [J\/cm2]CT Reactor E. coli L. innocua C. sporogenes129    (a)                                                                        (b)    (c)                                                                         (d)      (e)                                                                         (f) Figure 4.4: Fitting of survivors curves of microorganisms into log-linear with tail model in annular (AT) reactor for (a) 3.25%, (b) 2%, (c) 1%, (d) 0%, (e) red grape juice, and (f) watermelon juice  66.577.580 1 2 3 4Log10NtFluence [J\/cm2]E. coli L. innocuaC. sporogenes C. sporogenes_L-L-TE. coli_L-L-T L. innocua_L-L-T66.577.580 1 2 3 4Log10NtFluence [J\/cm2]E. coli L. innocuaC. sporogenes C. sporogenes_L-L-TE. coli_L-L-T L. innocua_L-L-T66.577.580 1 2 3 4Log10NtFluence [J\/cm2]E. coli L. innocuaC. sporogenes C. sporogenes_L-L-TE. coli_L-L-T L. innocua_L-L-T66.577.580 1 2 3 4Log10NtFluence [J\/cm2]E. coli L. innocuaC. sporogenes C. sporogenes_L-L-TE. coli_L-L-T L. innocua_L-L-T024680 1 2 3 4Log10NtFluence [J\/cm2]E. coli L. innocuaC. sporogenes C. sporogenes_L-L-TE. coli_L-L-T L. innocua_L-L-T024680 1 2 3 4Log10NtFluence [J\/cm2]E. coli L. innocuaC. sporogenes C. sporogenes_L-L-TE. coli_L-L-T L. innocua_L-L-T24680 5 10 15 20Log10NtFluence [J\/cm2]E. coli L. innocuaC. sporogenes C. sporogenes_L-L-TE. coli_L-L-T L. innocua_L-L-T24680 5 10 15 20Log10NtFluence [J\/cm2]E. coli L. innocuaC. sporogenes C. sporogenes_L-L-TE. coli_L-L-T L. innocua_L-L-T130                                      (a)                                                                        (b)                                      (c)                                                                        (d)                                      (e)                                                                        (f) Figure 5.5: Fitting of survivors curves of microorganisms into log-linear with tail model in coiled (CT) reactor for (a) 3.25%, (b) 2%, (c) 1%, (d) 0%, (e) red grape juice, and (f) watermelon juice The log-linear with tail model has been observed previously [85,175] during PL treatment of iceberg lettuce, white cabbage and carrots, and liquid foods like orange, pineapple, and coconut water. It has been observed previously that with further increase of UV light intensity, the inactivation rate decreases. The tailing of the graphs in the kinetic models suggests that the application of PL beyond a certain point does not add to inactivation and leads to destruction of quality and nutritional parameters [85]. The long-linear with tail model advises on the residual microbial population (\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60). This value should essentially be as low as possible as this governs the shelf-life of the product after treatment. The obtained values of \ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60 were higher for AT reactor than CT reactor due to lower inactivation rates in the former. Also, the \ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60 values were also higher for the milk samples than juice samples, suggesting higher inactivation efficacy for juices than milk.24680 5 10 15 20Log10NtFluence [J\/cm2]E. coli L. innocuaC. sporogenes C. sporogenes_L-L-TE. coli_L-L-T L. innocua_L-L-T24680 5 10 15 20Log10NtFluence [J\/cm2]E. coli L. innocuaC. sporogenes C. sporogenes_L-L-TE. coli_L-L-T L. innocua_L-L-T024680 5 10 15 20Log10NtFluence [J\/cm2]E. coli L. innocuaC. sporogenes C. sporogenes_L-L-TE. coli_L-L-T L. innocua_L-L-T024680 5 10 15 20Log10NtFluence [J\/cm2]E. coli L. innocuaC. sporogenes C. sporogenes_L-L-TE. coli_L-L-T L. innocua_L-L-T131  Table 5.3: Values of constant obtained by fitting the survivor curves for milk, grape, and watermelon juice in AT reactor Product Parameters Model Log-linear R2 RMSE Weibull R2 RMSE Log-linear plus tail R2 RMSE E. coli Milk (0%) \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.641 0.837  0.105  _ 0.977  0.035  1.616 0.949  0.055  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 7.271 7.557 7.408 \ud835\udeff [J\/cm2] _ 2.665 _ \ud835\udc5d [-] _ 0.453 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 6.504 Milk (1%) \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.596 0.808  0.112  _ 0.976  0.045  1.767 0.960  0.057  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 7.302 7.599 7.459 \ud835\udeff [J\/cm2] _ 2.931 _ \ud835\udc5d [-] _ 0.428 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 6.620 Milk (2%) \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.568 0.830  0.099  _ 0.978  0.040  1.561 0.960  0.053  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 7.406 7.672 7.537 \ud835\udeff [J\/cm2] _ 3.405 _ 132  \ud835\udc5d [-] _ 0.441 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 6.726 Milk (3.25%) \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.641 0.826  0.203  _ 0.978  0.182  1.616 0.970  0.183  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 7.271 7.557 7.408 \ud835\udeff [J\/cm2] _ 2.665 _ \ud835\udc5d [-] _ 0.453 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 6.504 Grape juice \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 3.414 0.600 1.053 _ 0.850 0.646 13.020 0.963 0.318  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 6.005 8.240 7.780  \ud835\udeff [J\/cm2] _ 0.022 _  \ud835\udc5d [-] _ 0.378 _  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 3.012 Watermelon juice \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 3.621 0.615 1.082 _ 0.868 0.633 13.169 0.967 0.319  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 5.896 8.262 7.690  \ud835\udeff [J\/cm2] _ 0.019 _  \ud835\udc5d [-] _ 0.377 _  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 2.681 Table 5.3: Continued 133  L. innocua Milk (0%) \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.720 0.745  0.156  _ 0.885  0.103  2.311 0.964  0.048  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 7.476 7.764 7.716 \ud835\udeff [J\/cm2] _ 2.166 _ \ud835\udc5d [-] _ 0.494 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 6.686 Milk (1%) \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.702 0.761  0.153  _ 0.903  0.101  2.206 0.981  0.060  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 7.486 7.768 7.710 \ud835\udeff [J\/cm2] _ 2.278 _ \ud835\udc5d [-] _ 0.493 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 6.709 Milk (2%) \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.633 0.759  0.136  _ 0.903  0.087  2.132 0.970  0.052  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 7.475 7.736 7.684 \ud835\udeff [J\/cm2] _ 2.780 _ \ud835\udc5d [-] _ 0.485 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 6.780 Table 5.3: Continued 134  Milk (3.25%) \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.720 0.740  0.148  _ 0.884  0.109  2.311 0.967  0.078  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 7.476 7.764 7.716 \ud835\udeff [J\/cm2] _ 2.166 _ \ud835\udc5d [-] _ 0.494 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 6.686 Grape juice \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 3.456 0.685 0.886 _ 0.888 0.527 10.043 0.982 0.214  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 6.391 8.241 7.725  \ud835\udeff [J\/cm2] _ 0.042 _  \ud835\udc5d [-] _ 0.423 _  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 3.010 Watermelon juice \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 3.517 0.674 0.924 _ 0.885 0.550 10.685 0.977 0.246  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 6.319 8.279 7.747  \ud835\udeff [J\/cm2] _ 0.035 _  \ud835\udc5d [-] _ 0.413 _  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 2.959 C. sporogenes Table 5.3: Continued 135  Milk (0%) \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.605 0.698  0.149  _ 0.859  0.100  2.342 0.962  0.045  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 7.475 7.744 7.714 \ud835\udeff [J\/cm2] _ 2.906 _ \ud835\udc5d [-] _ 0.471 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 6.823 Milk (1%) \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.598 0.698  0.150  _ 0.864  0.102  2.339 0.973  0.052  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 7.477 7.747 7.714 \ud835\udeff [J\/cm2] _ 2.955 _ \ud835\udc5d [-] _ 0.467 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 6.833 Milk (2%) \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.566 0.715  0.138  _ 0.868  0.097  2.173 0.964  0.060  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 7.492 7.736 7.703 \ud835\udeff [J\/cm2] _ 3.413 _ \ud835\udc5d [-] _ 0.476 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 6.877 Milk (3.25%) \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.605 0.702  0.147  _ 0.855  0.108  2.342 0.965  0.072  Table 5.3: Continued 136  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 7.475 7.744 7.714 \ud835\udeff [J\/cm2] _ 2.906 _ \ud835\udc5d [-] _ 0.471 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 6.823 Grape juice \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 3.389 0.728 0.782 _ 0.905 0.463 9.098 0.979 0.216  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 6.620 8.233 7.784  \ud835\udeff [J\/cm2] _ 0.062 _  \ud835\udc5d [-] _ 0.451 _  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 3.265 Watermelon juice \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 3.425 0.717 0.812 _ 0.893 0.500 9.482 0.983 0.197  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 6.603 8.227 7.833  \ud835\udeff [J\/cm2] _ 0.061 _  \ud835\udc5d [-] _ 0.452 _  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 3.231  Table 5.4: Values of constant obtained by fitting the survivor curves for milk, grape, and watermelon juice in CT reactor Product Parameters Model Log-linear R2 RMSE Weibull R2 RMSE Log-linear plus tail R2 RMSE Table 5.3: Continued 137  E. coli Milk (0%) \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 6.786 0.722  0.690  _ 0.892  0.437  1.465 0.979  0.231  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 0.575 8.132 7.738 \ud835\udeff [J\/cm2] _ 0.472 _ \ud835\udc5d [-] _ 0.458 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-]  _ 3.835 Milk (1%) \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 7.099 0.769  0.628  _ 0.878  0.459  1.254 0.968  0.247  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 0.596 8.118 7.841 \ud835\udeff [J\/cm2] _ 0.828 _ \ud835\udc5d [-] _ 0.535 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 3.817 Milk (2%) \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 7.269 0.774  0.618  _ 0.865  0.479  1.208 0.959  0.275  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 0.594 8.165 7.968 \ud835\udeff [J\/cm2] _ 1.015 _ \ud835\udc5d [-] _ 0.566 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 3.962 138  Milk (3.25%) \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 7.307 0.780  0.612  _ 0.863  0.485  1.163 0.958  0.279  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 0.598 8.148 7.960 \ud835\udeff [J\/cm2]  1.109 _ \ud835\udc5d [-]  0.583 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-]  _ 3.907 Grape juice \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.762 0.492 1.481 _ 0.890 0.691 3.793 0.960 0.419  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 4.402 8.349 7.044  \ud835\udeff [J\/cm2] _ 0.006 _  \ud835\udc5d [-] _ 0.275 _  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 1.193 Watermelon juice \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.758 0.466 1.554 _ 0.865 0.781 4.246 0.964 0.402  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 4.388 8.398 7.295  \ud835\udeff [J\/cm2] _ 0.006 _  \ud835\udc5d [-] _ 0.274 _  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 1.240 L. innocua Table 5.4: Continued 139  Milk (0%) \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 6.956 0.766  0.609  _ 0.895  0.414  1.305 0.972  0.243  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 0.569 8.042 7.756 \ud835\udeff [J\/cm2] _ 0.745 _ \ud835\udc5d [-] _ 0.508 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 3.965 Milk (1%) \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 7.106 0.696  0.732  _ 0.814  0.577  1.372 0.940  0.349  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 0.573 8.153 7.988 \ud835\udeff [J\/cm2] _ 0.804 _ \ud835\udc5d [-] _ 0.522 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 4.045 Milk (2%) \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 7.312 0.745  0.666  _ 0.834  0.541  1.229 0.946  0.326  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 0.588 8.179 8.042 \ud835\udeff [J\/cm2] _ 1.074 _ \ud835\udc5d [-] _ 0.573 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 4.045 Milk (3.25%) \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 7.368 0.765  0.620  _ 0.842  0.513  1.136 0.945  0.317  Table 5.4: Continued 140  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 0.577 8.134 8.012 \ud835\udeff [J\/cm2] _ 1.272 _ \ud835\udc5d [-] _ 0.598 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 4.092 Grape juice \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.862 0.545 1.509 _ 0.860 0.837 3.117 0.951 0.497  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 5.047 8.594 7.316  \ud835\udeff [J\/cm2] _ 0.019 _  \ud835\udc5d [-] _ 0.327 _  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 1.064 Watermelon juice \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.870 0.555 1.493 _ 0.870 0.805 3.126 0.966 0.415  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 5.070 8.620 7.346  \ud835\udeff [J\/cm2] _ 0.020 _  \ud835\udc5d [-] _ 0.329 _  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 1.038 C. sporogenes Milk (0%) \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 7.212 0.788  0.542  _ 0.890  0.398  1.093 0.962  0.261  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 0.536 8.072 7.834 Table 5.4: Continued 141  \ud835\udeff [J\/cm2] _ 1.113 _ \ud835\udc5d [-] _ 0.550 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 4.264 Milk (1%) \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 7.295 0.743  0.637  _ 0.836  0.512  1.180 0.944  0.318  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 0.559 8.151 7.998 \ud835\udeff [J\/cm2] _ 1.111 _ \ud835\udc5d [-] _ 0.564 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 4.198 Milk (2%) \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 7.440 0.764  0.601  _ 0.836  0.506  1.068 0.939  0.327  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 0.557 8.150 8.034 \ud835\udeff [J\/cm2] _ 1.415 _ \ud835\udc5d [-] _ 0.607 _ \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 4.238 Milk (3.25%) \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 7.423 0.786  0.557  _ 0.859  0.455  1.046 0.961  0.257  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 0.551 8.125 8.000 \ud835\udeff [J\/cm2] _ 1.441 _ \ud835\udc5d [-] _ 0.607 _ Table 5.4: Continued 142  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 4.244 Grape juice \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.968 0.631 1.419 _ 0.861 0.871 2.886 0.984 0.293  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 5.852 8.802 7.901  \ud835\udeff [J\/cm2] _ 0.061   \ud835\udc5d [-] _ 0.396   \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 1.038 Watermelon juice \ud835\udc58\ud835\udc5a\ud835\udc4e\ud835\udc65 [cm2\/J] 0.973 0.621 1.454 _ 0.861 0.881 2.970 0.990 0.231  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5c [-] 5.731 8.782 7.859  \ud835\udeff [J\/cm2] _ 0.054 _  \ud835\udc5d [-] _ 0.391 _  \ud835\udc59\ud835\udc5c\ud835\udc5410\ud835\udc41\ud835\udc5f\ud835\udc52\ud835\udc60[-] _ _ 0.919  Table 5.4: Continued 143  5.5 Conclusion This chapter discusses the PL based inactivation of inoculated microorganisms in milk and fruit juices. Overall, PL seems a great processing method for juices. Log reduction of >5 was achieved in both the juice types, meeting the FDA requirements, which has been defined as the acceptability criterion for any emerging process. However, to extend it to process milk, combination treatments like minimal heating (60-65 oC), other non-thermal methods need to be tested and validated. Further, the effects on the nutritional and quality changes in milk and juices need to be examined to get a holistic view of the PL processing of these liquid foods. It was observed that the log-linear tail model fitted the inactivation data of microorganisms. This indicates that PL fluence is applicable only up to a certain value after which further fluence application tends to negatively affect the quality and nutritional parameter.             144  Chapter 6: Effect of pulsed UV light treatment of physico-chemical and nutritional parameters of milk 6.1 Summary Apart from microbial inactivation, it is also important to understand the effect of PL processing on the quality and nutritional parameters of milk. Therefore, in this chapter, effect of PL treatment on pH, colour parameters, vitamins C and B2, lipid and protein oxidation in milk were evaluated. It was observed that the pH did not change to a great extent in milk after PL treatment, although the change was significant (p < 0.05). The colour parameter L* also did not change for milk samples after PL treatment. The a* values for milk increased as the PL fluence increased meaning the redness of milk increased while, b* values decreased after PL treatment significantly (p < 0.05). This could be attributed to the photodegradation of vitamin B2. The decrease in b* values was the greatest for skim milk and the least in 3.25%. Vitamins B2 and C also degraded with PL fluence, with higher degradation in case of CT reactor. Lipid oxidation was significantly (p < 0.05) higher than control, with the values being highest for 3.25% fat samples treated in CT reactor. Protein oxidation was also observed in PL-treated milk.  6.2 Introduction PL is a promising technology for inactivation of microorganisms in milk and milk products. However, an important consideration for any food processing technology is how it affects the physico-chemical, nutritional, and sensory properties of milk. Lipids and proteins in milk are quite susceptible to changes arising from light. Lipids in milk lead to off-flavour generation auto-oxidation phenomenon, due to light. This leads to changes in aroma of PL-treated milk samples. Slight aroma changes have been reported in PL-treated goat milk [83] possibly due to photo-chemical changes in milk components and lipid oxidation. Krishnamurthy et al. [51] observed increase in milk temperature up to 38 \u00b0C by continuous PL treatment of milk leading to fouling as well as alterations in milk quality. However, they added that adequately designed pulsed UV treatment process of milk does not induce lipid oxidation. Elmnasser et al. [115] carried out PL treatment of milk and did not notice any lipid oxidation and changes in amino acid sequence and composition of proteins after the PL treatment. Kasahara et al. [83] noticed that PL induced photo-chemical changes in milk lipids, proteins, and oxidation of vitamins. They also observed a slight decrease in viscosity, pH and density of milk after PL application. Proteins in milk undergo 145  changes such as oxidation of amino acids to yield protein carbonyls after UV treatment. Primary amino acids prone to oxidation include, methionine, tryptophan, histidine [176]. Formation of dityrosine is also a marker for changes in protein after UV-C treatment [177]. These changes have been extensively studied for milk and cheese subjected to UV exposure. However, such analysis has not been carried out for PL-treated milk. Changes in photosensitive vitamins like vitamin B2 and C is also a marker for assessing the extent of photo-induced changes in milk. The kinetics of these changes compared to the kinetics of microbial inactivation in milk by PL can play a role in effective process design.  Therefore, the main objective of this experiment is to assess the changes in the physico-chemical properties (pH, colour parameters), and nutritional parameters (vitamin C, B2) and lipids and protein oxidation in milk.  6.3 Materials and methods 6.3.1 Materials Milk samples of varying fat % (0%, 1%, 2% and 3.25%) from Dairyland (Dairyland, Saputo Dairy products, BC, Canada) were purchased from local grocery store. The thermophysical and optical properties of the milk samples were calculated and shown in Table 6.1. Milk was stored at refrigerated temperature (4 oC) in dark to reduce exposure to light and treated with pulsed light after bringing the temperature to 25 oC. For chemical analyses, riboflavin, and L-ascorbic acid standard for obtaining standard curves were purchased from Sigma Aldrich (St. Louis, MO, United States). Metaphosphoric acid (33% w\/v) was purchased from Alfa Aesar (Mississauga, ON, Canada). HPLC Grade methanol, 2 M hydrochloric acid and acetic acid were procured from VWR Chemicals (Mississauga, ON, Canada). Chloroform, ethanol, and ethyl acetate were purchased from Fisher Scientific (Fair Lawn, NJ, United States). Reagent grade methanol, ferrous sulphate, ammonium thiocyanate, 10 M hydrochloric acid and guanidine hydrochloride were purchased from Sigma Aldrich (St. Louis, MO, United States). Barium chloride and trichloroacetic acid were purchased from Ward\u2019s Science (St. Catherine\u2019s, ON, Canada). 2,4\u2013dinitrophenyl hydrazine (DNPH) was procured from Spectrum Chemicals (New Brunswick, NJ, Canada). Ultrapure water (18 M\u03a9) was used for dilution and making reagents for analyses. 146  6.3.2 Pulsed UV light treatments PL treatment system consisted of a cylindrical xenon flashlamp emitting high-intensity light pulses in UV-Visible-IR region (~200-1100 nm), provided by Solaris Disinfection Inc. (Mississauga, ON, Canada). The lamp dimensions were: length, \ud835\udc3f = 62 cm; diameter, \ud835\udc51\ud835\udc3f = 9 mm. The lamp emitted 30 J of energy per pulse. Considering a pulse width of 100 ms (since light fluctuation due to pulsation was hardly noticeable beyond 9 pulses\/s), the lamp sleeves emitted energy intensity per pulse of 17122.12 mW\/cm2. Treatments were carried out in both the type of reactor configurations, viz. AT and CT reactors. Milk samples in volumes 100 mL and 400 mL for AT and CT reactor were transferred into a stainless steel container with lid. Milk was pumped through the reactors using a peristaltic pump (Masterflex L\/S model 7554-90, Cole-Parmer Instruments, IL, United States) with a variable speed drive. Flow rates of  14.32, 32.48, 49.63, 64.13 and 74.90 L\/h were used and when the milk reached the inlet, the PL system was flashed (pulse frequency used were 1,3 and 5 Hz) until all the milk was collected from outlet in a bottled lined with aluminum foil. There was a total of 15 PL treatments (PL1-PL15) and control milk sample for all the milk fat% tested. The energy per pulse at distances 1.5 and 2 cm (for CT and AT rectors, respectively) from lamp were measured along the length of the lamp to calculate average energy per pulse, using a pyroelectric head sensor (PE80BF-DIF-C, Ophir-Spiricon LLC, UT, United states), with Nova II display (Ophir-Spiricon LLC, UT, United states) placed under the lamp. The average energies were 53 and 43 mJ\/cm2. All the treatments were carried out in triplicates. After treatment, the milk samples were kept refrigerated (4 oC) until analyses. The experimental design for the PL processing of milk is shown in the Table 6.1. 147  Table 6.1: Experimental design of the PL treatment of milk Pulse frequency [Hz] Flow rate [L\/h] Treatment Fluence [J\/cm2]a Reynolds number (\ud835\udc75\ud835\udc79\ud835\udc86) 0% 1% 2% 3.25% AT CT AT CT AT CT AT CT AT CT 1 14.3 PL1 0.703 4.399 30.97 282.16 32.08 293.45 34.43 313.66 25.25 230.07 32.5 PL2 0.309 1.573 70.25 640.02 72.77 665.64 78.09 711.48 57.28 521.89 49.6 PL3 0.203 1.029 107.35 978.12 111.21 1017.26 119.34 1087.32 87.54 797.57 64.1 PL4 0.157 0.797 138.71 1263.78 143.69 1314.35 154.19 1404.87 113.10 1030.50 74.9 PL5 0.135 0.683 162.01 1476.11 167.83 1535.18 180.10 1640.91 132.17 1203.64 3 14.3 PL6 2.109 10.708 30.97 282.16 32.08 293.45 34.43 313.66 25.25 230.07 32.5 PL7 0.929 4.719 70.25 640.02 72.77 665.64 78.09 711.48 57.28 521.89 49.6 PL8 0.609 3.089 107.35 978.12 111.21 1017.26 119.34 1087.32 87.54 797.57 64.1 PL9 0.471 2.390 138.71 1263.78 143.69 1314.35 154.19 1404.87 113.10 1030.50 74.9 PL10 0.403 2.046 162.01 1476.11 167.83 1535.18 180.10 1640.91 132.17 1203.64 5 14.3 PL11 3.515 17.846 30.97 282.16 32.08 293.45 34.43 313.66 25.25 230.07 32.5 PL12 1.549 7.866 70.25 640.02 72.77 665.64 78.09 711.48 57.28 521.89 49.6 PL13 1.014 5.149 107.35 978.12 111.21 1017.26 119.34 1087.32 87.54 797.57 64.1 PL14 0.785 3.984 138.71 1263.78 143.69 1314.35 154.19 1404.87 113.10 1030.50 74.9 PL15 0.672 3.410 162.01 1476.11 167.83 1535.18 180.10 1640.91 132.17 1203.64 aTotal fluence (\ud835\udc39\ud835\udc5c) was calculated as Average energy per pulse \u00d7 reactor residence time (\ud835\udc61\ud835\udc5f\ud835\udc52\ud835\udc60,\ud835\udc61\u210e)148  6.3.3 Analyses 6.3.3.1 pH measurements The pH measurements were carried out in a benchtop pH meter (Accumet AE150, Saint-Laurent, Quebec, Canada), which was previously calibrated using calibration standard buffers (pH 4.0, 7.0 and 10.0). Small aliquot of sample (5 mL) was taken in a small beaker and the pH meter probe was dipped into it. The pH was recorded when the reading stabilized. The measurements were carried out in triplicates.  6.3.3.2 Colour measurements The colour of the control milk samples before treatment and the PL-treated was measured using a benchtop colorimeter (HunterLab, model LabScanTM XE Plus, Hunter Associates Laboratory, Reston, VA, USA). Colour of the samples was recorded as the CIE L*a*b* tristimulus colour coordinates parameters (L*\u2013Lightness, a*\u2013red\/green, and b*\u2013blue\/yellow). From the colour parameters, the total colour difference (\u2206\ud835\udc38) was calculated for all samples using the Eq. (6.1) [135]. All the measurements were carried out in triplicates. \u2206\ud835\udc38 = \u221a(\ud835\udc3f\u2217 \u2212 \ud835\udc3f\ud835\udc5c\u2217 )2 + (\ud835\udc4e\u2217 \u2212 \ud835\udc4e\ud835\udc5c\u2217)2 + (\ud835\udc4f\u2217 \u2212 \ud835\udc4f\ud835\udc5c\u2217)2                                       (6.1) where, \ud835\udc3f\ud835\udc5c\u2217 , \ud835\udc4e\ud835\udc5c\u2217  and \ud835\udc4f\ud835\udc5c\u2217 are the colour parameters of the control milk sample. 6.3.3.3 Vitamin C determination Vitamin C content in milk samples was determined by following the method of Guneser and Yuceer, [178]. In brief, an aliquot of 5 mL milk sample was added with 5 mL of 5.6% (w\/v) metaphosporic acid  solution dropwise to lower the pH to 3.0. The mixture was then vortexed properly and centrifuge at 1000g for 15 min at 10 oC. Around 1.5 mL of the resulting supernatant was filtered through polyether sulfone (PES) syringe filters (0.22 \u03bcm) and placed in an autosampler vial. The mobile phase used was water: methanol: acetic acid in the ratio of 95: 5: 0.1 (v\/v) at a flow rate of 1 mL\/min. Then, 25 \u03bcL of sample was eluted using Zorbax SB-C18 column with HPLC system (Agilent 1100 system, Agilent Technologies, Santa Clara, CA, USA) and detected by diode array detector (DAD) at 254 nm. Chromatograms were analyzed with ChemStation Software version B.04.03. For standard curve development, 0-100 mg\/L of L-ascorbic acid standard solutions were injected into the HPLC system. Area versus concentration was plotted for the standard solutions to obtain the regression equation. Results were expressed as mg L-ascorbic acid \/L milk. Figure A2-1 shows the standard curve for vitamin C standard solution. Figure A2-3 149  and A2-4 show chromatogram for determination of vitamin C content in vitamin C standard solution and milk sample, respectively. Peak is shown at retention time of around 2 min.  6.3.3.4 Vitamin B2 determination Vitamin B2 content in milk samples were determined following the method of Ashoor et al. [179]. In brief, an aliquot of 10 mL milk sample was added with  50% (v\/v) glacial acetic acid solution dropwise to lower the pH to 3.0. The mixture was then stirred properly and centrifuge at 15000 rpm for 15 min. The supernatant was transferred into another 25 mL volumetric flask and the sediment was washed with 5 mL 2% (w\/v) acetic acid. The wash water was combined and centrifuged at the same condition. Then the supernatant was added to the flask and the volume was made up to 25 mL with 2% (v\/v) acetic acid. Around 1.5 mL of the resulting mixture was filtered through polyether sulfone (PES) syringe filters (0.22 \u03bcm) and placed in an autosampler vial. The mobile phase used was water: methanol: acetic acid in the ratio of 68: 32: 0.1 (v\/v) at a flow rate of 1 mL\/min. Then, 25 \u03bcL of sample was eluted using Zorbax SB-C18 column with HPLC system (Agilent 1100 system, Agilent Technologies, Santa Clara, CA, USA) and detected by diode array detector (DAD) at 270 nm. Chromatograms were analyzed with ChemStation Software version B.04.03. For standard curve development, 0-6 mg\/L of riboflavin standard solutions were injected into the HPLC system. Area versus concentration was plotted for the standard solutions to obtain the regression equation. Results were expressed as mg riboflavin\/L milk. Figure A2-2 shows the standard curve for vitamin B2 standard solution. Figure A2-5 and A2-6 show chromatogram for determination of vitamin B2 content in vitamin B2 standard solution and milk sample, respectively. Peak is shown at retention time of around 5.7 min. 6.3.3.5 Lipid oxidation Lipid oxidation in milk samples was determined in terms of formation of primary lipid oxidation products following the method of \u00d8stdal et al. [180] and Pereda et al. [181]. A 2 mL aliquot of milk sample was taken in a centrifuge tube and it was added with 2 mL methanol. Then, 4 mL of chloroform was added into the tube and the tube was vortexed for 30 s and centrifuged for 10 min at 12000g. The tube was then kept undisturbed to allow phase separation. For assay preparation, a solution of  Fe (II)\/thiocyanate in methanol: chloroform was made (50 mL of 32.7 mM BaCl2 was added with 50 mL of 36 mM FeSO4 and 2 mL of 10 M HCl and the mixture was continuously stirred using magnetic stirrer. The resultant mixture solution was filtered to remove precipitated 150  Ba3(PO4)2.  Thereafter, 500 \u03bcL of the clear solution was added with 500 \u03bcL of 3.94 M NH4SCN and 49 mL of 1:1 (v\/v) methanol: chloroform). 1 mL of the solution was added with 1 mL bottom chloroform phase (from milk samples) in a cuvette and the reaction (oxidation of ferrous, Fe2+ to ferric, Fe3+ ion by hydroperoxides in the presence of ammonium thiocyanate to yield ferric thiocyanate) was allowed to proceed for 5 min. The absorbance of the coloured ferric thiocyanate was measured at 500 nm using UV\/Vis spectrophotometer (UV-1800 Shimadzu, Kyoto, Japan). The results were expressed in terms of absorbance units [-]. 6.3.3.6 Protein oxidation Protein oxidation in milk samples was determined in terms of formation of protein carbonyls which serve as markers for protein oxidation in milk. It was determined spectrophotometrically following the method of Fenaille et al. [176]. For sample preparation, an aliquot of milk sample (equivalent to 2 mg milk proteins; based on 3.2 g\/100 mL in milk, 62.5 \u03bcL of milk was calculated to contain 2 mg protein) was pipetted out into a centrifuge tube, and it was added with 1 mL of 10 mM 2,4-dintrophenylhydrazine (DNPH) in 2 M HCl. The tube was vortexed for 30 s and incubated for 30 min. Then milk proteins were precipitated with 10% (w\/v) trichloroacetic acid and recovered by centrifugation at 7500g\/5 min. After centrifugation was discarded and protein pellet was washed with 50:50 (v\/v) ethanol: ethyl acetate. For assaying, the pellet was redissolved in 1 mL of 6 M guanidine hydrochloride (pH 2.3). Protein carbonyls were determined by UV\/Vis spectrophotometry at 370 nm using a UV\/Vis spectrophotometer (UV-1800 Shimadzu, Kyoto, Japan). Molar extinction coefficient of 220, 000 L\/(mol-cm) was used to calculate the protein carbonyl content in milk which was expressed as nM of Protein Carbonyl\/mg protein. 6.3.4 Statistical analysis The results were expressed as mean \u00b1 standard deviation (except for pH and colour parameters). The results were analyzed using two-way ANOVA and Fisher\u2019s least significant difference (LSD) was carried out as post-hoc test to determine if there was a significant difference (p < 0.05) between treatment means arising from flow rate, frequency, and their interaction. One-way ANOVA with LSD was conducted to see the effect of treatments on pH and colour parameters. The statistical analyses were carried out using SPSS software version 27 (IBM\u00ae Corp., Armonk, NY, USA) and Microsoft Excel (Microsoft Corporation, Redmond, WA, USA) with XLSTAT package. 151  6.4 Results and discussion 6.4.1 Effect on pH of milk The pH of the control, HTST pasteurized and PL-treated milk samples were measured and the data is presented in the Table 6.2. The PL-treated samples showed a statistically significant (p < 0.05) increase in the pH values compared to the control sample. The increase was evident regardless of the fat % of milk that was treated. The trend was not clear for the increase in pH of milk and the increase in pH was not significantly (p > 0.05) different among milk of different fat %. The increase in pH of milk could be attributed to the release of dissolved CO2 during pumping process in the reactor which can raise of pH of milk. HTST pasteurization of the milk samples did not significantly (p > 0.05) increase the pH as compared to the control.  152  Table 6.2: Effect of different treatments on pH of milk samples Treatment AT CT 3.25% 2% 1% 0% 3.25% 2% 1% 0% Control 6.67 \u00b1 0.0252e 6.71 \u00b1 0.012ef 6.7 \u00b1 0.0153f 6.82 \u00b1 0.00577d 6.67 \u00b1 0.0252d 6.71 \u00b1 0.012de 6.7 \u00b1 0.0153f 6.82 \u00b1 0.00577d HT 6.69 \u00b1 0.0153e 6.71 \u00b1 0.0186ef 6.71 \u00b1 0.01def 6.72 \u00b1 0.00577e 6.69 \u00b1 0.0153d 6.71 \u00b1 0.0186de 6.71 \u00b1 0.01ef 6.72 \u00b1 0.00577e PL1 6.72 \u00b1 0.01de 6.69 \u00b1 0.02f 6.82 \u00b1 0.0267a 6.82 \u00b1 0.0233d 6.84 \u00b1 0.0133bc 6.69 \u00b1 0.0208e 6.81 \u00b1 0.0186b 6.82 \u00b1 0.0233d PL2 6.77 \u00b1 0.01bd 6.76 \u00b1 0.00882cd 6.82 \u00b1 0.0208a 6.86 \u00b1 0.00667bd 6.84 \u00b1 0.00333bc 6.74 \u00b1 0.0133bcd 6.81 \u00b1 0.012bc 6.83 \u00b1 0.012d PL3 6.8 \u00b1 0.0208abc 6.75 \u00b1 0.00333d 6.84 \u00b1 0.0115a 6.88 \u00b1 0.00577abc 6.84 \u00b1 0.00333bc 6.75 \u00b1 0.00333bcd 6.82 \u00b1 0.0133b 6.87 \u00b1 0.00882ad PL4 6.82 \u00b1 0.00882ab 6.76 \u00b1 0.00333cd 6.82 \u00b1 0.00882ab 6.89 \u00b1 0.0153ab 6.83 \u00b1 0.00333c 6.73 \u00b1 0.0115cd 6.84 \u00b1 0.0153b 6.9 \u00b1 0.0203ab PL5 6.82 \u00b1 0.00882ab 6.76 \u00b1 0.00577bd 6.8 \u00b1 0.00333ac 6.88 \u00b1 0.01abc 6.84 \u00b1 0.00882bc 6.75 \u00b1 0.012bcd 6.83 \u00b1 0.00882b 6.87 \u00b1 0.0203ad PL6 6.82 \u00b1 0.02ab 6.71 \u00b1 0.0233ef 6.76 \u00b1 0.0219bcd 6.75 \u00b1 0.0473e 6.93 \u00b1 0.00577a 6.72 \u00b1 0.0153de 6.95 \u00b1 0.012a 6.75 \u00b1 0.0458e 153  Values are means \u00b1 SEM, n = 3 per treatment group.  Means in a column without a common superscript letter differ (p < 0.05) as analyzed by one-way ANOVA and the LSD test. PL7 6.79 \u00b1 0.0133abc 6.75 \u00b1 0.00882d 6.79 \u00b1 0.00667ac 6.84 \u00b1 0.00333cd 6.86 \u00b1 0.00882b 6.73 \u00b1 0.0153cd 6.84 \u00b1 0.0318b 6.84 \u00b1 0.00577cd PL8 6.8 \u00b1 0.0328abc 6.73 \u00b1 0.00882de 6.8 \u00b1 0.0186ac 6.87 \u00b1 0.00882abc 6.84 \u00b1 0.00882bc 6.74 \u00b1 0.00882bcd 6.83 \u00b1 0.01b 6.84 \u00b1 0.01cd PL9 6.81 \u00b1 0.0203abc 6.76 \u00b1 0.00577bd 6.71 \u00b1 0.0153def 6.9 \u00b1 0.00577a 6.83 \u00b1 0.00333c 6.76 \u00b1 0.00577ac 6.72 \u00b1 0.0328ef 6.89 \u00b1 0.00577abc PL10 6.83 \u00b1 0.0348a 6.76 \u00b1 0.00333cd 6.71 \u00b1 0.00333ef 6.89 \u00b1 0.00333ab 6.83 \u00b1 0.00333c 6.74 \u00b1 0.00333bcd 6.73 \u00b1 0.0219ef 6.88 \u00b1 0.012abc PL11 6.78 \u00b1 0.0186abc 6.78 \u00b1 0.00333bc 6.71 \u00b1 0.0404def 6.87 \u00b1 0.00577abc 6.95 \u00b1 0.00333a 6.76 \u00b1 0.0115ac 6.73 \u00b1 0.0306ef 6.85 \u00b1 0.00577bd PL12 6.77 \u00b1 0.00667bd 6.78 \u00b1 0.00333bc 6.76 \u00b1 0.026ce 6.89 \u00b1 0.00333ab 6.85 \u00b1 0.00882bc 6.77 \u00b1 0.0186ac 6.76 \u00b1 0.0291cde 6.88 \u00b1 0.01abc PL13 6.8 \u00b1 0.0321abc 6.75 \u00b1 0.00333d 6.76 \u00b1 0.026ce 6.91 \u00b1 0.00577a 6.84 \u00b1 0.00577bc 6.76 \u00b1 0.00667ac 6.74 \u00b1 0.0167df 6.91 \u00b1 0.00577a PL14 6.78 \u00b1 0.0153abc 6.79 \u00b1 0.00333b 6.73 \u00b1 0.0133def 6.9 \u00b1 0.00577a 6.84 \u00b1 0.00577bc 6.77 \u00b1 0.00882ab 6.73 \u00b1 0.0153ef 6.86 \u00b1 0.0265ad PL15 6.75 \u00b1 0.0203cd 6.82 \u00b1 0.00333a 6.79 \u00b1 0.0153ac 6.91 \u00b1 0.0115a 6.85 \u00b1 0.012bc 6.79 \u00b1 0.0173a 6.79 \u00b1 0.00667bd 6.9 \u00b1 0.0186ab Table 6.2: Continued 154  6.4.2 Effect on colour of milk The effect of PL processing on L*, a*, b* and \u2206\ud835\udc38 values of milk samples are given in Table 6.3, 6.4, 6.5 and 6.5, respectively. For all milk samples the colour was significantly (p < 0.05) affected. The L* value which shows the changes in lightness of samples milk samples was affected significantly (p < 0.05) regardless of fat %. The decrease in L* did not show any trend; however, the decrease in L* was more as the fat % decreased (decrease in L* in 0% > 3.25%). The milk samples with different fat % were significantly (p < 0.05) different in terms of decrease in L* value. The decrease was significantly (p < 0.05) higher in case of milk samples treated in CT reactor. HTST treated milk samples also showed a decrease in L* value compared to control and the decrease was greater in milk of lower fat %. These changes and browning could be related to the Maillard reaction happening between amino acids (especially lysine residues) and carbonyl groups of reducing sugars (lactose) in milk which produces brown-coloured melanoidins pigments as reported by Hu et al. [182].  155  Table 6.3: Effect of different treatments on L* of milk samples Treatment AT CT 3.25% 2% 1% 0% 3.25% 2% 1% 0% Control 81.4 \u00b1 0.212a 78.8 \u00b1 0.414a 74.7 \u00b1 0.406a 72 \u00b1 0.0208a 81.4 \u00b1 0.212ab 78.8 \u00b1 0.414a 74.7 \u00b1 0.406d 72 \u00b1 0.0208a HT 80.4 \u00b1 0.212ac 78.2 \u00b1 0.461ab 72.5 \u00b1 0.312ce 70.3 \u00b1 0.107b 80.4 \u00b1 0.212cd 78.2 \u00b1 0.461ab 72.5 \u00b1 0.312f 70.3 \u00b1 0.107b PL1 80.9 \u00b1 0.455ab 77.6 \u00b1 0.354bc 73 \u00b1 0.838be 65.5 \u00b1 0.322fgh 80.7 \u00b1 0.335ac 78 \u00b1 0.589ac 74.9 \u00b1 0.351cd 65.9 \u00b1 0.0186ce PL2 80.1 \u00b1 0.462bcd 76 \u00b1 0.24d 72.4 \u00b1 0.223ce 64.8 \u00b1 0.154i 80.9 \u00b1 0.255ac 77.4 \u00b1 0.432bcd 76 \u00b1 0.197ab 65.5 \u00b1 0.18de PL3 80.2 \u00b1 0.491bcd 76.9 \u00b1 0.593cd 72.7 \u00b1 0.0484be 66.2 \u00b1 0.172de 80.5 \u00b1 0.132cd 77.5 \u00b1 0.0839bcd 75.5 \u00b1 0.143ad 65.8 \u00b1 0.175ce PL4 79.4 \u00b1 0.512cd 77 \u00b1 0.133cd 73.1 \u00b1 0.111bc 64.9 \u00b1 0.162hi 80.7 \u00b1 0.199ac 77.3 \u00b1 0.353bcd 76.1 \u00b1 0.047a 65.3 \u00b1 0.172e PL5 79.6 \u00b1 0.446cd 76.9 \u00b1 0.287cd 72.9 \u00b1 0.156be 65.9 \u00b1 0.367eg 80.8 \u00b1 0.264ac 77.3 \u00b1 0.342bcd 75.6 \u00b1 0.213abc 65.9 \u00b1 0.0186ce PL6 79.9 \u00b1 0.372bcd 77 \u00b1 0.205cd 73 \u00b1 0.461be 67.6 \u00b1 0.0376c 78.4 \u00b1 0.138e 77.3 \u00b1 0.348bcd 75.5 \u00b1 0.0643ad 66.3 \u00b1 0.896ce 156  Values are means \u00b1 SEM, n = 3 per treatment group.  Means in a column without a common superscript letter differ (p < 0.05) as analyzed by one-way ANOVA and the LSD test.PL7 80.2 \u00b1 0.246bcd 76.3 \u00b1 0.203d 72 \u00b1 0.292de 66.3 \u00b1 0.237de 81.1 \u00b1 0.346ac 76.6 \u00b1 0.146d 75.2 \u00b1 0.195bd 65.6 \u00b1 0.239ce PL8 80.2 \u00b1 0.227bcd 76.8 \u00b1 0.472cd 72 \u00b1 0.39e 66.1 \u00b1 0.244def 80.7 \u00b1 0.195ac 77.4 \u00b1 0.197bcd 75.8 \u00b1 0.256ab 65.8 \u00b1 0.199ce PL9 79.3 \u00b1 0.399cd 76.3 \u00b1 0.369d 73.1 \u00b1 0.0437bcd 65.5 \u00b1 0.259fgh 81.4 \u00b1 0.0296a 76.6 \u00b1 0.0436d 73.1 \u00b1 0.0437ef 65.6 \u00b1 0.146ce PL10 79.9 \u00b1 0.253bcd 76.7 \u00b1 0.187cd 73.3 \u00b1 0.203bc 65.4 \u00b1 0.205gi 80.5 \u00b1 0.32cd 77.4 \u00b1 0.563bcd 73.3 \u00b1 0.203ef 66.6 \u00b1 0.598c PL11 79.9 \u00b1 0.191bcd 76.6 \u00b1 0.261d 73 \u00b1 0.761be 66.6 \u00b1 0.375d 77.7 \u00b1 0.264e 76.9 \u00b1 0.439cd 73 \u00b1 0.761ef 66.4 \u00b1 0.291cd PL12 79.7 \u00b1 0.316cd 78.5 \u00b1 0.367ab 72.7 \u00b1 0.0426be 66.4 \u00b1 0.215de 79.9 \u00b1 0.57d 77.8 \u00b1 0.59ac 72.7 \u00b1 0.0426f 66.3 \u00b1 0.116cd PL13 79.4 \u00b1 0.618cd 76.7 \u00b1 0.433cd 72.9 \u00b1 0.215be 66.4 \u00b1 0.215de 80.6 \u00b1 0.318cd 77 \u00b1 0.149cd 72.9 \u00b1 0.215ef 66 \u00b1 0.222ce PL14 79 \u00b1 0.732d 76.2 \u00b1 0.112d 73.1 \u00b1 0.441bcd 67.5 \u00b1 0.0458c 80.6 \u00b1 0.137bcd 76.9 \u00b1 0.44cd 73.1 \u00b1 0.441ef 66.2 \u00b1 0.653ce PL15 80 \u00b1 0.366bcd 76.6 \u00b1 0.217d 73.6 \u00b1 0.383ab 67.4 \u00b1 0.119c 80.9 \u00b1 0.264ac 77 \u00b1 0.244cd 73.6 \u00b1 0.383e 65.9 \u00b1 0.515ce Table 6.3: Continued 157  The a* values for milk samples decreased as per their fat%, which could be attributed to the presence of colour pigments, namely \u03b2-carotene, which is fat soluble and imparts yellow-orange tint to milk [153]. Absence of milk fat increases the a* value for samples.  HTST processing  significantly (p < 0.05) increased the a* value and the increase was highest in 0% fat milk. Similar trend was observed in PL-treated milk samples. The a* value of milk samples increased with PL fluence (with lower flow rates and higher frequencies; interaction was not significant (p > 0.05) for AT reactor, while it was significant (p < 0.05) for the CT reactor). The increase in a* value was significantly higher (p < 0.05) for milk of lower fat% for both the reactors. Overall, the CT reactor treated samples showed significantly (p < 0.05) greater increase in a* value compared to AT, due to higher fluence arising from higher residence times. The increase in a* value was almost similar to HTST processed samples. While an explanation is lacking which can explain decrease in a*, same could be attributed to Maillard browning reaction. However, this warrants further investigation. Contrastingly, Hu et al. [182] did not observe any increase in a* in milk after UV-C processing. 158  Table 6.4: Effect of different treatments on a* of milk samples Treatment AT CT 3.25% 2% 1% 0% 3.25% 2% 1% 0% Control -2.3 \u00b1 0.01h -2.9 \u00b1 0.02g -3.61 \u00b1 0.0267k -5.68 \u00b1 0.0208j -2.3 \u00b1 0.01g -2.9 \u00b1 0.02i -3.61 \u00b1 0.0267f -5.68 \u00b1 0.0208l HT -1.99 \u00b1 0.0608ade -2.13 \u00b1 0.0133b -3.48 \u00b1 0.00882hj -4.13 \u00b1 0.00882a -1.99 \u00b1 0.0608e -2.13 \u00b1 0.0133gh -3.48 \u00b1 0.00882e -4.13 \u00b1 0.00882bc PL1 -2.01 \u00b1 0.0451bcde -2.34 \u00b1 0.0862c -3.43 \u00b1 0.0203gh -4.9 \u00b1 0.0265d -1.97 \u00b1 0.0265e -1.81 \u00b1 0.0404be -3.33 \u00b1 0.0418cd -4.17 \u00b1 0.012c PL2 -2.06 \u00b1 0.0219dg -2.72 \u00b1 0.0208f -3.4 \u00b1 0.0203fh -5.06 \u00b1 0.0384g -1.94 \u00b1 0.0208e -1.89 \u00b1 0.0674cef -3.38 \u00b1 0.0318d -4.63 \u00b1 0.113gi PL3 -2.17 \u00b1 0.0433gh -2.59 \u00b1 0.0802df -3.44 \u00b1 0.0333ghi -5.15 \u00b1 0.0153hi -2.08 \u00b1 0.0186f -1.97 \u00b1 0.118eg -3.49 \u00b1 0.0167e -4.72 \u00b1 0.0186ij PL4 -2.14 \u00b1 0.0608fg -2.67 \u00b1 0.0503ef -3.52 \u00b1 0.012ij -5.13 \u00b1 0.0145h -2.11 \u00b1 0.0219f -2.07 \u00b1 0.0231fgh -3.64 \u00b1 0.00882f -4.8 \u00b1 0.024jk PL5 -2.11 \u00b1 0.0433eg -2.96 \u00b1 0.0656g -3.52 \u00b1 0.026j -5.21 \u00b1 0.0186i -2.13 \u00b1 0.02f -2.2 \u00b1 0.127h -3.59 \u00b1 0.00577f -4.89 \u00b1 0.0173k PL6 -1.92 \u00b1 0.01ac -2.11 \u00b1 0.0219b -3.17 \u00b1 0.0458ab -4.64 \u00b1 0.00667b -1.75 \u00b1 0.00667bc -1.69 \u00b1 0.0487bc -3.1 \u00b1 0.00667a -4.03 \u00b1 0.0115b 159  Values are means \u00b1 SEM, n = 3 per treatment group.  Means in a column without a common superscript letter differ (p < 0.05) as analyzed by one-way ANOVA and the LSD test.PL7 -1.93 \u00b1 0.00333ac -2.58 \u00b1 0.0549de -3.24 \u00b1 0.026bd -5 \u00b1 0.0203fg -1.74 \u00b1 0.0418bc -1.81 \u00b1 0.085be -3.03 \u00b1 0.0145a -4.46 \u00b1 0.0203e PL8 -1.87 \u00b1 0.0404a -2.65 \u00b1 0.0252df -3.22 \u00b1 0.0133bc -5.02 \u00b1 0.0367fg -1.85 \u00b1 0.0173d -1.94 \u00b1 0.123deg -3.62 \u00b1 0.0306f -4.49 \u00b1 0.0351ef PL9 -2.03 \u00b1 0.0208cdef -2.64 \u00b1 0.0458df -3.37 \u00b1 0.0133fg -5 \u00b1 0.0318fg -1.96 \u00b1 0.0208e -2.03 \u00b1 0.0802fgh -3.37 \u00b1 0.0133d -4.53 \u00b1 0.0153eg PL10 -2.02 \u00b1 0.0667cdef -2.87 \u00b1 0.0742g -3.36 \u00b1 0.0176efg -4.98 \u00b1 0.0252ef -1.98 \u00b1 0.01e -2.07 \u00b1 0.0533fgh -3.36 \u00b1 0.0176d -4.62 \u00b1 0.00577gi PL11 -1.98 \u00b1 0.0517ad -1.92 \u00b1 0.0265a -3.11 \u00b1 0.0203a -4.66 \u00b1 0.012b -1.81 \u00b1 0.0296cd -1.45 \u00b1 0.0669a -3.04 \u00b1 0.0173a -3.79 \u00b1 0.0265a PL12 -1.89 \u00b1 0.0693ab -2.52 \u00b1 0.00333d -3.24 \u00b1 0.0681bd -4.78 \u00b1 0.0233c -1.48 \u00b1 0.0328a -1.66 \u00b1 0.0759ab -3.24 \u00b1 0.0681b -4.33 \u00b1 0.0289d PL13 -2.01 \u00b1 0.0726bcde -2.58 \u00b1 0.00882de -3.32 \u00b1 0.024df -4.74 \u00b1 0.0219c -1.74 \u00b1 0.0115bc -1.61 \u00b1 0.0404ab -3.32 \u00b1 0.024bd -4.47 \u00b1 0.00333e PL14 -2.06 \u00b1 0.0208dg -2.64 \u00b1 0.0361df -3.28 \u00b1 0.0493cde -4.91 \u00b1 0.01d -1.81 \u00b1 0.0219cd -1.74 \u00b1 0.0613bcd -3.28 \u00b1 0.0493bc -4.57 \u00b1 0.0416fgh PL15 -2.07 \u00b1 0.0186dg -2.71 \u00b1 0.0367ef -3.32 \u00b1 0.0115df -4.91 \u00b1 0.0328de -1.71 \u00b1 0.00667b -1.78 \u00b1 0.0551be -3.32 \u00b1 0.0115bd -4.64 \u00b1 0.00577hi Table 6.4: Continued 160  The b* values for milk samples decreased as per their fat%. HTST processing  significantly (p < 0.05) increased the b* value and the increase was highest in 3.25% fat milk, while the change was not significant (p > 0.05) in 1% and 0% fat milk. Opposite trend was observed in PL-treated milk samples. The b* value of milk samples decreased with PL fluence (lower flow rates and higher frequencies; interaction was not significant (p > 0.05) for AT reactor, while it was significant (p < 0.05) for the CT reactor). The decrease in b* value was significantly higher (p < 0.05) for milk of lower fat% for both the reactors. Overall, the CT reactor treated samples showed significantly (p < 0.05) higher decrease in b* value compared to AT, due to higher fluence arising from higher residence times. The decrease in b* value can be attributed to the loss in riboflavin in milk due to light. The riboflavin imparts a greenish yellow tinge to milk serum and when it degrades under light, that imparts a bluish tinge to milk [183]. Effect of PL on riboflavin content of milk is warranted and should be tested. Contrastingly, Hu et al. [182] did not observe any increase in b* in milk after UV-C processing.161  Table 6.5: Effect of different treatments on b* value of milk samples Treatment AT CT 3.25% 2% 1% 0% 3.25% 2% 1% 0% Control 6.07 \u00b1 0.0817b 5.42 \u00b1 0.0536b 3.65 \u00b1 0.136a -1.31 \u00b1 0.00667a 6.07 \u00b1 0.0817b 5.42 \u00b1 0.0536b 3.65 \u00b1 0.136a -1.31 \u00b1 0.00667a HT 7.4 \u00b1 0.304a 7.25 \u00b1 0.145a 3.46 \u00b1 0.024a -1.21 \u00b1 0.00333a 7.4 \u00b1 0.304a 7.25 \u00b1 0.145a 3.46 \u00b1 0.024a -1.21 \u00b1 0.00333a PL1 4.44 \u00b1 0.0721fg 2.6 \u00b1 0.0426gh 0.43 \u00b1 0.0961g -4.9 \u00b1 0.0219f 4.8 \u00b1 0.0296cd 3.73 \u00b1 0.11e 1.05 \u00b1 0.00882cd -4 \u00b1 0.0426g PL2 4.92 \u00b1 0.0437d 2.76 \u00b1 0.127fg 1.64 \u00b1 0.109e -4.34 \u00b1 0.0481d 4.71 \u00b1 0.3d 3.87 \u00b1 0.0351de 1.36 \u00b1 0.0971bc -3.81 \u00b1 0.0503f PL3 5.42 \u00b1 0.0265c 2.81 \u00b1 0.0233fg 1.95 \u00b1 0.0338de -4.2 \u00b1 0.0348cd 4.47 \u00b1 0.501def 4.36 \u00b1 0.155cd 1.27 \u00b1 0.00577bd -3.44 \u00b1 0.0145d PL4 5.44 \u00b1 0.0498c 3.24 \u00b1 0.0521de 2.17 \u00b1 0.0371cd -4.06 \u00b1 0.127bc 4.49 \u00b1 0.166de 3.92 \u00b1 0.295de 1.51 \u00b1 0.0971bc -3.33 \u00b1 0.0265cd PL5 5.53 \u00b1 0.0321c 3.69 \u00b1 0.112c 2.68 \u00b1 0.0208b -4 \u00b1 0.0426b 5.38 \u00b1 0.319c 4.79 \u00b1 0.315c 1.72 \u00b1 0.0186b -3.16 \u00b1 0.00333b PL6 2.35 \u00b1 0.0521i 2.44 \u00b1 0.0493h -2.58 \u00b1 0.17i -5.2 \u00b1 0.0186g 3.83 \u00b1 0.129fgh 2.82 \u00b1 0.15fh -0.633 \u00b1 0.0176g -4.16 \u00b1 0.0265h 162  Values are means \u00b1 SEM, n = 3 per treatment group.  Means in a column without a common superscript letter differ (p < 0.05) as analyzed by one-way ANOVA and the LSD test.PL7 4.16 \u00b1 0.0557h 2.59 \u00b1 0.0318gh 0.95 \u00b1 0.0115f -4.87 \u00b1 0.0666f 3.73 \u00b1 0.0503gh 2.75 \u00b1 0.201fh 0.33 \u00b1 0.0681f -3.64 \u00b1 0.0722e PL8 4.49 \u00b1 0.0348ef 2.77 \u00b1 0.0669fg 2.12 \u00b1 0.0433cd -4.54 \u00b1 0.0328e 4.23 \u00b1 0.219dg 2.81 \u00b1 0.218fh 0.567 \u00b1 0.385ef -3.44 \u00b1 0.0736d PL9 4.82 \u00b1 0.0351d 3.18 \u00b1 0.121e 2.33 \u00b1 0.0333c -4.31 \u00b1 0.0338d 4.43 \u00b1 0.174def 2.99 \u00b1 0.0825fg 1.17 \u00b1 0.133cd -3.61 \u00b1 0.0338e PL10 4.72 \u00b1 0.0841de 3.44 \u00b1 0.0603d 2.2 \u00b1 0.0751cd -4.16 \u00b1 0.0265c 4.69 \u00b1 0.094d 3.21 \u00b1 0.0664f 1.37 \u00b1 0.0808bc -3.25 \u00b1 0.012bc PL11 1.46 \u00b1 0.0467j 1.89 \u00b1 0.0549j -3.66 \u00b1 0.0833j -5.5 \u00b1 0.0521h 3.62 \u00b1 0.146gh 2.44 \u00b1 0.233h -0.823 \u00b1 0.304g -4.6 \u00b1 0.0581i PL12 4.03 \u00b1 0.102h 2.17 \u00b1 0.0458i -0.497 \u00b1 0.0561h -5.1 \u00b1 0.0473g 3.48 \u00b1 0.012h 2.69 \u00b1 0.226gh -0.363 \u00b1 0.0991g -4.03 \u00b1 0.0186g PL13 4.2 \u00b1 0.0808gh 2.42 \u00b1 0.0865h 0.16 \u00b1 0.243g -4.88 \u00b1 0.0406f 3.39 \u00b1 0.153h 2.78 \u00b1 0.101fh 0.16 \u00b1 0.243f -4.07 \u00b1 0.041gh PL14 4.29 \u00b1 0.0889fh 2.69 \u00b1 0.0348g 0.803 \u00b1 0.169f -4.62 \u00b1 0.0557e 3.76 \u00b1 0.364gh 2.42 \u00b1 0.21h 0.803 \u00b1 0.169de -3.62 \u00b1 0.0557e PL15 4.18 \u00b1 0.0265gh 2.94 \u00b1 0.0833f 0.987 \u00b1 0.174f -4.6 \u00b1 0.0581e 4.01 \u00b1 0.106egh 2.96 \u00b1 0.0578fg 0.887 \u00b1 0.207de -3.43 \u00b1 0.0153d Table 6.5: Continued 163  The HTST processed milk samples showed a significant increase in \u2206\ud835\udc38 values compared to control. The values of \u2206\ud835\udc38 were in increasing trend as the fat% decreased. For the PL-treated milk samples, the \u2206\ud835\udc38 significantly (p < 0.05) increased as the PL fluence increased. Milk of different fat% were significantly (p < 0.05) differently from each other in terms of \u2206\ud835\udc38 values and the increase were higher in case of CT reactor that AT, owing to higher residence time allowing for more photo-chemical changes in milk. The colour change was significantly (p < 0.05) higher than the HTST processed samples (except milk samples with fat 3.25% (PL1, PL2, PL3) and 2% (PL1) treated in AT reactor. The maximum value of  \u2206\ud835\udc38 (~2.7) was achieved in case of milk samples with 0% fat treated at higher fluence using CT reactor. The value was however, less than what could be perceived by human eye.164  Table 6.6: Effect of different treatments on total colour difference (\u2206\ud835\udc6c) of milk samples Treatment AT CT 3.25% 2% 1% 0% 3.25% 2% 1% 0% Control 0 \u00b1 0f 0 \u00b1 0g 0 \u00b1 0f 0 \u00b1 0h 0 \u00b1 0j 0 \u00b1 0h 0 \u00b1 0j 0 \u00b1 0d HT 1.3 \u00b1 0.129de 1.47 \u00b1 0.0816ef 1.49 \u00b1 0.1e 1.52 \u00b1 0.0263g 1.3 \u00b1 0.129df 1.47 \u00b1 0.0816g 1.49 \u00b1 0.1gh 1.52 \u00b1 0.0263c PL1 1.22 \u00b1 0.089e 1.47 \u00b1 0.0834ef 1.82 \u00b1 0.106cd 2.66 \u00b1 0.0565ab 1.36 \u00b1 0.0658d 1.8 \u00b1 0.0322bd 1.81 \u00b1 0.0239e 2.7 \u00b1 0.00418a PL2 1.35 \u00b1 0.204ce 1.79 \u00b1 0.0621acd 1.81 \u00b1 0.0579cd 2.77 \u00b1 0.0247a 1.16 \u00b1 0.0432efg 1.8 \u00b1 0.0454bd 1.54 \u00b1 0.0454fg 2.7 \u00b1 0.0343a PL3 1.4 \u00b1 0.232ce 1.48 \u00b1 0.185ef 1.77 \u00b1 0.00916d 2.5 \u00b1 0.0335e 1.06 \u00b1 0.054ghi 1.75 \u00b1 0.00871cde 1.37 \u00b1 0.0163h 2.63 \u00b1 0.0315ab PL4 1.61 \u00b1 0.0979acd 1.56 \u00b1 0.0351def 1.63 \u00b1 0.0425de 2.72 \u00b1 0.0291ab 0.966 \u00b1 0.08hi 1.68 \u00b1 0.0466df 1.42 \u00b1 0.0183gh 2.71 \u00b1 0.0373a PL5 1.43 \u00b1 0.111bce 1.44 \u00b1 0.11f 1.64 \u00b1 0.0293de 2.54 \u00b1 0.0691cde 0.92 \u00b1 0.0864i 1.56 \u00b1 0.0839fg 1.15 \u00b1 0.0585i 2.61 \u00b1 0.000394ab PL6 1.65 \u00b1 0.0939ac 1.82 \u00b1 0.01ac 2.16 \u00b1 0.0381ab 2.32 \u00b1 0.00373f 2.19 \u00b1 0.0228b 1.89 \u00b1 0.0432bc 2.51 \u00b1 0.0339b 2.67 \u00b1 0.132a PL7 1.64 \u00b1 0.0475acd 1.93 \u00b1 0.0516ab 2.07 \u00b1 0.0543ab 2.5 \u00b1 0.0448e 1.44 \u00b1 0.0114d 1.94 \u00b1 0.0153ab 1.68 \u00b1 0.00697ef 2.73 \u00b1 0.0396a 165  Values are means \u00b1 SEM, n = 3 per treatment group.  Means in a column without a common superscript letter differ (p < 0.05) as analyzed by one-way ANOVA and the LSD test.PL8 1.49 \u00b1 0.099ae 1.82 \u00b1 0.123ac 2.03 \u00b1 0.129b 2.52 \u00b1 0.0463e 1.34 \u00b1 0.0414de 1.78 \u00b1 0.0221cd 1.38 \u00b1 0.0542h 2.68 \u00b1 0.0304a PL9 1.63 \u00b1 0.12acd 1.88 \u00b1 0.0701ac 1.72 \u00b1 0.0303d 2.64 \u00b1 0.0471bc 1.14 \u00b1 0.0126fh 1.8 \u00b1 0.0248bd 1.45 \u00b1 0.0163gh 2.68 \u00b1 0.0244a PL10 1.44 \u00b1 0.0493bce 1.75 \u00b1 0.0481bcd 1.65 \u00b1 0.032de 2.64 \u00b1 0.0369bd 1.29 \u00b1 0.0872df 1.63 \u00b1 0.0965ef 1.45 \u00b1 0.039gh 2.49 \u00b1 0.103b PL11 1.7 \u00b1 0.065ac 1.96 \u00b1 0.0815ab 2.23 \u00b1 0.0149a 2.55 \u00b1 0.0576cde 2.44 \u00b1 0.041a 2.07 \u00b1 0.0356a 2.76 \u00b1 0.0401a 2.7 \u00b1 0.0391a PL12 1.78 \u00b1 0.0465ab 1.68 \u00b1 0.0743ce 2.13 \u00b1 0.0218ab 2.51 \u00b1 0.038e 1.65 \u00b1 0.113c 1.93 \u00b1 0.0428b 2.16 \u00b1 0.0167c 2.64 \u00b1 0.0177ab PL13 1.84 \u00b1 0.126a 1.85 \u00b1 0.0888ac 1.99 \u00b1 0.0727bc 2.52 \u00b1 0.035de 1.46 \u00b1 0.063cd 1.93 \u00b1 0.04b 1.99 \u00b1 0.0727d 2.66 \u00b1 0.0366a PL14 1.81 \u00b1 0.208a 1.99 \u00b1 0.0536a 1.82 \u00b1 0.1cd 2.27 \u00b1 0.00612f 1.41 \u00b1 0.0418d 1.88 \u00b1 0.073bc 1.82 \u00b1 0.1e 2.61 \u00b1 0.104ab PL15 1.59 \u00b1 0.0589acd 1.82 \u00b1 0.0309ac 1.73 \u00b1 0.0923d 2.27 \u00b1 0.024f 1.44 \u00b1 0.0211d 1.82 \u00b1 0.0371bd 1.71 \u00b1 0.0813e 2.65 \u00b1 0.0872ab Table 6.6: Continued 166  6.4.3 Effect on vitamin C content The vitamin C content of milk samples are presented in Figures 6.1 and 6.2 for AT and CT reactors, respectively. The vitamin C content in 3.25, 2, 1 and 0% fat milk were 22.10 \u00b1 1.160, 20.54 \u00b1 0.232, 21.21 \u00b1 0.462, and 21.49 \u00b1 0.426 mg\/L, which were significantly (p < 0.05) higher than reported values in literature [184,185]. The obtained value of vitamin C content in HTST processed milk of 3.25, 2, 1 and 0% fat were 10.78 \u00b1 0.174, 10.55 \u00b1 0.087, 10.42 \u00b1 0.290, and 10.21 \u00b1 0.174 mg\/L, that were significantly (p < 0.05) lower than control. The reduction in vitamin C after heat treatment was almost 50%. The reduction of vitamin C in milk due to the heat sensitivity of vitamin as it quickly oxidizes, thereby reducing the vitamin C content [178]. The vitamin C content of PL-treated milk decreased with the applied fluence. The effect of frequency and flow rate was statistically significant (p < 0.05), however, the interaction between them was significant only for milk of 0% fat. Also, the loss in vitamin C was greater in CT reactor treated samples than AT reactor treated ones, owing to greater residence time in the CT reactors. The effect of fat% was not significant for AT reactor while it was significant for CT reactor, with 0% fat milk showing a significantly higher loss of vitamin C than 3.25% fat milk treated by PL. The loss in vitamin C was almost same (52%) as HTST sample for 0% fat milk treated in CT reactor. The loss in vitamin C of milk was also observed by Guneser and Yuceer [178], which could be ascribed to the light-sensitive behavior of vitamin C. Intense PL treatments at higher fluences thus prove to be disadvantageous as they drastically lower the vitamin C content of milk and should be avoided.  (a) 5791113151719212314.3 32.5 49.6 64.1 74.9Vitamin C [mg\/L]Flow rate [L\/h]AT Reactor 1 Hz 3 Hz 5 Hz+^&#**^***167   (b)  (c) 5791113151719212314.3 32.5 49.6 64.1 74.9Vitamin C [mg\/L]Flow rate [L\/h]AT Reactor 1 Hz 3 Hz 5 Hz* **+^&#%**5791113151719212314.3 32.5 49.6 64.1 74.9Vitamin C [mg\/L]Flow rate [L\/h]AT Reactor 1 Hz 3 Hz 5 Hz**+^&#%168   (d) Figure 6.1: Effect of PL treatment on vitamin C content of milk samples of different fat% in AT reactor. (a) 3.25%, (b) 2%, (c) 1%, and (d) 0%. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols  (a) 5791113151719212314.3 32.5 49.6 64.1 74.9Vitamin C [mg\/L]Flow rate [L\/h]AT Reactor 1 Hz 3 Hz 5 Hz***+^&#%**579111315171921232514.3 32.5 49.6 64.1 74.9Vitamin C [mg\/L]Flow rate [L\/h]CT Reactor 1 Hz 3 Hz 5 Hz+^&#** **^*169   (b)  (c) 579111315171921232514.3 32.5 49.6 64.1 74.9Vitamin C [mg\/L]Flow rate [L\/h]CT Reactor 1 Hz 3 Hz 5 Hz+^&#* ** **#579111315171921232514.3 32.5 49.6 64.1 74.9Vitamin C [mg\/L]Flow rate [L\/h]CT Reactor 1 Hz 3 Hz 5 Hz+^&* ****++170   (d) Figure 6.2: Effect of PL treatment on vitamin C content of milk samples of different fat% in CT reactor. (a) 3.25%, (b) 2%, (c) 1%, and (d) 0%. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols 6.4.4 Effect on vitamin B2 content The vitamin B2 content of milk samples are presented in Figures 6.3 and 6.4 for AT and CT reactors, respectively. The vitamin B2 content in 3.25, 2, 1 and 0% fat milk were 0.417 \u00b1 0.002, 0.403 \u00b1 0.006, 0.397 \u00b1 0.038, and 0.381 \u00b1 0.023 mg\/L, which were significantly (p < 0.05) lower than reported values in literature [186]. The obtained value of vitamin B2 content in HTST processed milk of 3.25, 2, 1 and 0% fat were 0.402 \u00b1 0.006, 0.387 \u00b1 0.008, 0.375 \u00b1 0.007, and 0.359 \u00b1 0.013 mg\/L, respectively, which were significantly (p < 0.05) lower than control. The vitamin B2 in milk after heat treatment was not greatly affected and more than 94% was still retained. The heat stability of vitamin B2 was shown earlier in literature [183]. The vitamin B2 content of PL-treated milk decreased with the applied fluence. The effect of frequency and flow rate was statistically significant (p < 0.05). Also, the loss in vitamin B2 was greater in CT reactor treated samples than AT reactor treated ones, owing to greater residence time in the CT reactors. 579111315171921232514.3 32.5 49.6 64.1 74.9Vitamin C [mg\/L]Flow rate [L\/h]CT Reactor 1 Hz 3 Hz 5 Hz+^&#*****+171  The effect of fat% was significant for both AT and CT reactor, with no clear trend on the effect of fat%. Overall, the retention of vitamin B2 in milk after PL treatment was almost 43%, which was significantly lower than HTST sample (>94% retention). The loss in vitamin B2 of milk was also observed by Sheraz et al. [183], which could be ascribed to the light-sensitive behavior of vitamin B2. The loss in vitamin B2 by photodegradation under UV light has been studied in detail. Riboflavin absorbs light of wavelength 223, 267, 373 and 444 nm in UV-Vis region and gets degraded into compounds like lumichrome, lumiflavin, formylflavin, carboxymethylflavin and so on by following the photoreduction, photoaddition and photodealkylation pathways [183]. The loss in riboflavin and formation of photoproducts led to the increased blueness of samples as mentioned in the previous section (6.4.2).  (a) 00.050.10.150.20.250.30.350.414.3 32.5 49.6 64.1 74.9Vitamin B2 [mg\/L]Flow rate [L\/h]AT Reactor 1 Hz 3 Hz 5 Hz+^&#*****%172   (b)  (c) 00.050.10.150.20.250.30.350.414.3 32.5 49.6 64.1 74.9Vitamin B2[mg\/L]Flow rate [L\/h]AT Reactor 1 Hz 3 Hz 5 Hz+^&#*****%00.050.10.150.20.250.30.350.414.3 32.5 49.6 64.1 74.9Vitamin B2[mg\/L]Flow rate [L\/h]AT Reactor 1 Hz 3 Hz 5 Hz+^&# *%****173   (d) Figure 6.3: Effect of PL treatment on vitamin B2 content of milk samples of different fat% in AT reactor. (a) 3.25%, (b) 2%, (c) 1%, and (d) 0%. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols  (a) 00.050.10.150.20.250.30.350.414.3 32.5 49.6 64.1 74.9Vitamin B2[mg\/L]Flow rate [L\/h]AT Reactor 1 Hz 3 Hz 5 Hz+^&# *%****00.050.10.150.20.250.30.350.414.3 32.5 49.6 64.1 74.9Vitamin B2[mg\/L]Flow rate [L\/h]CT Reactor 1 Hz 3 Hz 5 Hz+^&*****% #174   (b)  (c) 00.050.10.150.20.250.30.350.414.3 32.5 49.6 64.1 74.9Vitamin B2[mg\/L]Flow rate [L\/h]CT Reactor 1 Hz 3 Hz 5 Hz+^&*# * ***%00.050.10.150.20.250.30.350.414.3 32.5 49.6 64.1 74.9Vitamin B2[mg\/L]Flow rate [L\/h]CT Reactor 1 Hz 3 Hz 5 Hz+^** **^#&175   (d) Figure 6.4: Effect of PL treatment on vitamin B2 content of milk samples of different fat% in CT reactor. (a) 3.25%, (b) 2%, (c) 1%, and (d) 0%. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols 6.4.5 Lipid oxidation The lipid oxidation for the milk samples of 3.25, 2 and 1% fat treated in AT reactor and CT reactor are shown in Figure 6.5 (a)-(c) and 6.6 (a)-(c). The lipid oxidation was quantified in terms of formation of primary lipid oxidation products such as hydroperoxides which form from the peroxide free radicals that react with unsaturated fatty acid molecules. These hydroperoxides eventually decompose and form aldehydes and ketones (secondary oxidation products) [181]. The base values of lipid oxidation in terms of absorbance units [-] for 3.25, 2 and 1% milk were 1.179 \u00b1 0.022, 1.082 \u00b1 0.114, 1.043 \u00b1 0.013 [-], respectively which decreased as the fat content of milk decreased. HTST processing of milk increased the lipid oxidation of milk and values of 1.306 \u00b1 0.034, 1.213 \u00b1 0.014, 1.126 \u00b1 0.024 [-] were obtained for 3.25, 2 and 1% milk. The lipid oxidation was significant (p < 0.05) for milk of fat% 3.25 and 2%, treated in AT reactor, whereby, a higher fluence value showed higher lipid oxidation.  The lipid oxidation was significantly lower (p < 0.05) 00.050.10.150.20.250.30.350.414.3 32.5 49.6 64.1 74.9Vitamin B2[mg\/L]Flow rate [L\/h]CT Reactor 1 Hz 3 Hz 5 Hz+^& ** ***%#176  in case of 1% fat milk than other fat% milk. Similar trend was observed in case of CT reactor; however, the lipid oxidation in case of CT reactor was significantly (p < 0.05) higher than AT reactor. Milk samples after PL treatment exhibited a change in aroma like metallic, barn or cardboard type, which could be attributed to the formation of lipid oxidation product (E,E)\u22122,4-nonadienal that gives off a barn-like aroma in dried milk [187]. Singlet oxygen generated during photo-chemical reactions also react with lipid molecules to form cascade of oxidation products with off-flavours. Measurement of aroma and sensory evaluation seem an interesting avenue for research in the future.  (a)  (b) 00.511.522.5314.3 32.5 49.6 64.1 74.9Absorbance [-]Flow rate [L\/h]AT Reactor 1 Hz 3 Hz 5 Hz+*%^#&*00.511.522.5314.3 32.5 49.6 64.1 74.9Absorbance [-]Flow rate [L\/h]AT Reactor 1 Hz 3 Hz 5 Hz+***^^&**%177   (c) Figure 6.5: Effect of PL treatment on lipid oxidation content of milk samples of different fat% in AT reactor. (a) 3.25%, (b) 2%, (c) 1%. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols  (a) 00.511.522.5314.3 32.5 49.6 64.1 74.9Absorbance [-]Flow rate [L\/h]AT Reactor 1 Hz 3 Hz 5 Hz+*** ^#^&*00.511.522.5314.3 32.5 49.6 64.1 74.9Absorbance [-]Flow rate [L\/h]CT Reactor 1 Hz 3 Hz 5 Hz+***^# &%178   (b)  (c) Figure 6.6: Effect of PL treatment on lipid oxidation content of milk samples of different fat% in CT reactor. (a) 3.25%, (b) 2%, (c) 1%. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols 00.511.522.5314.3 32.5 49.6 64.1 74.9Absorbance [-]Flow rate [L\/h]CT Reactor 1 Hz 3 Hz 5 Hz+***^#&*%00.511.522.5314.3 32.5 49.6 64.1 74.9Absorbance [-]Flow rate [L\/h]CT Reactor 1 Hz 3 Hz 5 Hz++^#&***179  6.4.6 Protein oxidation The protein oxidation for the milk samples of 3.25, 2, 1 and 0% fat treated in AT reactor and CT reactor are shown in Figure 6.7 (a)-(d) and 6.8 (a)-(d). The protein oxidation was quantified in terms of formation of protein carbonyls which are generated due to the changes in individual amino acids. The base values of protein oxidation in terms of protein carbonyl formation were for 8.956 \u00b1 0.393, 9.194 \u00b1 0.950, 9.290 \u00b1 0.577, and 9.344 \u00b1 1.878 nM Protein Carbonyls\/mg protein for 3.25, 2, 1 and 0% milk, respectively which increased as the fat content of milk decreased. HTST processing of milk increased the protein oxidation of milk and values of 16.85 \u00b1 2.373, 17.24 \u00b1 1.948, 17.84 \u00b1 0.532, and 18.27 \u00b1 0.63 nM Protein Carbonyls\/mg protein were obtained for 3.25, 2, 1 and 0% milk, respectively. The formation of protein carbonyls in PL-treated milk samples increased significantly (p < 0.05) as the PL fluence increased. The effect of fat% was also significant (p < 0.05) and as the fat% decreased the protein oxidation was higher. Similar observations were made by Scheidegger et al. [177], whereby skim milk showed higher protein carbonyl formation compared to whole milk after UV light exposure. This suggests that lipids in milk offer stability to the photooxidation of proteins [177]. Also, formation of protein carbonyls was significantly (p < 0.05) higher in CT reactor treated samples compared to AT reactor. Protein carbonyls formation in PL-treated milk is suggested to be related to the oxidation of amino acids tryptophan, histidine, and methionine. These protein carbonyls are also related to off-flavour generation in UV-C irradiated milk samples [177].  0510152025303540455014.3 32.5 49.6 64.1 74.9nM Protein Carbonyls\/mg proteinFlow rate [L\/h]AT Reactor 1 Hz 3 Hz 5 Hz+**^# && ****180  (a)  (b)  (c) 0510152025303540455014.3 32.5 49.6 64.1 74.9nM Protein Carbonyls\/mg proteinFlow rate [L\/h]AT Reactor 1 Hz 3 Hz 5 Hz+**^#&^**0510152025303540455014.3 32.5 49.6 64.1 74.9nM Protein Carbonyls\/mg proteinFlow rate [L\/h]AT Reactor 1 Hz 3 Hz 5 Hz+* *^#%&**181   (d) Figure 6.7: Effect of PL treatment on protein oxidation content of milk samples of different fat% in AT reactor. (a) 3.25%, (b) 2%, (c) 1%, and (d) 0%. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols  (a) 0510152025303540455014.3 32.5 49.6 64.1 74.9nM Protein Carbonyls\/mg proteinFlow rate [L\/h]AT Reactor 1 Hz 3 Hz 5 Hz+** ^#%&***010203040506014.3 32.5 49.6 64.1 74.9nM Protein Carbonyls\/mg proteinFlow rate [L\/h]CT Reactor 1 Hz 3 Hz 5 Hz+**^#%**&182   (b)  (c) 010203040506014.3 32.5 49.6 64.1 74.9nM Protein Carbonyls\/mg proteinFlow rate [L\/h]CT Reactor 1 Hz 3 Hz 5 Hz+** ^#%***010203040506014.3 32.5 49.6 64.1 74.9nM Protein Carbonyls\/mg proteinFlow rate [L\/h]CT Reactor 1 Hz 3 Hz 5 Hz+**^#&***%183   (d) Figure 6.8: Effect of PL treatment on protein oxidation content of milk samples of different fat% in CT reactor. (a) 3.25%, (b) 2%, (c) 1%, and (d) 0%. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols 6.4.7 Multivariate analysis   The multivariate data analysis for processing of milk were carried out using principal component analysis and datasets were transformed into two mutually orthogonal principal components (PC1 and PC2). The Figures 6.9 (a) and (b) represent the rotated biplot for 3.25% milk treated in AT and CT reactor, respectively. 01020304050607014.3 32.5 49.6 64.1 74.9nM Protein Carbonyls\/mg proteinFlow rate [L\/h]CT Reactor 1 Hz 3 Hz 5 Hz+**^#%**&184   (a)  185   (b) Figure 6.9: Rotated biplot for the principal components for various analyses carried out for milk (3.25%) in (a) AT and (b) CT reactor In Figure 6.9 (a), described for PL-treated milk in AT reactor, the components PC1 and PC2 can describe 49.99 and 30.22% of variabilities in the dataset, respectively, with overall description of more than 80.21% of variability. The biplot responses such as vitamin C, vitamin B2, lipid and protein oxidation, pH, L*, a*, b* etc. denote the place they occupy in the quadrants. The treatments (PL5, and PL15) fall on the same quadrant as vitamin C when they are correlated, meaning these treatments show higher vitamin C retention. Vitamin B2 and \u2206\ud835\udc38 fall in a quadrant opposite which lie PL6, PL11, PL12, PL13 which are negatively correlated meaning they affect vitamin B2 negatively and positively affect the colour change. As can be seen, the treatments PL2, PL9, PL10 lie closer to responses E. coli, L. innocua and C. sporogenes inactivation, meaning their effect was minimal in inactivation. On the other hand, the treatments, having higher fluences (PL6, PL7, PL11, PL13, PL14, PL15 lie on the adjacent quadrant meaning their effect on inactivation was higher. Treatments PL6, PL7, PL11 lie closer to lipid and protein oxidation, meaning their effect 186  on these parameters were prominent. Control was also far from other treatments, which means the control was not similar in terms of affecting the parameters studied. Control was also in the same quadrant as vitamin C, meaning it shows the maximum  level of vitamin C. As HT sample showed highest inactivation of microorganisms, it lies opposite to the place occupied by the test microorganisms. Vitamin B2 and b* were highly correlated. \u2206\ud835\udc38 was negatively correlated vitamin B2 and b* values. This suggests that high vitamin B2 retention is obtained when colour difference is lower. In Figure 6.9 (b), described for PL-treated milk in CT reactor, the components PC1 and PC2 describe more than 85.69% of the variability in the dataset with PC1 describing 53.23% and PC2 describing 32.47% of the variability. The treatments (PL2, PL5, and PL9) fall on the same quadrant as vitamin C when they are correlated, meaning these treatments show higher vitamin C retention. Vitamin B2 and \u2206\ud835\udc38 fall in a quadrant opposite which lie PL6, PL11, which are negatively correlated meaning they affect vitamin B2 negatively. As can be seen, the treatments PL1, PL2, PL3, PL7, PL8 lie closer to responses E. coli, L. innocua and C. sporogenes inactivation, meaning their effect was minimal in inactivation. On the other hand, the treatments, having higher fluences (PL11, PL13, PL14, PL15 lie on the opposite quadrant meaning their effect on inactivation was higher. Treatments PL6, PL11 lie closer to lipid and protein oxidation, meaning their effect on these parameters were prominent. Control was also far from other treatments, which means the control was not similar in terms of affecting the parameters studied. Control was also in the same quadrant as vitamin C, meaning it shows the maximum  level of vitamin C. HT sample was also far from all other samples, meaning it affected the measured parameters differently that the PL treatments. As HT sample showed highest inactivation of microorganisms, it lies opposite to the place occupied by the test microorganisms. Vitamin B2 and b* were highly correlated. \u2206\ud835\udc38 was negatively correlated vitamin B2 and b* values. This suggests that high vitamin B2 retention is obtained when total colour difference is lower. 6.5 Conclusion These results show that milk\u2019s quality and nutritional parameters were affected after PL treatment. These were indicated in terms of changes in pH, colour (a*, b*), vitamin C and B2 content, lipid and protein oxidation. All the parameters were significantly affected by PL processing. However, these changes were lesser than the changes imparted by HTST processing. During HTST 187  processing, colour was slightly redder, while vitamin C was also greatly affected. PCA analysis revealed that the analyzed parameters were hugely dependent on the PL processing parameters. Since, the quality and nutritional parameters were affected after PL treatment, this necessitates drawing a balance between microbial inactivation (beneficial) and quality\/nutritional changes (undesirable) in milk.  188  Chapter 7: Effect of pulsed UV light treatment of physico-chemical and nutritional parameters of red grape and watermelon juice 7.1 Summary In this chapter, the effects of PL processing of the quality and nutritional parameters of red grape and watermelon juice are mentioned. For, PL-treated juices, pH changed slightly for both type of juices. For grape juice, the L* increased for the samples and a* decreased significantly (p < 0.05). Also, anthocyanins in juice samples decreased significantly (p < 0.05). In watermelon juice samples, the L* and b* did not vary greatly, while a* values decreased significantly (p < 0.05). The lycopene content was also observed to decrease in the watermelon juice samples. In grape juice samples, trans-resveratrol content also decreased significantly (p < 0.05). In watermelon juice, the vitamin C also decreased significantly (p < 0.05). The decrease in TPC and antioxidant capacity was also significant (p < 0.05) in both the type of juice.  7.2 Introduction PL has been shown to inactivate various microorganisms inoculated in these studies in grape juice and watermelon juice. However, it is also important to know how PL affects the nutritional and sensory properties of these juice samples. This will help in gaining knowledge about the acceptability of the PL-treated products. These juices are good source of polyphenol compounds and antioxidants. The susceptibility of these compounds to photodegradation during PL has been explored by Wiktor et al. [122]. It would be interesting to see these compounds are affected by PL. Watermelon juice is a good source of beneficial phytochemicals such as phenolic compounds and antioxidants and vitamin C.  It is also a good source of lycopene, which has been attributed to several health benefits like promotion of cardiovascular health, lowering of low density lipoprotein cholesterol, prevention of neurodegenerative disorders like Alzheimer\u2019s and Parkinson\u2019s disease [188]. Grapes are rich source of stilbenes like trans-resveratrol, which has a positive role in human health as anticarcinogenic agent and promoter of cardiovascular health [189]. However, this compound is mostly present in skin of grapes and very less quantity is present in the expressed juice from grape. Nevertheless, it is important to know how these compounds are affected by PL processing. This will help in adequate designing of the PL process for these juices with an aim to maximize the benefits of microbial inactivation and minimizing undesirable damage to phytochemicals compounds and sensory properties. 189  Therefore, the main objective of this experiment is to ascertain the effects of PL on the pH, colour, levels of phytochemicals and nutritional compounds (anthocyanins, phenolics, antioxidants, vitamin C, trans-resveratrol) in grape and watermelon juices.  7.3 Materials and methods 7.3.1 Materials Red grape juice of Concord variety (Welch Foods Inc., Concord, MA, United States) and watermelon juice (Simply Orange Juice Company, Apopka, FL, United States) were purchased from local grocery store. The thermophysical and optical properties of the juice samples were calculated and shown in Table 5.1. Juices were stored at refrigerated temperature (4 oC) in dark to reduce exposure to light and treated with pulsed light after bringing the temperature to 25 oC. For chemical analyses, resveratrol, and L-ascorbic acid standard for obtaining standard curves were purchased from Enzo Biochem, Inc. (Farmingdale, NY, United States) and Sigma Aldrich (St. Louis, MO, United States), respectively. Metaphosphoric acid (33% w\/v), 2,2-diphenyl-1-picrylhydrazyl (DPPH) and gallic acid purchased from Alfa Aesar (Mississauga, ON, Canada). HPLC Grade methanol, acetone and sodium acetate were procured from VWR Chemicals (Mississauga, ON, Canada). Folin-Ciocalteau (FC) reagent was purchased from Merck KGaA (Darmstadt, Germany). Potassium chloride, phosphoric acid, Potassium phosphate, and hexane were purchased from Sigma Aldrich (St. Louis, MO, United States). USP grade (95%) ethanol and butylated hydroxy toluene were purchased from Ward\u2019s Science (St. Catherine\u2019s, ON, Canada). Ultrapure water (18 M\u03a9) was used for dilution and making reagents for analyses. 7.3.2 Pulsed UV light treatments PL treatment system consisted of a cylindrical xenon flashlamp emitting high-intensity light pulses in UV-Visible-IR region (~200-1100nm), provided by Solaris Disinfection Inc. (Mississauga, ON, Canada). The lamp dimensions were: length, \ud835\udc3f = 62 cm; diameter, \ud835\udc51\ud835\udc3f= 9 mm. The lamp emitted 30 J of energy per pulse. Considering a pulse width of 100 ms (since light fluctuation due to pulsation was hardly noticeable beyond 9 pulses\/s), the lamp sleeves emitted energy intensity per pulse of 17122.12 mW\/cm2. Treatments were carried out in both the type of reactor configurations, viz. AT and CT reactors. Juice samples in volumes 100 mL and 400 mL for AT and CT reactor were transferred into a stainless steel container with lid. Juice was pumped through the reactors using a peristaltic pump (Masterflex L\/S model 7554-90, Cole-Parmer Instruments, IL, United 190  States) with a variable speed drive. Flow rates of  14.32, 32.48, 49.63, 64.13 and 74.90 L\/h were used and when the juice reached the inlet, the PL system was flashed (pulse frequency used were 1,3 and 5 Hz) until all the juice was collected from outlet in a bottled lined with aluminum foil. There was a total of 15 PL treatments (PL1-PL15) and control juice sample. The energy per pulse at distances 1.5 and 2 cm (for CT and AT rectors, respectively) from lamp were measured along the length of the lamp to calculate average energy per pulse, using a pyroelectric head sensor (PE80BF-DIF-C, Ophir-Spiricon LLC, UT, United states), with Nova II display (Ophir-Spiricon LLC, UT, United states) placed under the lamp. The average energies were 53 and 43 mJ\/cm2. All the treatments were carried out in triplicates. After treatment, the juice samples were kept refrigerated (4 oC) until analyses. The experimental design is shown in the Table 7.1. Table 7.1: Experimental design of the PL treatment of red grape and watermelon juice Pulse frequency [Hz] Flow rate [L\/h] Treatment Fluence [J\/cm2]a Reynolds number (\ud835\udc75\ud835\udc79\ud835\udc86) Red grape juice Watermelon juice AT CT AT CT AT CT 1 14.3 PL1 0.703 4.399 22.5 205.2 35.4 322.5 32.5 PL2 0.309 1.573 51.1 465.4 80.3 731.6 49.6 PL3 0.203 1.029 78.1 711.2 122.7 1118.1 64.1 PL4 0.157 0.797 100.9 918.9 158.6 1444.6 74.9 PL5 0.135 0.683 117.8 1073.3 185.2 1687.3 3 14.3 PL6 2.109 10.708 22.5 205.2 35.4 322.5 32.5 PL7 0.929 4.719 51.1 465.4 80.3 731.6 49.6 PL8 0.609 3.089 78.1 711.2 122.7 1118.1 64.1 PL9 0.471 2.390 100.9 918.9 158.6 1444.6 74.9 PL10 0.403 2.046 117.8 1073.3 185.2 1687.3 5 14.3 PL11 3.515 17.846 22.5 205.2 35.4 322.5 32.5 PL12 1.549 7.866 51.1 465.4 80.3 731.6 49.6 PL13 1.014 5.149 78.1 711.2 122.7 1118.1 64.1 PL14 0.785 3.984 100.9 918.9 158.6 1444.6 74.9 PL15 0.672 3.410 117.8 1073.3 185.2 1687.3 191  aTotal fluence (\ud835\udc39\ud835\udc5c) was calculated as Average energy per pulse \u00d7 reactor residence time (\ud835\udc61\ud835\udc5f\ud835\udc52\ud835\udc60,\ud835\udc61\u210e) 7.3.3 Analyses 7.3.3.1 pH measurements The pH measurements were carried out in a benchtop pH meter (Accumet AE150, Saint-Laurent, Quebec, Canada), which was previously calibrated using calibration standard buffers (pH 4.0, 7.0 and 10.0). Small aliquot of sample (5 mL) was taken in a small beaker and the pH meter probe was dipped into it. The pH was recorded when the reading stabilized. The measurements were carried out in triplicates. 7.3.3.2 Colour measurements The colour of the control juice samples before treatment and the PL-treated was measured using a benchtop colorimeter (HunterLab, model LabScanTM XE Plus, Hunter Associates Laboratory, Reston, VA, USA). Colour of the samples was recorded as the CIE L*a*b* tristimulus colour coordinates parameters (L*\u2013Lightness, a*\u2013red\/green, and b*\u2013blue\/yellow). From the colour parameters, the total colour difference (\u2206\ud835\udc38) was calculated for all samples using the Eq. (7.1) [132]. All the measurements were carried out in triplicates. \u2206\ud835\udc38 = \u221a(\ud835\udc3f\u2217 \u2212 \ud835\udc3f\ud835\udc5c\u2217 )2 + (\ud835\udc4e\u2217 \u2212 \ud835\udc4e\ud835\udc5c\u2217)2 + (\ud835\udc4f\u2217 \u2212 \ud835\udc4f\ud835\udc5c\u2217)2                                     (7.1) where, \ud835\udc3f\ud835\udc5c\u2217 , \ud835\udc4e\ud835\udc5c\u2217  and \ud835\udc4f\ud835\udc5c\u2217 are the colour parameters of the control sample. 7.3.3.3 Vitamin C determination (Watermelon juice) Vitamin C content in watermelon juice samples was determined by HPLC following the method of Patras et al. [190]. Twenty \u03bcL samples were injected onto a Zorbax SB-C18 column with HPLC system (Agilent 1100 system, Agilent Technologies, Santa Clara, CA, USA). Juice samples (25 mL) were pipetted into centrifuge tubes (50 mL) added with 5 mL of 2.5% (w\/v) metaphosphoric acid. Sample tubes were centrifuged at 2000  rpm for 10 min at 4 \u00b0C (Thermo Scientific Sorvall, UK). Around 1.5 mL of the supernatant was passed through polyether sulfone (PES) syringe filters (0.22 \u03bcm) and placed in an autosampler vial. 25 mM KH2PO4 (adjusted to pH 3.0 with phosphoric acid) and methanol at 80:20 ratio was used as the mobile phase at a flow rate of 1 mL\/min. Eluate was monitored and detected by diode array detector (DAD) at 245 nm. Chromatograms were analyzed with ChemStation Software version B.04.03. Standard curve was obtained by using standard L-ascorbic acid solutions (1-100 mg\/L). Results were reported as mg\/L of watermelon juice. All measurements were carried in triplicate. Figure A3-3 shows chromatogram for 192  determination of vitamin C content in watermelon juice sample. Peak is shown at retention time of around 2 min. 7.3.3.4 trans-Resveratrol content determination (red grape juice) The trans-Resveratrol content in red grape juice samples was determined by following the method of Hasan et al. [191]. The grape juices (1.5 mL) were filtered through a 0.22 \u03bcm PES syringe filters and collected into HPLC vials. 20 \u03bcL of filtered samples were injected into HPLC system Agilent 1100 system, Agilent Technologies, Santa Clara, CA, USA) with a Zorbax SB-C18 column and a DAD UV detector at 320 nm. The mobile phase included\u2014methanol:20 mM phosphoric acid (20:80) and methanol:20 mM phosphoric acid (80:20) at 70:30 (v\/v) ratio with a flow rate at 1 mL\/min. Standard  curve was obtained by using standard trans-Resveratrol solutions (1-10 mg\/L) to determine the regression equation. Results were reported as mg\/L of grape juice. All measurements were carried in triplicate. Figure A3-2 shows the standard curve for trans-Resveratrol standard solution. Figure A3-4 and A3-5 show chromatogram for determination of trans-Resveratrol content in trans-Resveratrol standard solution and grape juice sample, respectively. Peak is shown at retention time of around 17 min. 7.3.3.5 Total phenolics content determination Total phenolics content (TPC) in the juice samples was determined by following the method of Derradji-Benmeziane et al., [192]. In brief, 5 mL juice and mixed with 10 mL of 80% (v\/v) ethanol and centrifuged for 15 min at 3000 rpm. For TPC assay, 20 \u03bcL of centrifuged sample were taken in 96-well microplates and added with 100 \u03bcL of 10% (v\/v) FC reagent and 80 \u03bcL of 7.5 % w\/v of NaCO3. The plates were incubated for 30 min in the dark. After incubation, the plate absorbances were measured at 765 nm in a microplate spectrophotometer (Infinite Pro M200 series, TecanTM, M\u00e4nnedorf, Switzerland). For blank, distilled water was added. Standard for gallic acid solution (0-0.1 mg\/mL) were read for absorbances and the standard curve was developed to get the regression equation. Sample absorbance values were converted into TPC in terms of mg GAE\/100 mL juice. Figure A3-1 shows the standard curve for gallic acid standard solutions. 7.3.3.6 Antioxidant activity determination Antioxidant activity in juice samples was determined by following the method of Nile et al., [193]. Standard curve was obtained using \u00b5M Trolox standard (0-25 \u03bcL) and 20 \u03bcL of 1mM DPPH solution in methanol in a 96-well microplate, by filling the wells with methanol up to 200 \u03bcL. 193  Same extracts as TPC analyses were taken (0-50 \u03bcL) in the microplates and filled with DPPH and methanol. For blank, 200 \u00b5L methanol was taken. The plates were covered with aluminum foil and incubated at room temperature for 10 minutes. Absorbances were read at 519 nm in a microplate spectrophotometer (Infinite Pro M200 series, TecanTM, M\u00e4nnedorf, Switzerland) and blank corrected. The inhibition (%) of sample\/Trolox was calculated using the equation 7.2. % \ud835\udc3c\ud835\udc5b\u210e\ud835\udc56\ud835\udc4f\ud835\udc56\ud835\udc61\ud835\udc56\ud835\udc5c\ud835\udc5b =  [1 \u2013 (\ud835\udc34\ud835\udc4f\ud835\udc60\ud835\udc5c\ud835\udc5f\ud835\udc4f\ud835\udc4e\ud835\udc5b\ud835\udc50\ud835\udc52 \ud835\udc5c\ud835\udc53 \ud835\udc60\ud835\udc4e\ud835\udc5a\ud835\udc5d\ud835\udc59\ud835\udc52 \ud835\udc5c\ud835\udc5f \ud835\udc47\ud835\udc5f\ud835\udc5c\ud835\udc59\ud835\udc5c\ud835\udc65\ud835\udc34\ud835\udc4f\ud835\udc60\ud835\udc5c\ud835\udc5f\ud835\udc4f\ud835\udc4e\ud835\udc5b\ud835\udc50\ud835\udc52 \ud835\udc5c\ud835\udc53 \ud835\udc50\ud835\udc5c\ud835\udc5b\ud835\udc61\ud835\udc5f\ud835\udc5c\ud835\udc59)]  \u00d7  100                        (7.2) Where, control = sample blank; Inhibition (%) versus concentration of Trolox or sample were plotted and linear regression was used to find the slope of the line. Sample antioxidant activity was  expressed as Trolox Equivalent Antioxidant Capacity (\ud835\udc47\ud835\udc38\ud835\udc34\ud835\udc36) as per equation 7.3. \ud835\udc47\ud835\udc38\ud835\udc34\ud835\udc36 (\ud835\udf07\ud835\udc3f \ud835\udc47\ud835\udc5f\ud835\udc5c\ud835\udc59\ud835\udc5c\ud835\udc65\ud835\udf07\ud835\udc3f \ud835\udc57\ud835\udc62\ud835\udc56\ud835\udc50\ud835\udc52) =\ud835\udc60\ud835\udc59\ud835\udc5c\ud835\udc5d\ud835\udc52 \ud835\udc60\ud835\udc4e\ud835\udc5a\ud835\udc5d\ud835\udc59\ud835\udc52\ud835\udc60\ud835\udc59\ud835\udc5c\ud835\udc5d\ud835\udc52 \ud835\udc47\ud835\udc5f\ud835\udc5c\ud835\udc59\ud835\udc5c\ud835\udc65                                          (7.3) 7.3.3.7 Anthocyanin content determination (Red grape juice) Total monomeric anthocyanin content (TAC) in the juice samples was determined by following the method of Nile et al. [193]. For the assays, 1 mL of Juice and 4 mL of pH adjustment buffer (pH 1.0 by 0.025 M KCl; pH 4.5 by 0.4 M Sodium acetate; pH was adjusted with phosphoric acid). The sample absorbances were measured at 520 nm  and 700 nm using a microplate spectrophotometer (Infinite Pro M200 series, TecanTM, M\u00e4nnedorf, Switzerland). The results of TAC can be expressed as mg cyanidin-3-glucoside\/ L sample is given by: TAC (mg\/L)  =\ud835\udc34\u00d7\ud835\udc40\ud835\udc4a\u00d7\ud835\udc37\ud835\udc39\u00d7103\ud835\udf00\u00d7\ud835\udc59                                           (7.4) Where, \ud835\udc34 = [(\ud835\udc34520\ud835\udc5b\ud835\udc5a \u2013  \ud835\udc34700\ud835\udc5b\ud835\udc5a)\ud835\udc5d\ud835\udc3b 1.0 \u2013 (\ud835\udc34520\ud835\udc5b\ud835\udc5a \u2013 \ud835\udc34700\ud835\udc5b\ud835\udc5a)\ud835\udc5d\ud835\udc3b 4.5]; \ud835\udc40\ud835\udc4a= molecular weight for cyanidin-3-glucoside (cyd-3-glu), [449.2 g\/mol]; DF = dilution factor established in D; \ud835\udc59 = pathlength, [cm]; \ud835\udf00 =  molar extinction coefficient for cyd-3-glu, [26,900 L\/(mol-cm)]; and 103 = conversion factor for mg from g. 7.3.3.8 Lycopene content determination (watermelon juice) Lycopene content in watermelon samples was determined spectrophotometrically by following the method of Olms-Oliu et al. [194]. In brief, 0.6 g of watermelon juice was taken in centrifuge tubes and added to 5 mL of 0.05% (w\/v) solution of butylated hydroxytoluene (BHT) in acetone, 10 mL of hexane and5 mL of  ethanol (USP-grade 95%). The mixture was centrifuged at 320g for 15 min at 4 \u00b0C. Thereafter, 3 mL of distilled water was added to the tubes and the mixture was allowed to rest for phase separation. The upper hexane layer was measured for absorbance using a microplate 194  spectrophotometer (Infinite Pro M200 series, TecanTM, M\u00e4nnedorf, Switzerland) at 503 nm. The blank well was filled hexane. The lycopene content of juice sample was expressed as mg\/100 mL juice and calculated as: \ud835\udc3f\ud835\udc66\ud835\udc50\ud835\udc5c\ud835\udc5d\ud835\udc52\ud835\udc5b\ud835\udc52 \ud835\udc50\ud835\udc5c\ud835\udc5b\ud835\udc61\ud835\udc52\ud835\udc5b\ud835\udc61 (\ud835\udc5a\ud835\udc54100\ud835\udc5a\ud835\udc3f) =\u2206503\u00d7\ud835\udc40\ud835\udc4a\u00d7\ud835\udc37\ud835\udc39\u00d71000\ud835\udf00\u00d7\ud835\udc3f                              (7.5) Where, \ud835\udc40\ud835\udc4a= molecular weight of lycopene, [536.9 g\/mol]; DF = dilution factor; \ud835\udc59 = pathlength, [cm]; \ud835\udf00 =  molar extinction coefficient for lycopene, [172,000 L\/(mol-cm)]. 7.3.4 Statistical analysis The results were expressed as mean \u00b1 standard deviation. The results were analyzed using two-way ANOVA and Fisher\u2019s least significant difference (LSD) was carried out as post-hoc test to determine if there was a significant difference (p < 0.05) between treatment means. The statistical analyses were carried out using SPSS software version 27 (IBM\u00ae Corp., Armonk, NY, USA) and Microsoft Excel (Microsoft Corporation, Redmond, WA, USA) with XLSTAT package. 7.4 Results and discussion 7.4.1 Effect on pH of juices The pH of the watermelon juices was not affected and the changes in pH were not significant (p > 0.05) after PL treatment as compared to the control sample for both type of reactors (AT and CT) (Table 7.2). This is consistent with the results reported by Feng et al. [195], where UV-C treatment watermelon juice did not change the pH immediately after treatment and until 25 days.  In a study, Mu\u00f1oz et al. [76], observed that the pH level of apple juice was not significantly affected by PL. It is apparent that since the pH of the PL-treated juice was not affected, the overall quality of the juice should be retained. The pH of the red grape juice after PL treatment was slightly changed in AT reactor, though the changes were significantly different (p < 0.05) than control. The results are given in Table 7.3. For CT reactor, the pH of the juices was not significantly different (p > 0.05) than control. The results agree with Wang et al. [128], who observed that the pH of grape juice was not affected after PL treatment in spiral continuous reactor after 90 pulse application. 195  Table 7.2: Effect of PL treatments on pH and colour of watermelon juice Treatments pH L* a* b* \ud835\udf1f\ud835\udc6c  AT CT AT CT AT CT AT CT AT CT Control 5.02 \u00b1 0.00882 5.02 \u00b1 0.00882 42.2 \u00b1 0.0219df 42.2 \u00b1 0.0219 4.18 \u00b1 0.00882a 4.18 \u00b1 0.00882a 3.89 \u00b1 0.0115a 3.89 \u00b1 0.0115a 0 \u00b1 0h 0 \u00b1 0j PL1 5.1 \u00b1 0.0393 5.12 \u00b1 0.0318 42 \u00b1 0.128efg 42.1 \u00b1 0.532 2.85 \u00b1 0.0498g 1.8 \u00b1 0.057h 2.9 \u00b1 0.024fgh 1.9 \u00b1 0.0133ij 1.3 \u00b1 0.0233b 1.79 \u00b1 0.0271bc PL2 5.1 \u00b1 0.0546 5.05 \u00b1 0.0306 42.9 \u00b1 0.196b 42.3 \u00b1 0.314 3.64 \u00b1 0.0643bd 2.55 \u00b1 0.0581cde 2.95 \u00b1 0.139eg 2.3 \u00b1 0.0777gh 1.15 \u00b1 0.0079cd 1.53 \u00b1 0.0176egh PL3 5.08 \u00b1 0.0437 5.08 \u00b1 0.0133 41.4 \u00b1 0.0186h 41.7 \u00b1 0.503 3.5 \u00b1 0.0338cde 2.54 \u00b1 0.0384cde 3.36 \u00b1 0.0561bcd 2.43 \u00b1 0.0426eg 1.07 \u00b1 0.0169de 1.54 \u00b1 0.0286egh PL4 5.08 \u00b1 0.0513 5.12 \u00b1 0.0379 42.8 \u00b1 0.517bc 42.5 \u00b1 0.183 3.65 \u00b1 0.315bd 2.62 \u00b1 0.0348bd 3.52 \u00b1 0.113b 2.81 \u00b1 0.145bc 1.09 \u00b1 0.102de 1.4 \u00b1 0.0183hi PL5 5.07 \u00b1 0.0296 5.05 \u00b1 0.00882 42.2 \u00b1 0.0379df 42.7 \u00b1 0.484 3.9 \u00b1 0.0265ab 2.81 \u00b1 0.059b 3.82 \u00b1 0.0448a 3.02 \u00b1 0.0825b 0.548 \u00b1 0.0281g 1.36 \u00b1 0.0378i PL6 5.09 \u00b1 0.0348 5.12 \u00b1 0.0458 42.1 \u00b1 0.0874df 42.4 \u00b1 0.449 2.69 \u00b1 0.0876gh 1.65 \u00b1 0.148h 2.62 \u00b1 0.0809h 1.65 \u00b1 0.133kl 1.4 \u00b1 0.0187b 1.85 \u00b1 0.0563b PL7 5.12 \u00b1 0.041 5.14 \u00b1 0.0448 42.6 \u00b1 0.0203bd 41.8 \u00b1 0.745 3.03 \u00b1 0.0437fg 2.17 \u00b1 0.0862g 2.77 \u00b1 0.0517gh 2 \u00b1 0.0694ij 1.29 \u00b1 0.0208bc 1.73 \u00b1 0.0178bd PL8 5.1 \u00b1 0.0513 5.13 \u00b1 0.0664 41.9 \u00b1 0.289efg 42.1 \u00b1 0. 732 3.38 \u00b1 0.172df 2.34 \u00b1 0.0623eg 3.11 \u00b1 0.185def 2.34 \u00b1 0.0895fgh 1.12 \u00b1 0.0513d 1.62 \u00b1 0.0412dg 196  PL9 5.08 \u00b1 0.0491 5.06 \u00b1 0.0318 41.4 \u00b1 0.0252h 43.5 \u00b1 0.733 3.55 \u00b1 0.0742bde 2.46 \u00b1 0.0635def 3.21 \u00b1 0.0384ce 2.57 \u00b1 0.0524def 1.11 \u00b1 0.0296d 1.62 \u00b1 0.142def PL10 5.16 \u00b1 0.0593 5.16 \u00b1 0.0233 42.2 \u00b1 0.0291df 41.9 \u00b1 0.722 3.76 \u00b1 0.0681bc 2.75 \u00b1 0.0549bc 3.52 \u00b1 0.118b 2.71 \u00b1 0.0677cd 0.749 \u00b1 0.0816f 1.46 \u00b1 0.0154gi PL11 5.06 \u00b1 0.0273 5.08 \u00b1 0.00882 43.5 \u00b1 0.1a 43.7 \u00b1 1.03 2.44 \u00b1 0.145h 1.35 \u00b1 0.0578i 2.61 \u00b1 0.211h 1.54 \u00b1 0.0961l 1.6 \u00b1 0.0244a 2.03 \u00b1 0.117a PL12 5.12 \u00b1 0.0493 5.15 \u00b1 0.0233 41.9 \u00b1 0.0133efg 42.1 \u00b1 0.336 2.83 \u00b1 0.0839g 1.83 \u00b1 0.0777h 2.73 \u00b1 0.118gh 1.79 \u00b1 0.0669jk 1.34 \u00b1 0.0507b 1.79 \u00b1 0.022bc PL13 5.06 \u00b1 0.0203 5.14 \u00b1 0.0504 41.6 \u00b1 0.0674gh 42.4 \u00b1 0.431 2.84 \u00b1 0.194g 2.14 \u00b1 0.0561g 3.02 \u00b1 0.024eg 2.12 \u00b1 0.0737hi 1.3 \u00b1 0.0702b 1.67 \u00b1 0.0131cde PL14 5.08 \u00b1 0.0458 5.04 \u00b1 0.0153 42.4 \u00b1 0.0321cde 42.6 \u00b1 0.6 3.27 \u00b1 0.142ef 2.3 \u00b1 0.171fg 3.42 \u00b1 0.0814bc 2.46 \u00b1 0.104eg 1.02 \u00b1 0.0592de 1.6 \u00b1 0.00964dg PL15 5.08 \u00b1 0.0233 5.1 \u00b1 0.0219 41.8 \u00b1 0.00577fh 41.8 \u00b1 0.0351 3.47 \u00b1 0.114cde 2.48 \u00b1 0.0902def 3.43 \u00b1 0.0611bc 2.64 \u00b1 0.0569ce 0.962 \u00b1 0.0497e 1.47 \u00b1 0.0341fgi Values are means \u00b1 SEM, n = 3 per treatment group.  Means in a column without a common superscript letter differ (p < 0.05) as analyzed by one-way ANOVA and the LSD test. Table 7.3: Effect of PL treatments on pH and colour of red grape juice Treatments pH L* a* b* \ud835\udf1f\ud835\udc6c AT CT AT CT AT CT AT CT AT CT Control 3.37 \u00b1 0.0273ab 3.37 \u00b1 0.0273 10.2 \u00b1 0.0536fg 10.2 \u00b1 0.0536ij 20.1 \u00b1 1.81i 20.1 \u00b1 1.81j 13.1 \u00b1 0.509a 13.1 \u00b1 0.509b 0 \u00b1 0i 0 \u00b1 0k Table 7.2: Continued 197  PL1 3.43 \u00b1 0.0289a 3.34 \u00b1 0.0176 13.5 \u00b1 0.209bc 13.3 \u00b1 0.0643ef 28.2 \u00b1 0.36bc 30.8 \u00b1 0.0757cd 9.02 \u00b1 0.662cde 11.5 \u00b1 0.116cdg 3.1 \u00b1 0.0869c 3.35 \u00b1 0.00973c PL2 3.34 \u00b1 0.026bd 3.34 \u00b1 0.0176 11.1 \u00b1 0.0318ef 11.7 \u00b1 0.424h 26.3 \u00b1 0.186de 27.8 \u00b1 0.335fg 8.34 \u00b1 0.0296dg 10.5 \u00b1 0.157gi 2.81 \u00b1 0.0248d 2.87 \u00b1 0.0677ef PL3 3.32 \u00b1 0.024bd 3.34 \u00b1 0.0208 11.6 \u00b1 0.229de 10.4 \u00b1 0.266ij 24.4 \u00b1 0.278f 25.4 \u00b1 0.0252h 7.96 \u00b1 0.0608fg 11.4 \u00b1 0.0786dgh 2.61 \u00b1 0.0281ef 2.35 \u00b1 0.0123hi PL4 3.33 \u00b1 0.04bd 3.32 \u00b1 0.0176 8.3 \u00b1 0.699h 10 \u00b1 0.364j 22.3 \u00b1 0.154gh 23.6 \u00b1 0.271i 8.41 \u00b1 0.0717cdg 9.59 \u00b1 0.0633ij 2.36 \u00b1 0.0386gh 2.22 \u00b1 0.0496ij PL5 3.36 \u00b1 0.026b 3.32 \u00b1 0.0145 9.68 \u00b1 0.514g 10.1 \u00b1 0.295ij 21.5 \u00b1 0.133hi 22.2 \u00b1 0.431i 6.79 \u00b1 0.12h 9.59 \u00b1 0.424ij 2.55 \u00b1 0.0136f 2.04 \u00b1 0.0449j PL6 3.36 \u00b1 0.0219bd 3.35 \u00b1 0.0176 13.9 \u00b1 0.897ab 16.1 \u00b1 0.543b 29.5 \u00b1 0.16b 32.6 \u00b1 0.142b 8.26 \u00b1 0.488dg 12.5 \u00b1 0.203bc 3.36 \u00b1 0.073b 3.71 \u00b1 0.0487b PL7 3.34 \u00b1 0.0219bd 3.34 \u00b1 0.0145 11.9 \u00b1 0.415de 12 \u00b1 0.0897gh 26.8 \u00b1 0.423cd 30.2 \u00b1 0.593de 10.2 \u00b1 0.178b 12.1 \u00b1 0.0561bde 2.74 \u00b1 0.0725de 3.2 \u00b1 0.0943cd PL8 3.36 \u00b1 0.0208bc 3.32 \u00b1 0.0203 11.7 \u00b1 0.118de 11.8 \u00b1 0.183h 24.9 \u00b1 0.218ef 27.4 \u00b1 0.201fg 8.97 \u00b1 0.437cdf 11.8 \u00b1 0.0821cdef 2.55 \u00b1 0.091f 2.75 \u00b1 0.0428f PL9 3.33 \u00b1 0.0176bd 3.31 \u00b1 0.0186 10.8 \u00b1 0.345ef 12.2 \u00b1 0.163gh 24 \u00b1 0.261f 25.5 \u00b1 0.191h 9.39 \u00b1 0.573bc 11.1 \u00b1 0.109egh 2.35 \u00b1 0.0543h 2.47 \u00b1 0.0448gh PL10 3.31 \u00b1 0.0145bd 3.33 \u00b1 0.0176 9.5 \u00b1 0.21g 10.9 \u00b1 0.0426i 22.3 \u00b1 0.0862gh 23.5 \u00b1 0.118i 7.94 \u00b1 0.151fg 10.4 \u00b1 0.946hi 2.37 \u00b1 0.0233gh 2.11 \u00b1 0.138j 198  PL11 3.32 \u00b1 0.0173bd 3.32 \u00b1 0.0145 14.6 \u00b1 0.32a 17.6 \u00b1 0.17a 31.5 \u00b1 0.0964a 34.4 \u00b1 0.128a 9.2 \u00b1 0.395bd 14.7 \u00b1 0.487a 3.58 \u00b1 0.0349a 4.03 \u00b1 0.0192a PL12 3.29 \u00b1 0.0186d 3.32 \u00b1 0.0306 12.4 \u00b1 0.142cd 14.7 \u00b1 0.296cd 27.5 \u00b1 0.0433cd 31.9 \u00b1 0.476bc 8.39 \u00b1 0.0581cdg 12.3 \u00b1 0.451bd 3 \u00b1 0.00223c 3.55 \u00b1 0.0784b PL13 3.3 \u00b1 0.00882cd 3.31 \u00b1 0.0208 12.5 \u00b1 0.418cd 14 \u00b1 0.221de 23.6 \u00b1 0.388fg 29 \u00b1 0.494ef 8.64 \u00b1 0.503cdg 10.7 \u00b1 0.303gh 2.49 \u00b1 0.0527fg 3.15 \u00b1 0.0902d PL14 3.31 \u00b1 0.00882bd 3.33 \u00b1 0.0306 11.2 \u00b1 0.0611ef 14.9 \u00b1 0.134c 22.5 \u00b1 0.175gh 26.9 \u00b1 0.515gh 7.91 \u00b1 0.147g 9.13 \u00b1 0.214j 2.41 \u00b1 0.012gh 3.03 \u00b1 0.0523de PL15 3.31 \u00b1 0.0186bd 3.32 \u00b1 0.0203 10.8 \u00b1 0.0551ef 12.6 \u00b1 0.17fg 23.7 \u00b1 0.393fg 25.7 \u00b1 0.0736h 8.13 \u00b1 0.254efg 10.8 \u00b1 0.219fgh 2.49 \u00b1 0.00515fg 2.56 \u00b1 0.0167g Values are means \u00b1 SEM, n = 3 per treatment group.  Means in a column without a common superscript letter differ (p < 0.05) as analyzed by one-way ANOVA and the LSD test.Table 7.3: Continued 199  7.4.2 Effect on colour of juices The results of changes in colour parameters of watermelon juice due to PL are given in Table 7.2. The L* parameter (which shows the lightness-darkness characteristic of sample) is significantly (p < 0.05) affected for AT reactor treatments, especially for PL11 treatment; however, the changes were very slight. Effect of pulse frequency was not significant (p > 0.05), however, the interaction between flow rate and pulse frequency was significant (p < 0.05). For CT reactor, the L* value was not significantly (p > 0.05) affected after PL treatments. This is in accordance with the findings of Feng et al. [195], who did not observe any change in sample L* after UV-C treatment. The a* (redness value) was significantly decreased compared to control a* value for the watermelon juice for both the reactors, with the values decreasing with decreasing flow rate and increasing pulse frequency. The a* values were significantly (p < 0.05) lower for CT reactor than AT reactor. The decrease in a* value could be related to the photodegradation and isomerization of lycopene pigment [196]. The b* value decreased significantly (p < 0.05) with fluence for both the reactors, which could be related to the decrease in red colour of the watermelon juice samples. The decrease in b* was significantly higher (p < 0.05) for CT reactor than AT. The total colour difference (\u2206\ud835\udc38) values increased with PL fluence and were significantly higher (p < 0.05) for CT reactor compared to AT reactor, and value of around 2 was obtained which cannot be perceived by human eye. Overall, for watermelon juice, the colour change was significant but the difference in colour was not perceivable. The colour of the red grape juice was significantly (p < 0.05) affected by PL treatment especially at lower flow rates and higher frequency (Table 7.3). The L* value was significantly higher (p < 0.05) at higher pulse frequency and lower flow rate (PL1, PL 6, PL 11) which are directly related to imparting of higher fluences to samples. In case of  the samples treated in AT reactor, the interaction of flow rate and frequency was not significant (p > 0.05). There was a significant (p < 0.05) interaction of flow rate and frequency for CT reactors. The increase in L* value is in contrast with the observations of Wang et al. [128], which could be related to the difference in the samples used. The increase in the L* values at higher fluences is attributed to the decrease in the anthocyanins by light energy, which leads to sample become lighter.  Also, it could be related to the inactivation of browning enzymes like polyphenol oxidase [128]. The a* value of the PL-treated samples decreased significantly (p < 0.05) with PL treatment compared to control. The a* 200  value denotes the decrease in redness value of juice, that can also be attributed to the degradation of anthocyanins [197]. The b* value of the PL-treated juice samples significantly decreased (p < 0.05) with flow rate and increasing frequency (interaction significant, (p < 0.05)). However, the results are in contrast with Wang et al. [128], who observed that the b* did not change with PL fluence. The total colour difference (\u2206\ud835\udc38) increased significantly with higher pulse frequency and decreasing flow rates (p < 0.05; interaction significant). The \u2206\ud835\udc38 values were significantly higher (p < 0.05) for CT reactor compared to AT reactor, and value of around 4 was obtained which shows that the sample colours changed which could be visually seen by human eye. Overall, the colour of red grape juice was affected and the changes were higher in case of CT reactor, which were perceivable with naked eye. 7.4.3 Effect on vitamin C content in watermelon juice The vitamin C content of PL-treated watermelon juice for AT and CT reactor is given Figure 7.1 and 7.2. The vitamin C content of fresh watermelon juice was 49.16 \u00b1 0.922 mg\/L juice, which was higher compared to values reported in literature [194]. The vitamin C content was significantly (p < 0.05) lowered at higher fluences (lower flow rates, and higher pulse frequency); however, there was no significant interaction of frequency and flow rates (p > 0.05). The decrease in vitamin C can be attributed to its light sensitivity and follows kinetics similar to thermal degradation [198]. The loss of vitamin C was significantly (p < 0.05) higher for CT reactor than AT reactor. The retention of vitamin C was 53% at highest fluence for AT reactor compared to 29% in case of CT reactor. Contrastingly, for cut watermelon, the vitamin C loss was not significant (p > 0.05) [199], which could be due to the surface treatment for cut fruit compared to clear watermelon juice (in this study). Also, the loss in vitamin C of 71% is significantly higher than milk (52%) (chapter 5), which could be ascribed to the higher light penetration in clear juice comparted to milk [178]. The vitamin C level is an indicator for nutritional value of food and thus the retention of vitamin C during processing is related to the retention of nutritional value of food. Therefore, proper designing of a reactor should consider the minimization of nutrient loss with maximization of beneficial effects. 201   Figure 7.1: Changes in vitamin C content in PL-treated watermelon juice in annular (AT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols  Figure 7.2: Changes in vitamin C content in PL-treated watermelon juice in coiled (CT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data 0510152025303540455014.3 32.5 49.6 64.1 74.9Vitamin C [mg\/L Juice]Flow rate [L\/h]AT Reactor 1 Hz 3 Hz 5 Hz+^&#%* ****051015202530354014.3 32.5 49.6 64.1 74.9Vitamin C [mg\/L Juice]Flow rate [L\/h]CT Reactor 1 Hz 3 Hz 5 Hz+^&#%*****202  and other data for the same flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols 7.4.4 Effect on trans-resveratrol content in red grape juice The trans-resveratrol content of PL-treated red grape juice for AT and CT reactor is given Figure 7.3 and 7.4. The resveratrol content of fresh watermelon juice was 1.655 \u00b1 0.048 mg\/L juice, which was lower than reported value for the grapes. The higher resveratrol in grapes compared to the juice is because the resveratrol is mostly concentrated in the skin of the grapes which are discarded during the juicing process. The loss in trans-resveratrol content was significant (p < 0.05)  for both type of reactors, with the content decreasing with PL fluence. The interaction between frequency and flow rate was significant (p > 0.05) for both the reactors. The loss in trans-resveratrol was significantly (p < 0.05) greater for CT reactor than AT reactor. The loss in trans-resveratrol could be attributed to the photodegradation and isomerization into cis-resveratrol [200].  Figure 7.3: Changes in trans-Resveratrol content in PL-treated red grape juice in annular (AT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols 00.20.40.60.811.21.41.614.3 32.5 49.6 64.1 74.9trans-Resveratrol [mg\/L]Flow rate [L\/h]AT Reactor1 Hz 3 Hz 5 Hz^&#%+*****203   Figure 7.4: Changes in trans-Resveratrol content in PL-treated red grape juice in coiled (CT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols 7.4.5 Effect on total phenolics content in juices The total phenolics content for watermelon juice and red grape juice treated in AT and CT reactor  are given in figures 7.5-7.8. The TPC level of control red grape juice sample was 112.0 \u00b1 1.537 mg GAE\/100 mL juice, which was about 2.5 times higher than the value reported in literature [128]. These phenolic compounds are important indicator for food quality as their level change during processing, and thus their retention is important with respect to PL processing of juices. The TPC level of grape juice decreased significantly (p < 0.05) with PL fluence for both reactors (with decreasing flow rates and increasing frequency; interaction between flow rate and frequency was significant (p < 0.05)). The loss in TPC was significantly higher (p < 0.05) in CT than AT reactor, which can be attributed to greater turbulence, and higher residence time. Contrastingly, Wang et al. [128] reported that TPC levels were not affected after PL processing. However, Noci et al. [201] reported a decrease in TPC levels of apple juice after UV-C processing. Wiktor et al. [122] also reported degradation of phenolic compounds beyond a critical fluence of 3.82 J\/cm2. The TPC level of control watermelon juice was significantly lower than control grape juice. The value was calculated to be 5.469 \u00b1 0.431 mg\/100 mL juice, which was significantly lower than the 00.20.40.60.811.21.414.3 32.5 49.6 64.1 74.9trans-Resveratrol [mg\/L]Flow rate [L\/h]CT Reactor1 Hz 3 Hz 5 Hz^&#%+*****204  values reported by Feng et al. [195]. The TPC level was significantly (p < 0.05) reduced by PL processing in both the reactors and the reduction was greater in CT reactor than AT. This is in contrary to the study by Feng et al. [195], who observed no significant changes in TPC of watermelon juice.  Figure 7.5: Changes in total phenolics content in PL-treated red grape juice in annular (AT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols  70758085909510010511011514.3 32.5 49.6 64.1 74.9TPC [mg GAE\/100 mL]Flow rate [L\/h]AT Reactor1 Hz 3 Hz 5 Hz^&#% ****+*70758085909510010511014.3 32.5 49.6 64.1 74.9TPC [mg GAE\/100 mL]Flow rate [L\/h]CT Reactor1 Hz 3 Hz 5 Hz^&#% **+** *205  Figure 7.6: Changes in total phenolics content in PL-treated red grape juice in coiled (CT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols  Figure 7.7: Changes in total phenolics content in PL-treated watermelon juice in annular (AT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; Symbols +, ^, &, and # indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols  012345614.3 32.5 49.6 64.1 74.9TPC [mg GAE\/100 mL]Flow rate [L\/h]AT Reactor1 Hz 3 Hz 5 Hz*+^&#** * *^012345614.3 32.5 49.6 64.1 74.9TPC [mg GAE\/100 mL]Flow rate [L\/h]CT Reactor1 Hz 3 Hz 5 Hz++&#^ *****206  Figure 7.8: Changes in total phenolics content in PL-treated watermelon juice in coiled (CT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; Symbols +, ^, &, and # indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols 7.4.6 Effect on antioxidant activity in juices  The antioxidant activity of red grape juice and watermelon juice was measured in terms of Trolox equivalents by adopting the DPPH assay. The Figures 7.9 and 7.10 show the changes in levels of antioxidants in red grape juice in AT and CT reactor, respectively. The antioxidant activity of fresh grape juice was 0.489 \u00b1 0.027 \u03bcL Trolox equivalent\/\u03bcL Juice. The antioxidant activity of grape juice decreased significantly (p < 0.05) after PL treatment, with significant (p < 0.05) interaction between flow rate and pulse frequency for both reactors. The difference in the antioxidant level decrease in grape juice processed in AT and CT reactor is not significant (p > 0.05). The Figures 7.11 and 7.12 show the changes in levels of antioxidants in watermelon juice in AT and CT reactor, respectively. The antioxidant activity of fresh grape juice was 0.489 \u00b1 0.027 \u03bcL Trolox equivalent\/\u03bcL Juice. The antioxidant activity of watermelon juice decreased significantly (p < 0.05) after PL treatment, with no significant (p > 0.05) interaction between flow rate and pulse frequency for both reactors. The difference in the antioxidant level decrease in watermelon juice processed in AT and CT reactor is not significant (p > 0.05).  00.10.20.30.40.50.614.3 32.5 49.6 64.1 74.9\u03bcL Trolox equivalent\/\u03bcL JuiceFlow rate [L\/h]AT Reactor 1 Hz 3 Hz 5 Hz^#%+*****&207  Figure 7.9: Changes in antioxidant activity of PL-treated red grape juice in annular (AT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols  Figure 7.10: Changes in antioxidant activity of PL-treated red grape juice in coiled (CT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols  00.10.20.30.40.50.614.3 32.5 49.6 64.1 74.9\u03bcL Trolox equivalent\/\u03bcL JuiceFlow rate [L\/h]CT Reactor 1 Hz 3 Hz 5 Hz^&#%+*****00.050.10.150.20.250.30.350.40.4514.3 32.5 49.6 64.1 74.9\u03bcL Trolox equivalent\/\u03bcL JuiceFlow rate [L\/h[AT Reactor1 Hz 3 Hz 5 Hz*+^&#**+208  Figure 7.11: Changes in antioxidant activity of PL-treated watermelon juice in annular (AT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, and # indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols  Figure 7.12: Changes in antioxidant activity of PL-treated watermelon juice in coiled (CT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, and # indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols 7.4.7 Effect on anthocyanin content in red grape juice The total monomeric anthocyanins content for PL-treated red grape juice is given in Figure 7.13 and 7.14 for AT and CT reactor, respectively.  The total monomeric anthocyanin content of control red grape juice was 11.748 \u00b1 0.422 mg cynidine-3-glucoside\/L juice, which is somewhat closer to the value reported by Tiwari et al. [202]. The anthocyanin content decreased for red grape juice with applied PL fluence however, the effect of pulse frequency was not significant (p > 0.05) in case of samples treated in AT reactor, while effect of flow rate was significant (p < 0.05). In CT reactor, the total anthocyanin content was also decreased significantly (p < 0.05) with decreasing flow rate, and increasing pulse frequency, however, the interaction between flow rate and pulse frequency were not significant. Wang et al. [128] also reported a decrease in anthocyanin in grape 00.050.10.150.20.250.30.350.414.3 32.5 49.6 64.1 74.9\u03bcL Trolox equivalent\/\u03bcL JuiceFlow rate [L\/h]CT Reactor 1 Hz 3 Hz 5 Hz*+^&#****+209  juice, especially cynidine-3-glucoside with PL fluence. Caminiti et al. [203] also reported a decrease in cynidine-3-glucoside after PL treatment in apple juice. The decrease in anthocyanin content can also be related to the decrease in antioxidant activity and the redness values of grape juice after PL treatment.  Figure 7.13: Changes in total anthocyanin content in PL-treated red grape juice in annular (AT) reactor. A * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, # and % indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols  0246810121414.3 32.5 49.6 64.1 74.9mg cyanidin-3-glucoside\/ L Flow rate [L\/h]AT Reactor1 Hz 3 Hz 5 Hz^&#%+*0246810121414.3 32.5 49.6 64.1 74.9mg cyanidin-3-glucoside\/ L Flow rate [L\/h]CT Reactor 1 Hz 3 Hz 5 Hz+^&*+#** *210  Figure 7.14: Changes in total anthocyanin content in PL-treated red grape juice in coiled (CT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, and # indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols 7.4.8 Effect on lycopene content in watermelon juice The lycopene content of watermelon juice treated in AT and CT reactor is shown in Figures 7.15 and 7.16. The lycopene content of control watermelon juice sample was 5.360 \u00b1 0.098 mg\/100 mL juice, which was slightly higher than the values reported by Feng et al. [195]. The lycopene content was significantly (p < 0.05) decreased in watermelon juice after PL treatment for both AT and CT reactor, with effect being significant (p < 0.05) for both frequency and flow rate. However, the interaction between flow rate and pulse frequency was not significant (p > 0.05). The decrease in lycopene content was significantly (p < 0.05) higher in CT reactor than AT reactor. Feng et al. [195] reported that the lycopene content was slightly affected but the difference was not significant (p > 0.05). Lycopene stability (loss or gain) after PL treatment needs exploration. In a study, Liu et al. [204] observed enhanced formation of lycopene after radiation treatment. Moreover, However, Chen et al. [196] reported that UV exposure time and intensity has an antagonistic effect on lycopene stability. This differences could be related to the form as well as the matrices in which lycopene is present. In clear watermelon juice, lycopene apparently gets photodegraded by oxidation and isomerization into cis-lycopene.  Figure 7.15: Changes in lycopene content in PL-treated watermelon juice in annular (AT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups 012345614.3 32.5 49.6 64.1 74.9Lycopene content [mg\/100 mL]Flow rate [L\/h]AT Reactor 1 Hz 3 Hz 5 Hz+^ *****&#%211  (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, and # indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols  Figure 7.16: Changes in lycopene content in PL-treated watermelon juice in coiled (CT) reactor. A * symbol with the bar beneath denotes significant difference (p < 0.05) across all three experimental groups (pulse frequency) at each flow rate; An isolated * symbol denotes significant difference (p < 0.05) between the data and other data for the same flow rate; Symbols +, ^, &, and # indicate significant difference (p < 0.05) between the data marked by these at certain flow rates all other flow rates having different symbols 7.4.9 Multivariate analysis The multivariate data analysis for processing of red grape juice and watermelon juice were carried out using principal component analysis and datasets were transformed into two mutually orthogonal principal components (PC1 and PC2). The Figures 7.17 (a) and (b) represent the rotated biplot for grape juice treated in AT and CT reactor, respectively.  012345614.3 32.5 49.6 64.1 74.9Lycopene content [mg\/100 mL]Flow rate [L\/h]CT Reactor 1 Hz 3 Hz 5 Hz+^*&#% ****212   (a) 213   (b) Figure 7.17: Rotated biplot for the principal components for various analyses carried out for grape juice in (a) AT and (b) CT reactor.  In Figure 7.17(a), described for PL-treated grape juice in AT reactor, the components PC1 and PC2 describe more than 80.65% of the variability in the dataset with PC1 describing 69.04% and PC2 describing 11.61% of the variability. The biplot responses such as TPC, pH, L*, a*, b* etc. denote the place they occupy in the quadrants and the treatments (PL1, PL2, and so on) fall on the same quadrants when they are correlated. As can be seen, the treatments PL2, PL3, PL4, PL5, PL10, PL15 lie closer to responses E. coli, L. innocua and C. sporogenes inactivation, meaning their effect was minimal in inactivation. On the other hand, the treatments, having higher fluences PL1, PL6, PL11) lie on the opposite quadrant meaning their effect on inactivation was higher. Also, higher fluences affect the colour parameter greatly, and lie closer to them. They also lie in opposite quadrant to TPC, trans-resveratrol, anthocyanins, antioxidant activity, meaning these 214  parameters are greatly affected at higher fluences. Control was also far from other treatments, which means the control was not similar in terms of affecting the parameters studied. TPC, anthocyanins, trans-resveratrol, antioxidant activity were highly correlated (r ~ 0.9). \u2206\ud835\udc38 was also highly correlated (r > 0.8) with L* and a* values. A similar biplot was obtained for PL-treated grape juice in CT reactor (Figure 7.17(b)). The components PC1 and PC2 can describe 75.14 and 12.66% of variabilities in the dataset, respectively. Treatments PL1, PL6, PL7, PL11, lie in the opposite quadrant to inactivation of E. coli, L. innocua and C. sporogenes, meaning they have greater impact on their inactivation. PL4, PL5, PL2, PL10 lie closer to the spaces occupied by TPC, antioxidant activity, trans-resveratrol, anthocyanins, meaning they do not affect these parameters to a great extent. TPC, anthocyanins, trans-resveratrol, antioxidant activity were highly correlated (r ~ 0.75). \u2206\ud835\udc38 was also highly correlated (r > 0.89) with L* and a* values.  (a) 215   (b) Figure 6.18: Rotated biplot for the principal components for various analyses carried out for watermelon juice in (a) AT and (b) CT reactor In Figure 7.18(a), described for PL-treated watermelon juice in AT reactor, the components PC1 and PC2 describe more than 84.98% of the variability in the dataset with PC1 describing 75.58% and PC2 describing 9.61% of the variability. As can be seen, the treatments PL2, PL3, PL4, PL5, PL15 lie closer to responses E. coli, L. innocua and C. sporogenes inactivation, meaning their effect was minimal in inactivation. On the other hand, the treatments having higher fluences PL1, PL6, PL7, PL13) lie on the opposite quadrant meaning their effect on inactivation was higher. Also, higher fluences affect the colour parameter \u2206\ud835\udc38 greatly and lie closer to them. They also lie in opposite quadrant to TPC, vitamin C, lycopene content, antioxidant activity, meaning these parameters are greatly affected at higher fluences. Control was also far from other treatments, which means the control was not similar in terms of affecting the parameters studied. TPC, vitamin 216  C, lycopene content, antioxidant activity were highly correlated (r ~ 0.87). \u2206\ud835\udc38 was also negatively correlated (r < -0.8) with a* and b* values. A similar biplot was obtained for PL-treated watermelon juice in CT reactor (Figure 7.18(b)). The components PC1 and PC2 can describe 79.00 and 11.69% of variabilities in the dataset, respectively. Treatments PL1, PL6, PL13, lie in the opposite quadrant to inactivation of E. coli, L. innocua and C. sporogenes, meaning they have greater impact on their inactivation. PL3, PL4, lie closer to the spaces occupied by TPC, vitamin C, lycopene content, antioxidant activity, meaning they do not affect these parameters to a great extent. TPC, vitamin C, lycopene content, antioxidant activity were highly correlated (r ~ 0.93). \u2206\ud835\udc38 was also negatively correlated (r < -0.9) with a* and b* values. 7.5 Conclusion For the juice samples, PL treatment hugely affected the quality parameters and differed as per the sample. Colour parameters which are one of the indicators of sensory changes were affected. Total colour difference in grape juice sample >3 which meant that the changes were visible to human eye. However, the total colour difference was <3 for watermelon juice. Anthocyanins and lycopene content were also reduced in the grape juice and watermelon juice, respectively. TPC and antioxidant capacity of juices were also reduced. Vitamin C was reduced greatly in watermelon jice, while trans-resveratrol in grape juice was degraded. PCA analysis revealed that the analyzed parameters were hugely dependent on the PL processing parameters. However, optimized treatment will help in minimizing these changes.       217  Chapter 8: Conclusion, significance, limitations, and future directions The overarching objective of the doctoral research project was to develop a continuous PL processing system for liquid foods like milk and fruit juices. These products are still processed by using conventional techniques like HTST processing, and newer techniques like high-pressure processing. Application of light-based technologies like UV-C processing has been limited to the treatments of wastewater and drinking water. UV-C processing has also been extended and studied for treatment of milk and juices. Technologies like SurePure Turbulator\u2122 swirl tubes have been introduced for processing of milk. In line with UV-C processing, the PL technology also has the potential to process liquid foods. However, this novel area of research is still emerging. PL has been shown to be an excellent surface treatment process, owing to its superficial nature. For extending this technology to processing of liquid foods and thereby replicating and scaling-up, a proper characterization, modeling and validation of the PL process is necessary. 8.1 Modeling and simulation of PL processing The PL Lamp which had a cylindrical shape (length = 60 cm, diameter = 9 mm) emitted light energy with values of around 100 mJ\/cm2 per pulse at the lamp surface. The energy was minimum at either ends of the lamp and the maximum energy was measured at the lamp center longitudinally. The lamp energy was modeled in air and liquids (water, water+red dye, water+green dye, skim milk) of different optical properties (absorption coefficient, UV transmittivity) taken in petri dish by varying the liquid thickness and distance from lamp and moving the petri dish along the lamp.  The lamp energy was modeled using a three-parameter symmetric Gaussian model, which predicted lamp energy at any point in space in the liquids or air. The energy values decreased in the order as water > water+red dye > water+green dye > skim milk. Two different types of PL reactors, annular, AT and coiled, CT reactors were designed and fabricated using quartz glass. They were characterized by their hydrodynamic behavior (calculation of Reynolds number, Dean number (for CT reactor), RTD parameters) for five different flow rates (14-75 L\/h). In terms of hydrodynamics, the CT reactor performed better than AT and offered greater turbulence in the liquid flow profile. The AT reactor also had lower residence time than CT reactor. The UV dose delivered in the reactors for processing conditions (flow rate = 14-75 L\/h; pulse frequency =1-5 Hz) was estimated for both the reactors using potassium iodide-iodate actinometry. It was observed that the delivered UV-dose were significantly higher (p < 0.05) in CT reactor than AT reactor. 218  Lastly, CFD analysis and simulation was carried out for the PL process. The velocity and flow profiles were visualized for various processing conditions for both the reactors. Also, the volume averaged irradiance and fluence rates were calculated for the reactors for the aforementioned liquids. The results of CFD simulation confirmed that the fluence rates decreased in the same order as measured and predicted by the Gaussian model. The process uniformity was also visualized for both the reactors and it was observed that due to a thin profile (1 mm), AT reactor is able to deliver light energy without much attenuation of light. On the other hand, CT reactor, due to a thicker profile (9 mm), there was greater attenuation of light. However, since additional turbulence was imparted due to secondary flow in CT reactor, the treatment was more uniform. Combing the lamp energy prediction models with the CFD particle tracking was used to calculate delivered fluence in model liquids for different processing conditions. These results were further tested and validated using challenge microorganisms in the model liquids and real liquid foods. 8.2 Validation using challenge microorganisms in model and real liquid foods The PL process was modeled using hydrodynamics, CFD techniques and the model was validated using challenge microorganisms which were inoculated with E. coli, L. innocua, and C. sporogenes into the model liquids. The liquid foods were processed by PL under the same flow rates (14-75 L\/h) and pulse frequency (1-3 Hz). The inactivation of microorganisms in liquids foods varied as water > water+red dye > water+green dye > skim milk. E. coli was most susceptible and followed by L. innocua and C. sporogenes. The microbial survivor curves were modeled using log-linear, Weibull and log-linear plus tail model. The log-linear plus tail model fitted the survivor curves best in terms of R2 and RMSE error values. Biodosimetry was carried out using collimator tube protocol by inoculating the model liquids with challenge microorganisms to ascertain the D-value of the microorganisms in different liquids. It was found that the order of D-values followed the opposite order as water < water+red dye < water+green dye < skim milk, meaning it was easier to inactivate microorganisms in water than milk. Also, the D-value for C. sporogenes was the highest followed by L. innocua and E. coli. This confirms that the microbial inactivation in liquid foods by PL depends greatly on the optical properties of liquids. The microbial inactivation in real liquid foods was carried out by selecting three different liquid foods based on their differences in optical properties and pH values\u2014Milk (0, 1, 2, 3.25% fat), red grape juice and watermelon juice. The liquid foods were inoculated with the aforementioned 219  challenge microorganisms and treated using PL under the same conditions of flow rates and pulse frequency in both AT and CT reactors. The microbial inactivation followed the order as watermelon juice > grape juice > milk. The obtained log reduction in AT reactor was >5 logs for both the juice types while for milk up to 1 log reduction as achieved. In case of CT reactor, >7 logs reduction was achieved in juices, and reduction of >4 logs were achieved for milk. Fat% significantly affected the microbial inactivation due to its scattering properties\u2014skim milk showed the greatest inactivation while 3.25% milk showed the least inactivation. The inactivation kinetics were modeled using three different kinetic models and log-linear with tail model fitted the inactivation data best. Inactivation of bacterial strains in milk was compared against HTST treatment of milk, in which case >5 logs inactivation was achieved. Overall, PL seems a great processing method for juices. However, to extend it to process milk, combination treatments like minimal heating, other non-thermal methods need to be tested and validated. 8.3 PL-treated milk\u2014changes in nutritional and quality parameters The changes in quality parameter in milk was tested by measuring the changes in pH and colour parameters (L*, a* and b*). The pH did not change to a great extent in milk after PL treatment, although the change was significant (p < 0.05). The colour parameter L* also did not change for milk samples after PL treatment. The a* values for milk increased as the PL fluence increased meaning the redness of milk increased, which could be attributed to the Maillard browning. On the other hand, b* values decreased after PL treatment significantly (p < 0.05). This could be attributed to the development of photo-byproducts after photodegradation of vitamin B2. The decrease in b* values was the greatest for skim milk and the least in 3.25%. However, the total colour difference (\u2206\ud835\udc38) was less than 3, meaning the changes in colour were not perceptible.  Vitamins B2 and C also degraded with PL fluence and they were reduced up to 57 and 52% in the samples, with higher degradation in case of CT reactor. However, faster HTST pasteurization, the vitamin B2 content did not change much, which the vitamin C was degraded toa great extent, higher than PL. Lipid oxidation was measured for the milk samples in terms of primary oxidation products. Lipid oxidation was significantly (p < 0.05) higher than control, with the values being highest for 3.25% fat samples treated in CT reactor. Protein oxidation (measured in terms of protein carbonyls) was also observed in PL-treated milk. These results show that milk\u2019s quality 220  and nutritional parameters were affected after PL treatment, which necessitates drawing a balance between microbial inactivation (beneficial) and quality\/nutritional changes (undesirable) in milk.  8.4 PL-treated juices\u2014changes in nutritional and quality parameters For, PL-treated juices, pH changed slightly for both type of juices. For grape juice, the L* increased for the samples and a* decreased significantly (p < 0.05). These changes could be related to the inactivation of browning enzymes like polyphenol oxidase due to PL and also anthocyanin degradation. This was confirmed by measurement of anthocyanins in juice samples, which decreased significantly (p < 0.05), and decreased up to 60%.  In watermelon juice samples, the L* and b* did not vary greatly, while a* values decreased significantly (p < 0.05). The decrease in a* could be related to decrease in lycopene content which was also observed (up to 50%). In grape juice samples, trans-resveratrol content also decreased significantly (p < 0.05). In watermelon juice, the vitamin C also decreased significantly (p < 0.05). The decrease in TPC and antioxidant capacity was also significant (p < 0.05) in both the type of juice. However, the decrease was higher in watermelon juice, that could be related to different polyphenolic compounds in the juices which have different susceptibility to degradation by light. Parameter optimization will lead to balancing of the necessary microbial inactivation compared to the undesirable quality and nutritional changes in juices. 8.5 Significance  The significance of the research lies in the fact that the characterization of the PL reactors in this study lays down the basic framework which could be useful for future research in the field of pulsed light technology. These experiments could be replicated and verified by researchers working in the same area. Also, since replication of the processes is necessary (which has been lacking mostly till now, due to poorly measured or defined parameters like fluence, fluence rates, irradiance values), the methodology could be adopted for design and development of the continuous PL reactors.  Secondly, these results show that the PL technology is a potential technology that can be advised to and adopted by juice and dairy industry. Since, the food industries still rely on conventional technologies, the PL process could be a viable alternative for these industries. These protocols and experiments could be useful for more thorough studies for drawing treatment protocols, design and scale-up and trials in pilot-scale and will help in getting regulatory approvals.  221  8.6 Limitations  Some of the key limitations of the research were: a) Flow rates  used were still not high enough compared to what is used for conventional technologies like HTST processing. To be cost-effective, the product throughput should be higher.  b) The CFD modeling was carried out for the three-dimensional model developed  for the reactor. While great care was taken to make the geometry representative of the real reactors. However, the reactor  geometry was simpler compared to the real reactors, since designing of channels and closed conduits is complex and challenging. Intricate designing would also involve difficulty in meshing operation, which was not considered for simplicity and could lead to slight errors in prediction of the variables like velocity and flow rates. c) The challenge microorganisms selected were surrogates of the pathogenic variants but had almost similar susceptibility to light-based inactivation. More pathogenic strains should be tested in the PL treatments that were carried out. d) The liquid foods (milk and fruit juices) tested were obtained from the grocery stores nearby. It would  have been worthwhile to test on products which were obtained freshly. In this direction raw milk and freshly extracted juice could have been viable alternatives to their pasteurized or thermally processed counterparts that were used in these studies. 8.7 Future research directions There are various aspects of processing that remained untouched and need to be looked at in the future. Some of the identified future directions that have arisen from the current research are mentioned as follows: a) Use of multiple lamps, higher pulse frequencies and flow rates for PL processing. This could help in the processing of larger volume of liquids simultaneously  b) Scaling-up the process for developing a pilot scale system followed by a commercial prototype system. This will help in adoption of the technology for processing of liquid foods c) Testing the process efficacy using more challenge microorganisms to extend the range of application. Also, using of different samples of milk, juices (containing natural microflora) for PL processing to see if the treatment can adequately pasteurize the samples containing those natural microflora.  222  d) Use of assistive technologies like mild heating, ultrasonication to enhance the microbial inactivation level.  e) Development of aseptic packaging system for processed liquid foods can help enhancing their shelf-life. f) Sensory evaluation of the treated samples using discriminative and affective sensory tests. This  will help in wider acceptability of the PL technology.  g) The treated samples can also be tested for generation of volatile compounds due to PL. Some of the  compounds that are formed could be peroxides (lipid oxidation products), aldehydes (hexanal, butanal etc.), methyl ketones, oxidative products from amino acids. These volatiles could be measured using gas chromatographic techniques. 223  Bibliography 1. Pollock, A. M., Pratap Singh, A., Ramaswamy, H. S., & Ngadi, M. O. (2017). Pulsed light destruction kinetics of L. monocytogenes. LWT-Food Science and Technology, 84, 114-121. 2. Rahman, M. S. (2007). Food Preservation: Overview. In M. Shafiur Rahman (2nd ed.), Handbook of food preservation (pp. 3-17). Florida: CRC Press. 3. Singh, A. P., Singh, A., & Ramaswamy H. S. (2017). Heat transfer phenomena during thermal processing of liquid particulate mixtures\u2013A Review, Critical reviews in food science and nutrition, 57(7), 1350-1364. 4. Palmieri, L., & Cacace, D. (2005). High Intensity Pulsed Light Technology. In Da-wen Sun (ed.), Emerging Technologies for Food Processing (pp. 279-306). Cambridge: Academic Press, Elsevier Ltd. 5. Krishnamurthy, K., Tewari, J. C., Irudayaraj, J., & Demirci, A. (2010). Microscopic and spectroscopic evaluation of inactivation of Staphylococcus aureus by pulsed UV light and infrared heating. Food and Bioprocess Technology, 3, 93\u2013104. 6. Dunn, J.; Ott, T.; Clark, W. Pulsed-light treatment of food and packaging. Food Technology, 1995, 49, 95-98. 7. Dunn, J. (1996). Pulsed light and pulsed electric field for foods and eggs. Poultry Science, 75, 1133-1136. 8. Buchovec, I., Paskeviciute, E., & Luksiene, Z. B. (2010). Photosensitization-based inactivation of food pathogen Listeria monocytogenes in vitro and on the surface of packaging material. Journal of Photochemistry and Photobiology. B, Biology, 99, 9-14. 9. Woodling, S. E., & Moraru C. I. (2007). Effect of spectral range in surface inactivation of Listeria innocua using broad-spectrum pulsed light. Journal of Food Protection, 70, 909-916. 10. Rajkovic, A., Tomasevic, I., Smigic, N., Uyttendaele, M., Radovanovic, R., & Devlieghere, F. (2010). Pulsed ultraviolet light as an intervention strategy against Listeria monocytogenes and Escherichia coli O157:H7 on the surface of a meat slicing knife. Journal of Food Engineering, 100, 446-451. 11. FDA (1996). Code of Federal Regulations. 21CFR179.41. 224  12. Barbosa-Canovas, G. V., Schaffner, D. W., Pierson, M., & Zhang, Q. H. (2000). Pulsed light technology. Journal of Food Science, 65(Supplement), 82-85. 13. G\u00f3mez-L\u00f3pez, V. M., Ragaerta, P., Debeverea, J., & Devliegherea, F. (2007). Pulsed light for food decontamination: a review. Trends in Food Science & Technology, 18, 464-473. 14. Elmnasser, N., Guillou, S., Leroi, F., Orange, N., Bakhrouf, A., & Federighi, M. (2007). Pulsed-light system as a novel food decontamination technology: a review. Canadian Journal of Microbiology, 53, 813-821. 15. Oms-Oliu, G., Mart\u00edn-Belloso, O., & Soliva-Fortuny, R.  (2010). Pulsed Light Treatments for Food Preservation. A Review. Food and Bioprocess Technology, 3, 13-23. 16. Tomasevic, I., & Rajkovic, A. (2015). The Sensory Quality of Meat, Game, Poultry, Seafood and Meat Products as Affected by Intense Light Pulses: A Systematic Review. Procedia Food Science, 5, 285-288. 17. Heinrich, V., Zunabovic, M., Varzakas, T., Bergmair, J., & Kneifel, W. (2015). Pulsed Light Treatment of Different Food Types with a Special Focus on Meat: A Critical Review. Critical Reviews in Food Science and Nutrition, 56, 591-613. 18. Bhavya, M. L., & Hebbar, H. U. (2017). Review Pulsed light processing of foods for microbial safety. Food Quality and Safety, 01, 187\u2013202. 19. Koutchma, T.; Forney, L. J.; Moraru, C. I. Ultraviolet Light in Food Technology: Principles and Applications. Florida: CRC Press, Taylor & Francis group, USA, 2009. 20. Abida, J., Rayees, B., & Masoodi, F. A. (2014). Pulsed light technology: a novel method for food preservation. International Food Research Journal, 21, 839-848. 21. Ortega-Rivas, E. (2012). Pulsed light Technology. In E. Ortega-Rivas (ed.), Non-thermal Food Engineering Operations (pp. 263-272). New York: Springer. 22. Bintsis, T., Litopoulou\u2010Tzanetaki, E., & Robinson, R. K. (2000). Existing and potential applications of ultraviolet light in the food industry\u2013a critical review. Journal of the Science of Food and Agriculture, 80(6), 637-645. 23. G\u00f3mez-L\u00f3pez, V. M., & Bolton, J. R. (2016). An approach to standardize methods for fluence determination in bench-scale pulsed light experiments. Food and Bioprocess Technology, 9(6), 1040-1048. 225  24. G\u00f3mez-L\u00f3pez, V. M. (2012). Pulsed Light Technology. In Da-Wen Sun (ed.), Handbook of Food Safety Engineering (pp. 643-665). Singapore: Blackwell Publishing Ltd. 25. MacGregor, S. J., Rowan, N. J., McIlvaney, L., Anderson, J. G., Fouracre, R. A., & Farish, O. (1998). Light inactivation of food related pathogenic bacteria using a pulsed power source. Letters in Applied Microbiology, 27, 67-70. 26. Hillegas, S. L., & Demirci, A. (2003). Inactivation of Clostridium sporogenes in clover honey by pulsed UV-light treatment. Agricultural Engineering International, V. Manuscript FP 03 009. 27. Zou, X. Y., Lin, Y. L., Xu, B., Cao, T. C., Tang, Y. L., Pan, Y., Gao, Z.- C., & Gao, N. Y. (2019). Enhanced inactivation of E. coli by pulsed UV-LED irradiation during water disinfection. Science of the Total Environment, 650, 210-215. 28. Krishnamurthy, K., Demirci, A., & Irudayaraj, J. (2004). Inactivation of Staphylococcus aureus by pulsed UV-light sterilization. Journal of Food Protection, 67, 1027-1030. 29. Holck, A. L., Liland, K. H., Dr\u00f8mtorp, S. M., Carleh\u00f6g, M., & McLeod, A. (2018). Comparison of UV-C and Pulsed UV Light Treatments for Reduction of Salmonella, Listeria monocytogenes, and Enterohemorrhagic Escherichia coli on Eggs. Journal of Food Protection, 81, 6-16. 30. Huang, Y., & Chen, H. (2014). A novel water-assisted pulsed light processing for decontamination of blueberries. Food Microbiology, 40, 1-8. 31. Bialka, K.L., & Demirci, A. (2008). Efficacy of Pulsed UV-Light for the Decontamination of Escherichia coli O157:H7 and Salmonella spp. on Raspberries and Strawberries. Journal of Food Science, 73, 201-207. 32. Hiramoto, T. Method of sterilization. US Patent 4,464,336, 1984. 33. Dunn, J. E.; Clark, R. W.; Asmus, J. F.; Pearlman, J. S.; Boyer, K.; Painchaud, F.; Hofmann, G. A.; Weniger. Methods for preservation of foodstuffs. US Patent 4,871,559, 1989. 34. Wekhof, A., Trompeter, F. J., & Franken, O. (2001). Pulsed UV disintegration (PUVD): a new sterilisation mechanism for packaging and broad medical-hospital applications. The First International Conference on Ultraviolet Technologies, Washington D.C., USA. 35. Wekhof, A. (2002). Sterilising packaging and preserving foodstuffs with pulsed light. Newsletter of International UV Association, 4 (5). 226  36. Jun, S., Irudayaraj, J., Demirci, A., & Geiser, D. (2003). Pulsed UV-light treatment of corn meal for inactivation of Aspergillus niger spores. International Journal of Food Science and Technology, 38, 883-888. 37. Fine, F., & Gervais, P. (2004). Efficiency of pulsed UV light for microbial decontamination of food powders. Journal of Food Protection, 67, 787-792. 38. Sonenshein, A. L. (2003). Killing of Bacillus spores by high-intensity ultraviolet light. Xenon Corporation, Woburn, USA. 39. Rajkovic, A., Tomasevic, I., De Meulenaer, B., & Devlieghere, F. (2017). The effect of pulsed UV light on Escherichia coli O157: H7, Listeria monocytogenes, Salmonella Typhimurium, Staphylococcus aureus and staphylococcal enterotoxin A on sliced fermented salami and its chemical quality. Food Control, 73, 829\u2013837. 40. Solaris Lytbot 1.0. https:\/\/solarisrobots.com\/lytbot1\/ 41. Wang, T., MacGregor, S. J., Anderson, J. G., & Woolsey, G. A. (2005). Pulsed ultra-violet inactivation spectrum of Escherichia coli. Water Research, 39, 2921-2925. 42. Ramos-Villarroel, A.Y., Mart\u00edn-Belloso, O., & Soliva-Fortuny, R. (2011). Bacterial inactivation and quality changes in fresh-cut avocado treated with intense light pulses. European Food Research and Technology, 233, 395\u2013402. 43. Rowan, N. J., MacGregor, S. J., Anderson, J. G., Fouracre, R. A., McIlvaney, L., & Farish, O. (1999). Pulsed-light inactivation of food-related microorganisms. Applied and Environmental Microbiology, 65, 1312-1315. 44. Anderson, J. G., Rowan, N. J., MacGregor, S. J., Fouracre, R. A., & Farish, O. (2000). Inactivation of food-borne enteropathogenic bacteria and spoilage fungi using pulsed-light. IEEE Transactions on Plasma Science, 28, 83-88. 45. G\u00f3mez-L\u00f3pez, V. M., Devlieghere, F., Bonduelle, V., & Debevere, J. (2005). Factors affecting the inactivation of microorganisms by intense light pulses. Journal of Applied Microbiology, 99, 460-470. 46. G\u00f3mez-L\u00f3pez, V. M.; Koutchma, T.; Linden, K. Ultraviolet and pulsed light processing of fluid foods. In Novel thermal and non-thermal technologies for fluid foods, Cullen, P.J., Tiwari, B.K., Valdramidis, V.P., Eds., London: Academic Press, UK, 2012, pp. 185-223. 227  47. Luk\u0161iene, \u017d.  (2005). New approach to inactivate harmful and pathogenic microorganisms: photosensitization. Food Technology and Biotechnology, 43, 411\u2013418. 48. Luk\u0161iene, \u017d., Buchovec, I., & Pa\u0161kevi\u010di\u016bt\u0117, E. (2009). Inactivation of food pathogen Bacillus cereus by photosensitization in vitro and on the surface of packaging material. Journal of Applied Microbiology, 107, 2037-2046. 49. Nicorescu, I., Nguyen, B., Moreau-Ferret, M., Agoulon, A., Chevalier, S., & Orange, N. (2013). Pulsed light inactivation of Bacillus subtilis vegetative cells in suspensions and spices. Food Control, 31, 151-157. 50. Xu, W., & Wu, C. (2016). The impact of pulsed light on decontamination, quality, and bacterial attachment of fresh raspberries. Food Microbiology, 57, 135\u2013143. 51. Krishnamurthy, K., Demirci, A., & Irudayaraj, J. M. (2007). Inactivation of Staphylococcus aureus in milk using flow-through pulsed UV-light treatment system. Journal of Food Science. 72, 233\u2013239. 52. Takeshita, K., Shibato, J., Sameshima, T., Fukunaga, S., Isobe, S., Arihara, K., et al. (2003). Damage of yeasts induced by pulsed light irradiation. International Journal of Food Microbiology, 85, 151-158. 53. Ramos-Villarroel, A. Y., Aron-Maftei, N., Mart\u00edn-Belloso, O., & Soliva-Fortuny, R. (2012b). The role of pulsed light spectral distribution in the inactivation of Escherichia coli and Listeria innocua on fresh-cut mushrooms. Food Control, 24, 206-213. 54. Macias\u2010Rodriguez, B., Yang, W., Schneider, K., & Rock, C. (2014). Pulsed UV light as a postprocessing intervention for decontamination of hard\u2010cooked peeled eggs. International Journal of Food Science & Technology, 49(11), 2472-2480. 55. Roberts, P., & Hope, A. (2003). Virus inactivation by high intensity broad spectrum pulsed light. Journal of Virological Methods, 110, 61-65. 56. Dunn, J. E.; Clark, R. W.; Bushnell, A. H.; Salisbury, K. J. Deactivation of organisms using high-intensity pulsed polychromatic light. US Patent 6228332, 2001. 57. Marquenie, D., Geeraerd, A. H., Lammertyn, J., Soontjens, C., Van Impe, J. F., Michiels, C. W., & Nicola\u00ef, B. N. (2003b). Combinations of pulsed white light and UV-C or mild heat treatment to inactivate conidia of Botrytis cinerea and Monilia fructigena. International Journal of Food Microbiology, 85, 185-196. 228  58. Lagunas-Solar, M. C., Pi\u00f1a, C., MacDonald, J. D., & Bolkan, L. (2006). Development of pulsed UV light processes for surface fungal disinfection of fresh fruits. Journal of Food Protection. 69, 376-384. 59. Uesugi, A. R., Woodling, S. E., & Moraru, C. I. (2007). Inactivation kinetics and factors of variability in the pulsed light treatment of Listeria innocua cells. Journal of Food Protection, 70, 2518-2525. 60. Maclean, M., MacGregor, S. J., Anderson, J. G., & Woolsey, G. (2008). High-intensity narrow-spectrum light inactivation and wavelength sensitivity of Staphylococcus aureus. FEMS Microbiology Letters, 285, 227\u2013232. 61. Hierro, E., Manzano, S., Ord\u00f3\u00f1ez, J.  A., de la Hoz, L., & Fern\u00e1ndez, M. (2009). Inactivation of Salmonella enterica serovar enteritidis on shell eggs by pulsed light technology. International Journal of Food Microbiology, 135, 125-130. 62. Sauer, A., & Moraru, C. I. (2009). Inactivation of Escherichia coli ATCC 25922 and Escherichia coli O157:H7 in apple juice and apple cider, using pulsed light treatment. Journal of Food Protection, 72, 937-944. 63. Jean, J., Morales-Rayas, R., Anoman, M.-N., & Lamhoujeb, S. (2011). Inactivation of hepatitis A virus and norovirus surrogate in suspension and on food-contact surfaces using pulsed UV light (pulsed light inactivation of food-borne viruses). Food Microbiology, 28, 568-572. 64. Palgan, I., Caminiti, I. M., Mu\u00f1oz, A., Noci, F., Whyte, P., Morgan, D. J., Cronin, D. A., & Lyng, J. G. (2011). Effectiveness of High Intensity Light Pulses (HILP) treatments for the control of Escherichia coli and Listeria innocua in apple juice, orange juice and milk. Food Microbiology, 28, 14-20. 65. Cheigh, C. I., Hwang, H. J., & Chung, M. S. (2013). Intense pulsed light (IPL) and UV-C treatments for inactivating Listeria monocytogenes on solid medium and seafoods. Food Research International, 54, 745\u2013752. 66. Luk\u0161iene, \u017d. , Buchovec, I. & Viskelis, P. (2013). Impact of high-power pulsed light on microbial contamination, health promoting components and shelf life of strawberries. Food Technology and Biotechology, 51, 284-292. 229  67. Art\u00edguez M. L., & de Mara\u00f1\u00f3n, I. M. (2015a). Inactivation of spores and vegetative cells of Bacillus subtilis and Geobacillus stearothermophilus by pulsed light. Innovative Food Science and Emerging Technologies, 28, 52-58. 68. Yi, J. Y., Bae, Y.-K., Cheigh, C.-I., & Chung, M.-S. (2017). Microbial inactivation and effects of interrelated factors of intense pulsed light (IPL) treatment for Pseudomonas aeruginosa. LWT - Food Science and Technology, 77, 52-59. 69. Arroyo, C., Dorozko, A., Gaston, E., O\u2019Sullivan, M., Whyte, P., & Lyng, J.  G. (2017). Light based technologies for microbial inactivation of liquids, bead surfaces and powdered infant formula. Food Microbiology, 67, 49-57. 70. Huang, Y., Ye, M., Cao, X., & Chen, H. (2017). Pulsed light inactivation of murine norovirus, Tulane virus, Escherichia coli O157:H7 and Salmonella in suspension and on berry surfaces. Food Microbiology, 61, 1\u20134. 71. Faghihzadeh, F., Anaya, N. M., Hadjeres, H., Boving, T. B., & Oyanedel-Craver, V. (2019). Pulse UV light effect on microbial biomolecules and organic pollutants degradation in aqueous solutions. Chemosphere, 216, 677-683.  72. Smith, W. L., Lagunas-Solar M. C., & Cullor, J. S. (2002). Use of pulsed ultraviolet laser light for the cold pasteurization of bovine milk. Journal of Food Protection, 65, 1480-1482. 73. Pataro, G., Mu\u00f1oz, A., Palgan, I., Noci, F., Ferrari, G., & Lyng, J.G. (2011). Bacterial inactivation in fruit juices using a continuous flow Pulsed Light (PL) system. Food Research International, 44, 1642-1648. 74. Caminiti, I. M., Palgan, I., Noci, F., Mu\u00f1oz, A., Whyte, P., Cronin, D. A., Morgan, D. J., & Lyng, J. G. (2011). The effect of pulsed electric fields (PEF) in combination with high intensity light pulses (HILP) on Escherichia coli inactivation and quality attributes in apple juice. Innovative Food Science and Emerging Technologies, 12, 118\u2013123. 75. Mu\u00f1oz, A., Palgan, I., Noci, F., Morgan, D. J., Cronin, D. A., Whyte, P., & Lyng, J.G. (2011). Combinations of high intensity light pulses and thermosonication for the inactivation of Escherichia coli in orange juice. Food Microbiology, 28, 1200\u20131204. 76. Mu\u00f1oz, A., Caminiti, I. M., Palgan, I., Pataro, G., Noci, F., Morgan, D. J. Denis A.Cronin, D. A., Whyte P., Ferrari, G., & Lyng, J. G. (2012). Effects on Escherichia coli inactivation 230  and quality attributes in apple juice treated by combinations of pulsed light and thermosonication. Food Research International, 45, 299-305. 77. Miller, B. M., Sauer, A., & Moraru, C. I. (2012). Inactivation of Escherichia coli in milk and concentrated milk using pulsed-light treatment. Journal of Dairy Science, 95, 5597\u20135603. 78. Ferrario, M., Alzamora, S. M., & Guerrero, S. (2013). Inactivation kinetics of some microorganisms in apple, melon, orange and strawberry juices by high intensity light pulses. Journal of Food Engineering, 118, 302\u2013311. 79. Anugu, A. K. Microbial inactivation and allergen mitigation of food matrix by pulsed ultraviolet light. Ph.D. dissertation, University of Florida, Florida 2013. URL: http:\/\/ufdc.ufl.edu\/UFE0045406\/00001 80. Innocente, N., Segat, A., Manzocco, L., Marino, M., Maifreni, M., Bortolomeoli, I., Ignat, A., & Nicoli, M. C. (2014). Effect of pulsed light on total microbial count and alkaline phosphatase activity of raw milk. International Dairy Journal, 39, 108\u2013112. 81. Art\u00edguez, M.L., & de Mara\u00f1\u00f3n, I. M. (2015b). Improved process for decontamination of whey by a continuous flow-through pulsed light system. Food Control, 47, 599\u2013605. 82. Ferrario, M., Alzamora, S. M., & Guerrero, S. (2015). Study of pulsed light inactivation and growth dynamics during storage of Escherichia coli ATCC 35218, Listeria innocua ATCC 33090, Salmonella Enteritidis MA44 and Saccharomyces cerevisiae KE162 and native flora in apple, orange and strawberry juices. International Journal of Food Science & Technology, 50, 2498\u20132507. 83. Kasahara, I., Carrasco, V., & Aguilar, L. (2015). Inactivation of Escherichia coli in goat milk using pulsed ultraviolet light. Journal of Food Engineering, 152, 43\u201349. 84. Preetha, P., Venugopal, A. P., Varadharaju, N., & Kennedy, Z. J. (2017). Inactivation of Escherichia coli in Tender Coconut (Cocos Nucifera L.) Water by Pulsed Light Treatment. International Journal of Current Microbiology and Applied Science, 6(7), 1453-1461. 85. Preetha, P., Pandiselvam, R., Varadharaju, N., Kennedy, Z. J., Balakrishnan, M., & Kothakota, A. (2021). Effect of pulsed light treatment on inactivation kinetics of Escherichia coli (MTCC 433) in fruit juices. Food Control, 121, 107547. 231  86. Vollmer, K., Chakraborty, S., Bhalerao, P. P., Carle, R., Frank, J., & Steingass, C. B. (2020). Effect of pulsed light treatment on natural microbiota, enzyme activity, and phytochemical composition of pineapple (Ananas comosus [L.] Merr.) juice. Food and Bioprocess Technology, 13, 1095-1109. 87. Chakraborty, S., Mahale, S., Dhar, R., & Basak, S. (2022). Development of a mixed fruit beverage and pulsed light treatment thereof to obtain a microbially safe and enzymatically stable product. Food Bioscience, 45, 101508. 88. Uesugi, A. R., & Moraru, C. I. (2009). Reduction of Listeria on Ready-to-Eat Sausages after Exposure to a Combination of Pulsed Light and Nisin. Journal of Food Protection, 72, 347-353.  89. Vimont, A., Fliss, I., & Jean, J. (2015). Efficacy and mechanisms of murine norovirus inhibition by pulsed-light technology. Applied and Environmental Microbiology, 81, 2950\u20132957. 90. Dunn, J. E.; Clark, R. W.; Asmus, J. F.; Pearlman, J. S.; Boyer, K.; Pairchaud, F.; Hofmann, G. Methods for preservation of foodstuffs. US Patent 5,034,235, 1991. 91. Pa\u0161kevi\u010di\u016bte, E., Buchovec, I., & Luk\u0161iene, \u017d. (2011). High-power pulsed light for decontamination of chicken from food pathogens: A study on antimicrobial efficiency and organoleptic properties. Journal of Food Safety, 31, 61-68. 92. Lasagabaster, A., Arboleya, J. C., & Mart\u00ednez de Mara\u00f1\u00f3n, I. (2011). Pulsed light technology for surface decontamination of eggs: impact on Salmonella inactivation and egg quality. Innovative Food Science and Emerging Technologies, 12, 124-128. 93. Musgrove, M. T., Jones, D. R., Northcutt, J. K., Cox, N. A., & Harrison, M. A. (2005). Shell rinse and shell crush methods for the recovery of aerobic microorganisms and Enterobacteriaceae from shell eggs. Journal of Food Protection, 68, 2144\u20132148. 94. Huffman, D. E., Slifko, T. R., Salisbury, K., & Rose, J. B. (2000). Inactivation of bacteria, virus and Cryptosporidium by a point-of-use device using pulsed broad spectrum white light. Water Research, 34, 2491-2498. 95. Bialka, K.L., & Demirci, A. (2007). Decontamination of Escherichia coli O157:H7 and Salmonella enterica on blueberries using ozone and pulsed UV-light. Journal of Food Science, 72, 391-396.  232  96. G\u00f3mez-L\u00f3pez, V. M., Devlieghere, F., Bonduelle, V., & Debevere, J. (2005b). Intense light pulses decontamination of minimally processed vegetables and their shelf-life. International Journal of Food Microbiology, 103, 79-89. 97. Ramos-Villarroel, A.Y., Aron-Maftei, N., Mart\u00edn-Belloso, O., & Soliva-Fortuny, R. (2012a). Influence of spectral distribution on bacterial inactivation and quality changes of fresh-cut watermelon treated with intense light pulses. Postharvest Biology and Technology, 69, 32\u201339. 98. Luk\u0161iene, \u017d., Buchovec, I., Kairyte, K., Pa\u0161kevi\u010di\u016bt\u0117, E., & Viskelis, P. (2012). High-power pulsed light for microbial decontamination of some fruits and vegetables with different surfaces. Journal of Food, Agriculture & Environment, 10, 162-167. 99. Valdivia-N\u00e1jar, C. G., Mart\u00edn-Belloso, O., & Soliva-Fortuny, R. (2018). Impact of pulsed light treatments and storage time on the texture quality of fresh-cut tomatoes. Innovative Food Science and Emerging Technologies, 45, 29-35. 100. Lopes, M. M., Silva, E. O., Canuto, K. M., Silva, L. M., Gall\u00e3o, M. I., Urban, L., Fernando Ayala-Zavala, J., & Miranda, M.R.A. (2016). Low fluence pulsed light enhanced phytochemical content and antioxidant potential of \u2018Tommy Atkins\u2019 mango peel and pulp. Innovative Food Science and Emerging Technologies, 33, 216\u2013224. 101. Proulx, J., Hsu, L.C., Miller, B.M., Sullivan, G., Paradis, K., & Moraru, C.I. (2015). Pulsed-light inactivation of pathogenic and spoilage bacteria on cheese surface. Journal of Dairy Science, 98, 5890\u20135898. 102. Can, F. O., Demirci, A., Puri, V. M., & Gourama, H. (2014). Decontamination of hard cheeses by pulsed UV-light. Journal of Food Protection, 77, 1723\u20131731. 103. Dunn, J.; Clark, W. R.; Asmus, J. F.; Pearlman, J. S.; Boyer, K.; Pairchaud, F.; Hofmann, G. Methods and apparatus for preservation of foodstuffs. US Patent Application No. WO 88\/03369, 1988. 104. Shuwaish, A., Figueroa, J. E., & Silva, J. L. (2000). Pulsed-light-treated prepackaged catfish fillets. 2000 IFT Annual Meeting, Dallas, USA. 105. Ozer, N. P., & Demirci, A. (2006). Inactivation of Escherichia coli O157:H7 and Listeria monocytogenes inoculated on raw salmon fillets by pulsed UV-light treatment. International Journal of Food Science and Technology, 41, 354-360. 233  106. Pa\u0161kevi\u010di\u016bte, E., & Luk\u0161iene, \u017d. (2009). High-power pulsed light for decontamination of chicken breast surface. Chemin\u0117 Technologija, 4, 53.  107. Keklik, N. M., Demirci, A., & Puri, V. M. (2010a). Decontamination of unpackaged and vacuum-packaged boneless chicken breast with pulsed UV-light. Poultry Science, 89, 570-581. 108. Hierro, E., Manzano, S., Ord\u00f3\u00f1ez, J.  A., de la Hoz, L., & Fern\u00e1ndez, M. (2011). Efficacy of pulsed light for shelf-life extension and inactivation of Listeria monocytogenes on ready-to-eat cooked meat products. Innovative Food Science and Emerging Technologies, 12, 275\u2013281. 109. Wambura, P., & Verghese, M. (2011). Effect of pulsed ultraviolet light on quality of sliced ham. LWT-Food Science and Technology, 44(10), 2173-2179. 110. Nicorescu, I., Nguyen, B., Chevalier, S., & Orange, N. (2014). Effects of pulsed light on the organoleptic properties and shelf-life extension of pork and salmon. Food Control, 44, 138\u2013145. 111. Manzocco, L., Maifreni, M., Anese, M., Munari, M., Bartolomeoli, I., Zanardi, S., Suman, M., & Nicoli, M.C. (2014). Effect of pulsed light on safety and quality of fresh egg pasta. Food and Bioprocess Technology, 7, 1973\u20131980. 112. Marquenie, D., Michiels, C. W., Van Impe, J. F., Schrevens, E., & Nicola\u00ef, B. N. (2003a). Pulsed white light in combination with UV-C and heat to reduce storage rot of strawberry. Postharvest Biology and Technology, 28, 455-461. 113. Valdivia-N\u00e1jar, C. G., Mart\u00edn-Belloso, O., Giner-Segu\u00ed, J., & Soliva-Fortuny, R. (2017). Modeling the Inactivation of Listeria innocua and Escherichia coli in Fresh-Cut Tomato Treated with Pulsed Light. Food and Bioprocess Technology, 10, 266-274. 114. Tonon, F., & Agoulon, A. (2003). Lumiere pulse, principe et application au cas des solutions liquides. Industries Agro-alimentaires, la conservation de demain, 4th ed., Talence, France. 115. Elmnasser, N., Dalgalarrondo, M., Orange, N., Bakhrouf, A., Haertl\u00e9, T., Federighi, M., & Chobert, J. M. (2008). Effect of pulsed-light treatment on milk proteins and lipids. Journal of Agricultural and Food Chemistry, 56, 1984-1991.  116. Fern\u00e1ndez, M., Ganan, M., Guerra, C., & Hierro, E. (2014). Protein oxidation in processed cheese slices treated with pulsed light technology. Food Chemistry, 159, 388-390. 234  117. Cantos, E., Esp\u00edn, J. C., & Tom\u00e1s-Barber\u00e1n, F. A. (2001). Postharvest induction modeling method using UV irradiation pulses for obtaining resveratrol-enriched table grapes: a new \u2018\u2018functional\u2019\u2019 fruit? Journal of Agricultural and Food Chemistry, 49, 5052-5058.  118. Aguil\u00f3-Aguayo, I., Charles, F., Renard, C.M., Page, D., & Carlin, F. (2013). Pulsed light effects on surface decontamination, physical qualities and nutritional composition of tomato fruit. Postharvest Biology and Technology, 86, 29\u201336. 119. Ag\u00fcero, M. V., Jagus, R. J., Mart\u00edn-Belloso, O., & Soliva-Fortuny, R. (2016). Surface decontamination of spinach by intense pulsed light treatments: Impact on quality attributes. Postharvest Biology and Technology, 121,118\u2013125. 120. Pataro, G., Sinik, M., Capitoli, M. M., Dons\u00ec, G. & Ferrari, G. (2015). The influence of post-harvest UV-C and pulsed light treatments on quality and antioxidant properties of tomato fruits during storage. Innovative Food Science and Emerging Technologies, 30, 103\u2013111. 121. Braga, T. R., Silva, E. O., Rodrigues, S., & Fernandes, F. A. (2019). Drying of mangoes (Mangifera indica L.) applying pulsed UV light as pretreatment. Food and Bioproducts Processing, 114, 95-102.  122. Wiktor, A., Mandal, R., Singh, A., & Pratap Singh, A. (2019). Pulsed Light treatment below a Critical Fluence (3.82 J\/cm2) minimizes photo-degradation and browning of a model Phenolic (Gallic Acid) Solution. Foods, 8(9), 380. 123. G\u00f3mez-L\u00f3pez, V. M., Pataro, G., Tiwari, B., Gozzi, M., Meireles, M. \u00c1. A., Wang, S., ... & Morata, A. (2021). Guidelines on reporting treatment conditions for emerging technologies in food processing. Critical Reviews in Food Science and Nutrition, 1-25. 124. Choi, M. S., Cheigh, C. I., Jeong, E. A., Shin, J. K., & Chung, M. S. (2010). Nonthermal sterilization of Listeria monocytogenes in infant foods by intense pulsed-light treatment. Journal of Food Engineering, 97, 504-509.  125. Hwang, H. J., Cheigh, C. I., & Chung, M. S. (2015). Relationship between optical properties of beverages and microbial inactivation by intense pulsed light. Innovative Food Science & Emerging Technologies, 31, 91-96. 126. Art\u00edguez, M. L., Lasagabaster, A., & de Mara\u00f1\u00f3n, I. M. (2011). Factors affecting microbial inactivation by Pulsed Light in a continuous flow-through unit for liquid products treatment. Procedia Food Science, 1, 786-791. 235  127. Ferrario, M., & Guerrero, S. (2016). Effect of a continuous flow-through pulsed light system combined with ultrasound on microbial survivability, color and sensory shelf life of apple juice. Innovative Food Science & Emerging Technologies, 34, 214-224. 128. Wang, B., Wei, W., Zhang, Y., Xu, H., & Ma, H. (2022). Decontamination and quality assessment of freshly squeezed grape juice under spiral continuous flow\u2010through pulsed light (SCFPL) treatment. Journal of Food Processing and Preservation, 46(1), e16186. 129. At\u0131lgan, M. R. (2013). Design of a continuous flow UV reactor for opaque liquid foods by using computational fluid dynamics (CFD). PhD Dissertation. \u0130zmir Institute of Technology 130. Koutchma, T., Parisi, B., & Unluturk, S. K. (2006). Evaluation of UV dose in flow-through reactors for fresh apple juice and cider. Chemical Engineering Communications, 193(6), 715-728. 131. Guerrero-Beltr\u00e9n, J. A., & Barbosa-C\u00e9novas, G. V. (2006). Inactivation of Saccharomyces cerevisiae and polyphenoloxidase in mango nectar treated with UV light. Journal of Food Protection, 69(2), 362-368. 132. Mandal, R., Wiktor, A., Mohammadi, X., & Pratap-Singh, A. (2021). Pulsed UV Light Irradiation Processing of Black Tea Infusions: Effect on Color, Phenolic Content, and Antioxidant Capacity. Food and Bioprocess Technology, 1-13. 133. Koutchma, T. (2009). Advances in ultraviolet light technology for non-thermal processing of liquid foods. Food and Bioprocess Technology, 2(2), 138-155. 134. Dean, W. R. (1927). XVI. Note on the motion of fluid in a curved pipe. The London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science, 4(20), 208-223. 135. Mandal, R., & Pratap-Singh, A. (2021). Characterization of continuous-flow pulsed UV light reactors for processing of liquid foods in annular tube and coiled tube configurations using actinometry and computational fluid dynamics. Journal of Food Engineering, 304, 110590. 136. Bohrerova, Z., Shemer, H., Lantis, R., Impellitteri, C. A., & Linden, K. G. (2008). Comparative disinfection efficiency of pulsed and continuous-wave UV irradiation technologies. Water research, 42(12), 2975-2982. 137. Rahn, R. O., Stefan, M. I., Bolton, J. R., Goren, E., Shaw, P. S., & Lykke, K. R. (2003). Quantum Yield of the Iodide\u2013Iodate Chemical Actinometer: Dependence on Wavelength and Concentration. Photochemistry and Photobiology, 78(2), 146-152. 236  138. Sozzi, D. A., & Taghipour, F. (2006). UV reactor performance modeling by Eulerian and Lagrangian methods. Environmental science & technology, 40(5), 1609-1615. 139. Duran, J. E., Mohseni, M., & Taghipour, F. (2015). Computational modeling of UV photocatalytic reactors: model development, evaluation, and application. Water Quality Research Journal of Canada, 50(1), 21-33. 140. Casado, C., Marug\u00e1n, J., Timmers, R., Mu\u00f1oz, M., & van Grieken, R. (2017). Comprehensive multiphysics modeling of photocatalytic processes by computational fluid dynamics based on intrinsic kinetic parameters determined in a differential photoreactor. Chemical Engineering Journal, 310, 368-380. 141. Younis, B. A., Mahoney, L., & Palomo, N. (2018). A novel system for water disinfection with UV radiation. Water, 10(9), 1275. 142. ISO 6731 (2010). [IDF 21:2010]: Milk, cream and evaporated milk \u2014 Determination of total solids content (Reference method). International Organization for Standardization, Geneva, Switzerland. 143. Hsu, L., & Moraru, C. I. (2011). Quantifying and mapping the spatial distribution of fluence inside a pulsed light treatment chamber and various liquid substrates. Journal of Food Engineering, 103(1), 84-91. 144. Gnielinski, V. (2010). G2 Heat transfer in concentric annular and parallel plate ducts. In e. V. VDI (Ed.) VDI Heat Atlas. Springer, Heidelberg, Berlin, Germany. pp. 701-708. 145. Ligrani, Phillip M. \"A Study of Dean Vortex Development and Structure in a Curved Rectangular Channel With Aspect Ratio of 40 at Dean Numbers up to 430\", U.S. Army Research Laboratory (Contractor Report ARL-CR-l44) and Lewis Research Center (NASA Contractor Report 4607), July 1994. (Retrieved on 10 August 2020). 146. United States Environmental Protection Agency (USEPA), 1996. Ultraviolet light disinfection technology in drinking water application \u2013 an overview. EPA 811-R-96-002. URL: http:\/\/www.epa.gov\/nscep\/index.html    (Retrieved on 5th March 2020) 147. M\u00fcller, A., Stahl, M. R., Greiner, R., & Posten, C. (2014). Performance and dose validation of a coiled tube UV-C reactor for inactivation of microorganisms in absorbing liquids. Journal of Food Engineering, 138, 45-52. 237  148. Jonasz, M., & Fournier, G. (2011). Light scattering by particles in water: theoretical and experimental foundations. Academic Press, Elsevier, London, United Kingdom. 149. Aernouts, B., Van Beers, R., Watt\u00e9, R., Huybrechts, T., Lammertyn, J., & Saeys, W. (2015). Visible and near-infrared bulk optical properties of raw milk. Journal of Dairy Science, 98(10), 6727-6738. 150. Refractive index database. https:\/\/refractiveindex.info\/?shelf=other&book=air&page=Ciddor 151. Properties of fused silica. https:\/\/www.sydor.com\/wp-content\/uploads\/2019\/06\/Tosoh-ES-ED-H-Series-Fused-Silica.pdf 152. Index of refraction. http:\/\/hyperphysics.phy-astr.gsu.edu\/hbase\/Tables\/indrf.html 153. Walstra, P., Walstra, P., Wouters, J. T., & Geurts, T. J. (2005). Dairy science and technology. CRC Press, Boca Raton, Florida. 154. Viscosity of air, dynamic and kinematic. https:\/\/www.engineersedge.com\/physics\/viscosity_of_air_dynamic_and_kinematic_14483.htm 155. Air - Dynamic and Kinematic Viscosity. https:\/\/www.engineeringtoolbox.com\/air-absolute-kinematic-viscosity-d_601.html?vA=25&units=C# 156. Thermal conductivity. http:\/\/hyperphysics.phy-astr.gsu.edu\/hbase\/Tables\/thrcn.html 157. Properties of fused quartz. https:\/\/technicalglass.com\/technical_properties\/ 158. Specific heat. http:\/\/hyperphysics.phy-astr.gsu.edu\/hbase\/thermo\/spht.html 159. Xu, C., Zhao, X. S., & Rangaiah, G. P. (2013). Performance analysis of ultraviolet water disinfection reactors using computational fluid dynamics simulation. Chemical Engineering Journal, 221, 398-406. 160. Crofcheck, C., Wade, J., Swamy, J. N., Aslan, M. M., & Meng\u00fc\u00e7, M. P. (2005). Effect of fat and casein particles in milk on the scattering of elliptically polarized light. Transactions of the ASAE, 48(3), 1147-1155. 161. Koutchma, T., Parisi, B., & Patazca, E. (2007). Validation of UV coiled tube reactor for fresh juices. Journal of Environmental Engineering and Science, 6(3), 319-328. 162. Choudhary, R., Bandla, S., Watson, D. G., Haddock, J., Abughazaleh, A., & Bhattacharya, B. (2011). Performance of coiled tube ultraviolet reactors to inactivate Escherichia coli 238  W1485 and Bacillus cereus endospores in raw cow milk and commercially processed skimmed cow milk. Journal of Food Engineering, 107(1), 14-20. 163. M\u00fcller, A., Orlowska, M., Kn\u00f6rr, M., Stahl, M. R., Greiner, R., & Koutchma, T. (2017). Actinometric and biodosimetric evaluation of UV-C dose delivery in annular, Taylor\u2013Coutte and coiled tube continuous systems. Food Science and Technology International, 23(3), 222-234. 164. Koutchma, T., & Parisi, B. (2004). Biodosimetry of Escherichia coli UV inactivation in model juices with regard to dose distribution in annular UV reactors. Journal of Food Science, 69(1), FEP14-FEP22. 165. Gay\u00e1n, E., Monfort, S., \u00c1lvarez, I., & Cond\u00f3n, S. (2011). UV-C inactivation of Escherichia coli at different temperatures. Innovative Food Science & Emerging Technologies, 12(4), 531-541. 166. Vashisht, P., Pendyala, B., Gopisetty, V. V. S., & Patras, A. (2021). Modeling and validation of delivered fluence of a continuous Dean flow pilot scale UV system: monitoring fluence by biodosimetry approach. Food Research International, 148, 110625. 167. Bialka, K. L., Demirci, A., & Puri, V. M. (2008). Modeling the inactivation of Escherichia coli O157: H7 and Salmonella enterica on raspberries and strawberries resulting from exposure to ozone or pulsed UV-light. Journal of Food Engineering, 85(3), 444-449. 168. John, D., & Ramaswamy, H. S. (2020). Comparison of pulsed light inactivation kinetics and modeling of Escherichia coli (ATCC-29055), Clostridium sporogenes (ATCC-7955) and Geobacillus stearothermophilus (ATCC-10149). Current Research in Food Science, 3, 82-91. 169. Lasagabaster, A., & de Mara\u00f1\u00f3n, I. M. (2013). Impact of process parameters on Listeria innocua inactivation kinetics by pulsed light technology. Food and Bioprocess Technology, 6(7), 1828-1836. 170. Geeraerd, A. H., Valdramidis, V. P., & Van Impe, J. F. (2005). GInaFiT, a freeware tool to assess non-log-linear microbial survivor curves. International Journal of Food Microbiology, 102(1), 95-105. 171. Bigelow, W. D., & Esty, J. R. (1920). The thermal death point in relation to time of typical thermophilic organisms. The Journal of Infectious Diseases, 602-617. 239  172. Mafart, P., Couvert, O., Gaillard, S., & Legu\u00e9rinel, I. (2002). On calculating sterility in thermal preservation methods: application of the Weibull frequency distribution model. International Journal of Food Microbiology, 72(1-2), 107-113. 173. Geeraerd, A. H., Herremans, C. H., & Van Impe, J. F. (2000). Structural model requirements to describe microbial inactivation during a mild heat treatment. International Journal of Food Microbiology, 59(3), 185-209. 174. Xu, F., Wang, B., Hong, C., Telebielaigen, S., Nsor-Atindana, J., Duan, Y., & Zhong, F. (2019). Optimization of spiral continuous flow-through pulse light sterilization for Escherichia coli in red grape juice by response surface methodology. Food Control, 105, 8-12. 175. Izquier, A., & G\u00f3mez-L\u00f3pez, V. M. (2011). Modeling the pulsed light inactivation of microorganisms naturally occurring on vegetable substrates. Food Microbiology, 28(6), 1170-1174. 176. Fenaille, F., Parisod, V., Visani, P., Populaire, S., Tabet, J. C., & Guy, P. A. (2006). Modifications of milk constituents during processing: A preliminary benchmarking study. International Dairy Journal, 16(7), 728-739. 177. Scheidegger, D., Pecora, R. P., Radici, P. M., & Kivatinitz, S. C. (2010). Protein oxidative changes in whole and skim milk after ultraviolet or fluorescent light exposure. Journal of Dairy Science, 93(11), 5101-5109. 178. Guneser, O. N. U. R., & Yuceer, Y. K. (2012). Effect of ultraviolet light on water-and fat-soluble vitamins in cow and goat milk. Journal of Dairy Science, 95(11), 6230-6241. 179. Ashoor, S. H., Seperich, G. J., Monte, W. C., & Welty, J. (1983). HPLC determination of riboflavin in eggs and dairy products. Journal of Food Science, 48(1), 92-94. 180. \u00d8stdal, H., Andersen, H. J., & Nielsen, J. H. (2000). Antioxidative activity of urate in bovine milk. Journal of Agricultural and Food Chemistry, 48(11), 5588-5592. 181. Pereda, J., Ferragut, V., Quevedo, J. M., Guamis, B., & Trujillo, A. J. (2008). Effects of ultra-high-pressure homogenization treatment on the lipolysis and lipid oxidation of milk during refrigerated storage. Journal of Agricultural and Food Chemistry, 56(16), 7125-7130. 182. Hu, G., Zheng, Y., Wang, D., Zha, B., Liu, Z., & Deng, Y. (2015). Comparison of microbiological loads and physicochemical properties of raw milk treated with single-240  \/multiple-cycle high hydrostatic pressure and ultraviolet-C light. High Pressure Research, 35(3), 330-338. 183. Sheraz, M. A., Kazi, S. H., Ahmed, S., Anwar, Z., & Ahmad, I. (2014). Photo, thermal and chemical degradation of riboflavin. Beilstein Journal of organic chemistry, 10(1), 1999-2012. 184. King, C. G., & Waugh, W. A. (1934). The effect of pasteurization upon the vitamin C content of milk. Journal of Dairy Science, 17(7), 489-496. 185. Scott, K. J., & Bishop, D. R. (1986). Nutrient content of milk and milk products: vitamins of the B complex and vitamin C in retail market milk and milk products. International Journal of Dairy Technology, 39(1), 32-35. 186. Hand, D. B., & Sharp, P. F. (1939). The riboflavin content of cow's milk. Journal of Dairy Science, 22(10), 779-783. 187. Kang, M. W., Chen, D., Ruan, R., & Vickers, Z. M. (2021). The effect of intense pulsed light on the sensory properties of nonfat dry milk. Journal of Food Science, 86(9), 4119-4133. 188. Przybylska, S. (2020). Lycopene\u2013a bioactive carotenoid offering multiple health benefits: a review. International Journal of Food Science & Technology, 55(1), 11-32. 189. Juan, M. E., Lamuela-Ravent\u00f3s, R. M., de la Torre-Boronat, M. C., & Planas, J. M. (1999). Determination of trans-resveratrol in plasma by HPLC. Analytical chemistry, 71(3), 747-750. 190. Patras, A., Tiwari, B. K., Brunton, N. P., & Butler, F. (2009). Modelling the effect of different sterilisation treatments on antioxidant activity and colour of carrot slices during storage. Food Chemistry, 114(2), 484-491. 191. Hasan, M. M., Yun, H. K., Kwak, E. J., & Baek, K. H. (2014). Preparation of resveratrol-enriched grape juice from ultrasonication treated grape fruits. Ultrasonics sonochemistry, 21(2), 729-734. 192. Derradji-Benmeziane, F., Djamai, R., & Cadot, Y. (2014). Antioxidant capacity, total phenolic, carotenoid, and vitamin C contents of five table grape varieties from Algeria and their correlations. OENO One, 48(2), 153-162. 241  193. Nile, S. H., Kim, D. H., & Keum, Y. S. (2015). Determination of anthocyanin content and antioxidant capacity of different grape varieties. Ci\u00eancia e T\u00e9cnica Vitivin\u00edcola, 30(2), 60-68. 194. Oms-Oliu, G., Odriozola-Serrano, I., Soliva-Fortuny, R., & Mart\u00edn-Belloso, O. (2009). Effects of high-intensity pulsed electric field processing conditions on lycopene, vitamin C and antioxidant capacity of watermelon juice. Food Chemistry, 115(4), 1312-1319. 195. Feng, M., Ghafoor, K., Seo, B., Yang, K., & Park, J. (2013). Effects of ultraviolet-C treatment in Teflon\u00ae-coil on microbial populations and physico-chemical characteristics of watermelon juice. Innovative Food Science & Emerging Technologies, 19, 133-139. 196. Chen, J., Shi, J., Xue, S. J., & Ma, Y. (2009). Comparison of lycopene stability in water-and oil-based food model systems under thermal-and light-irradiation treatments. LWT-Food Science and Technology, 42(3), 740-747. 197. Karaoglan, H. A., Keklik, N. M., & Develi Is\u0131kl\u0131, N. (2019). Degradation kinetics of anthocyanin and physicochemical changes in fermented turnip juice exposed to pulsed UV light. Journal of Food Science and Technology, 56(1), 30-39. 198. Tran, M. T. T., & Farid, M. (2004). Ultraviolet treatment of orange juice. Innovative Food Science & Emerging Technologies, 5(4), 495-502. 199. Art\u00e9s-Hern\u00e1ndez, F., Robles, P. A., G\u00f3mez, P. A., Tom\u00e1s-Callejas, A., & Art\u00e9s, F. (2010). Low UV-C illumination for keeping overall quality of fresh-cut watermelon. Postharvest Biology and Technology, 55(2), 114-120. 200. Silva, C. G., Monteiro, J., Marques, R. R., Silva, A. M., Mart\u00ednez, C., Canle, M., & Faria, J. L. (2013). Photochemical and photocatalytic degradation of trans-resveratrol. Photochemical & Photobiological Sciences, 12(4), 638-644. 201. Noci, F., Riener, J., Walkling-Ribeiro, M., Cronin, D. A., Morgan, D. J., & Lyng, J. G. (2008). Ultraviolet irradiation and pulsed electric fields (PEF) in a hurdle strategy for the preservation of fresh apple juice. Journal of Food Engineering, 85(1), 141-146. 202. Tiwari, B. K., Patras, A., Brunton, N., Cullen, P. J., & O\u2019donnell, C. P. (2010). Effect of ultrasound processing on anthocyanins and color of red grape juice. Ultrasonics sonochemistry, 17(3), 598-604. 242  203. Caminiti, I. M., Noci, F., Mu\u00f1oz, A., Whyte, P., Morgan, D. J., Cronin, D. A., & Lyng, J. G. (2011). Impact of selected combinations of non-thermal processing technologies on the quality of an apple and cranberry juice blend. Food Chemistry, 124(4), 1387-1392. 204. Liu, L. H., Zabaras, D., Bennett, L. E., Aguas, P., & Woonton, B. W. (2009). Effects of UV-C, red light and sun light on the carotenoid content and physical qualities of tomatoes during post-harvest storage. Food Chemistry, 115(2), 495-500. 243  Appendices Appendix A1 Supplemental information for Chapter 3:   Figure A1-1: PL equipment setup: Annular and coiled tube reactors, PL lamp and controller  Figure A1-2: PL flashlamp emission spectrum (200-670 nm) as provided by Solaris Disinfection Inc. 244   Figure A1-3: Standard curve for conductivity vs concentration of NaCl used in residence time distribution studies.  Figure A1-4: Predicted energy as per Eq. 3.21 versus experimental energy values    245   Figure A1-5: Residence time distribution (curve) showing NaCl concentration vs time for real and ideal reactors. For ideal reactor, the spread of the curve is narrower compared to the real reactor.              246  Appendix A2 Supplemental information for Chapter 6:  Figure A2-1: Standard curve for vitamin C standard solutions used for quantification of vitamin C in milk samples  Figure A2-2: Standard curve for vitamin B2 standard solutions used for quantification of vitamin B2 in milk samples  247   Figure A2-3: Chromatogram for determination of Vitamin C content in vitamin C standard solution. Peak is shown at retention time of around 2 min   Figure A2-4: Chromatogram for determination of Vitamin C content in milk samples. Peak is shown at retention time of around 2 min  Figure A2-5: Chromatogram for determination of Vitamin B2 content in vitamin B2 standard solution. Peak is shown at retention time of around 5.7 min  248   Figure A2-6: Chromatogram for determination of Vitamin B2 content in milk samples. Peak is shown at retention time of around 5.7 min                  249  Appendix A3 Supplemental information for Chapter 7:  Figure A3-1: Standard curve of gallic acid solutions for estimation of total phenolic content in juice samples  Figure A3-2: Standard curve for trans-Resveratrol standard solutions used for quantification of trans-Resveratrol in grape juice samples 250   Figure A3-3: Chromatogram for determination of Vitamin C content in watermelon juice samples. Peak is shown at retention time of around 2 min  Figure A3-4: Chromatogram for determination of trans-Resveratrol content in trans-Resveratrol standard solution. Peak is shown at retention time of around 17 min  Figure A3-5: Chromatogram for determination of trans-Resveratrol content in grape juice sample. Peak is shown at retention time of around 17 min  ","attrs":{"lang":"en","ns":"http:\/\/www.w3.org\/2009\/08\/skos-reference\/skos.html#note","classmap":"oc:AnnotationContainer"},"iri":"http:\/\/www.w3.org\/2009\/08\/skos-reference\/skos.html#note","explain":"Simple Knowledge Organisation System; Notes are used to provide information relating to SKOS concepts. There is no restriction on the nature of this information, e.g., it could be plain text, hypertext, or an image; it could be a definition, information about the scope of a concept, editorial information, or any other type of information."}],"Genre":[{"label":"Genre","value":"Thesis\/Dissertation","attrs":{"lang":"en","ns":"http:\/\/www.europeana.eu\/schemas\/edm\/hasType","classmap":"dpla:SourceResource","property":"edm:hasType"},"iri":"http:\/\/www.europeana.eu\/schemas\/edm\/hasType","explain":"A Europeana Data Model Property; This property relates a resource with the concepts it belongs to in a suitable type system such as MIME or any thesaurus that captures categories of objects in a given field. It does NOT capture aboutness"}],"GraduationDate":[{"label":"Graduation Date","value":"2022-11","attrs":{"lang":"en","ns":"http:\/\/vivoweb.org\/ontology\/core#dateIssued","classmap":"vivo:DateTimeValue","property":"vivo:dateIssued"},"iri":"http:\/\/vivoweb.org\/ontology\/core#dateIssued","explain":"VIVO-ISF Ontology V1.6 Property; Date Optional Time Value, DateTime+Timezone Preferred "}],"IsShownAt":[{"label":"DOI","value":"10.14288\/1.0416287","attrs":{"lang":"en","ns":"http:\/\/www.europeana.eu\/schemas\/edm\/isShownAt","classmap":"edm:WebResource","property":"edm:isShownAt"},"iri":"http:\/\/www.europeana.eu\/schemas\/edm\/isShownAt","explain":"A Europeana Data Model Property; An unambiguous URL reference to the digital object on the provider\u2019s website in its full information context."}],"Language":[{"label":"Language","value":"eng","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/language","classmap":"dpla:SourceResource","property":"dcterms:language"},"iri":"http:\/\/purl.org\/dc\/terms\/language","explain":"A Dublin Core Terms Property; A language of the resource.; Recommended best practice is to use a controlled vocabulary such as RFC 4646 [RFC4646]."}],"Program":[{"label":"Program (Theses)","value":"Food Science","attrs":{"lang":"en","ns":"https:\/\/open.library.ubc.ca\/terms#degreeDiscipline","classmap":"oc:ThesisDescription","property":"oc:degreeDiscipline"},"iri":"https:\/\/open.library.ubc.ca\/terms#degreeDiscipline","explain":"UBC Open Collections Metadata Components; Local Field; Indicates the program for which the degree was granted."}],"Provider":[{"label":"Provider","value":"Vancouver : University of British Columbia Library","attrs":{"lang":"en","ns":"http:\/\/www.europeana.eu\/schemas\/edm\/provider","classmap":"ore:Aggregation","property":"edm:provider"},"iri":"http:\/\/www.europeana.eu\/schemas\/edm\/provider","explain":"A Europeana Data Model Property; The name or identifier of the organization who delivers data directly to an aggregation service (e.g. Europeana)"}],"Publisher":[{"label":"Publisher","value":"University of British Columbia","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/publisher","classmap":"dpla:SourceResource","property":"dcterms:publisher"},"iri":"http:\/\/purl.org\/dc\/terms\/publisher","explain":"A Dublin Core Terms Property; An entity responsible for making the resource available.; Examples of a Publisher include a person, an organization, or a service."}],"Rights":[{"label":"Rights","value":"Attribution-NonCommercial-NoDerivatives 4.0 International","attrs":{"lang":"*","ns":"http:\/\/purl.org\/dc\/terms\/rights","classmap":"edm:WebResource","property":"dcterms:rights"},"iri":"http:\/\/purl.org\/dc\/terms\/rights","explain":"A Dublin Core Terms Property; Information about rights held in and over the resource.; Typically, rights information includes a statement about various property rights associated with the resource, including intellectual property rights."}],"RightsURI":[{"label":"Rights URI","value":"http:\/\/creativecommons.org\/licenses\/by-nc-nd\/4.0\/","attrs":{"lang":"*","ns":"https:\/\/open.library.ubc.ca\/terms#rightsURI","classmap":"oc:PublicationDescription","property":"oc:rightsURI"},"iri":"https:\/\/open.library.ubc.ca\/terms#rightsURI","explain":"UBC Open Collections Metadata Components; Local Field; Indicates the Creative Commons license url."}],"ScholarlyLevel":[{"label":"Scholarly Level","value":"Graduate","attrs":{"lang":"en","ns":"https:\/\/open.library.ubc.ca\/terms#scholarLevel","classmap":"oc:PublicationDescription","property":"oc:scholarLevel"},"iri":"https:\/\/open.library.ubc.ca\/terms#scholarLevel","explain":"UBC Open Collections Metadata Components; Local Field; Identifies the scholarly level of the author(s)\/creator(s)."}],"Supervisor":[{"label":"Supervisor","value":"Pratap-Singh, Anubhav","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/contributor","classmap":"vivo:AdvisingRelationship","property":"dcterms:contributor"},"iri":"http:\/\/purl.org\/dc\/terms\/contributor","explain":"A Dublin Core Terms Property; An entity responsible for making contributions to the resource.; Examples of a Contributor include a person, an organization, or a service."}],"Title":[{"label":"Title ","value":"Design, development, and validation of continuous-flow pulsed UV light systems for liquid food processing","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/title","classmap":"dpla:SourceResource","property":"dcterms:title"},"iri":"http:\/\/purl.org\/dc\/terms\/title","explain":"A Dublin Core Terms Property; The name given to the resource."}],"Type":[{"label":"Type","value":"Text","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/type","classmap":"dpla:SourceResource","property":"dcterms:type"},"iri":"http:\/\/purl.org\/dc\/terms\/type","explain":"A Dublin Core Terms Property; The nature or genre of the resource.; Recommended best practice is to use a controlled vocabulary such as the DCMI Type Vocabulary [DCMITYPE]. To describe the file format, physical medium, or dimensions of the resource, use the Format element."}],"URI":[{"label":"URI","value":"http:\/\/hdl.handle.net\/2429\/82156","attrs":{"lang":"en","ns":"https:\/\/open.library.ubc.ca\/terms#identifierURI","classmap":"oc:PublicationDescription","property":"oc:identifierURI"},"iri":"https:\/\/open.library.ubc.ca\/terms#identifierURI","explain":"UBC Open Collections Metadata Components; Local Field; Indicates the handle for item record."}],"SortDate":[{"label":"Sort Date","value":"2022-12-31 AD","attrs":{"lang":"en","ns":"http:\/\/purl.org\/dc\/terms\/date","classmap":"oc:InternalResource","property":"dcterms:date"},"iri":"http:\/\/purl.org\/dc\/terms\/date","explain":"A Dublin Core Elements Property; A point or period of time associated with an event in the lifecycle of the resource.; Date may be used to express temporal information at any level of granularity. Recommended best practice is to use an encoding scheme, such as the W3CDTF profile of ISO 8601 [W3CDTF].; A point or period of time associated with an event in the lifecycle of the resource.; Date may be used to express temporal information at any level of granularity. Recommended best practice is to use an encoding scheme, such as the W3CDTF profile of ISO 8601 [W3CDTF]."}]}