UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Radiant energy vacuum microwave microencapsulation of natural antimicrobials for a controlled release… Sáenz Garza, Natalia Edith 2013

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2013_spring_saenzgarza_natalia.pdf [ 3.61MB ]
Metadata
JSON: 24-1.0073556.json
JSON-LD: 24-1.0073556-ld.json
RDF/XML (Pretty): 24-1.0073556-rdf.xml
RDF/JSON: 24-1.0073556-rdf.json
Turtle: 24-1.0073556-turtle.txt
N-Triples: 24-1.0073556-rdf-ntriples.txt
Original Record: 24-1.0073556-source.json
Full Text
24-1.0073556-fulltext.txt
Citation
24-1.0073556.ris

Full Text

Radiant Energy Vacuum Microwave Microencapsulation of Natural Antimicrobials for a Controlled Release Application in Fresh-Cut Ambrosia Apples by  Natalia Edith Sáenz Garza B. Sc. Food Industry Engineering, Tecnológico de Monterrey, 2007  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY  in  The Faculty of Graduate Studies  (Food Science)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  January 2013 © Natalia Edith Sáenz Garza, 2013  Abstract Microencapsulation of active compounds provides protection, ease of handling and dosing, as well as a controlled release. Fresh-cut fruit has a limited shelf-life which could be improved using natural antimicrobials in a microencapsulated form to achieve a controlled release. Vanillin and hexanal are two antimicrobials that are compatible with the food model chosen in this study, namely fresh-cut apple slices. Both vanillin and hexanal were microencapsulated in β-cyclodextrin-pectin blends by radiant energy vacuum (REV) drying and freeze-drying (FD). Although the encapsulation efficiency of vanillin in the REV-processed β-cyclodextrin-pectin blends was high (66%), a fast release into aqueous media and the textural effect of the encapsulant debris on the apple slices limited its practical application. Hexanal’s microencapsulation yield was not as high as that of vanillin, however, it was improved from 20% to 47%, for the most microbially efficient formulation, using a pilot scale 3.6 kW REV dehydrator. Hexanal containing microcapsules were tested for inhibition of Penicillium expansum in potato dextrose agar and on fresh-cut apple slices. Inhibition of P. expansum was examined by measurement of radial growth on the surface of potato dextrose agar (PDA) plates incubated for 5 days at 25°C and 10 days at 12°C. Apple slices inoculated with P. expansum spores were stored in sealed glass jars for 15 days and 12 days at 5°C and 12°C respectively. Shelf-life of fresh-cut apple slices was evaluated as inhibition of mould and secondary browning relative to control samples. Hexanal release profiles were quantified using solid phase microextraction-gas chromatography (SPMEGC) sampling of the headspace. Spore germination on PDA was inhibited for 5 days and radial growth for 10 days with 5.4-6.2 μg hexanal/mL air. Although FD gave a higher encapsulation yield, REV encapsulation produced the greatest inhibition of mould in PDA. Un-encapsulated hexanal inhibited mould on apple but caused rapid browning and softening of apple due to  ii  phytotoxicity. Encapsulated hexanal delayed quality loss without apparent phytotoxicity. REV and FD hexanal microcapsules produced similar shelf-life for up to 15 days at 5°C. Hexanal microencapsulation for a headspace application could present a viable strategy to extend the short-term shelf-life of fresh-cut fruit.  iii  Preface A version of chapter 4 has been accepted for publication in Food Research International (Elsevier) for the Special Issue (Innovation for Better Food).  Sáenz Garza, N. Delaquis, P., Durance, T. (in press) Microencapsulation of Hexanal by Radiant Energy Vacuum Microwave-Molecular Inclusion for Controlled Release and Inhibition of Penicillium expansum in a Model System and on Apple Tissue. Food Research International.  I conducted the experiments and wrote the manuscript. Delaquis, P. and Durance, T. supervised the experiments and edited the manuscript.  iv  Table of Contents Abstract ................................................................................................................................ ii Preface ................................................................................................................................ iv Table of Contents .................................................................................................................v List of Tables ....................................................................................................................... ix List of Figures ...................................................................................................................... xi Glossary of Terms ............................................................................................................. xiv Acknowledgements .......................................................................................................... xvii Dedication ....................................................................................................................... xviii 1 Introduction .................................................................................................................... 1 2 Literature Review ........................................................................................................... 2 2.1 General aspects of microencapsulation ................................................................... 2 2.1.1  Microcapsule characteristics ............................................................................ 3  2.2 Microencapsulation processes ................................................................................ 4 2.2.1  Spray drying ..................................................................................................... 5  2.2.2  Freeze-drying or lyophilization ......................................................................... 6  2.2.3  REV technology................................................................................................ 8  2.3 Preservation of fresh-cut fruit ................................................................................ 10 2.3.1  Penicillium expansum..................................................................................... 15  2.3.2  Strategies to control P. expansum in apples. ................................................. 18  2.4 Natural antimicrobials proposed ............................................................................ 19 2.4.1  Vanillin as an antimicrobial ............................................................................. 19  2.4.2  Hexanal as an antimicrobial ........................................................................... 22  2.5 Attempts to microencapsulate vanillin and hexanal ............................................... 28 2.5.1  Vanillin microencapsulation ............................................................................ 28  2.5.2  Hexanal microencapsulation .......................................................................... 29 v  2.6 Wall materials selected: ......................................................................................... 30 2.6.1  Beta cyclodextrin (CD) ................................................................................. 30  2.6.2  Pectin ............................................................................................................. 32  2.6.3  Criteria for selection of wall materials ............................................................. 34  2.7 Controlled release ................................................................................................. 36 2.7.1  Volatile controlled release .............................................................................. 37  2.8 Instrumental techniques ........................................................................................ 38 2.8.1  Gas chromatography (GC) ............................................................................. 38  2.8.2  Solid phase microextraction (SPME) .............................................................. 39  3 Vanillin Project: Antimicrobial Controlled Release on Fresh-Cut Surface .................... 42 3.1 Research objectives .............................................................................................. 42 3.2 Hypothesis ............................................................................................................. 42 3.3 Methodology outline .............................................................................................. 42 3.4 Materials and methods .......................................................................................... 44 3.4.1  STAGE 1: Formulation and processing .......................................................... 44  3.4.2  STAGE 2: Quantification and yield analysis ................................................... 47  3.4.3  STAGE 3: Microcapsule physical characterization ......................................... 51  3.4.4  STAGE 4: Controlled release ......................................................................... 52  3.4.5  STAGE 5: Microbial inhibition ......................................................................... 53  3.4.6  STAGE 6: In vivo studies ............................................................................... 55  3.4.7  Statistical analysis .......................................................................................... 56  3.5 Results and discussion .......................................................................................... 57 3.5.1  STAGE 1: Formulation and processing .......................................................... 57  3.5.2  STAGE 2: Quantification and yield analysis ................................................... 58 vi  3.5.3  STAGE 3: Microcapsule physical characterization ......................................... 67  3.5.4  STAGE 4: Controlled release ......................................................................... 72  3.5.5  STAGE 5: Microbial inhibition ......................................................................... 74  3.5.6  STAGE 6: In vivo studies ............................................................................... 77  3.6 Summary of findings for the vanillin project ........................................................... 79 4 Hexanal Project: Volatile Controlled Release into Headspace .................................... 81 4.1 Research objectives .............................................................................................. 81 4.2 Hypothesis ............................................................................................................. 81 4.3 Methodology outline .............................................................................................. 82 4.4 Materials and methods .......................................................................................... 84 4.4.1  STAGE 1: Formulation and processing .......................................................... 84  4.4.2  STAGE 2: Quantification and yield analysis ................................................... 87  4.4.3  STAGE 3: Microcapsule physical characterization ......................................... 88  4.4.4  STAGE 4: Microbial inhibition ......................................................................... 89  4.4.5  STAGE 5: Controlled release studies and target microbial inhibition on  laboratory media (potato dextrose agar) .................................................................... 90 4.4.6  STAGE 6: Apple shelf-life studies .................................................................. 97  4.4.7  Statistical analysis ........................................................................................ 108  4.5 Results and discussion ........................................................................................ 108 4.5.1  STAGE 1: Formulation and processing ........................................................ 108  4.5.2  STAGE 2: Quantification and yield analysis ................................................. 110  4.5.3  STAGE 3: Microcapsule physical characterization ....................................... 121  4.5.4  STAGE 4: Target microbial inhibition ........................................................... 124  4.5.5  STAGE 5: Controlled release studies and target microbial inhibition on  laboratory media ...................................................................................................... 125 vii  4.5.6  STAGE 6: Apple shelf-life experiments ........................................................ 144  4.6 Summary of findings for the hexanal project ....................................................... 168 5 Conclusions ............................................................................................................... 170 References ...................................................................................................................... 173  viii  List of Tables Table 1: Studies that have demonstrated the antimicrobial activity of vanillin in different media. .................................. 21 Table 2: Hexanal MIC assays for P. expansum................................................................................................................... 25 Table 3: Structure, precursor and odour threshold for hexanal. ........................................................................................ 27 Table 4: Theoretical amounts of vanillin (db) in each vanillin treatment. ......................................................................... 50 Table 5: Vanillin standard curve coefficients. .................................................................................................................... 50 Table 6: Batch formulations for vanillin-CD-pectin blend prior to drying. ....................................................................... 57 Table 7: Mean moisture and water activity for the vanillin-CD-pectin matrices. ............................................................ 57 Table 8: Mean vanillin concentrations in the vanillin-CD-pectin microcapsules (db). ..................................................... 58 Table 9: Mean encapsulation efficiency and vanillin recovery (yield) of microcapsules. ................................................... 58 Table 10: Vanillin concentration, encapsulation efficiency and vanillin recovery (yield) of the “increased vanillin” blends. ........................................................................................................................................................................................... 61 Table 11: Formulation prior to drying for all the vanillin in CD-calcium-crosslinked pectin matrices. ............................ 63 Table 12: Vanillin in CD-calcium-crosslinked pectin matrices and in residual CaCl2 solution. ......................................... 64 Table 13: Vanillin concentration of the test samples according to the HPLC and FC assays. ............................................ 66 Table 14: Dielectric properties of the original vanillin-CD-pectin blends measured at 2.54 GHz at 23°C (n = 3 reps, 5 measurements/rep). .......................................................................................................................................................... 71 Table 15: MIC for vanillin in a liquid medium at pH 4 and 5 and at a storage temperature of 5°C and 25°C. .................. 74 Table 16: Microscopic and visual MIC measurements for the 1:15 and 1:20 microcapsules. ........................................... 75 Table 17: Volatile inhibition of P. expansum 1525 by vanillin in the headspace of PDA plates Incubated at 25°C. .......... 76 Table 18: Volatile inhibition of P. expansum 1525 by vanillin in the headspace of PDA plates incubated at 5°C. ............ 77 Table 19: Results of a preliminary trial to determine the quantity of pectin gel that can be applied to apple slices. ....... 78 Table 20: Theoretical vanillin concentration achieved on the surface of apple slices when administering an 8% pectin gel assuming the higher concentration microcapsules. .......................................................................................................... 79 Table 21: Component percentages for each formulation prior to processing. .................................................................. 84 Table 22:Coefficients and conditions for Antoine’s equation calculations. ....................................................................... 95 Table 23: Maximum amounts of ethanol and hexanal that can volatilize in a 0.95 liter jar at a given temperature. ...... 95 Table 24: Headspace hexanal concentration standard curve coefficients. ........................................................................ 96 Table 25: Modified “Rating scale for overall visual quality of fresh-cut produce” (Kader and Cantwell, 2005). .............101 Table 26: Average moisture content of the microcapsules for all 5 treatments (tested in duplicate). ............................109 Table 27: Average water activity of microcapsules for all 5 treatments (tested in duplicate). .......................................109 Table 28: P-values for ANOVAS in two separate runs. .....................................................................................................116  ix  Table 29: P-values for concentration (db) and yield of the different REV processing combinations, and freeze-drying processing times. .............................................................................................................................................................117 Table 30: P-values for concentration (db) and yield of the different freeze-drying processing times. ............................117 Table 31: Energetic requirements of the REV machines. .................................................................................................118 Table 32: Averages for the dielectric constant (e') and dielectric loss factor (e'') of the hexanal-CD-pectin blends at different moistures at 2.45 GHz. ......................................................................................................................................123 Table 33: MIC of vaporized hexanal against P. expansum at different temperatures.....................................................124 Table 34: Growth rates and lag phases of P. expansum exposed to the low dose treatments stored at 25°C. ...............129 Table 35: Growth rates and lag phases of P. expansum exposed to the high dose treatments stored at 25°C. .............129 Table 36: Growth rates and lag phases of of P. expansum exposed to the high dose treatments on PDA plates stored at 12°C..................................................................................................................................................................................131 Table 37: P-Values for microbial response data for samples incubated on PDA at 25°C. ................................................132 Table 38: P-values for the microbial response variables obtained upon incubation on PDA at 12°C. .............................138 2  Table 39: Coefficients of determination (R ) for the cubic behavior of the average hexanal release profiles at 25°C (n=3). .........................................................................................................................................................................................143 2  Table 40: Coefficients of determination (R ) for the cubic behavior of the average hexanal release profiles at 12°C (n=3). .........................................................................................................................................................................................143 Table 41: Percentage of germinated wounds in apple varying spore suspension concentration and temperature. Reprinted from (Baert et al. 2008) with permission from Elsevier ©. .............................................................................144 Table 42:% Inhibition results. ...........................................................................................................................................154 Table 43: Hexanal released (mg/g db) by the "spent" microcapsules and the percentage of the original load..............160  x  List of Figures Figure 1: Morphology of different types of microcapsules. Reprinted from (Gharsallaoui et al., 2007), with permission from Elsevier © .................................................................................................................................................................... 3 Figure 2: Vanillin chemical structure. Reprinted from (Walton et al., 2003) with permission from Elsevier © ................. 19 Figure 3: Molecular structure and microstructure of CD. Reprinted from (Del Valle, 2004) with permission from Elsevier © ........................................................................................................................................................................................ 31 Figure 4: Adsorption/desorption for SPME sampling. Reprinted from (Shirey, 2012), with permission from Elsevier © .. 41 Figure 5: Strategic plan for studies on the controlled release of vanillin. .......................................................................... 43 Figure 6: Interaction plot for vanillin encapsulation efficiency for varying amounts of pectin in the blend. ..................... 60 Figure 7: Mean efficiency of encapsulation (y-axis) obtained with microcapsules containing increasing amounts of vanillin in the CD-pectin blend. ........................................................................................................................................ 62 Figure 8: HPLC vanillin standard curve and linear regression equation. AUC stands for area under curves in milli absorbance units*sec. ....................................................................................................................................................... 65 Figure 9 : Scanning electron micrographs of REV processed 1:20 blend samples. Images a) and c) were captured at 40x (the scale bar measures 500 m ) while b) and d) at 700x (the scale bar measures 20 m). ........................................... 68 Figure 10: Scanning electron micrographs of freeze-dried 1:20 blend samples. Images a) and c) were captured at 40x (the scale bar measures 500 m ) while b) and d) at 700x (thescale bar measures 20 m). ............................................ 69 Figure 11: Scanning electron micrographs of the vanillin crystals in their original state. Image a) was captured at 60x (the scale bar measures 200 m) while image b) was captured at 200x (the scale bar measures 100 m). .................... 70 Figure 12: Controlled release profile (n=4) of vanillin in an aqueous medium for the 1:20 REV and FD formulation at 25°C (vertical bars show the standard deviation). ............................................................................................................. 72 Figure 13: Release profile of vanillin in an aqueous medium for the 1:20 REV formulation at 25°C (n=4) and 5°C (n=2). 73 Figure 14: Strategic plan for studies on the controlled release of hexanal. ....................................................................... 83 Figure 15: 3.6 kW REV model dehydrator similar to the one used for this project. ........................................................... 86 Figure 16: Septum modified jar system with inoculated PDA petri dish / apple slices. ..................................................... 92 Figure 17: 1) Cork and pin for inflicting the wounds. 2) slice dimensions and inoculation sites. ...................................... 98 Figure 18: Examples apple slice appearance for categories in visual evaluation rating. .................................................102 Figure 19: Puncture test setup with probe and wooden holder. ......................................................................................104 Figure 20: Puncture test force vs deformation curve. Reprinted from (Mehinagic et al., 2004) with permission from Elsevier ©.........................................................................................................................................................................105 Figure 21: TPA specimen sampling and setup. ................................................................................................................106 Figure 22: Texture profile analysis (TPA) Model. .............................................................................................................107 Figure 23: Mean hexanal concentration (n=3) in microcapsules (mg/g) dry basis . ........................................................111  xi  Figure 24: Mean hexanal yield (n=3) in microcapsules (%) dry basis. ..............................................................................112 Figure 25: Main effects plot for yield (%). ........................................................................................................................114 Figure 26: Interactions plot for yield (%)..........................................................................................................................115 Figure 27: Mean hexanal concentrations (mg/g) dry basis and mean hexanal yields (%) dry basis (n=3) for treatments “A” and “B” processed using the 3.6 kW REV model ENWL291111-E020. ......................................................................120 Figure 28: Scanning electron micrographs of REV hexanal-CD-pectin "B" formulation a) fresh at 40x, b) “spent” at 40x ,c) fresh at 700x, d) “spent” at 700x. ...............................................................................................................................121 Figure 29: Scanning electron micrographs of freeze-dried hexanal-CD-pectin "B" formulation a) fresh at 40x, b) “spent” at 40x , c) fresh at 700x, d) “spent” at 700x. ......................................................................................................122 Figure 30: Average radial growth (cm) of P. expansum vs. time (hours) at the dose of 14 mg hexanal/jar (n=3) at 25°C. .........................................................................................................................................................................................126 Figure 31: Average radial growth (cm) of P. expansum vs. time (hours) at the dose of 14 mg hexanal/jar (n=3) at 12°C. .........................................................................................................................................................................................127 Figure 32: Main effects plot for final radius measurement (cm) after 120 hours incubation on PDA at 25°C. ...............133 Figure 33: Interactions plot for final radius measurement (cm) at 120 hours incubation on PDA at 25°C. .....................134 Figure 34: Main effects plot for the growth rate or slope (cm/hour) upon incubation on PDA at 25°C. .........................135 Figure 35: Interactions plot for the growth rate or slope (cm/hour) upon incubation on PDA at 25°C. .........................136 Figure 36: Main effects plot for calculated lag (hours) upon incubation on PDA at 25°C. ..............................................137 Figure 37: Interactions plot for calculated lag (hours) upon incubation on PDA at 25°C. ...............................................137 Figure 38: Main effects plot for final radius measurement (cm) at 240 hours incubation on PDA at 12°C. ....................139 Figure 39: Main effects plot for lag phase (hours) upon incubation on PDA at 12°C. .....................................................139 Figure 40: Headspace hexanal concentration (averages (n=3) released from microcapsules made from formulations A and B in the headspace above PDA plates incubated at 25°C over time. ........................................................................141 Figure 41: Headspace hexanal concentration averages (n=3) released from microcapsules made from formulations A and B in the headspace above PDA plates incubated at 12°C over time. ........................................................................142 Figure 42: Visual quality rating vs. time for un-inoculated apple slices stored at 5°C. ....................................................147 Figure 43: Visual quality rating vs. time for P. expansum inoculated apple slices stored at 5°C. ....................................148 Figure 44: Visual quality rating vs. time for un-inoculated apple slices stored at 12°C. ..................................................149 Figure 45: Visual quality rating vs. time for P. expansum inoculated apple slices stored at 12°C. ..................................149 Figure 46: Visual quality rating (n=3) of apple slices at day 6 (144 hours). .....................................................................151 Figure 47: Visual quality rating (n=3) of apple slices at day 9 (216 hours). .....................................................................151 Figure 48: Visual assessment rating (n=3) of apple slices at day 12 (288 hours). ...........................................................152 Figure 49: Visual assessment rating (n=3) at day 15 (360 hours). ...................................................................................153  xii  Figure 50: Headspace hexanal concentrations (n=3) in the headspace above apple slices vs. time for microencapsulated treatments stored at 5°C. ................................................................................................................................................156 Figure 51: Visual assessment of apple slices stored at 5°C vs. time. ...............................................................................157 Figure 52: Headspace hexanal concentration (n=3) above apple slices stored at 12°C vs. time. ....................................158 Figure 53: Visual assessment vs. time of microencapsulated treatments at 12°C. .........................................................159 Figure 54: Fracturability-Texture analysis TPA parameter (asterisks show treatments that are significantly different from time zero (p<0.05), vertical bars stand for standard deviations). ...........................................................................163 Figure 55: Hardness 2 -Texture analysis TPA parameter (n=6) (asterisks show treatments that are significantly different from time zero (p<0.05), vertical bars stand for standard deviations). ...........................................................................164 Figure 56: Hardness 2 -Texture analysis TPA parameter (n=6) (asterisks show treatments that are significantly different from time zero, (p<0.05), vertical bars stand for standard deviations). ..........................................................................164 Figure 57: Chewiness -Texture analysis TPA parameter (n=6) (asterisks show treatments that are significantly different from time zero (p<0.05), vertical bars stand for standard deviations). ...........................................................................165 Figure 58: Ff -Texture analysis puncture parameter (n=9) (asterisks show treatments that are significantly different from time zero (p<0.05), vertical bars stand for standard deviations). ...........................................................................166 Figure 59: Ws -Texture analysis puncture parameter (n=9) (asterisks show treatments that are significantly different from time zero (p<0.05), vertical bars stand for standard deviations). ...........................................................................166 Figure 60: Wf -Texture analysis puncture parameter (n=9) (asterisks show treatments that are significantly different from time zero (p<0.05), vertical bars stand for standard deviations). ...........................................................................167  xiii  Glossary of Terms  ASP status: “fully up to date toxicology information has been sought” Beta cyclodextrin (CD): a cyclic derivative of starch, consisting of seven glucopyranose units. It is obtained from the industrial enzymatic treatment of maltodextrins using the enzyme glucosyltransferase. Botrytis cinerea :also known as gray mould rot, an important postharvest disease for pomme fruits. Controlled release. method by which one or more active agents are made available at a desired site and time and at a specific rate Core material/ internal phase/ fill: synonyms for the substance inside the microcapsules DVB: Divinylbenzene; polar material, used in composition of SPME fibers EAFUS: “Everything added to Food in the United States” Encapsulant/ Coating material /shell/ wall/ membrane/ carrier: synonyms for the encapsulation material or matrix Encapsulation load or payload: weight fraction of the active ingredient (db) in the encapsulated product (this includes the active ingredient that is within and on the surface of the microcapsules) Encapsulation yield or recovery: fraction of the initial active agent amount that is not lost in microencapsulation processing and is found in both entrapped and surface forms. Fresh–cut produce: minimally processed fruits and vegetables that have been physically altered (trimmed, peeled, washed, and/or cut) but remain fresh xiv  FID: Flame Ionization Detector. Gas Chromatography: separation technique that uses a gaseous mobile phase through an inmiscible solid stationary phase, capable of separating and quantifying volatile/gaseous samples. Hexanal: C6-aldehyde rapidly formed after the onset of mechanical damage in plant tissue derived from the oxidation of linoleic acid via the lipoxygenase activity Inclusion complex: molecular cage for entrapped compounds through hydrophobic interactions. JECFA: Joint FAO/WHO Expert Committee on Food Additives Mason Jar: hermetical glass container commonly used in canning to preserve food. Microencapsulation: technology of packaging solids, liquid, or gaseous materials in miniature, sealed capsules ranging in size from 1m to 1000m. Microencapsulation efficiency: fraction of the initial active agent amount that is fully entrapped in the obtained microcapsules Oral rat LD50: dose required to kill half of the rats that ingest the tested compound orally after a specified test duration. LD50 figures are frequently used as a general indicator of a substance's acute toxicity. Partition coefficient: partitioning of an aromatic compound between a food matrix and the gaseous phase above it under equilibrium conditions Patulin: a toxic metabolite (polar mycotoxin) produced by various species of Penicillium, P. expansum being the predominant one PDA: Potato Dextrose Agar. Medium commonly used to grow moulds and yeasts. PDMS: Polydimethylsiloxane; non polar material, used in SPME fiber composition  xv  Pectin: group of natural polymers of (1→4) linked -D galacturonic acid partly esterified with methanol. Penicillium expansum: also known as blue mould rot, is a white fungus that produces bluish-green spores and will grow in an injured site in pomiferous fruits Radiant Energy Vacuum (REV): patented technology that allows for drying of highly viscous materials at low temperatures by the use of microwaves and vacuum SPME: Solid Phase Microextraction; extraction device that relies on the equilibrium between the gaseous and liquid phase of a volatile to adsorb/quantify (when coupled with GC) trace amounts of the volatile without the need for solvent extraction. Trans-2-hexenal: aldehyde produced from linolenic acid in plant tissues by the activation of the lipoxygenase hydroperoxide lyase enzyme pathway Vanillic acid: (4-hydroxy-3-methoxybenzoic acid) a compound found in the vanilla pod and bean. Vanillin: (4-hydroxy-3-methoxybenzaldehyde), the major compound in natural vanilla.  xvi  Acknowledgements  This work was supported by the NSERC’s Industrial Postgraduate Scholarship.  I would like to thank my research supervisors Dr. Tim Durance and Dr. David Kitts for their endless patience, guidance, and support throughout my graduate student program.  I thank Dr. Brent Skura for inspiring and encouraging me to do a graduate degree in the first place. I thank Dr. Pascal Delaquis for providing insight into the world of fresh-cut produce, without which I would have been utterly lost.  Special thanks to Parastoo Yaghmaee and Reihaneh Noorbarkhsh for their instruction in the use of the EnWave’s equipment and facilities.  I would also like to thank Valerie Skura and Pedro Aloise for all their technical assistance at the UBC laboratory.  Finally, I would like to thank my mom, who was my moral support throughout the “ups” and “downs” of this research adventure and without whom I would have never have managed to fully accomplish it.  xvii  Dedication  To my mom, who never stopped believing in me, and to my dad, who inspired me to go as far as “my head would take me”.  xviii  1 Introduction Microencapsulation is a technology used to store solids, liquids, and gases in microsized, sealed capsules capable of releasing their contents in the presence of specific conditions or triggers (Desai & Park, 2005). The microencapsulation of food ingredients has been used to protect, preserve, mask, dilute, and stabilize active compounds for different purposes. The market for fresh-cut fruits and vegetables is a growing one, since consumers are interested in ready to eat (RTE), healthy foods and snacks. The distribution of fruit processed in this format, such as fresh-cut apple slices (Malus domestica Borkh.), is often challenging due to the limited shelf-life that can be achieved with available preservation technologies (Lanciotti et al., 2004). Growth of spoilage microorganisms, notably fungi such as Penicillium expansum, can lead to the rapid development of quality defects including discoloration, exudation, or development of off-odors. Natural antimicrobials have been investigated as a possible way to preserve freshcut apples. Vanillin has been studied as a possible candidate for this purpose, however, its diffusion into the plant tissue as well as losses due to binding to organic compounds have limited its efficacy. Microencapsulation could allow a controlled release on the surface of the apple slices (where these antimicrobials are most needed) to preserve microbial stability and increase shelf-life. Hexanal, a C6-aldehyde that is naturally present in apple, is a strong inhibitor of a wide range of microorganisms including P. expansum (Song et al., 1998; Neri et al., 2006). The high volatility of the compound favors application in closed packaging systems but solubility in tissues and degradation through reactions with plant components limit efficacy. Microencapsulation for sustained release could alleviate both problems. Radiant Energy Vacuum (REV) is a technology that permits the rapid drying of highly viscous materials at low temperatures. It can be used as a microencapsulation method, and would define the characteristics of the encapsulated material. REV 1  microencapsulation of vanillin and hexanal, with the purpose of a controlled release application on fresh-cut apple slices, could provide a viable preservation method leading to improvements in shelf-life.  2 Literature Review 2.1 General aspects of microencapsulation Microcapsules can be defined as small particles ranging in size from 1m to 1000m, that contain an active agent or core material surrounded by a coating or a shell. The encapsulated material is commonly referred to as the core material, internal phase, or fill. The coating material can be referred to as the wall, shell, wall material, membrane, carrier or encapsulant (Gharsallaoui et al., 2007; Madene et al., 2006). The diameter of commercial microcapsules usually ranges between 3-800 m, and their core weight or payload ranges from 10-90% (Thies, 1996). Gharsallaoui et al. (2007) in their review article “Applications of spray-drying in microencapsulation of food ingredients” proposed six main reasons to microencapsulate substances. These are:   Reduction of core reactivity to environmental factors (protection)    Decrease in core material transfer rate to the environment    Promotion of easier handling    Controlled release of the core material    Masking of core taste or odour    Dilution of the core material to be used in only small amounts  To add to this list: an important functional application in microencapsulation is increasing the solubility of the core substance. For example, complexation of vanillin with beta cyclodextrin (βCD) improves its solubility in water (Karathanos et al., 2007). Finally, microencapsulation can be also be used to separate a component within a mixture that would otherwise react with another (Desai & Park, 2005). 2  2.1.1 Microcapsule characteristics Two common microcapsule structures are the continuous core/shell microcapsule and the multinuclear microcapsule. The first often has a spherical geometry with a continuous core surrounded by a continuous coating. This type of microcapsule can also be referred to as a “single-particle structure” (Desai & Park, 2005). The second often has an irregular geometry and contains a number of small droplets of core material dispersed throughout a continuous shell (Thies, 1996). This multinuclear microcapsule is also known as the “aggregate structure” (Desai & Park, 2005). Depending on the physicochemical properties of the core, the composition of the wall material, as well as the nature of the microencapsulation process, the microcapsule structure might differ considerably (Gharsallaoui et al., 2007) and does not strictly need to be a continuous core/shell microcapsule or multinuclear microcapsule. Other possible microcapsule structures are depicted in Figure 1.  Figure 1: Morphology of different types of microcapsules. Reprinted from (Gharsallaoui et al., 2007), with permission from Elsevier ©  The parameters that are usually used to characterize microcapsules include: their size, morphology, surface charge, encapsulation efficiency, the process yield, their loading, and their release profile (Gibbs et al., 1999). 3  2.2 Microencapsulation processes Microencapsulation processes are commonly classified as either chemical or mechanical. Chemical processes include: complex coacervation, polymer-polymer incompatibility, in situ polymerization, in-liquid drying, molecular inclusion, among others. Mechanical processes include: spray drying, spray chilling, fluidized bed drying, centrifugal extrusion, freeze-drying, electrostatic deposition, etc. (Thies, 1996, Madene et al., 2006). The encapsulation process used determines the shape (films, spheres, particles, irregular), structure (porous or compact), and physical structure (amorphous or crystalline dehydrated solid; rubbery or glassy) of the matrices produced (Madene et al., 2006). These factors have an effect on the diffusion of the active agents and also external substances, for example oxygen. They are also important factors in determining the stability of the food product (Madene et al., 2006). The general process for most microencapsulation methods begins with the creation of an emulsion between the wall material (considering its molecular weight, conformation, chemical groups, physical state) and the flavour or active compound (considering its relative volatility, polarity, chemical groups). The emulsion is then subjected to either a chemical process or a mechanical process and the microcapsules are harvested (Madene et al., 2006). Since Radiant Energy Vacuum (REV) microencapsulation is a mechanical process, it was compared against the most similar process that: 1) has been used to microencapsulate similar substances, 2) produces a similar product, and 3) is able to handle solutions of high viscosity: freeze-drying. Spray drying may be one of the most common physical encapsulation methods, however it will not be considered as the comparison process for this work since pectin induces excessive viscosity and renders samples incompatible with the spraying process. It is important to consider that the proposed microencapsulation method is truly a mixture of a chemical process (molecular inclusion due to the cyclodextrin-antimicrobial complexation) and a mechanical process (REV drying / Freeze-drying). The mechanical process of homogenizing, prior to drying, also contributes to the droplet size and distribution within the matrix. 4  2.2.1 Spray drying In most microencapsulation methods that involve drying, the first step involves the creation of an emulsion of the core material and the encapsulant (Madene et al., 2006). They are homogenized with or without an emulsifier (Gharsallaoui et al., 2007). For spray drying, the core material must be dispersed in a concentrated solution (40-60% solids) of shell material (Thies, 1996). This solution must have a relatively low viscosity: thus the use of pectin in this work precludes this method from working well. CD on the other hand, has been successfully used to spray-dry-encapsulate flavours such as lemon oil (Madene et al., 2006). Water is the preferred solvent for spray drying, since the use of organic solvents raises risks of fire and toxicity concerns (Thies, 1996). The emulsion is then sprayed or spun off a rotating disk into a heated chamber, with the dehydrated capsules falling to the bottom, where they are harvested. Particles produced this way are usually 10-300 m in diameter and tend to have irregular geometries and a multinuclear structure (Thies, 1996; Madene et al., 2006). Spray-drying has several advantages as a microencapsulation process. The technology is well established; equipment is readily available, and it produces large amounts of microcapsules. Shell materials that are used for spray dying are usually food grade, water soluble, and not chemically cross-linked. This provides the added benefit that the capsules dissolve in water thereby leaving no debris after having released their core (Thies, 1996). Other important benefits of spray drying include low cost (compared to freeze-drying) and flexibility (Desai & Park, 2005). As was mentioned before, one of the constraints of using spray-drying technology is the need for concentrated, low viscosity solutions, since high viscosity interferes with the atomization process (Gharsallaoui et al., 2007). Another important constraint is the restriction to water soluble or water dispersible cores, since water is the preferred solvent. Core loading for spray dried capsules is normally less than 20-30% weight, and there is usually free or un-encapsulated material, as residue, over the capsules. This can result in 5  problems if the material is either prone to oxidation or if it has a distinctive odour (Thies, 1996; Madene et al., 2006). Low boiling point compounds with little solubility in water tend to volatilize from the capsules in the spray chamber, thus reducing the yield of this method considerably (Thies, 1996; Madene et al., 2006). Hexanal would fall in this category, making spray drying once again inappropriate for purposes of this work. Another important disadvantage of spray drying is the energy waste due to the unfeasibility of using all the heat in the drying chamber (Gharsallaoui et al., 2007). Huge amounts of heat are required to evaporate much larger amounts of water for coating materials that have low solubilities or are viscous at high concentrations. Consequently, the process is comparatively more expensive in terms of energy consumption (Desai & Park, 2005). It may also produce very fine powders (10-100 m in diameter), which require further processing, such as agglomeration (Madene et al., 2006).  2.2.2 Freeze-drying or lyophilization Freeze-drying is used mostly for the dehydration of heat sensitive materials that are unstable in aqueous solutions (Madene et al., 2006). It is a simple technique that involves mixing the core and the coating solutions, freezing them, and drying under vacuum (Desai & Park, 2005). Upon water crystallization, the non-frozen solution is viscous and the diffusion of flavours is delayed (Madene et al., 2006). Freeze-drying also maintains the shape of the microcapsules because the droplets become fixed in the emulsion during the freezing step (Madene et al., 2006). Freeze-drying is a simple technique, suitable for the encapsulation of aromatic materials (Desai & Park, 2005), which makes it suitable for the encapsulation of hexanal. A disadvantage of lyophilization however, is the long dehydration times required (commonly 20h to 72h) (Desai & Park, 2005). The viscosity of the materials is of no importance for freeze-drying, therefore pectin can be used as a microencapsulant. For these reasons, freeze-drying was chosen as the conventional drying method for comparison with the more novel REV technology.  6  The retention of volatile compounds during the freeze-drying process is dependent on the chemical nature of the matrix (Desai & Park, 2005). Some losses are to be expected, particularly due to the vacuum exerted in the system. According to Flink and Karel (1970), when volatiles are frozen and dried in carbohydrate mixtures, they tend to form “microregions” where the volatile compound and the carbohydrate are most concentrated. The crystallization of the water during freezing causes the formation of these “microregions”, and as the moisture in these sections decreases, the carbohydrate moieties become molecularly associated through hydrogen bonds (Flink & Karel, 1970). This association controls the permeability of both water and the volatile. As the moisture decreases, the ease with which the volatile is lost decreases as well until the moisture reaches a critical level where no more volatile losses occur even as the moisture level decreases (Flink & Karel, 1970). The rate of freezing is important in the formation of “microregions”. Rapid freezing (e.g. using liquid nitrogen) results in many tiny crystals and many small “microregions”. Slow freezing on the other hand results in fewer, larger, and more concentrated “microregions”. Larger “microregions” are less permeable to volatile losses during drying, thus the volatile retention is higher when samples have been slow frozen (Flink & Karel, 1970). Within “microregions”, the organic volatiles are in several states: 1) adsorbed on sites which can be competitively occupied by water, 2) adsorbed on specific sites that cannot be replaced by water, 3) entrapped as condensed aggregates (like droplets) in which most of the molecules are not saturating the internal walls of the encapsulant (Chirife et al., 1973). Freeze-drying is 50 times more expensive than spray drying (Madene et al., 2006), which makes it unappealing from an industrial point of view. As was mentioned before, the long processing times required are also restrictive.  7  2.2.3 REV technology Radiant Energy Vacuum (REV) drying is the process of using microwaves (also called “radiant energy” or “nonionizing radiation”) under vacuum to remove water while maintaining an open cellular structure in the dried product (Scaman & Durance, 2005). Electromagnetic microwave energy penetrates the sample material and is converted into thermal energy, serving as a rapid heating mechanism (Durance & Yaghmaee, 2011). Microwaves are not capable of breaking chemical bonds. The heating occurs primarily when the electric field polarizes both full and partial charges in the material, but the molecules are not capable of matching the high frequency reversals of the field, dissipating the power as heat (Durance & Yaghmaee, 2011). Dielectric property measurement is important since it helps characterize the materials as well as allowing a better description of the microwave heating/drying process. A material’s dielectric properties define its potential for heating when microwaves enter and transmit energy through it (Thostenson and Chou, 1999). Thus, these are important parameters to consider when looking for process optimization. Dielectric properties are mainly characterized by the dielectric constant or permittivity (e’) and the dielectric loss factor (e’’), which quantify the capacitative and conductive (i.e. heating) components of the dielectric response (Thostenson and Chou, 1999). The dielectric constant indicates the amount of energy stored as an electric field by a specific material relative to empty space. The loss factor shows how much of the microwave energy can be converted into heat (Durance & Yaghmaee, 2011). Dielectric properties of materials depend on the material’s chemical composition, physical structure, microwave frequency, and temperature. Free water and dissolved ions have the highest dielectric properties (Abassi & Rahimi, 2008). The microwaves are efficiently absorbed by “lossy” molecules, or molecules that efficiently convert microwaves into heat, such as water and other polar or charged molecules (Durance & Yaghmaee, 2011). The term “lossy” relates to the loss factor  8  fraction of the dielectric properties of a material, since the amount of thermal energy converted in the material is proportional to the loss factor. Vacuum enhances the drying rate by lowering the boiling point of water and creating a pressure gradient (Durance & Yaghmaee, 2011). Drying at low temperatures is useful for protecting components prone to oxidation or heat damage. This technology has been used successfully for the inactivation as well as for the preservation of enzymes and microorganisms, pharmaceutical processing and histochemical processing of samples (Scaman & Durance, 2005). One of the benefits of using microwave energy is the possibility of generating heat within the moist portion of the food. One of the drawbacks is the non-uniformity of the heating (Zhang et al., 2006). Another effect of microwaving is that the rapid mass transport (rapid conversion of liquid water into steam) causes “puffing” with the creation of pores (Zhang et al., 2006; Durance & Yaghmaee, 2011). Compared to other drying methods, vacuum microwave drying significantly shortens the drying period in several ways: selective heating of interior portions of the matrix, rapid energy dissipation throughout the material, relatively minor migration of water-soluble compounds, lower product temperatures, and more efficient drying in the falling rate period (Zhang et al., 2006). Abassi and Rahimi (2008) recently used differences in dielectric constants of shell and core materials (nearly ten times) to achieve microencapsulation. By exposing to the appropriate electromagnetic energy, Abassi and Rahimi (2008), selectively heated citric acid (the active agent with a higher dielectric constant than the encapsulant material) fusing the surrounding shell material and forming microcapsules upon cooling by creating a fused shell around the core material. The remaining shell material could be sieved or washed away. Their results show that through the use of this method, the efficiency of shell material encapsulation was: casein>inulin>carboxyl methyl cellulose >low methoxy pectin>sorbitol (Abassi & Rahimi, 2008). 9  2.3 Preservation of fresh-cut fruit Fresh-cut vegetables and fruits is a rapidly developing sector of the food industry, driven by consumer demand for nutrition and convenience when buying food products. It is an industry that has grown to approximately US$15 billion per year and accounts for 15% of all produce sales (Rojas Graü & Martin-Belloso, 2011). “The term minimally processed fruit refers to any type of fruit that has been physically altered from its original state (trimmed, peeled, washed, and/or cut) but remains in a fresh unprocessed state” (Olivas & Barbosa-Cánovas, 2005). Thus, fresh-cut fruits are minimally processed fruits that are presented to the consumer in a state in which they can be consumed immediately without further preparation or processing (Olivas & BarbosaCánovas, 2005). The benefits of consuming fresh-cut produce in terms of convenience include: 100% consumable product and a substantial decrease in labor for preparation and waste disposal (Rojas Graü & Martin-Belloso, 2011). Fresh-cut fruit provides consumers with the option of healthy snacks that are ready to eat and require no preparation time. Increasing the availability of fresh-cut fruit products in vending machines at schools or offices for example, could contribute to strategies meant to combat obesity and nutrition related illnesses (Olivas & Barbosa-Cánovas, 2005; Rojas Graü & Martin-Belloso, 2011). However, this approach is limited by the short microbial and quality shelf-life of fresh-cut fruit products due to excessive tissue softening and cutsurface browning (Soliva-Fortuny & Martín-Belloso 2003a). A shelf-life under refrigerated conditions that is long enough to make transport and retail feasible is needed (Ahvenainen, 1996), and for fresh-cut fruit it is usually limited to 5-7 days (Lanciotti et al., 2004). Therefore, there is a need to develop strategies to improve the microbiological quality and shelf-life of these products. During fresh-cut processing, produce undergoes substantial mechanical injury, due to cutting and peeling, which results in the exposure of nutrients that become available for microbial growth. Cutting and peeling also elevate the rate of respiration (1.2-7 fold), which increases the rate of the product’s deterioration (Ahvenainen, 1996; Lu et al., 2007; 10  Fonseca et al., 2002) by causing an accelerated consumption of sugars, lipids, and organic acids and by increasing the production of ethylene (Olivas & Barbosa-Cánovas, 2005). Ethylene induces accelerated ripening and senescence and is partially responsible for softening of tissues (Ahvenainen, 1996). Mechanical injury ruptures cells and releases enzymes which cause chemical deterioration such as off flavours, discolourations, and loss of firmness (Raybaudi-Massilia et al., 2007; Al-Ati & Hotchkiss 2002; Ahvenainen, 1996), the result being loss of sensory quality. A high content of fermentable sugars in fresh-cut fruits leads to microbial growth and the production of volatile compounds such as ethanol, adehydes, and ketones (Ragaert et al., 2011). Cutting and trimming notably accelerate the texture decline in fruit since structures such as the cell wall, middle lamella and cellular membrane undergo biochemical changes that lead to loss of cohesion between cells and weaken structures as a result. Water loss also promotes loss of turgor in cells, which leads to a “mushy” texture (Olivas & BarbosaCánovas, 2005). Using sharp knives diminishes damage to cells and decreases release of cellular contents and fluids (Ahvenainen, 1996; Ártes et al., 2007). Washing and sanitizing steps are important stages in the fresh-cut process since they remove dirt, pesticide residues and microorganisms responsible for quality loss and decay. These steps also serve to cool the product as well as remove cell exudates that could promote microbial growth (Beltrán 2005). A concentration of 50-200 mg of chlorine or citric acid per liter of washing solution is known to be effective for disinfecting pome fruits (Ahvenainen, 1996; Soliva-Fortuny & Martín-Belloso 2003a). The only disadvantage is the fact that increasing organic matter in wash water reacts with chlorine decreasing the concentration of its active form (hypochlorous acid) and reducing its efficacy. For this reason, careful monitoring of chlorine concentration and wash water pH levels is needed (Ártes et al., 2007). Cut-edge enzymatic browning is catalyzed by polyphenol oxidases (PPO) and/or phenol peroxidases (Toivonen & Brummel, 2008). Enzymatic browning requires four components: the oxidizing enzyme, copper, oxygen, and a suitable substrate (Ahvenainen, 1996). Physical stress, wounding or senescence compromises the cell compartment 11  integrity, releasing phenolic compounds (suitable substrates) from the cellular vacuole, and allowing them to mix with these enzymes (Toivonen & Brummel, 2008). PPOs take part in the hydroxylation of monophenols to diphenols, the oxidation of diphenols to quinones and the non-enzymatic polymerization of melanines, which are red, brown, or black pigments (Soliva-Fortuny & Martín-Belloso 2003a; Toivonen & Brummel, 2008; Ártes et al., 2007). The use of stainless steel knives reduces the risk of cut-surface browning avoiding the contribution of metallic ions like copper that have the ability to boost it (Ártes et al., 2007). Lipoxidase is another important enzyme that catalyzes the peroxidation reactions, resulting in malodorous aldehydes and ketones (Ahvenainen, 1996). PPO cut-edge enzymatic browning is one of the plant’s defense mechanisms against herbivores. The quinones formed are capable of binding to the plant proteins, reducing their digestibility and nutritive value, making the plant tissue less appetizing from a nutritional perspective (Queiroz et al., 2008). Cut-edge enzymatic browning should not be confused with secondary browning due to microbial growth. There are two ways to distinguish between them: 1) secondary browning is localized, while cut-edge browning is homogeneous in the tissue, and 2) the time of browning is different, cut-edge browning occurs within hours after cutting and secondary browning between 1-3 weeks of storage (Toivonen & Brummel, 2008). Certain fungi, like Botrytis cinerea are capable of producing laccases, fungal polyphenol oxidases (also called multicopper oxidases) in apple tissue (Brijwani et al., 2010). Toivonen et al. (2010) recently evaluated the effect of controlled atmosphere storage on fresh-cut apple slice quality in terms of secondary browning of the “Delicious” and “Spartan” varieties using a 4 level visual assessment scale. To inhibit cut-edge enzymatic browning, combinations of citric acid, ascorbic acid, and 4-hexylresorcinol have been used since sulphite usage is prohibited (Ahvenainen, 1996). Ascorbic acid is commonly used and is hypothesized to control PPO activity by reducing quinones to the native diphenols (Toivonen & Brummel, 2008) as well as lowering the pH of the system. Since optimum pH for PPO reactions is 5-7.5, lowering the pH is an effective method to inhibit the reaction (Queiroz et al., 2008). Ascorbic acid treatment is 12  capable of reducing cut-edge browning and increasing shelf-life only temporarily since it is known to soften the tissue and promote the growth of fungi (Gil, Gorny, & Kader, 1998). Modified atmospheres with lowered O2 levels are known to decrease enzymatic browning (Ártes et al., 2007) but promote anaerobic fermentation. In a more natural approach, the use of low pH, sugary solutions has also been employed with good results, for example dipping apple slices in pineapple juice to prevent browning (Ártes et al., 2007). Anti-browning solutions have no significant effect on secondary browning (Toivonen & Brummel, 2008). In recent years, outbreaks of food-borne illness have been associated with the consumption of vegetables and fruits (or unpasteurized products produced from them). Increased distribution over large distances and the new consumption patterns and practices have “undoubtedly contributed to this increase” (Beuchat & Ryu 1997). Since the pH of fruits is relatively low compared to other food systems, there is an effect to suppress bacterial growth, the most important microorganisms that cause postharvest wastage are fungi (Ayala-Zavala et al., 2008). Yet some particular strains of human pathogens such as E. coli O157:H7 have been known to survive and even grow at similar pH. In fact, E. coli O157:H7, Salmonella, Listeria innocua (as a surrogate for Listeria monocytogenes) increased more than two log units in 24 hours in Golden Delicious, Granny Smith, and Shampion variety fresh-cut apple plugs stored at 20-25°C. Only L. innocua was able to grow at 5°C (Alegre et al., 2010). The native flora of apples consists mainly of mesophilic bacteria, psychrotrophic bacteria, yeasts and moulds (Raybaudi-Massilia et al., 2007). Mesophilic bacteria counts of 103-106 cfu/g have been found in minimally processed products analyzed immediately after packaging (Lanciotti et al., 2004). These numbers inevitably increase during storage, and more rapidly if temperature abuse occurs. Several different strategies have been tested and used to decrease microbial and quality deterioration in fresh-cut apples. Modified atmospheres have been shown to enhance the shelf-life of apple slices with less than 5 log increase in mesophilic counts 13  after a month in refrigerated storage (Soliva-Fortuny et al., 2004). Unfortunately, these low oxygen and high carbon dioxide concentrations in modified atmospheres affect the flavor and aroma of the fruit by significantly reducing the production of aroma compounds such as acetate esters (Olivas & Barbosa-Cánovas, 2005). Compounds such as calcium chloride have been used to improve fruit texture during storage, since they cross link with pectins in the cell wall and middle lamella reinforcing cohesion between cells (Olivas & Barbosa-Cánovas, 2005). Chen et al. patented a formulation of calcium salts named “Nature Seal ®” to protect apple slices from changes in colour, texture, and taste (Olivas & Barbosa-Cánovas, 2005). The ripeness stage of the fruit when it undergoes processing is of extreme importance in its shelf-life. In Fuji apples, the use of partially ripe fruit, instead of fruit in an advanced ripeness stage, has been shown to increase post-cutting shelf-life (Rojas-Graü et al., 2007). 1-Methylcyclopropene (1-MCP), which has a similar structure to ethylene, has been proven to block the binding of superficial cell receptors to ethylene interfering with the ripening and delaying senescence (Soliva-Fortuny & Martín-Belloso 2003a). Storage temperature is the most important external factor that influences respiration, since biological reactions will increase 2-3 fold for every 10°C rise in cold chain disruptions (Fonseca et al., 2002). Different natural and synthetic antimicrobials such as benzoic acid, sodium benzoate, potassium sorbate, vanillin, lemongrass, oregano oil (Olivas & BarbosaCánovas, 2005; Rojas Graü et al., 2006) among others have been used to preserve freshcut apples with fairly good results. A combination of 1% N-acetyl-L-cysteine, 1% glutathione, 1% calcium lactate and 2.5% D-L-malic acid in dipping solution for Fuji apples resulted in shelf-life of at least 14 days (Raybaudi-Massilia et al., 2007). “However the diffusion of the preservatives into the fruit may decrease their effectiveness over time” (Olivas & Barbosa-Cánovas, 2005). This phenomenon reduces the availability of the compounds at the surface of the slice, where they are needed to inhibit microbial growth.  14  These compounds are also limited by the strong odours and tastes that they might impart to the fruit (Soliva-Fortuny & Martín-Belloso 2003a) Edible films containing antimicrobials retain preservatives at the surface of cut fruit, avoiding diffusion into the tissue (Olivas & Barbosa-Cánovas, 2005). Apple puree-alginate coatings with lemongrass, oregano oil, and vanillin (0.3% w/w) have been shown to be effective in preserving Fuji apple slices for two weeks storage under refrigerated conditions (Rojas Graü et al., 2006). Apple slice “wraps”, made of various concentrations of fatty acids, fatty alcohols, beeswax, and vegetable oil, were capable of reducing moisture loss and browning and preserving colour for 12 days at 5°C (McHugh and Senesi, 2000) An important limitation in the use of edible coatings and “wraps” is that they restrict oxygen diffusion, limiting respiration and thus inducing anaerobic respiration. Anaerobic respiration converts sugars to ethanol, which forms off-flavours. Prolonged exposure of fruit to anaerobic conditions can lead to cell death (Olivas & Barbosa-Cánovas, 2005). Other limitations encountered by edible coatings is the low adherence presented by the highly hydrophilic cut surfaces, that are wet, and exude liquids, likely compromising the integrity of the coating (Vargas et al., 2008). As for “wraps” and films, lipids and waxes are commonly used since they have good water barrier properties however they encounter poor sensory acceptance (undesirable color, taste and flavour) (Vargas et al., 2008). As a result, there is a need to find alternative means to effectively administer antimicrobials to fresh-cut fruit.  2.3.1 Penicillium expansum Penicillium expansum, commonly known as blue rot mould, is a fungus that produces bluish-green spores that can germinate and grow in the injured tissues of pomiferous fruits (Bliss, 2003). In the United States, this microorganism is the most significant cause of post-harvest decay in stored apples (Bliss, 2003). P. expansum spores 15  are naturally found in soil and may contaminate apples through soil, crop debris, field bins, equipment, and water handling systems. Rot begins as soon as the spores are capable of reaching a wound in the apple (Morales et al., 2010). Since fresh-cut apple slices are essentially “wounded” by processing, any spores left after the sanitization process are bound to cause problems. P. expansum is considered a “soft rot” since it produces substantial amounts of pectinesterase (Cole & Wood, 1961). Injured apple flesh turns soft, brown, and watery. The spores are able to survive for long periods of time in wooden containers, walls, cold room ceilings etc, where P. expansum may establish itself and even grow (SánchezVentura et al., 2008). In a study conducted in France, airborne P. expansum spore concentrations of 2 x 104 spores/m3 were measured in a non-disinfected apple warehouse after one month of storage while a previously disinfected warehouse had a concentration of 2.5 x 103 spores/m3. This shows that complete elimination of conidia is unlikely (Morales et al., 2010) in the fruit storage environment. Fungal growth occurs in two steps: spore germination and hyphal elongation to form a filamentous colony (Baert et al., 2008). A germinating spore will first activate, then swell and finally form a germ tube. When the germ tube is equal or larger than the largest swollen spore then the spore can be considered to have germinated. The germ tube initially begins increasing in length exponentially and then has a constant linear extension rate with lateral branches formed (Prosser, 1993). The rate at which a colony expands radially (Kt) is usually determined by the growth of the peripheral zone (Prosser 1993). Patulin (4-hydroxy-4H-furo[3,2c]pyran-2(6H)) is a toxic metabolite (polar mycotoxin) produced by various species of Penicillium, P. expansum being the predominant one (Bennet & Kilch, 2003). Patulin has been found mainly on the rotten parts of apples, but occasionally on fruits like pears, apricots, peaches and grapes P. expansum shows psychotrophic characteristics, being able to grow and produce patulin under refrigerated storage (Baert et al., 2007a). Acute symptoms of ingestion include: nervousness, convulsions, lung congestion, edema, hyperaemia, gastrointestinal tract distension, intestinal hemorrhage, and epithelial cell degeneration (De Souza et al., 2008). 16  Organic acid concentration and pH can significantly affect patulin accumulation when the apples are in cold storage. These parameters are dependent on degree of ripeness and apple variety. Morales et al. (2008a) found that Golden apples (which presented the lowest pH) accumulated the most patulin at 1°C. However, at 20°C, the varieties with the highest concentration of organic acids (Golden and Fuji apples) were the ones to accumulate the most patulin. Ascorbic acid can react with patulin molecules and effectively decrease the toxin concentrations (De Souza et al., 2008). However the ascorbic acid dips performed to inhibit browning usually occur during processing, and patulin production could still occur during the post-processing storage. P. expansum’s optimum growth occurs at 25°C (Baert et al., 2007b). Reducing temperature from 20 to 10°C or 4°C can stimulate patulin production. A further decrease of temperature to 1°C will decrease patulin production. It should be noted however that refrigeration temperature fluctuations during retail storage are common. Baert et al. (2007b) concluded that the induction of stress to a fungus, such as lowering temperature or lowering the oxygen levels will stimulate patulin production, however this property varies in different strains. P. expansum grows in more than 90% of inoculations done in apples when the inoculum is equal or higher than 2 x 104 spores (Baert et al., 2008). Baert et al. (2007b) recognized that most studies will use high inoculum levels, while most infections (in apples) are probably due to a low number of spores. In a study conducted in the Unites States, the dump water used to rinse apples was evaluated for P. expansum spore concentration during the packing season, and was found to range between 5.6 x 101 to 4.2 x 103 spores/ml (Spotts & Cervantes, 1993). Baert et al. (2008) concluded that “the real food matrix can inhibit the germination of spores in comparison to a synthetic or a simulation medium” (Baert et al., 2008). For this reason, shelf-life studies on non-inoculated food matrices must also be performed.  17  P. expansum is capable of producing tyrosinase during germination. This fungal enzyme is one of the causes for secondary browning (Toivonen & Brummel, 2008). When P. expansum infects apple tissue, it produces enzymes capable of degrading the hemicellulosic polysaccharides and xyloglucan from the cell wall. This degradation of the cell wall causes cell-wall loosening, increasing its porosity and thus allowing for full infection of the apple tissue (Miedes & Lorences, 2004). Infection of apple tissue by P. expansum, results in the activation of several defense mechanisms. One of these mechanisms is lignification which enhances mechanical strength of cell walls and deters fungal infection (Valentines et al., 2005; Toivonen & Brummel, 2008). Peroxidase (POX) enzyme is the terminal enzyme needed for the polymerization and synthesis of lignin in apple tissue (Toivonen & Brummel, 2008). Valentines et al. (2005) found a correlation between lignin content and P. expansum resistance in Golden Delicious apples, indirectly showing the role POX enzyme plays in this process.  2.3.2 Strategies to control P. expansum in apples. One of the approaches used to control the growth of P. expansum in stored apples involves the use of antagonistic microorganisms as forms of biocontrol. Calvo et al. (2007) employed the bacterium Rahnella aquatilis on Red Delicious apples that had been injured and inoculated with P. expansum. They observed a nearly 100% reduction in incidence of disease caused by P. expansum in apples stored at 15°C. In apples stored at 4°C, however, infections caused by P.expansum were reduced by only 40%. Sánchez-Ventura et al. (2008) used antagonistic yeasts like Candida incommunis to inhibit P. expansum growth in vitro and in vivo in apples. Sanzani et al. (2008) tested the effectiveness of phenolic compounds: including: esculetin, ferulic acid, quercetin, resveratrol, scopoletin, scoparone, and umbelliferone in the inhibition of blue mould growth and patulin production. The in vitro studies they conducted showed that quercetin and umbelliferone were the most effective phenolic  18  compounds, the first inhibiting growth and the second inhibiting patulin production. They were also effective in vivo when tested on Granny Smith and Golden Delicious apples.  2.4 Natural antimicrobials proposed 2.4.1 Vanillin as an antimicrobial The mode of action of most natural antimicrobials can be classified as one of the following: 1) reaction with the integrity of the cell membrane, 2) inactivation of essential enzymes, and/or 3) the destruction or inactivation of cell DNA (Fitzgerald et al., 2004). Vanillin falls in the first category. 2.4.1.1 Vanillin Vanillin (4-hydroxy-3-methoxybenzaldehyde) is a major component of natural vanilla extract. Natural vanilla is obtained from the bean or the pod of the tropical vanilla orchid Vanilla planifolia mainly. Figure 2 shows the chemical structure of vanillin. Although 1,200 tons of vanillin are produced per year, less than 1% is of natural origin (Walton et al., 2003). The remaining 99% is produced synthetically. The difference in economic value of natural vanillin vs. artificial vanillin is considerable.  Figure 2: Vanillin chemical structure. Reprinted from (Walton et al., 2003) with permission from Elsevier ©  19  Vanillin is not soluble in water beyond 1% (w/w) (Karathanos et al., 2007; Chattopadhyaya et al., 1998). Its relative volatility is 0.0004 mg/cm2/hour at 24°C (Cartwright, 1953). The effects of vanillin as a bacteriostatic agent are well known. It has been found to be active against both Gram-positive and Gram-negative bacteria as well as against yeasts and moulds (Walton et al., 2003). Table 1 shows the results and experimental conditions of some studies that have been conducted to measure the antimicrobial activity of vanillin in different media.  20  Table 1: Studies that have demonstrated the antimicrobial activity of vanillin in different media.  Inhibited microorganism  Medium  Listeria innocua  apple juice medium apple juice medium and apple juice respective media* apple puree  Listeria monocytogenes, E. coli O157:H7  Escherichia coli, Lactobacillus plantarum, L. innocua Saccharomyces cerevisiae, Zygosaccharomyces rouxii, Debaryomyces hansenii, Z. bailii Escherichia coli, Pseudomonas respective aeruginosa, Enterobacter media* aerogenes, Salmonella enterica subsp. Enteric serovar Newport, Candida Albicans, Lactobacillus casei, Penicillum expansum, Saccharomyces cerevisiae  Listeria innocua  Inoculated into apple slices  18 fungi isolated from stored fresh-cut mango  PDA** and uninoculated fresh-cut mango and pineapple  Concentration pH used 3000 ppm (19.7 3.3 and mM) 3.8 40mM 3.42 and 4 35mM  Temperature (°C) 30°C 4 and 15°C  (Corte et al., 2004) (Moon et al., 2006)  neutral pH NR  27°C  6-18mM  NR  Not reported  (Rupasinghe et al., 2006)  0.6% or (39.4 mM) (achieved 3 log reductions) MIC for fungi ranged 5-15.8 mM For fresh-cut fruit, 80mM treatment was effective  NR  4°C  (Rupasinghe et al., 2006)  PDA adjuste d to pH 5  25°C for PDA 5°C and 10°C for fresh-cut fruit  (Ngarmsak et al., 2007)  2000 ppm (13.1 mM)  NR  Reference  (Fitzgerald et al., 2004) (Cerrutti & Alzamora, 1996)  NR stands for not reported *Respective media traditionally employed to grow the corresponding microorganism tested **PDA stands for potato dextrose agar  Rupasinghe et al. (2006) carried out a study with four pathogenic or indicator bacteria as well as four spoilage species representative of broad categories that can be associated with fresh-cut apples. In this study, total aerobic counts in fresh-cut Empire and  21  Crispin apples, were reduced by 37% and 66% respectively (against controls) during 19 days of storage at 4°C with a vanillin concentration of 12 mM (Rupasinghe et al., 2006). Vanillin was capable of strongly inhibiting aerobic bacteria in Fuji apple slices dipped in 40-120 mM vanillin solutions and stored for 3 weeks at 4°C, and moulds for 2 weeks of storage (Chung et al., 2009) Vanillin has been found to be bacteriostatic rather than bacteriocidal. Bacteriostasis is probably the result of inhibition of cell respiration due to damage to the membrane integrity, although this damage is less severe than that caused by bacteriocidal compounds like carvacrol (Fitzgerald et al., 2004). In eukaryotic cells, the aldehyde moiety seems to play a key role in the antifungal activity of vanillin. Consequently, the oxidation of vanillin to vanillic acid results in a dramatic decrease in antifungal activity (Fitzgerald, 2005). Cultivar specific factors may play important roles in the effectiveness of vanillin as an antimicrobial for fresh-cut apples. These include: surface pH, levels of essential nutrients (vitamins, minerals, and nitrogen containing compounds) or natural phenolic compounds originally present in apple flesh (Rupasinghe et al., 2006). One limitation of using vanillin at minimum inhibitory concentrations is its strong flavour, yet this might be partially overcome by combining it with other synergistic antimicrobials, thus lowering the effective concentrations that are necessary to achieve the desired preservative effect (Walton et al., 2003). Concentrations beyond 12 mM of vanillin would cause fresh-cut apples to become unacceptable organoleptically (Rupasinghe et al., 2006).  2.4.2 Hexanal as an antimicrobial For the typical apple (Malus domestica Borkh.) most flavour develops during the ripening stage, and it is during this climacteric peak that most endogenous volatiles are produced (Dixon and Hewett, 2000). Over 300 volatile compounds have been identified in 22  the traditional apple bouquet. These compounds include alcohols, aldehydes, carboxylic esters, ketones, and ethers. The aldehydes that provide the green/sharp notes and grasslike smell in apples are: acetaldehyde, trans-2-hexenal, and hexanal (Dixon and Hewett 2000). Some compounds provide the typical apple aroma/flavour (like ethyl-2-methyl butanoate for example), others the aroma intensity (e.g. trans-2-hexenal), and others contribute to the aroma quality (e.g. ethanol) (Dixon and Hewett, 2000). Apple volatile production can be categorized in many ways, one being according to C 6 compound production. For example Cox’s Orange Pippin and Jonathan apples are known to produce 4-5 times more hexanal and 100 times more trans-2-hexenal than Golden Delicious apples (Dixon and Hewett, 2000). At the time of harvest for Golden Delicious apples, Willaert et al. (1983) measured 1.05-1.57 g of hexanal / 5 liters headspace/ 100 g apple, which during storage increased to to 15.77 g per 5 liters headspace/100 g apple.Hexanal was a quantitatively important compound in immature green apples, in conjunction with trans-2-hexenal representing 30% of the total volatile composition. Hexanal is one of the compounds responsible for the green flavour character of pre-climateric unripe apples. 2.4.2.1 Hexanal Hexanal is a C6-aldehyde that is rapidly formed after the onset of mechanical damage in plant tissue (Arimura et al., 2009) derived from the oxidation of linoleic acid via lipoxygenase activity (Goff and Klee, 2006; Hatanaka, 1993. It is often described as a “green”, “grassy”, “herbaceous” odour (Dixon and Hewett, 2000; Aprea et al., 2009) and dubbed one of the “green leaf volatiles” (Baldwin, 2010, Matsui, 2006). C6 volatiles are normally produced in small amounts by intact plants, but they are emitted in larger amounts by plant tissue that has been wounded (Kant et al., 2009). Tissue disruption allows for the hydrolysis of galactolipids from chloroplast membranes providing the free fatty acids from which the C6 volatiles are derived (Kant et al., 2009; Matsui, 2006). 23  C6-aldehydes, like hexanal, are precursors for aroma compounds like hexylacetate and hexylhexanoate, naturally present in fruit like apples, pears, strawberries, bananas, pineapple, and melons. For this reason, its use in packaging as an effective antifungal agent seems viable (Song et al., 1998). Hexanal has a molecular weight of 100.16 g / mol and a solubility in water at 50°C of 3.5 mg/g of water (molar fraction of 6.32 x 10-4) (Hertel et al., 2007). The antimicrobial effect of hexanal is not fully understood but this effect is known to be linked to the affinity of hexanal for the microbial membrane phospholipid bilayer (SolivaFortuny & Martín-Belloso 2003a; Lanciotti et al., 2003). It has been suggested that its hydrophobic properties are important for its fungicidal activity (Matsui, 2006). Some evidence suggests that the membrane disruption caused by hexanal is followed by leakage of electrolytes and reducing sugars and amino acids from the cells (Fan et al., 2006). Gram-negative bacteria tend to display resistance against many antimicrobial agents due to: 1) the outer membrane acting as an efficient permeability barrier, 2) the high content of cyclopropanic fatty acids of the inner membrane (Chang & Cronan, 1999). This is no obstacle however for small hydrophobic hexanal, which is capable of entering the cell via porins, without altering the outer membrane permeability (Lanciotti et al., 2004). Hexanal provides antimicrobial activity toward a wide range of microorganisms (Lanciotti et al., 2004; Kant et al., 2009). Added to fresh-cut apple slices in doses not higher than 100 ppm at 4°C, it was able to fully inhibit mesophilic bacteria, and at 15°C it strongly delayed the growth of mould, yeasts, mesophilic and psychrotrophic bacteria (Lanciotti et al., 2004). Hexanal has a significant inhibitory effect against pathogens like E. coli O157:H7, S. enteritidis, S. aureus, and L. monocytogenes when inoculated in both model (Nakamura & Hatanaka, 2002) and real systems like fresh-cut apple slices (Lanciotti et al., 2003). In apple slices, hexanal, at a level of 150 ppm, had a bactericidal effect on L. monocytogenes and significantly extended the lag phase of E. coli and S. enteritidis inoculated at levels of 104-105 cfu/g (Lanciotti et al., 2004).  24  Table 2 shows a list of different minimum inhibitory concentration derived from studies that have been performed using hexanal against P. expansum utilizing different media and variable conditions. Table 2: Hexanal MIC assays for P. expansum.  Concentration (l/liter or ppm) 100  Exposure time  Medium  Type of inhibition  Reference  48 hours at inoculation  PDA*  Song et al. , 1998  250  48 hours at inoculation 48 hours at inoculation Exposed for 48 h, two days after inoculation Exposed for 48 h, two days after inoculation 48 hours at inoculation  PDA*  hyphal growth (colony diameter) reduced by 50% compared to controls inhibition for 120 hours after treatment fungicidal, no recovery  450 250 450 250 and 450  30.7 73.8 196.8 98.4 40 mol/liter (4922 l/liter or ppm) 4 mg/liter (4.9 l/liter or ppm)  indefinite indefinite indefinite indefinite 24 hours 18 hours  PDA* PDA* PDA* Golden Delicious and Jonagold MEA** MEA** MEA** MEA** Whole apple fruit Gala apples  slowed down colony growth, for P. expansum 50% less than control Stopped colony growth  Song et al. , 1998 Song et al. , 1998 Song et al. , 1998 Song et al. , 1998  growth inhibited,  Song et al. , 1998  inhibited conidial germination inhibited mycelial growth fungicidal for spores fungicidal for mycelial growth Reduced spore viability by 94% Reduced to low levels at 15°C  Neri et al., 2006 Neri et al., 2006 Neri et al., 2006 Neri et al., 2006 Fan et al., 2006 Sholberg & Randall, 2007  * PDA stands for potato dextrose agar. **MEA stands for malt extract agar.  Hexanal showed fungistatic activity against B. cinerea spores on PDA stored at 23°C for 7 days at a concentration of 1.5 l/liter of air (Almenar et al., 2007). The efficacy of pure hexanal in the inhibition of fungi depends on its initial dosage, its vapour pressure in the system (effective concentration), as well as the specific fungus 25  tested (Almenar et al., 2007). It should also be noted that some studies allowed for a continued contact between hexanal and the inoculated media/fruit, while others performed a timed exposure followed by active-agent-free storage. Higher concentrations are required for the second scenario, since in this “fumigation” technique, the active agent must be concentrated enough to achieve initial fungicidal activity. In addition, some studies (e.g. Song et al., 1998 and Fan et al., 2006) used flow systems to maintain a constant hexanal concentration in the headspace while others (e.g. Sholberg & Randall, 2007) dosed initially and allowed the hexanal to be depleted by absorption/metabolism even if the “fumigation” time was the same (48 hours). Song et al. (1998) found an interesting “side effect” of the hexanal treatment. After only 0.5 hours of exposure to 250 l/liter of hexanal, both Jonagold and Golden Delicious had a significant aroma production detectable in the air leaving the chamber (not detected in the hexanal-free chamber of the control apples). Hexanal was converted to hexanol, and by hour 5 and hexylacetate production had begun. By hour 7.5, butyl hexanoate and hexyl hexanoate were detectable. After 24 hours traces of hexyl butanoate, 2-methyl-propylhexanoate, and butyl butanoate were also detected by GC/MS (Song et al., 1998; Fan et al., 2006). All of these compounds occur naturally in the apple aroma bouquet. Another interesting and positive side effect was encountered by Corbo et al. (2000) who reported that low levels of hexanal increased colour stability in apple slices for up to 16 days. This effect presumably resulted from hexanal’s bioconversion to hexanol in intact fruit tissues and its interaction with polyphenol oxidase either by either preventing its production (de novo enzyme synthesis) or by inhibiting the preformed enzyme. Lanciotti et al. (2004) reported the same beneficial effects on colour, however they hypothesized that hexanal may do so by targeting phenyl-alanine ammonia lyase, a key enzyme in the biosynthesis of polyphenols. Either way, hexanal’s possible use as an enzymatic browning inhibitor in conjunction with its antimicrobial activity is desirable. Hexanal’s antimicrobial activity is dependent on its vapour pressure, since it is an indirect measure of its hydrophobicity. The higher the vapour pressure the less likely hexanal will remain soluble in water and in a liquid phase, and the more likely it is to enter 26  the vapour phase. The higher hexanal’s vapour pressure, the higher its solubility in the hydrophobic cell membranes, (enhancing its antimicrobial activity) (Gardini et al., 1997; Lanciotti et al., 2004). Hexanal’s vapour pressure increases with temperature, thus the not uncommon cold chain disruptions could aid in making this antimicrobial more effective when it is needed the most. It is not uncommon that certain microorganisms are more resistant to antimicrobials when exposed to harsher conditions. In the case of P. expansum however, it would seem that  lower  temperatures  make  it  more  sensitive  to  antimicrobial  compounds.  Leepipattanawit et al. (1997) found that only 25% of the 2-nonanone needed at 23°C to control the growth of P. expansum in apples was needed when apples were stored at 5°C. Fan et al. (2006) encountered a similar situation where low temperatures also seemed to enhance the efficacy of hexanal with only 52% of the apples at 4°C developing P. expansum inoculated lesions when exposed to 5-7 mol/liter (0.5-0.7 mg/liter) hexanal, while at 98% of the apples developed lesions at 23°C. The structure, precursor and odour threshold in parts per billion for hexanal are shown in Table 3 (Goff and Klee, 2006, Dixon & Hewett, 2000). Table 3: Structure, precursor and odour threshold for hexanal.  Odour Volatile  Structure  Precursor  Threshold (ppb)  Hexanal  Linoleic acid  5  At high concentrations, exposure to hexanal vapour will cause phytotoxic symptoms in fruit tissue. In Gala and Golden Delicious apples, 12 mg/liter of hexanal for 48 hours at 1°C or 2 mg/liter for 48 hours at 20°C were phytotoxic showing scald like discolouration in 27  the intact fruit (Sholberg & Randall, 2007). Red Delicious apples were more resistant and only damaged by 15 mg/liter for 48 hours at 1°C. These results show that the higher vapour pressure of hexanal at higher temperatures (thus its higher concentration in the headspace) allows for a smaller dose to be phytotoxic and these results are apple-variety dependent. 2.4.2.2 Toxicological information for hexanal Hexanal has been listed by the Food and Drug Administration (FDA) as “Everything added to Food in the United States” (EAFUS) and has been reported to have “ASP” status: (fully up to date toxicology information has been sought (FDA, 2010)). Hexanal has an oral rat LD50 of 4.89 g/kg (Fisher Scientific, 2012). Volatile compounds added to a packaging atmosphere may be absorbed by the fruit tissue. Since fresh products are mainly water, like microbiological media, Almenar et al. (2007) concluded that a fraction of the applied hexanal may be absorbed by the media during testing and exert its activity during storage. Almenar et al. (2007) quantified 5.2-12 g of hexanal per mg PDA. However, they did not perform this test in real fruit tissue. Since fruit tissue has a different organic compound composition than microbiological media and can convert the absorbed hexanal into other aromatic metabolites, the hexanal absorbed by the tissue may not be as effective an antimicrobial agent as it is in media.  2.5 Attempts to microencapsulate vanillin and hexanal 2.5.1 Vanillin microencapsulation Vanillin, being a compound used as a model for core material behaviour, has already been encapsulated by spray drying in gum arabic and oxidized starches by (Chattopadhyaya et al., 1998), as well as in starches of different origins including: amaranth, colocasia, chenopodium and rice (Tari and Singhal, 2002). Vanillin has also been encapsulated in CD, one of the proposed shell materials for this project, by freeze-drying (Karathanos et al., 2007). This particular study used nuclear 28  magnetic resonance (NMR) to show that the stoichiometry of the inclusion complex formed by vanillin-CD is 1:1, and that the complexation protects vanillin against oxidation and increases its solubility in water. The binding constant value of the vanillin-CD complex is 1.11 x 104 M-1 (Divakar, 1990).  2.5.2 Hexanal microencapsulation CD has been shown to create an inclusion complex with hexanal, with a 10 fold better yield than and  cyclodextrins (Almenar et al., 2007). Hexanal was successfully encapsulated in a 1:1 M relationship with CD and maximum release was attained by day 2 of the storage. Encapsulation was done by mixing with heat, centrifugation and further conventional drying at 60°C (Almenar et al., 2007). However, when tested against the pure volatile in hermetical containers that contained PDA inoculated with Colletotrichum acutatum, 1.2 g of CD-hexanal inclusion complex (that would supposedly achieve the needed 1.1 l/l concentration in the air) had a slow release that ranged from 0.15 to 0.49 l/l (headspace concentration) and was only able to inhibit radial growth by 30% when compared to the control. The pure hexanal achieved a concentration of 1.1 l/l by day 1, and slowly decreased to 0.002  l/l by day 7, but achieved 100% inhibition (Almenar et al.,  2007). Although the purpose of this study was to examine the potential application of the hexanal-CD inclusion complex in the preservation of berry fruit, no data was provided supporting that tests on fruit were performed. Lanciotti et al. (2004) strongly support the viability of hexanal as a shelf-life extender for fresh-cut produce for the following reasons: its effectiveness at low levels, its natural occurrence in fruits, and the possibility of “using unregulated doses as flavouring agents”. These properties together with the enhancement of aromatic compounds synthesis in apples, colour retention, and the possibility of use at lower concentrations due to a sustained release from microcapsules, suggest that microencapsulated hexanal may be an excellent option for fresh-cut fruit shelf-life extension.  29  2.6 Wall materials selected: 2.6.1 Beta cyclodextrin (CD) Molecular  inclusion,  unlike  conventional  drying  methods,  leads  to  microencapsulation at the molecular level (Desai & Park, 2005; Chen et al., 1993). Beta cyclodextrin (CD) is a cyclic derivative of starch, consisting of seven glucopyranose units. It is obtained from the industrial enzymatic treatment of maltodextrins using the enzyme glucosyltransferase (Szente and Szejtli, 2004; Madene et al., 2006; Del Valle, 2004)). The external part of the molecule is hydrophilic, while the internal cavity is hydrophobic. This property enables the molecule to entrap apolar guest molecules through hydrophobic interactions in a “molecular cage” (Acosta, 2008). The cavity of the βCD molecule is about 0.65nm in diameter (Desai & Park, 2005). It is lined by hydrogen atoms and glycosidic oxygen bridges. CD has the shape of a truncated conical cylinder (Astray et al., 2009). When βCD is mixed with apolar compounds that fit dimensionally within its cavity, an “inclusion complex” forms with the guest molecule in the presence of water. If the guest molecule is the wrong size, it will not fit properly into the βCD cavity (Azala-Zavala et al., 2008; Del Valle, 2004). The encapsulated molecule is held in place by hydrogen bonding, Van der Waal forces, or by the entropy-driven hydrophobic effect (Gouin, 2004). The stability reached by the inclusion complex is therefore proportional to the covalent bonding achieved due to the spatial arrangement of the molecules (Astray et al., 2009). “For a complex to form, there must be a favourable net energetic driving force that pulls the guest into the cyclodextrin” (Del Valle, 2004). The molecular mass of CD is 1135 (Szente & Szejtli, 2004). CD’s solubility is 1.85 g/100 ml of water at room temperature (Chatjigakis et al., 1992; Astray et al., 2009). CD structure is that of a hollow truncated cone shaped molecule or “torus shape” (Gouin, 2004). Figure 6 shows the structure of βCD. In an aqueous solution, the cyclodextrin cavity, which is slightly apolar, will be occupied by energetically un-favored water molecules (since they are of polar nature), and will be readily substituted by the 30  appropriate “guest” molecules of lesser polarity than water (Szejtli, 1998). The steric hindrance, or the relative size of the guest molecule and the cavity (Ayala-Zavala, 2008), has an effect on the speed of complex formation. With larger guest molecules, formation is slow and decomposition of the inclusion complex can occur (Astray et al., 2009).  Figure 3: Molecular structure and microstructure of CD. Reprinted from (Del Valle, 2004) with permission from Elsevier ©  There are three basic methods for molecular encapsulation using βCD. The first consists of mixing an aqueous solution of βCD with the active material, followed by separation of the crystals formed and drying. Separation can be done using filtration methods (Madene et al., 2006). The second method, involves using a more concentrated CD suspension and following the same procedure. The third and most practical of the methods consists of kneading a concentrated paste of CD and water with the active compounds to form the inclusion complex. This method does not require a final drying step (Desai & Park, 2005). A cyclodextrin-complexation method using a ball mill has been patented (Desai & Park, 2005). The drying conditions of the wet complexes influence the crystallinity, flowing properties, and other mechanical properties of the end product (Szente & Szejtli, 2004). Important factors to consider when forming an inclusion complex are: the geometric compatibility between the host and guest, the structure, charge and polarity of the guest molecule, the effect of the solvent/medium, and the temperature at which the complexation is taking place (Astray et al., 2009). Guest molecules that are less polar than 31  water and have a neutral charge form a complex more efficiently with βCD than do polar or ionized molecules (Astray et al., 2009). CD has been approved for use in foods by the Food and Drug Administration (Partanen et al., 2002). It has been a GRAS ingredient since 1998 and can be added as a flavour carrier and a protectant at a level of 2% in numerous products (Szente and Szejtli, 2004; Astray et al., 2009). Cyclodextrins are considered non-toxic since they are not absorbed from the gastrointestinal tract (Del Valle, 2004). The loading efficiency is typically low (about 11%), but the encapsulation efficiency is high (Acosta, 2008). CD is superior to starch in humid conditions and is also more thermostable (Partanen et al., 2002) withstanding temperatures up to 200°C (Acosta, 2008). The presence of water or high temperatures is required to release guest molecules within the complex (Madene et al., 2006). Cyclodextrins will protect flavours through processes such as freezing, thawing, and microwaving (Del Valle, 2004). Complexing flavours with βCD have proven to be very effective in protecting against heat and evaporation (Astray et al., 2009). When vanillin is mixed with a solution of βCD, both the phenolic end and the aldehyde end are important in the complexation (Divakar, 1990). Within the torus shaped CD, the phenolic end is nearer to the narrower part of the torus shape, while the aldehyde end locates at the wider end of the CD molecule (Divakar, 1990). The maximum βCD level recommended in food by the Joint FAO/WHO Expert Committee on Food Additives (JECFA) is 5 mg/kg per day (Astray et al., 2009).  2.6.2 Pectin Pectin is the name given to a group of natural polymers of (1→4) linked -D galacturonic acid partly esterified with methanol. The percentage of methyl ester-links to galacturonic acid subunits, also known as the degree of esterification (DE), influences the 32  functional properties of pectin. There are two main classifications according to the degree of esterification: low-ester pectins (DE<50%) and high-ester pectins (DE≥50%) (Rolin & De Vries, 1990). High ester pectins form gels at low pH values when the water activity is reduced by the addition of sugars. Low-ester pectins on the other hand, form gels in the presence of calcium ions (Rolin & De Vries, 1990). Commercial pectin consists mainly of the partial methyl esters of polygalacturonic acids and their sodium, potassium, calcium and ammonium salts (Rolin & De Vries, 1990). The percentage of amidated galacturonic acid subunits is referred to as the degree of amidation (DA). 2.6.2.1 Solubility of pectin Pectin is soluble in pure water, but it is not soluble under conditions under which it can form a gel (Rolin & De Vries, 1990). Solutions of up to 10% (limited viscosity) pectin can be made by gradually adding pectin to water at not less than 60°C in an efficient mixer. 2.6.2.2 Rheology of pectic solutions Pectin and water mixtures are viscous; their viscosity depends on the presence of salts, particularly salts of calcium or other non-alkali metals, and on the pH. Other factors that affect a pectic solution’s viscosity are: the degree of esterification and the average molecular weight. Most pectins form thixotropic solutions in the presence of calcium. The stiffness of the gel formed depends on the pectin concentration, calcium concentration, pH, and pectin type. There is almost no difference between the setting and melting temperatures of low ester pectin gels, therefore it is usually possible to re-melt a low ester pectin, while this is usually not a possibility for high ester pectins (Rolin & De Vries, 1990). 2.6.2.3 Gelation of low ester pectin At higher DA, low ester pectins require more calcium ions to gel, as well as lower gelling temperatures. These gels can be re-melted, therefore they are considered to be thermoreversible. Gelation occurs immediately after the gelling conditions are introduced. 33  The mechanism of gelation consists of calcium bridging between the chains of galacturonic acid. The “egg-box” model has been used to explain the low-ester pectin/calcium ion gelation mechanism. Junction zones consist of pectin chains in helices with two subunits per turn. These two-fold helix structures are joined by calcium ions that bridge two opposing carboxyl opposing groups (Rolin & De Vries, 1990). Amidated pectins need less calcium than non-amidated pectins (Rolin & De Vries, 1990). Gel strength is positively correlated to the concentration of soluble solids and negatively correlated to the pH. The typical pH range of low-ester pectin solutions is 3-4.5 (Rolin & De Vries, 1990). In systems with less than 20% soluble solids, solely low ester pectin, gels have a large tendency for syneresis (Rolin & De Vries, 1990). Pectin produces stable emulsions at low concentrations. It has very good emulsifying properties due to the protein residues present in the pectin chains, as well as the high content of acetyl groups (Gharsallaoui et al., 2007). For years, research has been made to develop pectin as a drug delivery system (DDS) (Liu et al., 2007). Limitations for this approach include: a lack of reproducible performance in pectin formulations, and the large diversity of pectin molecular characteristics, which generate problems in quality control (Liu et al., 2007).  2.6.3 Criteria for selection of wall materials Important criteria that should be considered when selecting a coating material for an encapsulation process include: solubility, molecular weight, film forming and emulsifying properties, mechanical strength, compatibility with the food product, appropriate thermal or dissolution release, and cost (Gharsallaoui et al., 2007). Non-reactivity towards the encapsulated material during processing and prolonged storage, the ability to seal the active material, as well as release characteristics under particular conditions, and regulatory considerations must also be taken into account (Desai & Park, 2005). Important aspects to evaluate in the final microencapsulated product are: encapsulation efficiency, stability under different storage conditions, degree of protection 34  provided by the coating material, and surface characterization using scanning electron microscopy (SEM) (Gharsallaoui et al., 2007). Two parameters should be considered when assessing the efficiency of microencapsulation. The first parameter is the encapsulation load, or “payload” which is the weight fraction of the active ingredient (db) in the encapsulated product (this includes the active ingredient that is within and on the surface of the microcapsules). The second parameter is the core retention or “encapsulation efficiency” which is the fraction of the active ingredient (initially added to the emulsion) that is completely encapsulated in the final product (Acosta, 2008). Since vanillin is not very volatile, both the “payload” (surface + entrapped) and “encapsulation efficiency” (entrapped) can be measured. Hexanal being a volatile evaporates from the matrix if not entirely bound, thus the “payload” is equivalent to the encapsulation efficiency since it exclusively considers the hexanal that is completely encapsulated and thus retained. CD was selected as one of the coating materials due to its unique characteristic of creating a molecular inclusion complex. Its GRAS status, as well as previous success in encapsulating vanillin (Karathanos et al., 2007) make it an ideal encapsulating agent to test with REV technology. Pectin was selected as one of the coating materials since it is naturally present in fruit and thus it does not introduce completely foreign material. Use of other materials like proteins for example, could introduce complex issues associated with potential allergenic properties, and the possibility of rejection by some consumers (like vegetarians if animal proteins were to be used). As was mentioned earlier, pectin produces stable emulsions and it provides the viscosity required for the use of REV technology. Another important reason for choosing pectin is the possibility of exploiting the pectinase activity associated with the elevated metabolism of cut fruit, so that it may degrade the pectin matrix and release the core for its antimicrobial functions. Since pectinases are not volatile, this release would only work for the vanillin microcapsules that would be in direct contact with the fruit and not for the hexanal gaseous application.  35  2.7 Controlled release Controlled release is defined as the method by which one or more active agents are made available at a specific rate at the target location and time. (Pothakamury & BarbosaCánovas, 1995; Desai & Park, 2005). Factors that affect the release of a compound from a capsule include: the type and geometry of the capsule, the transfer from the matrix to the environment, and the degradation or dissolution of the matrix material (Madene et al., 2006). Releasing food additives at a desired time and location can broaden the application of food ingredients by ensuring optimal concentration and consequently providing the manufacturer with the economic benefit of requiring less raw materials (Desai & Park, 2005). For example, cyclodextrin-complexed antimicrobial agents have been incorporated into food packaging plastic films. The loss of volatile antimicrobial substances has been effectively reduced, showing an improvement in preservation of the food (Szente & Szejtli, 2004). Controlled release delivery systems are categorized according to the type of trigger (physical or chemical) needed to achieve release (Pothakamury & Barbosa-Cánovas, 1995). Many possible triggers can be used to initiate the release of a microcapsule’s content such as: pH changes, mechanical stress, temperature, enzymatic activity, time, osmotic force, or digestion of an enteric coating (Desai & Park, 2005). Active agents are usually released by diffusion, biodegradation, swelling, or osmotic pressure (Pothakamury & Barbosa-Cánovas, 1995). Diffusion in the microcapsules is controlled by the solubility of the active compound in the encapsulant and the permeability of the compound through the matrix (Madene et al., 2006). For example, when βCD comes in contact with water it partially releases the active compounds entrapped in it by the dissociation of the inclusion complexes in aqueous systems (Szente & Szejtli, 2004). The interaction between the βCD and its host is weakened by water, the complex losing its stability. Water molecules are capable of 36  interacting with the polar groups of the CD-guest complex, which causes a displacement of the guest compound from the interior of the cyclodextrin cavity (Ayala-Zavala et al., 2008). The compounds contained in the cavity will be released whenever a better substrate becomes available and replaces them in the cavity to form an even more stable complex with the cyclodextrin (Gouin, 2004).  2.7.1 Volatile controlled release Ayala-Zavala et al. (2008) propose the use of cyclodextrins as antimicrobial vehicles to coat onto the packaging of fresh-cut produce. The high relative humidity generated by the moisture loss by the product would trigger the release of the hydrophobic antimicrobial compounds encapsulated in the CD and coated onto the “active” package. They conclude that generation of AP (active packaging) material based on natural antimicrobial-CD complexes could be an alternative to modified or passive atmosphere packaging” (AyalaZavala et al., 2008). Polysaccharides may influence the volatility of aromatic compounds mainly by the following two mechanisms. The first one relies on a diffusion decrease following the Stokes-Einstein equation in which diffusion is inversely proportional to viscosity (Baines and Morris, 1987). The second mechanism involves interactions between the volatile compound and the polysaccharide, which could be: possible entrapment in microregions, complexation, encapsulation, and molecular interactions including hydrogen and covalent bonding and hydrophobic interactions (Godshall, 1997; Baines and Morris, 1987). The release of a volatile compound is determined by thermodynamic and kinetic factors. Partitioning of an aromatic compound between a food matrix and the gaseous phase above it under equilibrium conditions is described by the partition coefficient (defined as the ratio of the concentration of a volatile in gaseous phase and in food matrix). This partition coefficient is the thermodynamic component of volatile release. A compound with a high partition coefficient will distribute with more ease in the gaseous phase and will  37  likely have a low threshold value (Terta et al., 2006). The kinetic component consists of the rate at which equilibrium is obtained (Bakker, 1995; Secouard et al., 2003). Static headspace analysis provides information on the air/product partition coefficient, whereas dynamic headspace analysis helps determine mass transfer behavior and temporal release of aromatic compounds (Terta et al., 2006).  2.8 Instrumental techniques 2.8.1 Gas chromatography (GC) Gas chromatography is a powerful technique for separating and quantifying gaseous and volatile compounds. The sample is dissolved/transported by a mobile phase, usually helium gas, and forced through an immiscible stationary phase, which is fixed in a column of known dimensions (Rounds and Nielsen, 2003). The time taken by the compound being measured to travel through the stationary phase and reach the detector depends on its affinity to the column and is referred to as the “retention time”. The peak produced by the compound at that specific time can then be both identified and quantified. (Rounds and Nielsen, 2003). In a flame ionization detector (FID), the effluent from the column is directed into an air-hydrogen flame. The target compound produces ions and electrons upon pyrolization in the flame, which are then directed to a collector electrode. The resulting current (which is around 10-12 Amps) is measured by a high-impedance picoammeter. The signal shows as a peak in the chromatogram and can later be translated into an actual amount using a calibration curve. Since the FID is a mass-sensitive device (instead of concentrationsensitive), any changes in flow rate of the mobile gas have little effect on the detector response (Reineccius, 2003). The FID has a high sensitivity (~10 -13 g/s), a large linear response range (~10-7), and is the most useful general detector for most organic samples (Reineccius, 2003)  38  2.8.2 Solid phase microextraction (SPME) A solid phase microextraction (SPME) device consists of a hollow needle containing a small fused silica rod (1 cm length by 0.11 cm internal diameter) that has been coated with the polymeric phase, which consists of a sorbing material (Castro et al., 2008; Nerin et al., 2009, Shirey, 2012). The principle underlying the SPME device is the second law of thermodynamics, which states that the chemical potential of each compound should be equal throughout the system. In other words, the compound(s) are distributed until all the “corresponding partial molar free energies are the same in all parts of the system formed by the fiber and the sample” (Nerin et al., 2009). Namely, it relies on the equilibrium partition process between the fiber coating and the aqueous solution (Zhang & Pawliszyn, 1993). Several parameters affect the capacity and efficacy of SPME extraction. These include: the kind of polymeric coating of the fibre, the extraction temperature and time, the saline effect, the pH of the sample, the volume of the sample, the volume of the headspace, the agitation of the sample and the shape of the vial/container (Castro et al., 2008). The type of fiber used is dependent on the nature of the analyte (Castro et al., 2008). The thickness of the coating determines the fibre capacity for adsorbing the analyte, as well as the time required in the extraction time to reach equilibrium (Shirey, 2012). Polydimethylsiloxane (PDMS) is a non-polar material that has great affinity for apolar compounds but can be used to extract moderately polar compounds as well (Castro et al., 2008). Mixed phases, like polydimethylsiloxane/polydivinylbenzene (PDMS/DVB) and carboxen/PDMS are more polar than those of polyacrylate (PA), and are suitable for extracting more polar compounds like alcohols and ethers. Carboxen/PDMS have larger surface area, which makes them good for the extraction of volatile organic compounds (Castro et al., 2008). Hexanal, being an aldehyde, is slightly polar. Hydrogen bonds and dipole-dipole interactions are responsible for attaching the analyte to the fiber. This process is usually referred to as “extraction”, however “sorption” is 39  a better term (Nerin et al., 2009). The active points for sorption in the extracting fiber work as as scavengers of the analyte (Nerin et al., 2009). The particular SPME fiber used for the research described herein was a 55/30 m DVB/carboxen-PDMS needle. This fiber is appropriate for analytes C3-C20 (hexanal is C6) (Supelco, 2012). It is of the adsorbent type, which means it physically traps or chemically reacts/bonds with analytes (Shirey and Mindrup, 1999). This mixture of porous materials is capable of retaining analytes until energy or a solvent is applied. It is a bipolar fiber with a high surface area but a limited capacity and in which analytes are capable of competing for bonding sites (Shirey and Mindrup, 1999). Carboxen-PDMS fibers are best for the analysis of smaller analytes (MW<125) (Shirey, 2012), so this is an appropriate fiber to use for hexanal (MW = 100.16). Giuffrida et al. (2005) compared different SPME fibers for hexanal extraction (as a lipid oxidation indicator) in beef bouillons. They determined that DVB/CAR/PDMS showed a better performance in terms of recovery of the analyte, and best results were obtained at 37°C for 40 min (Giuffrida et al., 2005). An important aspect to consider in a qualitative and quantitative technique such as SPME is reproducibility. One fiber can perform multiple extractions obtaining reproducible results. A “new fiber should produce similar results compared to the replaced fiber” (Shirey, 2012). SPME fibers can be immersed in a stirred solution or can be used to sample the headspace above one. Carboxen/PDMS fibers work best in the heaspace mode since it reduces the amount of non-volatile compounds that contaminate the surface and end up in the GC. This is also another advantage of using SPME instead of a gas tight syringe for example, since oxygen and moisture injected into a GC column reduce the column lifetime (Zhang & Pawliszyn, 1993). Extraction is also faster when the analytes sampled are in the headspace (Shirey, 2012). In order to use the headspace SPME method, two analyte parameters are important to consider: the Henry’s constant (KH) and the partition coefficient of the analyte between 40  octanol and water (Kow). An practical equilibration time can still be reached, for compounds with small KH, if their Kow,values are small as well (Zhang &Pawliszyn, 1993). This is the case for hexanal, which has a KH of 208.33 cm3 x atm/mol (Sander, 1999), and a logKow of 1.89 (Spafiu et al., 2009). To put this into perspective, a PDMS coating is very effective if the log10Kow < 4, and the KH >90 atm x cm3/mol (Zhang & Pawliszyn, 1993). The PDMS coating is therefore an adequate material for hexanal as a heaspace measured analyte. Song et al. (1997) found that extraction times of 6-8 minutes were sufficient for apple samples. The amount of compound absorbed by the SPME fiber will decrease with temperature, however if equilibrium is reached, then there should be no significant effect on accuracy between 0-40°C (Arthur et al., 1992). Figure 4 depicts the adsorption/desorption process in SPME fiber testing. The two steps involved in the analysis are: 1) partitioning of the volatiles between the extraction phase and the sample, 2) desorption of the concentrated analytes into the GC (Risticevic et al., 2009).  Figure 4: Adsorption/desorption for SPME sampling. Reprinted from (Shirey, 2012), with permission from Elsevier ©  41  3 Vanillin Project: Antimicrobial Controlled Release on Fresh-Cut Surface 3.1 Research objectives The first objective of the vanillin project was to successfully encapsulate vanillin in CD and low ester pectin using REV technology. The second objective was to compare and contrast encapsulation efficiency, controlled release, and yield of REV produced microcapsules against the conventional method of freeze-drying as a control. The project aimed to demonstrate microbial inhibition of the target microorganism: P. expansum, as well as to quantify the controlled release profile of the produced microcapsules in liquid media.  3.2 Hypothesis It was hypothesized that vanillin would be successfully encapsulated in a blend of CD and pectin using REV processing. The REV produced microcapsules were expected to show at least equivalent yield and encapsulation efficiency as the freeze-dried ones. Vanillin-containing REV microcapsules were expected to release their contents in a controlled manner when exposed to liquid media. Finally, vanillin-containing REV microcapsules were expected to inhibit P. expansum spore germination.  3.3 Methodology outline A staged strategic plan was developed to achieve the research objectives of the vanillin experiments. Figure 5 summarizes the stages of this strategic plan. A more detailed explanation of the experiments that were performed is contained in the Materials and Methods of section within this chapter.  42  Figure 5: Strategic plan for studies on the controlled release of vanillin.  43  3.4 Materials and methods 3.4.1 STAGE 1: Formulation and processing 3.4.1.1 VanillinCD-pectin blend batch formulation Solutions containing vanillin (Sigma) and βCD (Wako Pure Chemical Industries) were prepared in a ratio of 1:10 by weight as recommended by Chattopadhyaya et al (1998), and by Tari et al. (2002). βCD was solubilized in water using a mechanical stirrer operated at 100°C for 2 min. The vanillin was solubilized in propylene glycol (Fisher Scientific) (4.6% by weight of total mixture) by stirring with mild heating. Once dissolved the vanillin was introduced to the βCD solution. Low methoxy pectin (Gum Technology’s Coyote Brand LM 50, Tucson, AZ 85737) was then added to achieve 1:15, 1:17.5 and 1:20 (by weight) vanillin to pectin ratios. The mixtures were mixed by hand before homogenization for 2 min at 11,000 rpm using an Ultra Turrax T25 basic homogenizer. Control mixtures were also prepared by the same procedures including the propylene glycol without vanillin. Altogether, this process was repeated six times to obtain 6 different batches or repetitions. One batch of solutions with increased vanillin-pectin ratios was made by doubling, tripling and quintupling the vanillin content in the 1:20 blend formulation while maintaining the rest of the ingredients constant. These formulations, termed “2:20” (1:10), “3:20”, and “5:20” (1:4), were processed in the same manner as the rest. 3.4.1.2 REV drying The mixtures were of high viscosity and almost immediately gelled to a stiff semisolid material that was cut into cubes of 3.5 to 5 g. Half of the original mixture was placed in the freezer at -50°C for eventual freeze-drying. Cubes that were not frozen were placed into custom-made open quartz cylindrical containers (3.1 cm diameter and 1.9 cm height) to be processed using a 900 W REV machine, (model VMD900W, EnWave corp., Richmond, BC). Cubes were dried for 12 to 20 min at 20torr and 200 W. The samples were treated until a range of 63 to 71 % of the water was removed (verified by differences in 44  weight). The dried matrices enclosed in an aluminum foil envelopes were ruptured using a hammer. The pieces were then mechanically ground using a mortar and pestle. The resulting powders were stored in BD Falcon 50 ml conical centrifuge tubes (Fischer Scientific), inside a desiccator. The small scale REV machine used for this experiment was model VMD900W (EnWave Corp., Vancouver, BC). This particular model was designed with a “travelingwave” microwave applicator in which a quartz tube is connected to a vacuum pump and a variable power, 2450 MHz microwave generator (Durance et al., 2006). “Cold” and “hot” spots in a microwave chamber are the result of attenuation of a microwave field strength by absorption of the “lossy” material, and standing waves caused by reflection and interference (Durance & Yaghmaee, 2011). The latter is dramatically reduced in traveling wave designs like this one, resulting in more even microwave fields (Durance & Yaghmaee, 2011). 3.4.1.3 Freeze-drying One half of each mixture was frozen at -60°C and dried in a freeze drier (Labconco Freeze Dryer 18) for 48 h at a condenser temperature of -50°C 150mtorr absolute pressure. Freeze-dried samples were stored in BD Falcon 50 ml conical centrifuge tubes (Fischer Scientific), inside a desiccator. 3.4.1.4 Moisture analysis Moisture analysis was performed in duplicate by drying under vacuum (27 in Hg)for 24 hours at 70°C, using a Shell Lab Vacuum oven (model No. 1430). 3.4.1.5 Water activity determination Water activity measurements were obtained in duplicate for each sample using a water activity meter (Aqualab ® Model Series 3 Decagon Devices Inc Washington USA).  45  3.4.1.6 REV-encapsulation of vanillin in CD-calcium-crosslinked pectin matrix Controlled release tests in aqueous media (see section 3.5.4) unfortunately showed that the maximum vanillin release of the microcapsules would occur within 24 hours, which would limit applications where the release needs to occur over many days or weeks, as is the case with the fresh-cut apple application. In an attempt to reduce the rate of vanillin release, the encapsulation approach was modified by cross-linking the pectin with calcium ions (ionic gelation). Aqueous solutions containing pectin, vanillin, propylene glycol/ethanol (to solubilize vanillin), and CD were homogenized for 2 min at 11,000 rpm using the Ultra Turrax T25 basic homogenizer. The formulations (% weight) are shown on Table 6. The vanillin was dissolved in the propylene glycol under mild heat or at room temperature if ethanol was employed. The homogenized mixture was poured into a 20 ml syringe with a 21 gauge needle. The syringe was then used to extrude droplets of the mixture into a 1 M CaCl2 (BDH Chemicals) stirred solution at room temperature. The droplets immediately formed visible opaque pseudo-spheres (keeping the original size of the droplets) upon contact with the CaCl2 solution. A maximum of 3.33 g pectin/mole of CaCl2 were mixed to ensure enough Ca2+ ions were available for cross-linking. Once the extrusion was completed, the pseudo-spheres were left to harden in the solution with gentle stirring for 3 hours at room temperature, as described by Jaya et al. (2009). After the 3 hour-hardening period, the pseudo-spheres were harvested by filtering the solution using #4 filter paper (rapid filtering, Whatman brand). The residual CaCl2 solution was retained to quantify the vanillin loss during hardening. The harvested pseudospheres were dried using the 900 W REV machine (model VMD900W, EnWave corp., Richmond, BC) for 5 minutes at 20torr and 100 W (to reduce the amount of spattering) and then for 25-30 min at the same pressure but increasing power to 200 W. Spheres were dried until 88-92 % of the water was removed in the samples (verified by differences in weight).  46  3.4.2 STAGE 2: Quantification and yield analysis 3.4.2.1 Surface vanillin extraction Unlike CD and pectin, vanillin is soluble in ethanol. It was therefore assumed that an extraction with absolute alcohol would remove all the vanillin that was not completely encapsulated by either or both of these compounds. Alcoholic extraction was performed as specified by Chattopadyaya et al. (1998) and by Tari and colleages (2003). Samples (0.200.25 g) of each treatment/control were washed in 25 ml of absolute ethanol and filtered using #1 Whatman filter paper. The filtered ethanol was then diluted to 50 ml in a volumetric flask. The remaining residue was then washed in water and put under mechanical stirring, with slight heating, until no visible particles remained. The volume was then made up to 100 ml using distilled water. Each batch was extracted in duplicate for samples and controls. The REV-dried calcium cross-linked pseudo-spheres were processed as described above, however the sample size was reduced to approximately 0.125 g of each sample/control due to limited yield. The samples were washed in 5 ml of absolute ethanol and diluted to 10 ml with distilled water. The remaining residue was insoluble in water (even when adding some ethanol) and had to be homogenized to rupture the microcapsules and achieve the release of the encapsulated vanillin. The same homogenizer employed in the batch formulation was used, at 1,000 RPM for a minute at room temperature. The homogenized sample was then diluted to 25 ml with water. Less dilution was required for analysis of the cross-linked capsules since the expected amount of vanillin was small. Residual CaCl2 solutions were diluted (1/10) in water prior to analysis by the Folin Ciocalteu method (detailed in the following section). The basic pH required for the method caused some of the CaCl2 to precipitate in. Consequently the reacted solutions were placed in 1.5 ml Eppendorf 1.5 ml tubes and spun at 10,000 rpm (relative centrifugal force of 7000) in a Microcentrifuge (Desaga MC2) for 5 min at room temperature immediately before measurement to eliminate interference during spectrophotometric analysis. 47  3.4.2.2 Colourimetric quantification method The Folin Denis colourimetric phenol assay was used to measure vanillin concentrations in the extraction samples as performed by Chattopadyaya et al.. (1998), Tari et al. (2002), and originally proposed by North (1949). Because production of the Folin-Denis reagent has been discontinued the method was adapted to permit for the Folin Ciocalteu Phenol Reagent. The latter is actually an improved version of the original reagent. It provides advantages including formulation with lithium sulfate which prevents precipitation and thus eliminates the need for filtration, and the formation of a bluer pigment with all phenols, particularly vanillin (Singleton & Rossi, 1965). The principle underlying the method is the transfer of electrons in alkaline medium from the phenolic compounds to phosphomolybdic / phosphotungstic acid complexes, which produces a blue colour (Singleton & Rossi, 1965). This colour can then be measured using a spectrophotometer. Standard curves were prepared following the procedure specified by Bärlocher and Graca (2005), starting from a 0.1 mg/ml stock solution. Dilutions were made to create a curve within the range of 0.01 mg/ml to 0.1 mg/ml of vanillin. One ml of each dilution was mixed with 5 ml of 2% Na2CO3 in 0.1 N NaOH. After five minutes 0.5 ml of a 1:2 FolinCiocalteu reagent-distilled water mixture was added and the mixtures were reacted for two hours. Absorbances were read on a Shimadzu UV-1700 PharmSpec spectrophotometer (model CPS-24OA). A spectrum analysis was conducted to identify the optimum wavelength for measurement (736 nm). Standard curves prepared with vanillin solubilized in water were performed in duplicate on separate days. Similar standard curves were also prepared with vanillin solubilized in ethanol (1 ml of vanillin solution with 50% ethanol). Extracted solutions (filtrate and residue) were analyzed by the same procedures but without further dilution. Filtrate solution concentrations were determined from standard curves prepared with vanillin solubilized in ethanol, while residue solution concentrations were determined from standard curves prepared with vanillin solubilized in water. Vanillin concentrations in the filtrate were considered to be derived from the surface, while  48  concentrations in the residue were considered to be derived from entrapped or encapsulated vanillin. 3.4.2.2.1 High performance liquid chromatography spectrophotometric methods.  (HPLC)  validation  of  the  An HPLC method was employed to determine the validity of the FC spectrophotometric assay detailed in the section above. Vanillin solutions of 0.02, 0.04, 0.06, 0.08 mg/ml were prepared from a stock solution consisting of 0.102 mg/ml and analyzed using the FC assay and a HPLC method adapted from Gumi et al., 2009 and Peña et al., 2009. Samples (5 l) were analyzed on a HPLC equipped with an Agilent Eclipse XDB-C18 column using a 70:30 ratio of MilliQ water:acetonitrile as the mobile phase. The flow rate was set at 1 ml/min, and the column temperature at 40°C. The analysis time was 4 min, and the vanillin concentration was determined at 229 nm with a retention time of 2.8 min. Eight additional vanillin solutions were prepared in water with similar concentrations as the ones usually encountered when quantifying microcapsule extractions using the FC assay, as well as concentrations on the lower and higher end of the standard curve range. Concentrations were measured by both the HPLC and the FC assay with the purpose of comparing quantification accuracy 3.4.2.2.2 Mass balance predictions for recovery and encapsulation efficiency calculations  The proportion of vanillin dry basis (db) in the original solutions was determined to calculate the recovery and encapsulation efficiency after processing. The boiling points of the propylene glycol-water solutions were expected to be greater than 100°C because they contained 7-8% propylene glycol (Curme & Johnston, 1952). Since both water and propylene  glycol  were  likely removed  during  moisture  analysis,  the  expected  concentrations of vanillin (db) in the samples were calculated assuming the calculated moisture was only water, although it was likely a mixture of both components. Consequently, the expected amounts of vanillin in each dried sample (db) would be the 49  proportion of the vanillin originally introduced into the mixtures, divided by the total amount of solids (excluding propylene glycol). Equation 1 shows the formula for this calculation.  Equation 1: Theoretical fraction of vanillin in mixture (db)  Table 4 shows the expected or theoretical amounts of vanillin (db) in each treatment in mg of vanillin per gram of dry powder.  Table 4: Theoretical amounts of vanillin (db) in each vanillin treatment.  Powdered extract 1:15  Theoretical vanillin concentration (mg/g dry powder) 38.46  1:17.5  34.09  1:20  31.29  Table 5 shows the experimental coefficients obtained for the standard vanillin water and alcohol based standard curves. Table 5: Vanillin standard curve coefficients.  Equation coefficients  Coefficient of Determination 2  CURVE  m  b  water based @ 736 nm  5.865  0.084  R 0.931  alcohol based @ 736 nm  4.593  0.042  0.965  50  The coefficients of determination for the water based and alcohol based curves were 0.931 and 0.965 respectively, which confirmed the linear relationship between absorbance and concentration for vanillin solutions between the range of 0.02 mg/ml and 0.1 mg/ml. The percentage (by weight) of original vanillin measured in the residue solution was referred to as “encapsulation efficiency”. The sum of the encapsulated and surface vanillin concentrations over the original amount in the sample was considered as “vanillin recovery” or yield. Equation 2 and Equation 3 show the calculations for “encapsulation efficiency” and “vanillin recovery” respectively.  Equation 2: Formula for encapsulation efficiency  Equation 3: Formula for vanillin recovery  3.4.3 STAGE 3: Microcapsule physical characterization 3.4.3.1 Scanning electron microscopy (SEM) Scanning electron microscopic (SEM) images were generated using a Hitachi S2500 microscope at Pacific Agri-Food Research Centre (4200 Highway #97, South, Summerland, British Columbia, V0H1Z0). Dried microcapsules were fixed onto an iron stub, using adhesive, and then made electrically conductive by sputter coating with a thin layer of gold. The images obtained ranged in magnification from 20 to 700x. 51  3.4.3.2 Dielectric property measurement The  dielectric properties  (at  2.45  Ghz)  of  the  1:15,  1:17.5  and  1:20  vanillin:pectin:CD blends were measured in their dry form in triplicate (5 measurements per replicate) using an Agilent Technologies E5071C ENA Series Network Analyzer (9 kHz to 4.5 GHz) coupled with Agilent 85070 E Dielectric Probe Kit (200 MHz to 50 GHz).  3.4.4 STAGE 4: Controlled release 3.4.4.1 Controlled release in aqueous media Controlled release of vanillin from microcapsules made with the 1:20 vanillin : pectin formulation was examined in duplicate for two separate runs (4 reps) for both the REV and freeze-dried forms. To assess the controlled release rate in aqueous media, 0.5 g of dried microcapsules were combined with 50 ml of distilled water and placed in a distilled water bath set at 25 ± 2°C (Thelco Model 4 Incubator). Sub-samples (0.5 ml) were removed at predetermined time intervals and replaced with 0.5 ml of distilled water, taking care to mix and re-suspend the remaining particles. The sub-samples were placed in 1.5 ml UltrafreeMC Millipore tubes with built-in 0.45 mm filters and were spun for 13 minutes at 11,000 rpm (RCF of 8470) in a Microcentrifuge (Desaga MC2) at room temperature. The filtered supernatants were diluted 1:4 and residual vanillin concentrations were measured using the Folin-Ciocalteu method as described above. The measurements were assumed to provide the total amount of vanillin released from both the surface and interior of the microcapsules at the time of sampling. Controls were processed in the same fashion and the supernatants were used as blanks for spectrophotometric analyses. Sub-samples were withdrawn every 15 minutes for the first hour, every hour for the first 6 hours and at 24 hours. The initial and final pH of the mixtures was pH 3.1 ± 0.16, measured using a Fisher Scientific Accumet Basic AB15 pH meter Vanillin release from REV 1:20 microcapsules was also examined at 5°C to assess the effect of temperature on release in aqueous media. One trial was performed at this temperature (2 reps) using the procedures described above with samples incubated in a Forma Scientific Incubator set at 5 ± 1°C. 52  3.4.5 STAGE 5: Microbial inhibition 3.4.5.1 Measurement of minimum inhibitory concentration (MIC) in pH adjusted liquid media The minimum inhibitory concentration (MIC) of vanillin required to inhibit P. expansum strain 1525 (apple isolate, Pacific Agri-Food Research Centre, Summerland, BC) was examined in a liquid growth medium at pH 4 and 5. Actively growing cultures of P. expansum strain 1525 on petri plates containing Potato Dextrose Agar (PDA, Difco) were used to prepare spore suspensions. Five ml of sterile water were added aseptically to the plates which were held at room temperature for 30 min with occasional swirling. Fluids were removed with a pipette and diluted with 15 ml of sterile water. The spore suspension was agitated on a vortex prior to inoculation or analysis. Spore concentrations were determined by the spread plate method. Suitably diluted aliquots (0.1 ml) were spread on the surface of duplicate in PDA plates which were incubated upside down at 25°C for 5 days prior to counting colonies. The spore suspension concentration was 9.8 x 105 cfu/ml MIC measurements were performed in Potato Dextrose Broth (PDB, HiMedia) prepared at 1.5 x the recommended concentration. The pH was adjusted to pH 4 and 5 (measured pH 3.99 and 4.89 respectively) using a 1M citric acid solution and the media were sterilized by passage through 0.45 m Fisherbrand syringe filter. Aliquots (0.3 ml) were dispensed in sterile 2 ml Eppendorf tubes and each was inoculated with 0.05 ml of P. expansum spore suspension. A stock solution of vanillin was prepared in water at 50°C to ensure complete dissolution and several further dilutions were prepared to achieve final test concentrations ranging from 2-20 mM and from 20-200 mM by addition of 0.25 ml of the diluted solutions to the Eppendorf tubes. The contents were mixed immediately after addition of vanillin and the tubes were placed in an incubator. Two separate assays were performed in triplicate at 5°C and 25°C. MIC values were measured at 25°C and 5°C after 5 and 10 days of incubation, respectively. Each tube was examined for the presence of hyphal growth. Where none was evident the medium was examined with a light microscope. The presence of swollen spores (indicative of germination) or mycelial development were considered to indicate lack of inhibition. 53  MICs were also measured with microcapsules from batches 4 and 5 of the 1:15 and 1:20 vanillin-CD-pectin blends to determine whether encapsulation efficiency and drying method influenced inhibition. The pH of the medium was adjusted to pH 4 and 5 (measured pH 3.97 and 4.83 respectively). Concentrations higher than 20 mM could not be tested in the assay system due to the viscosity of the solutions and the difficulty in accurate delivery with conventional dispensing equipment. The nature of the inhibition (fungistatic or fungicidal activity) was established using a recovery procedure. Samples (0.1 ml) of fluids from tubes without evidence of germination or mycelial growth were transferred to tubes containing 1.5 ml of fresh PDB to allow for recovery of spores that were not fully inactivated by the vanillin. Following incubation at 25°C for three days 0.01 ml of the contents were spread onto PDA plates which were incubated at 25°C for five days. Lack of colony development on the plates was considered indicative of fungicidal activity. 3.4.5.2 Minimum inhibitory concentration (MIC) of vanillin in the gas phase on solid media Vanillin has a melting point between 81-82°C. Its relative volatility at 24°C is 0.0004 mg/cm2/hour (Cartwright 1953). This volatility increases with temperature. The limit for detection of vanillin by gas chromatography is 2 mg/liter (Dyer and Martin 1980). The antimicrobial activity of vanillin in the gas phase against P. expansum 1525 spores on the surface of PDA was measured at 5°C and 25°C. Aliquots (0.1 ml) of a standardized P. expansum spore suspension were spread on the surface of PDA plates. Vanillin crystals (50, 114, 200, or 300 mg) were affixed to one side of double-sided tape (Scotch®) and attached to the lid of the petri plates. Vanillin concentrations were estimated assuming a total volume of 70 ml and 46.7 ml of headspace above a plate containing 20 ml of agar. Using these assumptions, it was calculated that the complete vaporization of 113 mg vanillin would yield a headspace concentration of 16 mM, equivalent to the MIC for vanillin in PDA broth. Parafilm tape was wrapped around each plate to prevent evaporation during incubation of the plates in an upright position at 25°C for 5 days, when the plates were examined for visual evidence of growth. Each analysis was performed twice. 54  For measurements at 5°C the amount of vanillin was increased to compensate for the lesser volatility. Because a maximum of 300 mg vanillin crystals could be attached to double sided tape, sterile 15 ml conical centrifuge tube lids were added to the plates to hold larger amounts of vanillin. A sterile lid was used to puncture a hole at the edge of the agar to accommodate the lid. This method was used to measure inhibition with 100, 200, 300, 400, and 500 mg of vanillin in plates for 10 days at 5°C using inoculation procedures described above.  3.4.6 STAGE 6: In vivo studies 3.4.6.1 Preliminary application of pectin to apple slices A preliminary trial was conducted to determine the amount of pectin that could adhere to apple slices, as well as their edibility after application. Apple slices were prepared essentially as described by Rojas Graü et al. (2007), but the three minute sanitation step was carried out with a 100 ppm sodium hypochlorite solution for three minutes rather than the 300 ppm used in the previous study. The apples were then rinsed and dried prior to coring and cutting into 16 slices/apple. The maximum number of apples processed at a time was limited to 6 to minimize time of exposure to the air. All slices were immersed for 1 min in a chilled (5°C) anti-browning solution consisting of 1% (w/v) ascorbic acid (Fisher Scientific). The slices were then drained to remove excess anti-browning solution. Low methoxy pectin (Gum Technology’s Coyote Brand LM 50, Tucson, AZ 85737) was used for the preliminary trial instead of the microencapsulated samples. The highest pectin concentration was associated with the 1:20 vanillin/CD/pectin blend where the microcapsules contained approximately 60% pectin. Assuming a 10% moisture content in the 1:20 microcapsules and approximately 85% recovery, the anticipated vanillin concentration was 2.49% (wb). Given an apple density of 0.576 g/ml (Rahman, 1995), the amount of 1:20 microcapsules that would have to be used to achieve a 6, 12, 16, 20, 24, and 28 mM vanillin concentration (based on an MIC of 14-16 mM, Table 15) were calculated per gram of apple. Since these treatments would consist of 60% pectin, the 55  numbers were adjusted to calculate the mass of 100% pectin required to simulate the specified mM vanillin concentrations. The corresponding amount of pectin was applied to the apple slices using a “sprinkling” or “breading” method. However the appearance of the slices was adversely affected even at the lowest (6 mM) concentration. The apple slices were packaged in sealed plastic packaging material (LDPE) and placed in refrigerated storage (2-4°C). The appearance of the apple slices improved after two days incubation as the pectin absorbed moisture and formed a transparent gel. However, even at the lowest concentration the slices had a distinctly “gummy” texture. It should also be noted here that a 1 min treatment with the 1% ascorbic acid solution did not inhibit browning satisfactorily. In an attempt to improve edibility and appearance, a second preliminary study on apple slices was performed wherein a pectin gel was applied with a brush rather than in powdered form. To improve the anti-browning step two ascorbic acid solutions containing 1% (w/v) and 2% (w/v) were applied to apples for 1 and 5 minutes by immersion. Increasing treatment time with the 1% (w/v) solution and increasing the concentration were both effective for the prevention of browning. Solutions containing 2%, 4%, 6%, 8% and 10% pectin were mixed by hand. The consistency of the solutions increased from that of a weak gel at 2% to a thick, paste-like material at 10%. Each was brushed onto four pre-weighted apple slices. The amount that adhered to the apple slices was calculated by weight difference. Adherence was best with the 8% pectin solution; however, small amounts of the gel were lost by drip loss prior to packaging.  3.4.7 Statistical analysis Data was analyzed using the General Linear Model Analysis of Variance with Minitab ® (Quality Plaza, 1829 Pine Hall Rd, State College PA, Version 16) and significant differences were calculated using Tukey’s honest significant difference (HSD) with a significance level of <0.05. 56  3.5 Results and discussion 3.5.1 STAGE 1: Formulation and processing 3.5.1.1 REV encapsulation of vanillin in CD-pectin matrix Table 6 shows the formulations that were used for the vanillin-CD-pectin matrices. Table 6: Batch formulations for vanillin-CD-pectin blend prior to drying.  Component  1:15  1:17.5  1:20  CD  9.5%  9.5%  9.5%  pectin  14.2%  16.5%  18.9%  vanillin  0.9%  0.9%  0.9%  water  70.9%  68.5%  66.2%  PG  5%  5%  5%  The mean moisture and water activity for the 6 batches of REV and freeze-dried vanillin-CD-pectin blends are given in Table 7. Table 7: Mean moisture and water activity for the vanillin-CD-pectin matrices.  REV  Freeze-dried  Sample Water activity Moisture content (%) Water activity Moisture content (%) 1:15  0.776 ± 0.055  7.15 ± 1.8  0.690 ± 0.049  7.18 ± 1.4  1:17.5  0.771 ± 0.048  7.03 ± 2.3  0.691 ± 0.046  5.86 ± 2.0  1:20  0.756 ± 0.049  6.70 ± 2.1  0.667 ± 0.041  4.95 ± 1.2  NOTE: Values are averages of 6 batches (2 reps each) ± standard deviation  Since the 6 batches had varying moisture percentages and water activities, the quantification calculations were based on each batch’s individual moisture content so the amounts quantified are on the same base (dry basis). 57  3.5.2 STAGE 2: Quantification and yield analysis 3.5.2.1 Encapsulation efficiency and recovery quantification Table 8 shows the mean vanillin concentrations for the FC assay quantification of the vanillin-CD-pectin blends. Table 8: Mean vanillin concentrations in the vanillin-CD-pectin microcapsules (db).  Vanillin concentration (mg/g) (db)  REV  FD  surface  13.8 ± 5.4  24.6 ± 5.4  entrapped  18.1 ± 3.5  9.3 ± 4.1  SUM  31.8 ± 5.2  33.9 ± 4.4  surface  10.2 ± 5.0  20.2 ± 4.1  entrapped  22.3 ± 4.6  11.4 ± 5.5  SUM  32.5 ± 3.3  31.6 ± 3.4  surface  7.9 ± 2.4  17.5 ± 5.4  entrapped  21.5 ± 2.5  10.9 ± 4.4  SUM  29.4 ± 3.3  28.3 ± 5.50  1:15  1:17.5  1:20  NOTE: Values are averages of 6 batches (3 measurements for each) ± standard deviation  Table 9 shows the average encapsulation efficiencies and vanillin recoveries or yield obtained with each formulation. Table 9: Mean encapsulation efficiency and vanillin recovery (yield) of microcapsules.  Encapsulation Efficiency (%)  1:15  Vanillin Recovery (%)  REV  Freeze-dried  REV  Freeze-dried  47 ±9.0 b  24 ± 11 c  83 ± 13.5  89 ± 12  93 ± 9.5  90 ± 9.6  91 ± 10.3  88 ± 17.0  a  1:17.5  63 ± 13.0  1:20  66 ± 7.6a  33 ± 15.7  c  34 ± 13.6 bc  NOTE: Values are averages of 6 batches ± standard deviation  Supercripts show significant differences at 0.05 significance level  A significant correlation (Pearson Correlation Coefficient of 0.616, p-value=0.000) was found between encapsulation efficiency and pectin concentration with REV processed 58  microcapsules. Hence the encapsulation efficiency of the samples appeared to increase with pectin concentration. In contrast, no such correlation was found for the freeze-dried samples. This result suggested that the increased pectin content improved the encapsulation of vanillin only in REV-dried microcapsules. The results of the analyses showed that microencapsulation method (FD vs REV) had a significant effect (p-value=0.000) on encapsulation efficiency, as did pectin concentration (p-value=0.000). Encapsulation efficiency of both the 1:17.5 and 1:20 REV blends were significantly different from all freeze-dried treatments. The REV 1:17.5 and 1:20 REV blends had on average double the amount of encapsulated vanillin as their freeze-dried counterparts. Freeze-dried treatments were not significantly different from each other independent of the pectin concentration. Only the highest pectin concentration (1:20 treatment) in the freeze-dried form was as efficient at encapsulating vanillin as the lowest pectin concentration in the REV form (1:15). The interaction between efficiency of vanillin encapsulation for the freeze-dried and REV microcapsules is shown in Figure 6. Although there is no significant interaction (P-value = 0.289) a clear trend was apparent. Increasing the pectin concentration improved encapsulation efficiency in all treatments, and the effect was more pronounced when REV was used to dry the capsules.  59  Figure 6: Interaction plot for vanillin encapsulation efficiency for varying amounts of pectin in the blend.  ANOVAs conducted on vanillin recovery data suggested that neither the drying method (p-value of 0.955) nor the pectin concentration (p-value of 0.308) had influence on the recovery. Table 10 shows the vanillin concentrations, encapsulation efficiencies, and vanillin recoveries that were measured in microcapsules made with excessive vanillin. The average results of the measurements obtained for 6 repetitions of blend 1:20 were also added to the table for comparative purposes.  60  Table 10: Vanillin concentration, encapsulation efficiency and vanillin recovery (yield) of the “increased vanillin” blends.  REV  Freeze-dried  Treatment  1:20  2:20  3:20  5:20  1:20  2:20  3:20  5:20  Entrapped  21.5 ± 2.5  44.3  45.1  55.9  10.9 ± 4.4  8.3  5.0  6.3  29.4 ± 3.3  61.30  92.8  140.5  28.3 ± 5.5  49.3  62.2  96  66 ± 7.6 a  71 a  50 bc  39 cd  34 ± 13.6 cd  13 de  6e  4e  91 ± 10.3  98  102  98  88 ± 17  79  74  72  vanillin (mg/g db) Total Concentration (mg/g db) Encapsulation Efficiency (%) Vanillin Recovery (%) NOTE: Values for the 1:20 blend are the averages of 6 batches (3 measurements each) ± standard deviation Values for the rest of the blends are the results of the quantification of one batch tested in duplicate (3 measurements per test) Letters show significant difference at <0.05 level in encapsulation efficiencies  As can be seen from Table 10, the total vanillin concentration (second row) was increased as expected, since the initial amount added to the formulation was higher (double  triple  and  five  times  higher).  The  concentration  of  vanillin  in  encapsulated/entrapped form also increased, although not as dramatically, since the proportion of total vanillin in this state (encapsulation efficiency) went down for all increased vanillin treatments with the exception of the 2:20 REV. Since the original blends were formulated so that the vanillin:CD mass ratio was 1:10 (to achieve the stoichiometric 1:1 ratio), this calculation relied exclusively on the CD’s encapsulation potential but did not take into consideration the pectin also present in the mix. As a proven encapsulant, pectin could contribute to a higher encapsulation efficiency acting additively with the CD when more vanillin was present in the mix, as shown by the higher encapsulation efficiency of the “2:20” REV blend. This, however, does not occur for the same formulation in the freeze-dried treatment. And although pectin has proven to be a 61  good encapsulant, it would probably also have a saturation limit in terms of how much vanillin it could encapsulate. This could explain why the encapsulation efficiency did decrease in treatments higher than “2:20”. Both the drying method and the vanillin concentration had a significant effect (pvalue =0.000 for both) in the ANOVA analysis, but there was no significant interaction between the factors (p-value=0.181). The relationship between efficiency of vanillin encapsulation and pectin content for the freeze-dried and REV microcapsules is shown in Figure 7  Figure 7: Mean efficiency of encapsulation (y-axis) obtained with microcapsules containing increasing amounts of vanillin in the CD-pectin blend.  The results of these experiments were somewhat inconclusive due to a lack of repetition (only one batch of the excessive vanillin formulations was created and quantified in duplicate). However they confirmed trends observed with the original blends: that encapsulation efficiency was higher with REV processing than freeze-drying, and that recovery is lower in freeze-dried microcapsules. 62  3.5.2.1.1 Encapsulation efficiency and vanillin recovery for vanillin in CD-calciumcrosslinked pectin matrix Four different formulations were prepared for the calcium crosslinking extrusion process. The first two: “3x PG 3h” and “3x ethanol 3h” differed exclusively in the solvent for the vanillin, the first containing propylene glycol and the second containing ethanol. The second two formulations: “3x PG only pectin” and “3x no PG only pectin” were devoid of CD, the first containing propylene glycol as a solvent and the second not containing any solvent for the vanillin. Table 11 shows the “wet” formulations of all four formulations prior to the drying step. Table 11: Formulation prior to drying for all the vanillin in CD-calcium-crosslinked pectin matrices.  3x PG  3x ethanol  3x PG only  3x no PG only  3h  3h  pectin  pectin  pectin  3.34%  3.33%  3.61%  3.81%  CD  7.52%  7.49%  0.00%  0.00%  vanillin  0.75%  0.75%  0.81%  0.86%  PG/ethanol  5%  5%  5%  0.00%  Note: When no propylene glycol or ethanol was used, the volume was replaced by water.  As was mentioned earlier, vanillin concentrations in the residual CaCl2 solutions were measured to assess losses during processing. Unfortunately most of the vanillin added to the blends (78-93%) was lost in the solution during the 3 hour hardening period. The remaining amount was mainly surface vanillin (un-encapsulated), as evidenced by measurements performed after the extraction/quantification process. Considerable losses during processing (extruding, homogenizing, transferring) where also evident when the total amount measured in both residual solution and samples was compared against the initial input. Table 12 shows amount of vanillin measured in residual solutions and in the samples, as well as the amount lost in processing.  63  Table 12: Vanillin in CD-calcium-crosslinked pectin matrices and in residual CaCl2 solution.  3x PG 3h  3x  3x PG only  3x no PG only  ethanol  pectin  pectin  3h Vanillin concentration in residual  0.5  0.6  0.6  0.5  Vanillin lost in solution (%)  78  90.2  92.6  79.1  Expected vanillin amount (mg/g)  24.5  15.8  4.8  17.6  Total quantified vanillin (mg/g) (db)  2.4  2.7  2.5  2.2  Surface vanillin in sample (mg/g)  2.4  2.2  2.5  2.2  0.02*  0.5  0.02*  0.02*  22.1  13.1  2.3  15.5  0.05  0.02  0.02  0.04  CaCl2 solution (mg/ml)  (db)  (db) Entrapped vanillin in sample (mg/g) (db) Vanillin lost in processing (mg/g) (db) Final vanillin/pectin ratio NOTE: * quantified as less than that amount  The highest loss of vanillin (92.6%) in the CaCl2 residual solution occurred for the treatment with propylene glycol and only pectin (devoid of CD). This suggested that the CD was retaining the vanillin, since the same 3 hours of hardening in the formulation (where vanillin was also dissolved in propylene glycol) that contained CD eliminated 78% of the vanillin. The vehicle of solubilization of the vanillin also seemed to play a role in the loss, since 90.2% was lost when the solvent was ethanol, vs. a 78% loss when the solvent was propylene glycol. It would seem that propylene glycol is aiding in the retention since removing it from the formulation and replacing it with no other solvent showed a loss of 79% of the vanillin. As was mentioned earlier, the dried cross-linked beads became insoluble in water and had to be ruptured using a homogenizer to release their contents for quantification. The homogenization process left debris in the solution. Since the crosslinked matrix 64  becomes insoluble, microcapsules present on the surface of the apple slices were likely to leave insoluble debris on the surface of the slices. The debris on the apples slices would make them unacceptable by consumers, making this process unfavourable for the apple slice surface application. Only one batch of each formulation was made since the considerable loss in the hardening step made this process inefficient. For these reasons, the cross-linking technique was discarded as a processing option. 3.5.2.2 HPLC validation of the FC spectrophotometric quantification method Figure 8 shows the standard curve that was created using vanillin standard solutions. 6000 y = 5102x + 3.647 R² = 1  AUC (mAU*S)  5000 4000 3000 2000 1000 0 0  0.2  0.4  0.6  0.8  1  1.2  Vanillin (ug)  Figure 8: HPLC vanillin standard curve and linear regression equation. AUC stands for area under curves in milli absorbance units*sec.  Vanillin concentrations obtained by chemical analysis using the FC assay and by HPLC are presented in Table 13. The table also shows the expected concentrations calculated from the amount of vanillin added to the stock solution..  65  Table 13: Vanillin concentration of the test samples according to the HPLC and FC assays.  Concentrations (mg/ml) Samples  weight and  FC assay  HPLC  dilution A  0.081  0.079  0.085  B  0.061  0.064  0.063  C  0.041  0.045  0.042  D  0.020  0.022  0.021  E  0.032  0.039  0.031  F  0.019  0.024  0.019  G  0.043  0.052  0.043  H  0.069  0.071  0.069  Vanillin concentrations measured by HPLC differed from calculated values by 1-5%. In contrast, the FC assay yielded values that differed by 2-23%. Hence the HPLC method provided more reliable measurements. The Pearson correlation coefficient for the 8 results obtained for the HPLC vs the FC assay was 0.9855. The critical value (df=6, two tailed) for a 99% confidence interval is 0.834, thus the correlation is established. When the values were graphed as x (FC assay) and y (HPLC) coordinates, the coefficient of determination (R2) was 0.9712, showing an excellent linear regression.  66  3.5.3 STAGE 3: Microcapsule physical characterization 3.5.3.1 Scanning electron microscopy (SEM) imagery Scanning electron micrographs of the microcapsules obtained through REV processing are shown in Figure 9, micrographs of the freeze-dried samples are shown in Figure 10 vanillin crystals are shown in Figure 11. Scanning electron microscopy of the microcapsules is difficult as they are mainly composed of non-conductive natural polymers. The encapsulants are also electron-beam sensitive, which is the reason why radiation damage artifacts may appear in some of the SEM generated micrographs (Rosenberg et al., 1985).  67  Figure 9 : Scanning electron micrographs of REV processed 1:20 blend samples. Images a) and c) were captured at 40x (the scale bar measures 500 m ) while b) and d) at 700x (the scale bar measures 20 m).  68  Figure 10: Scanning electron micrographs of freeze-dried 1:20 blend samples. Images a) and c) were captured at 40x (the scale bar measures 500 m ) while b) and d) at 700x (thescale bar measures 20 m).  69  Figure 11: Scanning electron micrographs of the vanillin crystals in their original state. Image a) was captured at 60x (the scale bar measures 200 m) while image b) was captured at 200x (the scale bar measures 100 m).  The crystalline structure of vanillin (Figure 11) was no longer evident when microencapsulated in the 1:20 blends. Although some differences can be seen in the shape of the particles that have been freeze-dried and REV-dried (Figures 10 and 9), they do not provide evidence to conclude differences in the microencapsulation structure.  70  3.5.3.2 Dielectric property measurement Table 14 shows the dielectric properties of the original vanillin-CD-pectin blends measured at 2.54 GHz. Table 14: Dielectric properties of the original vanillin-CD-pectin blends measured at 2.54 GHz at 23°C (n = 3 reps, 5 measurements/rep).  Control (no vanillin) Vanillin sample Formulation (vanillin: pectin) e' e'' e' e'' 1:15 1.99±0.01c 2.17±0.35 2.32±0.15 b 2.40±0.14 1:17.5 1.62 ±0.10d 2.38±0.15 2.31±0.12 b 2.87±0.66 cd 1:20 1.80 ±0.08 3.02±0.15 2.77±0.12 a 2.54±0.67 Note: Averages ± standard deviation. Letters show significant differences at a level of <0.05. The dielectric constant (e’) is important for predicting microwave reflection and penetration of the material. E’ is the real portion of a complex number while e” is the imaginary component. The dielectric constants were analyzed using a GLM ANOVA; the presence of vanillin and the pectin concentration both had a significant effect (pvalue=0.000). As shown by the letters in Table 14, all dielectric constants (permittivities) were higher for the microcapsules containing vanillin than for the blank blends. The loss factors were not significantly different between vanillin sample and control blends, the pectin concentration (p-value=0.130) and vanillin presence (p-value=0.692) having no significant effect. The difference between the dielectric constant of the vanillin sample and its corresponding blank blend should be the permittivity contribution of the vanillin. These differences were calculated and were found to be strongly correlated (Pearson correlation coefficient of 0.883, P-value =0.000) with the pectin concentration. The higher the pectin concentration used, the higher the difference in dielectric constants or the vanillin permittivity contribution (averaging 0.33, 0.69, and 0.97 for the 1:15, 1:17.5, and the 1:20 blends respectively). Since REV microcapsules containing higher pectin concentration had a significantly higher encapsulation efficiency, then permittivity contribution could be due to the microencapsulated state of vanillin, encapsulated vanillin being able to store energy in an 71  electric field (relative to empty space) better than unbound vanillin. Abassi & Rahimi (2008) concluded that the dielectric properties of materials depend on their chemical composition, physical structure, microwave frequency and temperature. Since the last two were constant for all samples and chemical composition was similar, then the higher permittivity contribution was likely due to differences in physical structure (more encapsulated vanillin).  3.5.4 STAGE 4: Controlled release 3.5.4.1 Controlled release in aqueous media Figure 12 shows the release profile of the 1:20 formulation in a liquid medium at 25°C for both the REV and the freeze-dried microcapsules. 100% 90%  % vanillin release  80% 70% 60% 50%  REV  40% 30%  FD  20% 10% 0%  0  20  40  60  80  100  120  time (hours) Figure 12: Controlled release profile (n=4) of vanillin in an aqueous medium for the 1:20 REV and FD formulation at 25°C (vertical bars show the standard deviation).  72  The release trend was the same for both REV and FD microcapsules. However, the 100% release for the FD capsules was achieved in two hours, while the maximum release for the REV capsules was only 80% and was achieved after 24 hours. Figure 13 compares the controlled release of vanillin from the REV 1:20 blend formulation at both 25°C and 5°C.  100% 90%  % vanillin release  80% 70% 60%  5°C  50%  25°C  40% 30% 20% 10% 0%  0  10  20 30 time (hours)  40  50  Figure 13: Release profile of vanillin in an aqueous medium for the 1:20 REV formulation at 25°C (n=4) and 5°C (n=2).  Since the microcapsules were placed in an aqueous medium at the temperature corresponding to their storage, it can be noted that the initial amount released by the capsules at time zero for the different temperatures was different. The amount released at time zero at 25°C water was higher as is expected due to the higher solubility of vanillin at  73  a higher temperature. Other than the initial concentration at time zero, the release trend for the 1:20 treatment REV capsules is the same at both assay temperatures.  3.5.5 STAGE 5: Microbial inhibition 3.5.5.1 MIC in pH adjusted liquid media Table 15 shows the minimum inhibitory concentration determined for vanillin in PDB at pH 4 and 5 at 25°C and 5°C. These pH conditions were chosen to compare the effect of pH on the vanillin MIC, since apple would have a pH close to 4. The spore suspension contained 6.5 x 105 cfu of P. expansum 1525/ml.  Table 15: MIC for vanillin in a liquid medium at pH 4 and 5 and at a storage temperature of 5°C and 25°C.  Storage Temperature  pH  MIC (mM)  5  4  16  5  5  16  25  4  14  25  5  16  (°C)  NOTE: For storage temperature of 5°C, MIC was determined at day 10 of storage, while at 25°C it was determined at day 5..  Recovery after treatment with vanillin was performed in vanillin-free liquid and solid media to determine the nature of the antimicrobial activity. Both methods revealed that spores germinated and that mycelial growth resumed after exposure to vanillin concentrations between 16-100 mM. Hence the inhibition can be characterized as fungistatic under these conditions. At concentrations from 100 to 150mM there was no evidence of germination or mycelial growth after 5 days at 25°C, indicative of fungicidal activity. It should be noted that concentrations above 150mM are clearly above practical levels for food applications. Consequently, the compound could only be employed as a fungistatic agent.  74  Table 16 summarizes the results of the antimicrobial assays conducted with the encapsulated 1:15 and 1:20 microcapsules. Table 16: Microscopic and visual MIC measurements for the 1:15 and 1:20 microcapsules.  Treatment Encapsulation efficiency 1:15 REV 56% 1:15 FD  34%  1:20 REV  65%  1:15 FD  16%  pH  MIC (mM) visual  4 5 4 5 4 5 4 5  20 15 20* 20* 20* 20 20 20  MIC (mM) microscopy 20* 15 20* 20* 20* 20* 20* 20*  NOTE: * Not even 20 mM was high enough to inhibit growth by day 5.  The highest concentration tested (20 mM) could not completely inhibit P. expansum 1525 germination and growth at 25°C. Higher concentrations were not tested due to the dispensing equipment constraints mentioned earlier in the Materials and Methods section.  3.5.5.2 Volatility MIC in solid media Vanillin was not volatile enough to exert inhibition of the P. expansum 1525 spores at either 5°C nor 25°C. Tables 17 and 18 show the effect of vanillin in the headspace above PDA on the growth of P. expansum at 5°C and 25°C, respectively. All concentrations were tested in duplicate and the presence/absence of growth was assessed visually. The highest concentration tested (300 mg) delayed, but could not inhibit, growth of the fungus at both temperatures. These results suggest that the limited volatility of vanillin at these temperatures would preclude the use of the compound as a gas phase antimicrobial.  75  Table 17: Volatile inhibition of P. expansum 1525 by vanillin in the headspace of PDA plates Incubated at 25°C.  Day 1 2 3 4 5  rep 1 2 1 2 1 2 1 2 1 2  C (+)  C(-)  (-)  (-)  (-)  (+/-)  (-)  (+)  (-)  (+)  (-)  (+)  50 mg (-) (-) (-) (-) (+) (+) (+) (+) (+) (+)  114 mg (-) (-) (-) (-) (+) (+) (+) (+) (+) (+)  200 mg (-) (-) (-) (-) (+) (+) (+) (+) (+) (+)  300 mg (-) (-) (-) (-) (-) (-) (+) (+/-) (+) (+)  NOTE:(-) means no visible germination/growth (+) establishes the positive visual detection of germination/growth 5 The spore suspension used in inoculation was 1.9 x 10 The positive control (C+) was un-inoculated and had no treatment (proof of aseptic technique) The negative control (C-) was inoculated but had no treatment  76  Table 18: Volatile inhibition of P. expansum 1525 by vanillin in the headspace of PDA plates incubated at 5°C.  Day 1 2 3 4 5 6 7 8 9 10  rep 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2  C(+)  C(-)  (-)  (-)  (-)  (-)  (-)  (-)  (-)  (-)  (-)  (+/-)  (-)  (+)  (-)  (+)  (-)  (+)  (-)  (+)  (-)  (+)  114 mg (-) (-) (-) (-) (-) (-) (-) (-) (+/-) (+/-) (+) (+) (+) (+) (+) (+) (+) (+) (+) (+)  200 mg (-) (-) (-) (-) (-) (-) (-) (-) (+/-) (+/-) (+) (+) (+) (+) (+) (+) (+) (+) (+) (+)  300 mg (-) (-) (-) (-) (-) (-) (-) (-) (+/-) (+/-) (+) (+) (+) (+) (+) (+) (+) (+) (+) (+)  400mg (-) (-) (-) (-) (-) (-) (-) (-) (+/-) (-) (+) (+) (+) (+) (+) (+) (+) (+) (+) (+)  500 mg (-) (-) (-) (-) (-) (-) (-) (-) (-) (-) (+) (+) (+) (+) (+) (+) (+) (+) (+) (+)  NOTE: :(-) means no visible germination/growth (+) establishes the positive visual detection of germination/growth 6 The spore suspension used in inoculation was 1.95 x 10 , which is higher than the usual suspensions. The positive control (C+) was un-inoculated and had no treatment (proof of aseptic technique) The negative control (C-) was inoculated but had no treatment  3.5.6 STAGE 6: In vivo studies 3.5.6.1 Preliminary application of pectin to apple slices The mean weight of pectin gel that could be adhered to the apple slices and the theoretical concentration of vanillin that could be delivered to the fruit surface are shown in Table 19. Since only the surface of the slice would actually be in contact with the treatment, the theoretical vanillin concentration could be calculated using only one third of the weight of the slice, assuming this would be the weight of the surface of the apple slice. 77  It should also be noted that the average weight of gel was actually somewhat smaller than the reported amount since a small amount of drip loss was observed after the weighing. Table 19: Results of a preliminary trial to determine the quantity of pectin gel that can be applied to apple slices.  Pectin concentration (w/v)  2%  4%  6%  8%  10%  Average slice weight (g) (4 reps)  18.26  17.81  20.40  17.93  19.79  Theoretical surface weight (1/3 of average weight) (g)  6.09  5.94  6.80  5.98  6.60  Average weight of attached gel (g/g of surface apple)  0.0071 0.0112 0.0126 0.0184 0.0114  Theoretical vanillin concentration (mg/g surface apple)  0.06  0.29  0.61  0.99  0.86  Theoretical vanillin concentration (mM)  0.24  1.09  2.29  3.74  3.26  The results of these analyses showed that the highest rate of adhesion to the slices was obtained with a pectin concentration of 8%, and this corresponds to the highest delivery of vanillin to the apple slices surface. Unfortunately, this concentration is likely too low to achieve inhibition considering that the MIC for vanillin on P. expansum in PDB was 14-16 mM. Furthermore, some of the vanillin delivered to the apple slice surface would remain entrapped in the microcapsules and therefore unavailable for inhibition. The experimental information above could be used to calculate the theoretical vanillin concentration that could be delivered if microcapsules containing higher vanillin concentrations were used (2:20, 3:30 and 5:20 blends) in the 8% pectin gel formulation. Table 20 contains the results of these calculations.  78  Table 20: Theoretical vanillin concentration achieved on the surface of apple slices when administering an 8% pectin gel assuming the higher concentration microcapsules.  Blend used Theoretical Vanillin concentration on apple surface (mM)  1:20 Blend 3.74  2:20 Blend 7.25  3:20 Blend 10.54  5:20 Blend 16.57  According to these results only the 5:20 blend could deliver vanillin concentrations close to the MIC. However Table 10 shows that the encapsulation efficiency for this blend was only 39%.  Hence it is unlikely that microcapsules could be made to deliver the  amount required to inhibit a fungus such as P. expansum.  3.6 Summary of findings for the vanillin project Vanillin was successfully encapsulated in a βCD-pectin matrix by both freeze-drying and REV technologies. It had significantly higher encapsulation efficiency in the REV processed form when compared to the freeze-dried form. The rate of vanillin release in aqueous media was fast, maximum release being obtained in 2 hours (100% release) for the freeze-dried microcapsules, and 24 hours (80%) release for the REV microcapsules. The temperature of the medium had no significant effect on the rate of release. Cross-linking of pectin in the matrix with calcium ions was attempted to decrease the rate of controlled release in aqueous media. The vanillin yield obtained from the crosslinked beads was very low however. The microcapsules became insoluble in water as well, which would mean left over debris on apple slices, making them unacceptable to consumers. A preliminary trial on apple slices was performed to determine the effects of pectin on vanillin MIC. The end result was apple slices with unappealing texture.  79  In order to achieve microcapsules with higher antimicrobial potency, the vanillin:pectin ratio was increased in formulation. Doubling this ratio resulted in increased encapsulation efficiency. However, further increments in vanillin input resulted in lower encapsulation efficiency. Studies performed showed that vanillin was insufficiently potent for an efficient gas phase antimicrobial application. Due to all of these limitations, there was a need to research a natural antimicrobial capable of both: exerting antimicrobial activity through volatile application (to avoid the surface inedibility problems), and achieving target inhibition at lower concentrations. Hexanal was proposed as a volatile alternative.  80  4 Hexanal Project: Volatile Controlled Release into Headspace 4.1 Research objectives The first objective of this project was to successfully encapsulate hexanal in a blend of βCD, and low ester pectin using REV technology. The second objective was to compare and contrast structure, morphology, and yield of REV microcapsules with those obtained by freeze-drying. Thirdly, tests were proposed to examine the volatile controlled release profile of hexanal in a humid headspace from the microcapsules produced by both methods. Another objective was to prove microbial inhibition of the target microorganism P. expansum in vitro, (on PDA), as well as in vivo (inoculated apple slices). The final objective was to conduct shelf-life studies of “Ambrosia” variety apple slices that had been packaged with a microcapsule containing sachet, documenting several quality attributes: visual assessment of the severity of secondary browning (associated with fungal germination), texture differences, and headspace hexanal concentration analysis.  4.2 Hypothesis It was hypothesized that hexanal would be successfully encapsulated in a blend of CD and pectin. A gradual release of hexanal was expected following the exposure of the microcapsules to a moist environment. It was predicted that treatments containing microencapsulated hexanal would prove to be effective in inhibiting and/or delaying germination of P. expansum in vitro, on PDA. Treatments containing microencapsulated hexanal were also expected to be effective in inhibiting and/or delaying germination of P. expansum in vivo, on inoculated apple slices. It was further hypothesized that treatments containing microencapsulated hexanal would yield a significantly longer shelf-life for apple slices than the positive control, thus 81  proving to be more effective in controlling microbial growth, than the direct injection of hexanal. Microencapsulation will allow for a more “stable” headspace hexanal concentration. It was assumed that the microencapsulated form of hexanal delivery would allow for a controlled release of the volatile into the headspace throughout the storage time as relative humidity in the jar triggers the release. A consistent hexanal amount in the headspace would hypothetically correlate with less microbial growth. Finally, a significant difference in texture between the treated slices vs. the control slices was expected. The hexanal-exposed slices were expected to show a smaller loss in firmness during storage than would the control slices.  4.3 Methodology outline A strategic plan was developed to order to achieve the research objectives, specific stages are shown in Figure 14. A more detailed explanation of specific experiments is contained in the Materials and Methods section of the chapter.  82  Figure 14: Strategic plan for studies on the controlled release of hexanal.  83  4.4 Materials and methods 4.4.1 STAGE 1: Formulation and processing 4.4.1.1 HexanalCD-pectin blend batch formulation Five different formulations of the encapsulants: low methoxy pectin (Gum Technology’s Coyote Brand LM 50, Tucson, AZ 85737) and CD hydrate (Sigma) and the active agent: hexanal (natural source, ≥95% purity, Sigma) were prepared. The first (A) was prepared with a 1:12 hexanal to β-CD ratio by weight, which is equivalent to a 1:1 molecular ratio necessary for the formation of a molecular inclusion complex (Almenar et al., 2007). The second formulation (B) was prepared with double the mass of βCD and the same proportion of the other components. In further formulations the amount of βCD was halved (Z), and eliminated (O). In one additional formulation, the mass of βCD remained constant but the mass of pectin was increased 1.6 x (P). βCD was dissolved in water at 100°C with stirring, hexanal and pectin were added and the mixtures were homogenized at 11,000 rpm for 2 min (Ultra Turrax T25 homogenizer). The ratio of each component (in % by weight) in each formulation prior to processing are contained in Table 21. Table 21: Component percentages for each formulation prior to processing.  Treament A  Treament B Treatment Z  Treatment P Treatment O  CD  20.0%  33.2%  11.1%  18.3%  0.0%  water  65.5%  54.6%  72.7%  60.1%  81.8%  hexanal  1.7%  1.5%  2.0%  1.6%  2.2%  pectin  12.8%  10.7%  14.2%  20.0%  16.0%  NOTE: Although the proportion (percentage) of hexanal seems to vary along the treatments, the absolute amount (volume) was kept consistent. In the same manner, the mass of CD in B is double the mass in A, halved in treatment Z and fully eliminated in treatment O. Treatment P has 1.6 times the mass of pectin contained in A.  84  4.4.1.2 Processing of the hexanalCD-pectin microencapsulated complexes Since the mixtures were of high viscosity and almost immediately gelled to a stiff semi-solid, they were cut into cubes of 3.5 to 5 g. One half of each mixture was frozen at -60°C and processed in a freeze drier (Labconco Freeze Dryer 18) for 48 h at condenser temperature of -50°C and 150mtorr absolute pressure. Cubes that were not frozen were placed into custom-made open quartz cylindrical containers (3.1 cm diameter and 1.9 cm height) to be processed in the REV machine, (model VMD900W, EnWave corp., Richmond, BC). Cubes were dried for 12 min at 20torr and 300 W, which equaled a total power of approximately 30.86 kJoules per gram of wet sample. The dried matrices, enclosed in foil envelopes were ruptured using a hammer. The pieces were then mechanically ground using a mortar and pestle  4.4.1.3 Moisture analysis Moisture analysis was performed in duplicate by drying under vacuum (30 in Hg) for 24 hours at 70°C using a Shell Lab Vacuum oven (model No. 1430).  4.4.1.4 Water activity determination Water activity readings were obtained in duplicate for each sample using a water activity reader Aqualab ® Model Series 3 (Decagon Devices Inc. Washington USA)  4.4.1.5 Scale-up microencapsulation trials To create larger samples, a 3.6 kW REV dehydrator, (model ENWL291111-E020, EnWave Corp 106-1668 Derwent Way Delta, BC) was used. This dehydrator has four 900 Watt magnetrons which activate randomly while the sample rotates inside the chamber making the drying homogeneous.  85  Two batches of approximately 228 g (wet sample) of each of the formulations “A” and “B” were dried activating 3 out of 4 magnetrons providing a total power of 1200 Watts for 45 min (54,000 W*min) under a vacuum of 20 torr absolute pressure. The maximum temperature during drying was 40°C. Figure 15 shows a picture of a 3.6 kW REV model dehydrator similar to the one used for this study.  Figure 15: 3.6 kW REV model dehydrator similar to the one used for this project.  86  4.4.2 STAGE 2: Quantification and yield analysis 4.4.2.1 Hexanal extraction and SPME quantification protocol One hundred mg of dried microcapsules were dissolved in 4 ml water in 10 ml vials (CL GLS Serum, Fisher Scientific) at 25°C (Thelco Laboratory incubator) for 15 min. Hexanal  released  into  the  headspace  was  measured  using  a  50/30  μm  PDMS/DVB/Carboxen SPME (solid phase microextraction) fiber (Supelco, 595 North Harrison Road, Bellefonte, PA) and gas chromatography (GC). The fiber was exposed to the headspace for 10 minutes and the trapped volatiles were desorbed for 5 minutes in the splitless injection port of a Shimadzu GC-14A gas chromatograph (Shimadzu Scientific Instruments, 7102 Riverwood Drive, Columbia, MD 21046, U.S.A.) equipped with an Agilent J & W DB-5 (30 m x 0.25 mm x 0.25 μm) column, using helium as the carrier gas. The gas chromatograph used for this study was equipped with a flame ionization detector (FID). SPME fibers were used for a maximum of 100 extractions in this work and replaced with new ones. All measurements done in this work were headspace measurements. Oven temperature was held at 40°C for 5 minutes (hexanal retention time at approx. 3.5 min), and increased by 40 celsius degrees per minute until a final temperature of 220°C was achieved. The injector and detector temperatures were set at 230°C. Concentrations were determined by comparison of measurements against a standard curve (R2=0.9805). Yield was defined as the fraction of added hexanal that was successfully encapsulated and recovered. 4.4.2.2 Validation of gaseous and liquid equilibrium for SPME quantification An experiment was carried out to verify and validate the SPME quantification procedure was conducted. The purpose was to show that equilibrium was achieved during 15 minutes of microcapsule extraction time (stirring in vial) prior to SPME adsorption. This was deemed necessary because equilibrium must be achieved to ensure the accuracy of quantification by SPME Treatment “P” was used because this formulation has the largest pectin content and the highest viscosity at time of extraction, which could delay release into headspace and a 87  represent a “worst case scenario”. One hundred mg of microcapsules (both REV and freeze-dried) were extracted in duplicate as described above for 15, 30, 60 and 120 minutes. Hexanal was quantified using SPME protocols once the extraction was complete. This experiment was performed twice on separate days. 4.4.2.3 Processing parameter optimization Processing parameters for both the REV and freeze-drying methods were optimized in two separate experiments. Formulations A and B were prepared in triplicate and randomly processed under the following REV conditions: 200 W for 12 min, 200 W for 18 min, 300 W for 12 min, and 300 W for 18 min. All REV processing was performed at 20 torr. The objective of this experiment was to determine the effect of time and power on the final hexanal concentration (db) and yield of the dried microcapsules. For the second experiment three batches of formulations P, O, and Z were freezedried at room temperature for 24, 48 and 72 hours. This experiment was meant to test the effect of freeze-drying process time on final concentration (db), yield, and moisture content.  4.4.3 STAGE 3: Microcapsule physical characterization 4.4.3.1 Scanning electron microscope (SEM) images Scanning electron microscope (SEM) images were generated using a Hitachi S4700 microscope (89 Galaxy Blvd. Suite14 Rexdale, Ontario, Canada, M9W 6A4), by the UBC Bio Imaging Facility Department of Botany (6270 University Blvd., Vancouver, B.C., V6T1Z4, Canada). A small amount of the dried microcapsules was fixed onto an iron stub and then made electrically conductive by sputter coating with a thin layer of gold-palladium. The images obtained ranged in magnification from 20 to 700x. Water absorption can significantly alter the physicochemical properties of powders (Rosenberg et al., 1985). Moisture acts a release trigger for hexanal from the microcapsules prepared for this work. Consequently both newly processed and hydrated 88  (12 days at 12°C in the presence of apple slices) microcapsules were examined by SEM. Microcapsules were observed under the microscope to try to obtain a “comprehensive understanding of water influence on product quality”, as was performed by Rosenberg et al. (1985). It is however likely that the “exposed” microcapsules could have released some of their moisture during gold coating due to vacuum exposure. 4.4.3.2 Dielectric property measurement Dielectric properties are highly affected by moisture content. For this reason, sample formulations were created and: left un-processed (moist form), fully dried (12 min at 300 W and 20 torr absolute pressure) or only dried half way (6 min at 300W and 20 torr absolute pressure) before measuring their dielectric properties. The dielectric properties (at 2.45 Ghz) of formulations “A” and “B” in their dry half processed, and moist form were measured in triplicate (5 measurements per rep) using an Agilent Technologies E5071C ENA Series Network Analyzer (9 kHz to 4.5 GHz) coupled with Agilent 85070 E Dielectric Probe Kit (200 MHz to 50 GHz).  4.4.4 STAGE 4: Microbial inhibition 4.4.4.1 Measurement of minimum inhibitory concentration (MIC) of hexanal in the gas phase on solid media  P. expansum conidia were harvested by swirling 5 ml of 0.1% sterile polysorbate 80 (Tween) on a sporulated potato dextrose agar (PDA) plate that had been previously incubated at 25°C for 7 days. The suspensions were passed through 8 layers of sterile cheese cloth and diluted by adding 15 ml of 0.1% sterile polysorbate 80 (Tween). The suspensions were passed through the layers of sterile cheese cloth to remove mycelia as done by Okull and LaBorde (2004). Spore concentrations were determined by serial dilution and plating on PDA. The minimum inhibitory concentration (MIC) of hexanal against P. expansum strain 1525 was examined at 12°C and 5°C. Petri plates (50 mm diameter) containing PDA were 89  inoculated in the geometric centre with 3.5 l of the spore suspension. The plates were aseptically placed inside sterile 1 liter mason jars. Sufficient hexanal to achieve concentrations ranging from 5.03-7.73 g hexanal/ml (assuming complete vaporization) was introduced into the jar bottoms, beside the inoculated petri dish (in duplicate). The jars were immediately closed with new canning lids. One half of the jars was stored at 12°C and the remainder at 5°C. Each plate was examined for evidence of growth after 10 days. Plates without evidence of growth were removed from the jars and  returned to an  incubator at 25°C. They were incubated for two weeks to determine if the spores survived exposure to hexanal and were able to recover.  4.4.5 STAGE 5: Controlled release studies and target microbial inhibition on laboratory media (potato dextrose agar) 4.4.5.1 GC/Antimicrobial in vitro assay for P. expansum at 25°C An enclosed model system was constructed from Kerr brand “self sealing” 950 ml wide mouth mason jars to permit examination of hexanal in the gas phase. The jar lids were adapted to accommodate small GC septa (Mandel Scientific, SG 041890), which were inserted in small holes (0.52 cm in diameter) drilled through the lid. General Electric “Premium Waterproof Silicone” was applied to the exterior and interior edges where the septum met the drilled lid to prevent leakage. The jars were sterilized in an oven at 300°F for 2.5 hours (CDC, 2008). A diagram of the model system is provided in Figure 16. P. expansum was grown for 7 days in 100 mm diameter PDA plates at 25°C to allow sporulation. Spores were harvested by swirling 5 ml of sterile water and further dilution in 15 ml of sterile water. Petri plates (50 mm diameter) containing PDA were inoculated in the geometric centre with 3.5 l of the spore suspension and were placed inside the mason jars as shown in Figure 16. Appropriate amounts of microcapsules were placed next to the petri dish, the jars were closed and placed in an incubator set at 25°C. Radial growth of the colony was measured using a caliper every 24 hours by visual inspection through the  90  bottom of the jar (both the jar and the agar dish were see through and the colony appeared opaque). Hexanal measurements were also performed daily. Pre-conditioned 50/30 m PDMS/DVB/Carboxen SPME needles were inserted through the septa in the lid and exposed to the headspace for 10 min inside the incubator. The needles were immediately desorbed for 5 minutes in the GC to measure hexanal concentrations in the headspace. The procedure for the preparation of calibration curves are contained in the Section 4.4.5.2 “Creating a headspace hexanal concentration calibration curve” of this manuscript. A randomized block design was used to test all treatments at two different dosages: low (4 mg hexanal/jar) and high (14 mg hexanal/jar) in triplicate. A negative control (containing no hexanal) was tested for comparison, as well as a positive control (the direct addition of un-microencapsulated hexanal in the corresponding dosage amount). A fresh spore suspension was made for each block and contained an average of 2.73 ± 2.27 x 105 cfu /ml, which was quantified by serial dilution and plating. Figure 16 shows a diagram of the model system.  91  Tyvek ® Sachet with microcapsules for apple slices  Microcapsules for PDA assay  Figure 16: Septum modified jar system with inoculated PDA petri dish / apple slices.  4.4.5.2 Creating a headspace hexanal concentration calibration curve SPME Quantification methods rely on the equilibrium reached between the liquid and gaseous phases of the sample. The detector creates a signal proportional to the amount of sample in the container, which can be determined from a calibration curve. Subsection 4.5.2.2 of this manuscript shows how the 15 minutes of stirring at room temperature were sufficient to extract hexanal from the microcapsules and reach equilibrium between the liquid and gaseous phases produced. However, if full evaporation of the sample is achieved, the signal being measured by the SPME fiber corresponds to the concentration in the headspace. In this way, SPME fibers can be used to quantify headspace concentrations. Almenar et al. (2007) used this technique in their work with hexanal. To create a headspace calibration curve, varying amounts of hexanal were volatilized within containers of the same dimensions and under the same conditions.  92  This was achieved by the addition of increasing amounts of hexanal to the model system containing PDA agar plates. All materials were prepared in a cold room to minimize losses of hexanal due to volatility during pipetting. PDA plates were included in the jars to ensure constancy in relative humidity and potential losses due to absorption of hexanal by the moisture in the agar. Once closed, the jars were warmed to 25°C in an incubator for 30 minutes prior to doing the SPME 10 min adsorption, 5 min desorption measurement. While higher volumes of hexanal could be dispensed with the needed precision, volumes less than 0.5 l (equivalent to 0.41 g hexanal/ml in the 950 ml jar) could not be delivered accurately by this method. Accordingly, a second set of calibration curves was prepared from hexanal diluted in ethanol to permit accurate measurement of low concentrations in the headspace. This was done ensuring that the amount of ethanol present in the solution was capable of full volatilization as well. Two standard curves (with ethanol dilution and without) were prepared at 25°C. For 12°C and 5°C calibration curves for the lower range alone were required since the lower vapour pressure of hexanal at lower temperature was not conducive to complete vaporization. Table 24 shows the standard curve coefficients and the coefficient of determinations obtained. Vaporization of hexanal and ethanol in an enclosed space can be described by Antoine’s equation which describes the relationship between vapour pressure and temperature for pure components. Equation 4 shows the two forms of Antoine’s equation using the logarithm based 10 and the natural logarithm.  93  Equation 4: Antoine's equation (both forms)  The equation was used to calculate the vapor pressure of both hexanal and ethanol at 25°C. The coefficients for hexanal at the temperature range between -56°C to 305.85°C were reported as A=7.34663, B=1588.31 and C=227.359 when using Antoine’s log base 10 equation and the pressure (P) is in mmHg and the temperature (T) in Celsius (IRCHE, 2011). Using this data the vapour pressure of hexanal at 25°C was calculated as: 11.29 mmHg. Substituting this vapour pressure into the ideal gas law, the number of moles(n) can be obtained. Using hexanal’s molecular weight (MW= 100.15 g/gmol), the maximum mass amount (m) of hexanal that can fully vaporize in a 0.95 liter container can be calculated. The calculations are shown below Equation 5. Equation 5: Ideal gas law  n= PV/RT n = (11.29221 mmHg) (0.95 liters) / (62.3637 L*mmHg*K-1*mol-1) (298.15 K) n = 0.000577 gmoles m= n*MW m = 0.057 grams = 57.79 mg The maximum amount of hexanal that can vaporize in the 950 ml mason jars at 25°C is 57.79 mg. The maximum amount used in the calibration curve measurements was 94  14.65 mg, which is well below the limit. This ensures full volatilization and confirms the reliability of the measurements. Table 22 shows the coefficients/conditions for Antoine’s equations. Table 23 shows the calculations for ethanol and hexanal at 12°C and 5°C that prove that all the maximum pipetted amounts are below the maximum limit that can vaporize at the given temperature. Table 22:Coefficients and conditions for Antoine’s equation calculations. Compound  A  B  C  Equation form  Reference  Temperature range for equation (°C)  Hexanal  7.35 1588.31 227.36  log base 10  (IRCHE, 2011)  (-)56-305.9  Ethanol  16.9 3795.17 230.92  natural log  (Poling, Prausnitz and O´Connell, 2001)  3 - 96  Table 23: Maximum amounts of ethanol and hexanal that can volatilize in a 0.95 liter jar at a given temperature. Compound  Units of Pressure and Temp  Hexanal  Ethanol  mmHg and °C mmHg and °C mmHg and °C kPa and °C  Ethanol Ethanol  Hexanal Hexanal  Temperature Vapour pressure  5  3.24  Max amount that can vaporize (mg) 17.8  Max amount pipette d (mg)  12  5.14  27.5  14.7  25  11.29  57.8  14.7  5  2.25  42.5  7.9  kPa and °C  12  3.57  65.9  7.9  kPa and °C  25  7.89  139.4  7.9  14.7  Verdic t  Full Vaporization Full Vaporization Full Vaporization Full Vaporization Full Vaporization Full Vaporization  The range of concentrations created by direct addition of hexanal into the containers was 0.4-14.7 g hexanal / ml of air. Seven points were tested in duplicate (total of 14 points in the curve). The range of concentrations tested by dilution in ethanol prior to 95  addition in the containers was 0.00004-0.407 g hexanal / ml of air. Five concentrations were measured in duplicate (total of 10 points in the curve). Table 24 contains the standard curve coefficients for all the headspace measuring standard curves produced. Table 24: Headspace hexanal concentration standard curve coefficients.  Equation coefficients  Coefficient of Determination 2  CURVE  m  b  Ethanolic dilution (lower range) at 25°C No ethanol (higher range) at 25°C Ethanolic dilution (lower range) at 12°C Ethanolic dilution (lower range) at 5°C  219265  7749.1  0.9158  15602  118357  0.9318  310195  6451.3  0.8677  306717  9187.8  0.9065  R  There was a noticeable change in slopes between the curves prepared at 25°C, by the different methodologies. This could be due to the fact that SPME fibers are known to “level off” over a certain concentration since the sites for bonding in the fiber become occupied and the capacity for adsorption is decreased (Shirey and Mindrup, 1999). Carboxen-PDMS fibers are known to have good linearity but saturate at high levels with little displacement (Shirey and Mindrup, 1999). Unfortunately, due to the large range of concentrations observed in the 25°C assay, both curves were required. 4.4.5.3 GC/Antimicrobial in vitro assay for P. expansum at 12°C The same assay described in section 4.4.5.1 was repeated at a storage temperature of 12°C. It was run for P. expansum 1525 using only treatment B (since the above assay showed it was the most effective treatment) and treatment A (to keep as a standard).  96  Protocols used to prepare a suspension of P. expansum spores were as described in Section 4.4.4.1. The testing time was increased from 5 to 10 days at 12°C.  4.4.6 STAGE 6: Apple shelf-life studies A 23 full factorial randomized block design (with a total of 4 blocks and 3 reps per treatment) was conducted in glass mason jars fitted with lids equipped with septa to permit sampling of the headspace. Slices prepared from Ambrosia cv. apples were used for these experiments. The factors analyzed were: 1) microencapsulation method, with levels corresponding to REV drying and freeze-drying (both using formulation “B”), 2) storage temperature, with levels corresponding to storage at 5°C (simulating refrigeration temperatures in retail) and 12°C (temperature abuse), and 3) the presence/absence of P. expansum spore inoculum. The response variables were: secondary browning using a visual assessment rating scale, texture measurements (comparing values at the beginning and end of the experiment), and the headspace hexanal concentration profile. The experiment contained two levels of controls. The first was a negative control, containing the inoculated/un-inoculated slices that received no treatment and thus allowed for normal growth of the inoculum or the natural microflora. The second was a positive control, with direct addition of hexanal to the jars in a concentration equivalent to that contributed by the microencapsules. Ambrosia apples were obtained from “Top Ten Produce” (4536 10th Ave W, Vancouver B.C.). The fruit was harvested in September 2011 in the Okanagan Valley of British Columbia and was stored under CO2 at 0-2°C in a controlled atmosphere (CA) storage facility for 90 days at the packing house before market distribution. The lot used in this work consisted of fruit between 170-200g. Apples free of visible defects weighing between 170-175g were selected for the experiments. 4.4.6.1 Preparation of apple slices and inoculation Apple slices were prepared as described by Rojas Graü et al. (2007) with a few modifications. The whole apples where immersed in a 100 ppm sodium hypochlorite 97  solution for three minutes, dried, and cut/cored using a very sharp 8 piece slicer (IKEA brand). All the slices were “wounded” using a sterile pinhead fixed to a cork (depicted in Figure 17) to make two 1.5 mm diameter holes in the flesh. The slices were then immersed for 5 min in a chilled (2-5°C) 2% (w/v) ascorbic acid solution (Sigma) and drained before they were placed in the jars. Inoculation was achieved by addition of 6 l of a 7.1 x 104 -7.0 x 105 spore/ml P. expansum spore suspension to each wound, thereby achieving a range of spore concentrations between 400-4,200 spores/site. The wounds easily accommodated the volume of inoculum and growth was clearly visible in control samples.  Figure 17: 1) Cork and pin for inflicting the wounds. 2) slice inoculation sites.  dimensions and  98  4.4.6.2 Delivery of the microcapsules The contents (bentonite) of Tyvek ® Bag Desiccants (Uline model No. S5167, Size 2, Uline Canada60 Hereford St., Brampton, ON L6Y-0N3) were removed under vacuum through a small opening made with scissors. Although sachets were not sterilized, even the negative control jars possessed an empty sachet, which made the possibility for contamination introduced by the sachet, consistent in all jars. Appropriate weights of microcapsules (determined by hexanal content) were placed inside the sachets and the cuts were heat-sealed. Where needed, the sachets were fastened to the inside walls of the glass jars using double sided tape (Scotch ®). For positive controls, open sachets were pasted onto the inside walls of the container and the required amount of hexanal was dispensed quickly into the sachet with a pipette, immediately before the jars were sealed. All sachets were inserted in the jars immediately after the slices were processed, inoculated, and placed in an upright position on the floor of the corresponding container. Experiments were carried out with three inoculated/un-inoculated slices per sterile 1 liter mason jar. The treatment formulation and dosage were fixed variables set at 112 mg hexanal / jar in the form of treatment “B”. In the PDA experiments, a dosage of 14 mg of hexanal / jar (1.5 mg of hexanal / g of water in agar) had not been enough to inhibit germination at 12°C for 10 days. Since the jars now contained three slices, of an average weight of 18.9 g each, and considering apples having 86% moisture content, the dose was increased to 2.3 mg of hexanal / g of water in apple slices (56% more than the highest concentration in the PDA experiments). In other words, three apple slices would weigh approximately 56.7 g out of which 48.7 g would be water. Since the dose was increased to 2.3 mg of hexanal / g of water in apple slices, the dose added was 112 mg hexanal per jar. 4.4.6.3 Visual assessment Secondary browning assessment using a rating scale is a useful tool for determining the visual quality of apple slices with more objectivity. For this experiment, the scale used was that of Kader and Cantwell (2005), which was originally used by Gil, Gorny, and Kader (1998) to evaluate apples exposed to different ascorbic acid and low oxygen atmospheres. The scale ranges from 9, “Excellent”, to 5 which marks the “limit of saleability”, 3 the “limit 99  of usability” and 1 which is “extremely poor”. For this experiment, more detail was used to describe each of the levels, and two levels were added to the end of the scale to achieve a better rating of the slices. It should be noted that although the levels in the scale were clearly defined, the natural decay and magnitude of each level and between each level was not necessarily equal or symmetrical. Thus it would not be strictly correct on a theoretical basis to consider scale levels as consecutive integers. Nevertheless, , assigning numbers to represent levels of decay allowed for simple statistical analysis of visual observations and so this practice was followed in spite of its theoretical shortcomings. The scale used is shown in Table 25. Figure 18 contains pictures of representative apple slices that would be scored according to the proposed rating categories. The slicer used for these experiments did not remove carpal tissue. Since the core and adjacent carpal tissues are more prone to browning than other parts of the fruit (Soliva-Fortuny & Martín-Belloso 2003a), slices showing only this type of browning were assigned a rating of “8”.  100  Table 25: Modified “Rating scale for overall visual quality of fresh-cut produce” (Kader and Cantwell, 2005). Rating  Description  9  Excellent, essentially no symptoms of deterioration*  8  Minimal carpal tissue browning  7  Good, minor symptoms of deterioration*, not objectionable  6  Symptoms of deterioration* evident, but still not objectionable  5  Fair, deterioration* evident, but not serious, limit of saleablility  4  Deterioration* more than evident, but still not serious  3  Poor, serious deterioration*, limit of usability  2  Browning spots at inoculation sites but NO detectable fungal growth yet Extremely poor, not usable, off odors, fungal decay  1 0  Full deterioration of slices (almost no healthy tissue left)  -1  Green discoloration caused by Penicillium expansum = sporulation Note: “deterioration* = increased browning discoloration, softening, water-soaked appearance, tissue  disintegration”  101  Figure 18: Examples apple slice appearance for categories in visual evaluation rating. Note: The positive control (C+) showed browning from the first 24 hours, therefore it was labeled a 1, and it did not deteriorate further since no fungal growth was detected by day 15.  102  All visual rating evaluations were performed by the same individual under the even lighting of a laminar flow hood. A high quality image (3-3.5 MB, 300 dpi per image) was recorded of each sample on every measurement. Visual evaluation was performed on days 1, 2, 3, 6, 9, and 12 for the samples stored at 12°C, and also on day 15 for the samples stored at 5°C. 4.4.6.4 Headspace hexanal concentration measurement To measure the hexanal heaspace concentration, a pre-conditioned 50/30 m PDMS/DVB/Carboxen SPME needle was inserted through the septum in the lid and exposed to the headspace for 10 min, inside the incubator, using a universal stand and tongs. It was immediately desorbed for 5 minutes in the GC, under the conditions detailed on Section 4.4.2.1. Headspace hexanal concentration was measured on days 1, 2, 3, 6, 9, 12, and 15 (for samples stored at 5°C) as well. Calibration curves were prepared at each temperature. 4.4.6.5 Texture analysis Texture analysis was performed to assess the changes in texture of apple slices at the different storage conditions and treatments. Texture analysis was performed using a Stable Microsystems TA XT2 Texture Analyzer (Stable Micro Systems Ltd., Vienna Court, Lammas Road, Godalming, Surrey GU7 1YL, UK), coupled with XTRAD software. Two different methods were employed to quantify texture parameters: a puncture method and a Texture Profile Analysis (TPA). 4.4.6.5.1 Puncture test The puncture test was based on the methods described by Mehinagic et al. (2004). A 4 mm cylindrical probe was pushed at a speed of 50 mm/min (0.83 mm/sec) to a depth of 10 mm through the peel of the apple slices immobilized in a wooden holder. A Force Deformation curve (force vs. deformation, see Fig. 20) was obtained from each analysis to enable the determination of several associated parameters. These included total puncture force (Fs), or the highest point in the curve and represents the force needed 103  to rupture the skin. The force at exactly 7 mm of deformation (Ff) is assumed to be the force needed to rupture the flesh after the skin has been broken. Following this pattern, the area under the curve from the origin to the highest peak is equal to the work required to rupture the apple skin (Ws), and the area under the curve from the origin to the 7 mm point is the work required to rupture the skin and flesh (Wf). Figure 19 depicts an apple slice immobilized in the wooden holder for the puncture test.  Figure 19: Puncture test setup with probe and wooden holder.  104  Figure 20: Puncture test force vs deformation curve. Reprinted from (Mehinagic et al., 2004) with permission from Elsevier ©.  4.4.6.5.2 Texture profile analysis (TPA) test The TPA test is based on the methods performed by Barreiro et al. (1998) and Soliva-Fortuny et al. (2003b) with modifications. A cylindrical flesh specimen of 1 cm height and 2 cm diameter (changed from the original 1.7 cm) was removed from the slices with a cork borer, penetrating on the side of the flesh, as shown in Figure 21. The cylinders obtained were trimmed to the 1 cm height and then compressed/decompressed at a speed of 20 mm/min obtaining a deformation of 2.5 mm (25% of the original height of the cylinder).  105  Figure 21: TPA specimen sampling and setup.  106  Figure 22: Texture profile analysis (TPA) Model.  A force vs. time plot with two distinct peaks was obtained for each sample cylinder (slices were tested in duplicate). The parameters measured were: Fracturability (the first peak), Hardness 1 (the highest peak), Hardness 2 (the peak attributed to the second compression), length 1 (the length from the origin to the highest peak), length 2 (the length from the beginning of the second compression to the second compression peak), and area 1 and 2 which are the areas under the respective compression/decompression peaks. Figure 22 provides a visual representation of these parameters. Using this information, “Cohesiveness” (how well a product withstands a second deformation relative to how it behaved under the first), “Springiness” (how well a product springs back after it has been compressed a first time), and “Chewiness” were calculated. Equations 6, 7, and 8, were used to perform the calculations. It should be noted that these calculated parameters were not compared against their actual sensory analysis score, but were meant to be used as indicators of changes in texture due to deterioration in both treatments and controls.  107  Equation 6: TPA "Cohesiveness" (dimensionless)  Equation 7: TPA "Springiness" (mm)  Equation 8: TPA:"Chewiness" (N*mm)  4.4.7 Statistical analysis Plots and regression analysis were performed using Microsoft Excel ®. Analysis of variance (ANOVAS) following the “General Linear Model”, general effect plots, and interaction plots were prepared using Minitab ® (Quality Plaza, 1829 Pine Hall Rd,State College PA, Version 16).This software was also used for paired comparisons: using Tukey’s honest significant test (HSD) for significant differences between treatments and Dunnett’s test for when comparing against the original conditions at time zero as the control.  4.5 Results and discussion 4.5.1 STAGE 1: Formulation and processing 4.5.1.1 REV encapsulation of hexanal in βCD-pectin matrix This microencapsulation procedure differs from that performed by Almenar et al. (2007) in: the drying method, REV and freeze-drying vs. conventional air drying at 60°C for 24h, the elimination of a centrifuging stage (step which decreases the yield since some of the hexanal is removed in the filtrate), and of course the addition of pectin as a viscosity agent and coating material. 108  The maximum temperature achieved during REV processing of the microcapsules was 40°C and room temperature during the freeze-drying. This was considerably lower than the air-drying at 60°C performed by Almenar et al. (2007) in their work on hexanal microencapsulation. Since hexanal is so volatile, these lower temperatures may contribute to a better yield. Almenar et al. (2007) did not report their actual hexanal microcapsule concentrations or yields, so this hypothesis can only be inferred. Tables 26 and 27 contain the average moisture content and water activity for the microcapsules prepared by REV and freeze-drying. The set processing parameters used (determined at Section 4.5.2.3 “Processing Parameter Optimization”) were 20 torr, 300 W for 12 minutes for REV samples and 48 hours of freeze-drying at room temperature.  Table 26: Average moisture content of the microcapsules for all 5 treatments (tested in duplicate).  A  B  Z  P  O  REV  6.0%  7.0%  19.8%  7.8%  19.5%  FD  0.1%  0.3%  0.6%  0.1%  1.6%  Table 27: Average water activity of microcapsules for all 5 treatments (tested in duplicate).  A  B  Z  P  O  REV  0.55  0.20  0.72  0.57  0.74  FD  0.06  0.06  0.49  0.49  0.21  Microcapsules that were freeze-dried had lower moisture contents and water activities than their REV counterparts. This does not affect the quantification calculations, 109  since they are all reported in dry weight basis. To avoid shelf-life issues, all treatments were stored in a -60°C freezer as soon as processing and grinding were accomplished. Hexanal losses during processing may also occur during the heating/dissolving of CD (although these losses would affect microcapsules made both by REV and FD). Cyclodextrin inclusion complexes can occur up to 200°C (Gibbs et al., 1999). If CD is in solution, more cyclodextrin molecules become available for complexation. Heat allows for better solubility of both the CD and the guest molecule (Del Valle, 2004). However heating can also destabilize the inclusion complex (Del Valle, 2004) and in the case of hexanal, could lead to losses during processing though vaporization  4.5.2 STAGE 2: Quantification and yield analysis 4.5.2.1 Encapsulation quantification and yield Figures 23 and 24 show the hexanal concentrations in the microcapsules (db) and the yields obtained from processing for all five treatments. The yield is a better measurement for comparison since it takes into account the hexanal dilution by the solids present due to the particular formulation.  110  Figure 23: Mean hexanal concentration (n=3) in microcapsules (mg/g) dry basis . NOTE: Vertical bars represent the standard deviation, letters show significant differences at a level of <0.05.  111  Figure 24: Mean hexanal yield (n=3) in microcapsules (%) dry basis. NOTE: Vertical bars represent the standard deviation, letters show significant differences at a level of <0.05.  With the exception of the standard treatment (A), the yield of freeze-dried microcapsules was consistently and significantly higher (p-value=0.000). This could possibly be due to the maximum temperature achieved during processing. Although REV processing never exceeded 40°C, freeze-drying was done at a maximum of room temperature (21-24°C), leading to some loss though vaporization. Pressure build-up leading to swelling/melting during the microwave process could also lead to release of encapsulated volatiles (Chen et al., 1993). This could explain why the same encapsulants in the matrix can be very efficient at encapsulating vanillin (which is not volatile) but not as efficient with hexanal, which is far more volatile and may be lost by diffusion through the swollen/puffed structure.  112  The higher encapsulation efficiency of vanillin could also be due to the size of the molecules. CD has a cavity size of 262 Å. Steric hindrance is an important parameter in the association constant (Goubet et al., 1998). The bulkier a guest is, the more stable the inclusion complex becomes (Goubet et al., 1998). Vanillin (MW = 152.15) is a slightly larger molecule than hexanal (MW = 100.15) and may therefore be better retained by the complex. Figure 24 shows large variation in some values. The SPME procedure used to quantify hexanal was a potential source of variation due to its manual nature. Using SPME/GC-MS to measure the volatiles in the headspace of raspberry juices, Aprea et al. (2008) reported coefficients of variation ranging from 4-25% (mean value of 11%) and considered them to show “good repeatability of the method considering the manual procedure adopted”. Uneven heating in the REV equipment could also explain some of the variablilty. The newer larger scale 3.6 KW REV model dehydrator improved the yield for treatment “B” in section 4.5.8, when compared to the treatment B samples produced using the 900W REV model. Figures 25 and 26 show the main effects plot and interactions plot for yield analysis.  113  Figure 25: Main effects plot for yield (%).  Figure 25 shows that the overall mean yield for the REV treatments was lower than that for the freeze-dried treatments (p-value = 0.000).  The formulation also had a  significant effect (p-value = 0.000); treatments A and O are the only ones that have a mean yield above the overall yield for all treatments (represented by the horizontal line in Figure 25).  114  Figure 26: Interactions plot for yield (%).  The ANOVA for Yield concluded (with at level of significance of =0.05) that there is an interaction between the factors: drying method and matrix formulation. Figure 26 shows that although the yield for the REV treatments was consistently smaller than for the freezedried forms, particularly for treatment B, where the yield was considerably better in the freeze-dried form, than its REV counterpart. This difference however is not observed when producing the batches “B” formulation with the newer 3.6 kW REV model machine.  115  4.5.2.2 Validation of gaseous and liquid equilibrium for SPME quantification Table 28 shows the F-values obtained from performing analysis of variance for the extractions performed for 15, 30, 60 and 120 minutes in two separate runs. Table 28: P-values for ANOVAS in two separate runs.  First run  Second run  Extraction time  Rep  Extraction time  Rep  Concentration (mg/g) db  0.517  0.467  0.742  0.407  Yield (%)  0.580  0.449  0.706  0.411  All p-values in Table 28 were greater than =0.05, therefore it can be concluded that the extraction time had no effect on the quantification/extraction method. In other words, a 15 minutes extraction was as good as 2 hours in terms of extracting hexanal. This information indicates that 15 minute extraction time is sufficient to obtain the maximum hexanal release possible in two hours and reach equilibrium. The p-values for the repetitions indicated that there was no significant difference between repetitions.  4.5.2.3 Processing parameter optimization The first experiment, run to test different power/time combinations for REV processing, was done for formulations A and B. The second experiment, testing different freeze-drying times, was done for treatments P, O, and Z. Tables 29 and 30 contain the pvalues derived from analysis of variance for the REV and freeze-drying processing parameters experiments respectively.  116  Table 29: P-values for concentration (db) and yield of the different REV processing combinations, and freeze-drying processing times.  p-values Factor/Interaction  Concentration (db)  Yield  Time  0.052  0.057  Power  0.116  0.077  Formulation  0.000  0.073  Repetition  0.079  0.209  Time*Power  0.647  0.650  Power*Formulation  0.660  0.403  Time*Formulation  0.109  0.074  Time*formulation*power  0.883  0.817  Table 29 shows that neither time, nor power, nor any of the interactions significantly affected the concentration or yield at a =0.05 level of significance. This suggested that the processing conditions for REV did not influence either response variable. Consequently, the standard REV processing conditions of 12 minutes at 300 W and 20 torr were used to prepare microcapsules for further experimentation.  Table 30: P-values for concentration (db) and yield of the different freeze-drying processing times.  p-values Factor/Interaction  Concentration (db)  Yield (%)  Time  0.654  0.808  Formulation  0.000  0.000  Repetitions  0.202  0.213  Time*formulation  0.446  0.291  117  Table 30 shows that freeze-drying time (1-3 days) had no effect on concentration or yield at =0.05 level of significance. Hence the formulation influenced both yield and concentration (which was expected), but processing time by either method was of no significance. A freeze-drying time of 48 hours was therefore used to prepare the future microcapsule batches.  4.5.2.4 Scale-up microencapsulation trials The microcapsule formulation chosen for shelf-life trials was formulation “B” with a 3:72:23 ratio of hexanal:CD:pectin prior to drying because it provided the strongest inhibition of P. expansum in assays performed at 12°C and 25°C in microbiological medium. Since large amounts of microcapsules were required to perform the experiments, the formulation was dried using the larger scale 3.6 kW REV model equipment. Table 31 provides a comparison of the energy requirements for the model VMD900W REV and the 3.6 kW REV model ENWL291111-E020 equipment. The larger scale 3.6 kW REV machine was capable of processing larger batches of wet formulation in a more energy efficient manner, with savings of up to 16.65 KJ per gram of wet sample (a 54% reduction in energy requirements). Table 31: Energetic requirements of the REV machines. REV model  Model VMD900W 3.6 kW REV model ENWL291111E020  Wet mass processed (g) at a time 7  Number of magnetrons powered  Power (Watts)  Time (min  1  300  12  Work per g of wet sample (KJ/g) 30.86  228  3  1200  45  14.21  The 3.6 kW REV model dehydrator was also able to improve the hexanal concentration of the “B” formulation, from 6.20 ± 3.50 mg/g (db) to 14.50 ± 1.01 mg/g (db). 118  Consequently, the yield was also improved, from 20 ± 11% to 47 ± 3%. Since the dosing for all microbial and shelf-life tests has been done based on hexanal concentration, this improvement was taken into consideration and the microcapsules were dosed accordingly. As was mentioned in a previously, the higher yield could be caused by more even heating of the sample in the 3.6 kW REV model machine. Figure 27 shows the hexanal concentration and yield in dried microcapsules from both A and B treatments processed using the 3.6 kW REV model dehydrator. Formulation “B”, which contained twice as much β-CD as formulation “A”, provided consistently greater yields in the 3.6 kW drier compared to the the smaller REV model. Since each hexanal molecule is expected to form an inclusion complex with one molecule of β-CD (Almenar et al., 2007) it was anticipated that the yield would be optimal with a 1:1 molecular ratio of hexanal: β-CD. Increasing the ratio may have enhanced the availability of β-CD for complex formation leading to improved hexanal retention.  119  Figure 27: Mean hexanal concentrations (mg/g) dry basis and mean hexanal yields (%) dry basis (n=3) for treatments “A” and “B” processed using the 3.6 kW REV model ENWL291111-E020. NOTE: Vertical bars represent the standard deviation.  120  4.5.3 STAGE 3: Microcapsule physical characterization 4.5.3.1 Scanning electron microscopy Figures 28 and 29 show scanning electron micrographs of formulation B in the “fresh” mode (6% moisture) and “spent” mode (25.3% moisture) microcapsules created by REV drying and freeze-drying respectively. The “spent” microcapsules were recovered after being exposed to apple slices at 12°C for 12 days.  Figure 28: Scanning electron micrographs of REV hexanal-CD-pectin "B" formulation a) fresh at 40x, b) “spent” at 40x ,c) fresh at 700x, d) “spent” at 700x.  121  Figure 29: Scanning electron micrographs of freeze-dried hexanal-CD-pectin "B" formulation a) fresh at 40x, b) “spent” at 40x , c) fresh at 700x, d) “spent” at 700x.  Visual inspection of the micrographs seems to show a slightly more porous structure in the freeze-dried form (Figure 29c) than in the REV form (Figure 28c). Tsami et al. (1999) found that freeze-drying pectin-sugar gels allowed for a higher porosity than when obtained using regular microwaving. They also found that the freeze-dried pectin matrix has a higher water sorption capacity than the microwave version. This could explain why the freezedried “spent” microcapsules had a higher final moisture content and were able to release almost 100% of the hexanal encapsulated, unlike the REV treated capsules. One of the 122  factors that affects volatile release is pore size (Goubet et al., 1998). Larger pores in the freeze-dried microcapsules could have led to a higher release. Tsami et al. (1999) also found that microwave treatment slightly altered the colour of the pectin matrix while the freeze-dried one remained the same colour as commercial pectin. This was also observed in both the vanillin and the hexanal CD-pectin blends produced in this study, the microcapsules produced by REV technology always being a shade darker than the ones produced by freeze-drying.  4.5.3.2 Dielectric properties of the dried vanillin microcapsules Table 32 contains the averages of the dielectric properties and moistures of the formulations A and B at different stages of processing (n=3, 5 measurements per rep). Table 32: Averages for the dielectric constant (e') and dielectric loss factor (e'') of the hexanal-CD-pectin blends at different moistures at 2.45 GHz.  Treatment A (standard)  Treatment B (double ΒCD of A)  Process stage / temperature  Moisture (%)  e'  e''  Moisture (%)  e'  e''  Initial stage (t=0min) at 20°C  66.38%  42.59a±1.31  9.93 x ±0.56  50.35%  39.67 b ±0.06  9.64 x ±0.26  Half process (t=6 min) at 26-30°C  23.30%  3.46 c ±0.12  0.86 z ±0.02  14.56%  2.54 d ±0.02  1.70 y ±0.03  Dry form (t=12 min) at 35-40°C  4.55%  1.63 d ±0.13  0.64 z ±0.09  2.08%  2.07 d ±0.01  1.02 yz±0.15  Note: The numbers are averages of 3 reps (5 measurements) ± standard deviation. Letters a,b,c,d show significant differences in e’ and x,y,z in e’’.  The dielectric properties were examined statistically to determine significant differences. Both moisture content (p=0.000 and p=0.000) and treatment formulation (p=0.001 and p=0.040) had a significant effect on the dielectric constant and the loss factor respectively. In terms of the sample’s capacity to reflect microwaves, this was highly 123  related to the moisture content [the higher the moisture content the higher the dielectric constant) until it probably made no difference (moisture content 2-4%) This is probably due to the water content contribution, water having an empirical dielectric constant of approximately 78 (at 23°C). In terms of the differences (all differences calculated using Tukey’s HSD test) in loss factor (the capacity to convert microwaves into heat), the formulation plays a significant role. Having twice the amount of CD in treatment B provided a higher loss factor (more efficient heating) at lower moisture contents (23-2%). From a processing point of view, having a lower loss factor (e'') in the dry form than in the moist form (as can be seen in Table 32) is beneficial. Since the loss factor defines how the sample converts microwaves to heat, the drier the matrix becomes, the less likely it is capable of producing heat. A less heated matrix decreases the probability of hexanal losses due to sample overheating.  4.5.4 STAGE 4: Target microbial inhibition 4.5.4.1 Hexanal minimum inhibitory concentration (MIC) 4.5.4.1.1 Volatility MIC on solid media Table 33 contains the minimum inhibitory concentrations of vaporized hexanal found for the P. expansum at different temperatures. Table 33: MIC of vaporized hexanal against P. expansum at different temperatures.  Inoculum concentration  Temperatur  Day of  MIC (g hexanal  e (°C)  assesmen  / ml air)  t 2.7 x 105 and 8.1 x 106  25  5  5.7  5 x 106  12  10  6.2  5  10  5.4  5 x 10  6  124  It was clear that the compound exerted strong antifungal effects at all temperatures. However the spores were capable of germination upon further incubation at 25°in the absence of hexanal. This suggests that hexanal exerted fungistatic, rather than fungicidal activity at concentrations equivalent to the MIC. The MIC measurements reported in this work (Table 33) were considerably lower than literature values. For example, a similar attempt to quantify the antifungal activity of hexanal in the gas phase against P. expansum on malt extract agar yielded an MIC of 30.7 l / liter (ppm) (Neri et al., 2006), which was considerably higher than the 5.4-6.2 g hexanal /ml air reported here. However it should be noted that differences in test strain and experimental system design will influence experimental outcomes preclude strict comparison of data. In both cases hexanal was found to be fungistatic, since the fungus was able to resume growth once it was removed from the surrounding atmosphere. It should also be noted that in all cases the MIC for hexanal against P. expansum was considerably lower than that for vanillin (2130-2434 ppm, Table 15). Hence it is clear that the latter is a weaker antifungal agent than hexanal.  4.5.5 STAGE 5: Controlled release studies and target microbial inhibition on laboratory media 4.5.5.1 GC/Antimicrobial in vitro assay for P. expansum at 25°C and 12°C 4.5.5.1.1 Radial growth vs time at 25°C and 12°C Of the formulations tested at low (4 mg / jar) and high doses (14 mg / jar), formulation B provided the greatest inhibition. The effect of formulation B on the radial growth of the fungus was therefore examined at the high dose and treatment A microcapsules were included as a standard for comparison Figures 30 and 31 show the average radial growth of P. expansum (n=3) for treatments A and B at a dose of 14 mg hexanal/ jar, for the storage temperatures of 25°C and 12°C respectively. 125  1.6 1.4  Radial growth (radius cm)  REV A  1.2 FD A  1 REV B  0.8 FD B  0.6  C (+)  0.4  control (-)  0.2 0 0  -0.2  50  100  150  TIME (hours)  Figure 30: Average radial growth (cm) of P. expansum vs. time (hours) at the dose of 14 mg hexanal/jar (n=3) at 25°C.  126  2.50 REV A FD A 2.00  Radial growth (radius cm)  REV B FD B C (+)  1.50  C (-)  1.00  0.50  0.00 0  50  100  150  200  250  300  TIME (hours) Figure 31: Average radial growth (cm) of P. expansum vs. time (hours) at the dose of 14 mg hexanal/jar (n=3) at 12°C.  At 25°C only treatment B (REV) (which contains double the amount of CD as the standard) was capable of fully inhibiting germination for 120 hours, as did the positive control. Two main variables can be used to characterize fungal colony development in addition to the colony diameter: the lag phase () and the growth rate (max). The lag phase may be defined as the time delay between germination and the development of hyphal growth. Morales et al. (2010) state that short lag phases lead to earlier apple decay. The growth rate defines the increase in the diameter of the colony per unit of time. Since colony radial growth tends to be linear the growth rate can be derived from the slope of a plot of colony diameter vs. time. The lag phase can be determined empirically (by taking note of 127  when the colony appears) or can be determined from the x- intercept of a linear regression curve. This method assumes that once the lag phase is over, the growth is immediate and occurs at maximum rate (Baert et al., 2007b). Inhibition (in percent) was defined as the reduction in growth of a colony compared with the control. Equation 9 shows the formula used to calculate this value Equation 9: Percent Inhibition  At 25°C the empirical lag phase was 24 hours for the control (Figure 29). This was extended by 1, 2 or, in the case of full inhibition, 5 days with the treatments. Only microcapsules prepared from freeze-dried P and REV Z formulations at low doses failed to extend the lag phase (data not shown). Microcapsules from treatment P and O in both REV and FD (high doses and REV P low dose) induced a 2 day lag phase extension. The lag phase was extended by one day with all remaining treatments. Tables 34 and 35 show the means (n=3) of the calculated slopes and length of the lag phases for the high and low doses, respectively at 25°C. The lag phase in these tables was the x-intercept determined by linear regression analysis. The last column shows the lag increase, which is the actual length of the lag phase calculated by subtracting the natural lag observed in the negative control from the treatment lag phase.  128  Table 34: Growth rates and lag phases of P. expansum exposed to the low dose treatments stored at 25°C.  Process  Controls  FD  REV  Treatment C (-) C (+) A B O P Z A B O P Z  Growth Rate (slope) cm/hour 0.012 0.0072 0.011 0.013 0.012 0.012 0.011 0.012 0.0043 0.014 0.0069 0.012  R  Lag (hours germination delay)  Lag increase (hours)  0.99 0.99 0.98 0.99 0.98 0.97 0.97 0.91 0.97 0.97 1.00 0.98  4.50 81.8 31.22 42.81 34.77 24.15 31.46 55.57 69.09 49.5 72.63 29.74  0.00 77.3 26.72 38.31 30.27 19.65 26.96 51.07 64.59 45.00 68.13 25.24  2  Table 35: Growth rates and lag phases of P. expansum exposed to the high dose treatments stored at 25°C.  Process  Controls  FD  REV  Treatmen t C (-) C (+) A B O P Z A B O P Z  Growth Rate (slope) cm/hour 0.012 0.000 0.0030 0.0047 0.013 0.0026 0.0035 0.0093 0.000 0.0067 0.0030 0.013  R2  Lag (hours germinatio n delay)  0.99 1.00 0.98 0.99 0.97 0.99 0.99 0.96 1.00 0.99 0.98 0.97  4.50 120.0 98.15 64.44 57.95 97.75 90.85 42.07 120.0 89.92 91.48 38.05  Lag increas e (hours) 0.00 120.0 93.65 59.94 53.45 93.25 86.35 37.57 120.0 85.42 86.98 33.54  129  At 25°C a delay in germination of at least 24 hours was evident for all treatments and doses. Inhibition of radial growth ranged from 23-77% for the low doses, to 20-100% at the high doses. Only microcapsules from treatment REV B provided complete inhibition similar to that obtained with the positive control applied at the high dose. At 12°C neither microcapsules from treatments A or B or the positive control were capable of fully inhibiting (100%) germination and growth for 10 days. Maximum inhibition in the positive control was calculated as 91% after 240 hours. In contrast, inhibition ranged between 27-44% with the treatments and the lag phase was extended 1-3 days. The inoculum used for the experiment conducted at 12°C was considerably larger than that used at 25°C. It was an average of 2.4±2.3 x 107 cfu/ml. Morales et al. (2010) proved that increasing inoculum size decreases the lag phase, while the actual growth rate is not affected. Baert et al. (2008) applied different levels of P. expansum spores onto apple tissue. They reported a growth rate of 0.0236 ± 0.002 cm/hour and a lag phase of 54 ± 6 hours at 25°C, after addition of 20l of a 2 x 105 spore/ml suspension. This is close to the growth rate reported here for negative controls (0.0123 cm/hour) on PDA. The lag phase on PDA was much shorter at 25°C however, likely because PDA is a more hospitable environment for germination than apple tissue. Table 36 contains the averages (n=3) of the calculated slope and lag response variables for the high dose at 12°C storage.  130  Table 36: Growth rates and lag phases of of P. expansum exposed to the high dose treatments on PDA plates stored at 12°C.  Process  Controls FD REV  Treatment C (-) C (+) A B A B  Growth Rate (slope) cm/hour 0.0083 0.0018 0.0075 0.0065 0.0062 0.0058  R2  Lag (hours germination delay)  0.9867 0.9845 0.9832 0.9872 0.9709 0.9877  7.05 204.97 52.84 40.59 59.97 61.82  Lag increase (hours) 0 197.91 45.78 33.54 52.91 54.77  In apple tissue a growth rate of 0.0163 ± 0.003 cm/hour and a lag phase of 192 ± 49 hours at 12°C was reported for a P. expansum spore suspension of 2 x 106 spores/ml (Baert et al., 2008). On PDA (Table 36) the negative control growth rate was slower at 0.0083 cm/hour, but the lag phase was considerably shorter at only 7.05 hours. Once again, the neutral pH and comparatively un-stressful conditions in the microbiological medium likely promoted spore germination. Sanzani et al. (2009) exposed a Penicillium expansum colony on agarised apple juice medium to 10 g/ml of phenolic compounds for 8 days at 16°C storage. After 8 days the colony of the negative control had reached a diameter of 5.2 ± 0.64 cm while the best antifungal agent tested, quercetin, provided a reduction to 4.4 ± 1.0 cm, or only 15% inhibition. In the present work the negative control on PDA had a radial colony diameter of 1.58 ± 0.14 cm while the best treatment (REV A) led to 51% inhibition (0.82 ± 0.21 cm) after 8 days at 12°C. While these results were derived from studies performed in different model systems and at different temperatures, they do support the observation that hexanal is a potent antifungal compound. Tables 37 and 38 show the p-values obtained from analysis of variance for the factors: drying process, formulation, dose, replicates, and interactions between the 25°C data and the 12°C data respectively 131  Table 37: P-Values for microbial response data for samples incubated on PDA at 25°C.  P-values Growth rate (slopes in cm/hour)  FACTOR/response variable  Final radius measurement (cm) 5 days  Lag phase (hours)  Process Treatment Dose  0.124 0.005 0.000  0.371 0.001 0.000  0.186 0.069 0.000  Rep Process*Treatment Process*Dose Treatment*Dose Process*Treatment*Dose  0.394 0.007 0.039 0.840 0.162  0.003 0.000 0.028 0.323 0.163  0.000 0.008 0.033 0.901 0.039  Whether the microcapsules were freeze-dried or REV-dried did not influence growth of the fungus on PDA but the treatment (formulation used to make the microcapsules) significantly influenced the growth parameters. As expected, dose had a significant effect on growth but the replicates did not (which provides statistical validity). In all three response variables, there is an interaction between the process and the treatment formulation, as well as between the process and the dose. However, the interaction between the treatment and the dose (both significant factors individually) is nonexistent in the microbial data. Figures 32 through 35 show the main effects plots and interaction plots for each of the microbial response variables for a storage temperature of 25°C. Figures 36 through 37 show the same response variable plots, this time for the storage temperature of 12°C  132  Figure 32: Main effects plot for final radius measurement (cm) after 120 hours incubation on PDA at 25°C.  Figure 32 shows that the smallest final colony radii were achieved with treatment B and that final radii for treatments A, B and P were below the overall average. Treatment Z was the least effective in terms of final colony radius.  133  Figure 33: Interactions plot for final radius measurement (cm) at 120 hours incubation on PDA at 25°C.  Figure 33 shows the interaction of process and treatment, providing evidence that a combination REV processing with formulation B achieved the best result.  134  Figure 34: Main effects plot for the growth rate or slope (cm/hour) upon incubation on PDA at 25°C.  In Figure 34, the slopes show the same results as the maximum radii in terms of tendencies.  135  Figure 35: Interactions plot for the growth rate or slope (cm/hour) upon incubation on PDA at 25°C.  In Figure 35, the interaction between treatment and process is apparent since growth rate was reduced with microcapsules from all treatments that were dried by REV, with the exception of the combinations REV Z and REV A. Also interesting, when comparing drying process to dose, how at the high doses, REV processing produced a faster growth rate of the P. expansum colonies than did the FD process. However, at the lower dose, the opposite occurred.  136  Figure 36: Main effects plot for calculated lag (hours) upon incubation on PDA at 25°C.  Figure 37: Interactions plot for calculated lag (hours) upon incubation on PDA at 25°C.  137  Table 38 contains the p-values for the microbial response variables obtained at an incubation temperature of 12°C. Table 38: P-values for the microbial response variables obtained upon incubation on PDA at 12°C.  FACTOR/response variable  Final radius measurement (cm) at 240h  Process Treatment Rep Process*Treatment  0.031 0.073 0.189 0.693  p-values Growth rate (cm/hour)  0.084 0.167 0.854 0.588  Lag phase (hours) 0.036 0.363 0.008 0.230  At 12°C storage the mode of drying process had a significant effect (<0.05) on the final colony radius at day 10 and on the lag phase, but not the growth rate. However formulation did not significantly influence these response variables. Figures 38 and 39 show the main effects plots for the final colony radius diameter at 240 h and the main effects plot for the lag phase at 12°C storage. Since no significant effects or interactions were found in the growth rates, main effects and interactions plots were not included in this work.  138  Figure 38: Main effects plot for final radius measurement (cm) at 240 hours incubation on PDA at 12°C.  Figure 39: Main effects plot for lag phase (hours) upon incubation on PDA at 12°C.  139  4.5.5.1.2 Headspace hexanal concentration at storage temperatures of 25° and 12°C.  Headspace concentration profiles were prepared by plotting the measurements derived from SPME analysis against time. The profiles were then compared against the radial growth plots to understand how fluctuations in headspace concentrations affected germination and radial growth of the colonies. Figures 40 and 41 show the headspace hexanal concentration profile obtained with the high dose (14 mg/jar) of microcapsules prepared from formulations A and B during incubation at 25°C and 12°C. It should be noted that the gaps in the release profiles are due to SPME readings that were out of the range of the calibration curve. They occurred mainly at time zero, when the hexanal concentration was very low.  140  Headspace hexanal concentration (mcg hexanal/ml of air)  3.5 REV A  3  FD A 2.5 REV B 2  FD B  1.5  1  0.5  0 0  20  40  60 80 TIME (hours)  100  120  140  Figure 40: Headspace hexanal concentration (averages (n=3) released from microcapsules made from formulations A and B in the headspace above PDA plates incubated at 25°C over time.  It is clear that the higher rate of release from FD microcapsules from treatment A resulted in a slower growth rate and final colony radius. This was the case with the higher doses obtained with treatments A, Z, and P (data not shown). For treatment B, only the REV-dried microcapsules were capable of fully inhibiting growth despite similar release profiles for the FD and REV microcapsules.  141  0.70 Headspace hexanal concentration (mcg hexanal / ml air)  REV A FD A  0.60  REV B  0.50  FD B  0.40 0.30 0.20 0.10 0.00 0  50  100  150 200 TIME (hours)  250  300  Figure 41: Headspace hexanal concentration averages (n=3) released from microcapsules made from formulations A and B in the headspace above PDA plates incubated at 12°C over time.  The apparent hexanal headspace concentration profile is a result of several phenomena occurring simultaneously: release from the microcapsules, absorption by the aqueous agar, and inactivation through interaction with the germinating conidia. In a real food system (such as apple slices), hexanal concentrations would also be affected by metabolism or reactions with native enzymes. Sholberg and Randall (2007) reported that hexanal concentration in a fumigation chamber with whole apples declined slowly and could be described by a third-order polynomial equation. Although the profiles reported so far did not include interactions with a food component, the interaction could be characterized using a similar approach. Table 39 and 40 contain the coefficients of determination (R2) obtained for the cubic regressions of all the profiles.  142  Table 39: Coefficients of determination (R2) for the cubic behavior of the average hexanal release profiles at 25°C (n=3).  Drying method REV FD  Coefficients of determination (R2) B Z P  Dose  A  High Low High Low  0.987 0.981 0.993 1.000  0.992 0.971 0.922 1.000  1.000 1.000 0.909 0.997  0.951 0.999 0.974 1.000  O 0.722 1.000 0.976 1.000  Table 40: Coefficients of determination (R2) for the cubic behavior of the average hexanal release profiles at 12°C (n=3).  Drying method REV FD  Coefficients of determination (R2) A B 0.892 0.868 0.909 0.891  Another important aspect to consider when comparing results obtained at the two incubation temperatures is that the vapour pressure of hexanal is greater at 25°C than 12°C. Consequently, the antimicrobial efficiency increases despite a corresponding acceleration in the germination and growth rate of the target microorganism  143  4.5.6 STAGE 6: Apple shelf-life experiments Two separate experiments were carried out with apple slices. The first was carried out over 21 days, but this proved too long and the samples were too deteriorated for analysis. Another important limitation was the size of the P. expansum spore suspension used to inoculate the apple slices. The original concentration of 10 6-107 spores/ml (1.6x104 -1.1 x 105 spores per site) was too high to observe inhibitory effects. A similar observation was reported by Morales et al. (2008b) who inoculated Golden apples with 104 and 106 P. expansum conidia/ml suspensions and found out that “the lag phase was significantly lower in more concentrated inocula”, indicating that the spores germinated at a faster rate at a higher inoculum concentration. Inoculum concentration should not be too low, however. Baert et al. (2008) found that when inoculating 20 l of 102, 103, 104, 105, 106 cfu/ml P.expansum spore suspensions in apple wounds for 15 days at 4, 12, and 25°C, the inocula needed to be above 2 x104 spores/ml (400 spores/wound) to achieve 93% germination rate or higher at 4°C. Lower concentrations would be inhibited by the apple pH and natural environment. Table 41 shows their results. Table 41: Percentage of germinated wounds in apple varying spore suspension concentration and temperature. Reprinted from (Baert et al. 2008) with permission from Elsevier ©.  144  Thus inoculum concentration must be above the 104 cfu/ml level so that any inhibition can in fact be attributed to the treatments and not to antimicrobial effects associated with natural environment and components present in the apple tissue For the second run, the storage temperatures were on average 4.8°C ± 0.3 (n=20) and 12.1°C ± 0.1 (n=20). The P. expansum inoculum spore suspension inocula were 7.1 x 104 - 7.0 x 105 cfu/ml (400-4200 cfu/site), which was still enough to overcome the natural environment of the apple flesh and not so concentrated that growth was impossible to combat.  4.5.6.1 Visual assessment results As expected, inoculated apple slices decayed more rapidly than un-inoculated controls, and the process was faster at 12°C than 5°C. By day 12 (288 h), the only treatment without evidence of fungal growth or secondary browning (browning caused by germination of spores but no visible mycelia) was the positive control, which had received 112 mg of non-microencapsulated hexanal. It should be noted, however, that the high dose of hexanal in the headspace induced browning of the tissue that was visible within 24 hours. After 48 hours the browning was consistent throughout the apple tissue (as can be seen in Figure 18, image labeled “C+”) and did not change further over time. The slices also appeared soggy, deteriorated and had a “rubbery” texture, and they were given a visual assessment rating of “1”. The browning, sogginess, and rubbery texture were a consequence of phytotoxic effects induced by hexanal. Thavong et al. (2010) reported the phytotoxic effects of hexanal on intact longan fruit at a dose of 900l hexanal/ liter of air for two hours. They reported electrolyte leakage of the pericarp as well as a pericarp that was more reddish brown and less intense in colour. Thavong et al. (2010) hypothesized the electrolyte leakage caused by the hexanal led to the release of phenolics from vacuoles into the cytoplasm where PPO and POD enzymes may have induced the synthesis of browning compounds. 145  Total phenolics were not quantified in this project since there was no evidence that P. expansum induces changes in apple phenolics. Cole & Wood (1961) found that the phenolic concentration of healthy apple tissue that had been allowed to brown was the same as the concentration in tissues affected by P. expansum. Another reason total phenolics were not quantified was that the ascorbic acid treatment given to slices, was capable of interfering with the assay, and in addition would have been an anti-browning agent in the treated samples. Figures 42 and 43 show the visual assessment rating of slices that were either not inoculated or inoculated with P. expansum upon storage at 5°C. Although deterioration was evident by day 6 (144 hours) it was not until day 9 (216 hours) that the visual quality of negative controls deteriorated further with inoculated slices than un-inoculated slices. In the inoculated slices, visual quality scores were significantly higher (Tukey’s HSD <0.05) with freeze-dried and REV microcapsules than the negative controls on days 9 (216 h) and 12 (288 h), but not on day 15 (360 h). In the un-inoculated slices, visual quality scores were also significantly higher (p<0.05) with freeze-dried and REV-dried microcapsules than the negative controls on days 12 (288 h) and 15 (360 h). Assuming limits of “saleability” and “useability” for ratings “5” and “3”, respectively, according to the scale of Kader and Cantwell (2005), the inoculated slices stored at 5°C with freeze-dried and REV microcapsules (average rating of 5.0±1.0 and 4.0±1.0 respectively, not significantly different <0.05) were still considered “sellable” day 12 (288 h). However low scores (1.7 ± 1.2 at 5°C and 1.0 ± 0.0 at 12°C) at day 15 (360 h) showed they were no longer considered “usable”. The negative control was also of poor quality at day 12 (1.7 ± 1.2) and was scored very poorly by day 15 (1.0 ± 0.6). As for un-inoculated apple slices stored at 5°C, the means scores (8.3 ± 0.3) indicated they were not only still “sellable” but between “Excellent” and “Good” quality at day 12 (288 h) when both the freeze-dried and REV microcapsules were applied. At day 15 (360 h) the mean rating was 6.0 ± 1.0 and 5.7 ± 2.3 (for the freeze-dried and REV microcapsules respectively, which not significantly different), a score just slightly above the 146  “sellable” limit. The negative control was also considered sellable on day 12 (5.5 ± 1.8) and completely spoiled by day 15 (2.0 ± 1.0).  9.00 8.00  Visual assessment rating  7.00 6.00 5.00 REV B  4.00  FD B  3.00 C (+)  2.00  C (-)  1.00 0.00 0 -1.00  2  4  6  8  10  12  14  16  TIME (days)  Figure 42: Visual quality rating vs. time for un-inoculated apple slices stored at 5°C.  147  9.00 8.00  Visual assessment rating  7.00 6.00 REV B  5.00 FD B  4.00  C (+)  3.00  C (-)  2.00 1.00 0.00 -1.00  0  5  10  15  20  TIME (days)  Figure 43: Visual quality rating vs. time for P. expansum inoculated apple slices stored at 5°C.  Figures 44 and 45 show the visual assessment ratings given through time to the slices that were not inoculated and inoculated with P. expansum respectively and were stored at 12°C.  148  9.00 8.00  Visual assessment rating  7.00 REV B  6.00 5.00  FD B  4.00 C (+)  3.00 C (-)  2.00 1.00 0.00 -1.00  0  2  4  6  8  10  12  14  TIME (days)  Figure 44: Visual quality rating vs. time for un-inoculated apple slices stored at 12°C. 9.00 8.00  Visual assessment rating  7.00 6.00  REV B  5.00 FD B  4.00 C (+)  3.00  C (-)  2.00 1.00 0.00 0 -1.00  2  4  6  8  10  12  14  TIME (days)  Figure 45: Visual quality rating vs. time for P. expansum inoculated apple slices stored at 12°C.  149  For inoculated apple slices stored at 12°C, treatments with the freeze-dried and REV-dried microcapsules and negative controls were not significantly different (p<0.05) and there was no evidence of microbial growth after 6 days (144 h). By day 3 (72 h), however, the mean rating (8.8 ± 0.3) for REV treated slices was significantly different from the negative controls, unlike slices treated with freeze-dried microcapsules, which were not significantly different. Mean ratings for un-inoculated apple slices stored at 12°C were not significantly different from the negative control at day 6 (144 h), with mean ratings for 5.0 ± 1.0 for the freeze-dried and 4.3 ± 0.6 for REV-dried microcapsules. By day 9 (216 h) mean ratings were still significantly different from the negative control and green discoloration associated with growth of fungi became evident. Figures 45-48 show the mean ratings obtained for the different treatments at days 6, 9, 12, and 15. It should be noted that the samples stored at 12°C were not tested on day 15, since they were opened and used for texture measurements on day 12, as specified in the materials and methods.  150  Figure 46: Visual quality rating (n=3) of apple slices at day 6 (144 hours). NOTE: Vertical bars represent the standard deviation, letters show significant differences at a level of =0.05.  Figure 47: Visual quality rating (n=3) of apple slices at day 9 (216 hours). NOTE: Vertical bars represent the standard deviation, letters show significant differences at a level of =0.05.  151  Figure 48: Visual assessment rating (n=3) of apple slices at day 12 (288 hours). NOTE: Vertical bars represent the standard deviation, letters show significant differences at a level of =0.05.  Figures 46-49 show the average ratings obtained for the different treatments at days 6, 9, 12, and 15. It should be noted that the 12°C stored samples were not tested on day 15, since they were opened and tested for texture on day 12.  152  Figure 49: Visual assessment rating (n=3) at day 15 (360 hours). NOTE: Vertical bars represent the standard deviation, letters show significant differences at a level of =0.05.  4.5.6.1.1 Assessment of P. expansum inhibition in wounded apple slices P. expansum was inoculated in two wounds per slice as can be seen in Figure 17. This yielded a total of 6 wound inoculation sites per jar (three slices per jar) with the possibility for germination. The time at which a colony was visually detected in the wounds was documented for all samples. This occurred at day 6 (144 h) for apple slices stored at 12°C and day 15 (360 h) for apple slices stored at 5°C. Colony diameter was impossible to measure accurately and only the presence/absence of visual growth was recorded. Percent inhibition was defined as the percentage of wounds without visible mycelial growth at a given point in time. Equation 10 shows the calculation. For example, if there is evidence of growth in three of the wound sites, the calculated value is 50% at that particular point in time.  153  Equation 10: Formula to Calculate % Inhibition  Table 42 contains the % inhibition results obtained at day 6 at 12°C and day 15 at 5°C. Table 42:% Inhibition results.  Treatment  day 6 at 12°C  day 15 at 5°C  C(+)  100 ± 0.0%a  100 ± 0.0%a  FD  72.2 ± 35%ab  11.1 ± 19% b  REV  33.3 ± 17%bc  38.8 ± 25%b  C(-)  0.0 ± 0.0%c  0.0 ± 0.0%b  NOTE: Values are averages ± standard deviation.  Supercripts show significant differences at 0.05 significance, but only within a point in time.  It should be noted that the variations for the microencapsulated treatment inhibition percentages (Table 42) were large, therefore no statistical difference could be found between them. In conclusion, visual assessments showed that un-inoculated apple slices treated with hexanal-containing microcapsules (produced both by REV and freeze-drying methods) were still “sellable” after 15 days (360 h) at 5°C even though the negative control was nearing “poor” quality rating scores. This produced a shelf-life of slightly more than two weeks associated with the microencapsulated treatments. Different shelf-lives are reported for commercial apple slices. McDonald’s “Apple Dippers” are reports to have a 14 day shelf-life and “new shipments arrive three times a week to replenish the stocks” (Meyer, 2011). Italy’s “La Veneta” products can be stored for “10 days or more if preserved at 4°C” (Anonymous, 2011). Finally Sun Rich Fresh Foods Inc., a local company in Richmond, B.C., reports 8 days for apple slices in their “Dry Pack & Retail Portfolio” (Sun Rich Fresh Foods Inc., 2012) 154  It should be noted however, that these commercial apple products are treated with calcium salts such as Natureseal ® to crosslink the pectin in the tissue and retain better texture for longer time. Natureseal ® was not used for the experiments described in this work. Another important factor to consider is that fresh-cut apple slices are not sold in hermetically sealed glass containers, which would promote anaerobic fermentation which may cause further deterioration of the tissue. Commercial packaging systems employ materials that allow for some permeation of oxygen. Despite the factors mentioned above, the shelf-life achieved at 5°C with microencapsulated hexanal with un-inoculated apple slices was approximately 15 days. This is more than the shelf-life reported by the three commercial producers mentioned above. The use of Natureseal® and adequate oxygen permeable packaging could possibly lead to additional improvements in shelf-life. Further research should focus on the application of these factors to improve the shelf-life for commercial applications of hexanal containing, CD-pectin microcapsules.  4.5.6.2 Visual assessment results and headspace hexanal concentration/release profiles. In an attempt to understand how hexanal released from the microcapsules improved the visual quality of the apple slices, release profiles were determined by measuring the hexanal concentration in the headspace. Figures 50 and 51 show the hexanal release profiles and the visual assessment ratings over time for apple slices stored at 5°C. It should be noted that the significant difference letters assigned to individual points in the plots are only comparable for specific points in time, for example two hexanal concentrations for different treatments at day 6 (144 h), but not two hexanal concentrations at day 6 and day 9 (216 h).  155  Figure 50: Headspace hexanal concentrations (n=3) in the headspace above apple slices vs. time for microencapsulated treatments stored at 5°C. (Note: Letters show significant differences (=0.05) at specific times only).  156  Figure 51: Visual assessment of apple slices stored at 5°C vs. time. (Note: Letters show significant differences (=0.05) at specific times only).  As shown in Figure 50, a decrease in headspace hexanal concentration after day 6 coincided with a decline in visual assessment scores (Figure 51), a trend which became quite apparent by day 9. Although hexanal concentrations in the headspace of jars that received the freeze-dried microcapsules were higher than in jars with REV microcapsules, there was no significant difference in the visual assessment scores. Figures 52 and 53 show the hexanal release profiles and the visual assessment scores obtained at 12°C. Although the quality of non-inoculated apple slices was retained slightly better with the freeze-dried microcapsules (which coincides with higher hexanal concentrations in the headspace), the difference was not statistically significant (<0.05). Hexanal concentrations began decreasing at day 3 (72 h), which explains why deterioration was evident by day 6 (144 h). 157  Figure 52: Headspace hexanal concentration (n=3) above apple slices stored at 12°C vs. time. (Note: Letters show significant differences (=0.05) at specific times only).  158  Figure 53: Visual assessment vs. time of microencapsulated treatments at 12°C. (Note: Letters show significant differences (=0.05) at specific times only).  As expected, the hexanal concentrations at 12°C were higher than at 5°C, since the vapour pressure of hexanal is higher at the higher temperature.  4.5.6.3 “Spent” microcapsule quantification On an interesting note, the maximum headspace concentrations measured at 5°C and 12°C were 0.09 and 0.14 g hexanal / ml of air, respectively. Both these concentration “maximums” were observed with the freeze-dried microcapsules for non-inoculated treatments. Although the highest obtained, these concentrations were well below the maximum attainable according to the Antoine and ideal gas law equation calculations of 159  Table 23 (18.7 g/ml air at 5°C and 28.9 g/ml air at 12°C), which was expected given a dosage of 112 g/ml of air. This prompted the question of how much hexanal contained in the microcapsules was actually released into the headspace. In order to find out, sachets from the jars at days 12 and 15 were opened and the remaining hexanal in “spent” microcapsules was measured according to the protocol in section 4.4.2.1. In order to keep the proportions of sample and extraction solvent (water) the same, the calibration curve, had to be slightly extended in the lower range to include the new lower concentration “spent” samples (data not shown). The calibration curve had a coefficient of determination (R2) of 0.963. Table 43 shows the residual hexanal concentrations in the microcapsules as well as the percentage of hexanal released (of the original amount per gram of dry sample). Table 43: Hexanal released (mg/g db) by the "spent" microcapsules and the percentage of the original load.  5°C  12°C  Hexanal released (mg/g) db  Amount released (%)  REV (n=12)  10.22±1.07c  71±7c  FD (n=12)  14.03±1.14b  78±6b  REV (n=9)  9.72 ±0.87c  67±6c  FD (n=9)  16.58 ±0.71a  92±4a  Note: Letters show significant differences at a significance level of =0.05.  Hence both storage temperature and drying method had a significant effect (pvalue=0.002 and p-value=0.000 respectively) on hexanal release from the microcapsules. However the presence/absence of an inoculum did not significantly affect (p-value=0.527) the amount of hexanal released. Freeze-dried microcapsules appeared capable of releasing significantly more hexanal than the REV-dried microcapsules, which correlated with significantly higher 160  headspace hexanal concentrations observed in Figures 50 and 52. However this was not enough to induce significant differences in the visual assessment rating (Figures 51 and 53) Another interesting observation is that the maximum headspace hexanal concentration observed at 12°C in a container with apples was 0.14 g hexanal /ml of air when the original dose was 112g hexanal /ml, while a smaller hexanal dose (14.7g hexanal /ml) in the same container with PDA resulted in a maximum headspace hexanal concentration of 0.58 g hexanal /ml of air at the same temperature. In order to explain how a smaller dose led to a larger maximum headspace concentration, several phenomena must be considered. As mentioned in the literature review, hexanal is known to be converted to hexanol, 5 hexylacetate, butyl hexanoate, hexyl hexanoate, hexyl butanoate, 2-methyl-propyl-hexanoate, and butyl butanoate (Song et al., 1998) when metabolized by apple tissue. Sholberg & Randall (2007) reported that hexanal concentrations in fumigated apples were reduced from 3 mg/liter hexanal to 0.5 mg/hexanal within 24 hours, presumably due to uptake by the fruit. It is also known that water is capable of absorbing/dissolving some of the hexanal (Almenar et al., 2007), although this factor was considered when calculating the microcapsule dose. The hexanal in the microcapsules was dosed per gram of water in the sample. In addition, other vaporized compounds may be released by the fruit into the headspace of a closed container. These gases may compete for the binding sites in the SPME fiber, thus allowing for hexanal to go under-quantified. It should be noted that nonwater soluble components may compete for binding sites in the fiber, but water/moisture in the system does not since hexanal has poor affinity for water. Water in the system therefore should not affect its partition or its vapour pressure (Gardini et al., 1997). 4.5.6.4 Texture analysis Texture analysis was used to document apple flesh and skin deterioration, and the results were analyzed using analysis of variance and paired comparisons employing Dunnett’s test to verify whether texture parameters for the treated slices at day 12 (12°C 161  storage) and day 15 (5°C storage) changed over time. The “time zero” measurements (used as the “control” for Dunnett’s test) were a combination of 12 repetitions taken from samples of each of the four batches Figures 54, 55, 56, and 57, show the influence of the treatment on the Fracturability, Hardness 1, Hardness 2, and Chewiness parameters for the TPA testing method. The parameters Cohesiveness and Springiness did not change over time (data not shown). In terms of Fracturability, the results show that temperature had a significant effect (P=0.000), and the values obtained were significantly smaller for the samples stored at 12°C than at 5°C. Furthermore, the positive control consistently lacked Fracturability. This was possibly due to the “rubbery” texture that excess hexanal caused in the apple tissue and peel. As can be seen in figure 54, all apple slices held at 12°C differed significantly (<0.05) from the time zero control value of 68.87 ± 17.74 N. At 5°C, only the “rubbery” positive controls were significantly different. Although the Fracturability of apple slices stored with microencapsulated hexanal appeared to be better preserved, differences with the negative controls were not significant.  162  Figure 54: Fracturability-Texture analysis TPA parameter (asterisks show treatments that are significantly different from time zero (p<0.05), vertical bars stand for standard deviations).  The same trend was seen with the Hardness 1 and Hardness 2 values (Figure 55 and 56), where all apple slices stored 12°C stored were significantly less “hard” (=0.05) than the original apple slices at time zero during both compressions. The “rubbery” positive controls stored at 5°C were again the only apple slices that lost their “hardness” during the first compression, but on the second compression the negative inoculated control were also significantly less “hard”. Once again treatment with the microcapsules provided the best values, and with one exception, they were not significantly different from one another or the negative control. An exception was noted with the Hardness 2 test (force to perform the second compression) where apple slices stored with freeze-dried microcapsules were significantly harder (=0.05) than the negative control, but not its REV counterpart. The same cannot be said for the REV treatment. This coincides with the fact that the freezedried microcapsules were shown to release more hexanal into the headspace than the REV microcapsules, very likely allowing for the sample to retain its firmness by inhibiting the growth of P. expansum more efficiently.  163  Figure 55: Hardness 2 -Texture analysis TPA parameter (n=6) (asterisks show treatments that are significantly different from time zero (p<0.05), vertical bars stand for standard deviations).  Figure 56: Hardness 2 -Texture analysis TPA parameter (n=6) (asterisks show treatments that are significantly different from time zero, (p<0.05), vertical bars stand for standard deviations).  164  Figure 57: Chewiness -Texture analysis TPA parameter (n=6) (asterisks show treatments that are significantly different from time zero (p<0.05), vertical bars stand for standard deviations).  The Chewiness parameter was most affected by the microencapsulation treatments. Once again storage at 12°C led to deterioration of apple slice texture (all treatments led to significantly less “chewiness” than the original value of 44.9 ± 14.7 N*mm). However, at 5°C both the “rubbery” positive control and the negative control (inoculated and noninoculated) showed significant deterioration in terms of Chewiness, while treatment with the microcapsules did not. This provided strong evidence that the release of hexanal from the microcapsules improved the retention of fresh apple slice texture attributes. Figures 58, 59, and 60 depict the results for the parameters Ff, Ws, and W f of the puncture method.  165  Figure 58: Ff -Texture analysis puncture parameter (n=9) (asterisks show treatments that are significantly different from time zero (p<0.05), vertical bars stand for standard deviations).  Figure 59: Ws -Texture analysis puncture parameter (n=9) (asterisks show treatments that are significantly different from time zero (p<0.05), vertical bars stand for standard deviations).  166  Figure 60: Wf -Texture analysis puncture parameter (n=9) (asterisks show treatments that are significantly different from time zero (p<0.05), vertical bars stand for standard deviations).  The puncture method of texture analysis was not as efficient in finding significant differences among the treatments. The method also proved impractical for the measurement of texture in samples showing evidence of deterioration. Slices stored at 12°C were so seriously degraded that the puncture test parameters were extremely variable. For this reason only results of the studies performed at 5°C are reported and considered in this manuscript. There was no significant difference in the force needed to rupture the skin (F s) (data not shown) for any treatment and they were not significantly different from the original time zero values. This was probably due to the fact that microbial deterioration occurred mainly on the flesh, which had more exposed nutrients and in some cases had been inoculated with P. expansum spores. In terms of the force required to rupture the flesh once the skin had been broken (F f) as well as the work required to rupture the skin (W s) (Figures 58 and 59), only the “rubbery” positive controls showed a significant difference from the original values at time zero. The 167  “rubbery” positive controls seem to need significantly more work (<0.05) to puncture the skin, but once broken the force needed to penetrate the deteriorated flesh was less than the original values. Finally, in terms of the work required to rupture both skin and flesh (W f) (Figure 60) the “rubbery” positive controls required significantly less work than the apple slices at time zero. This is also true for the inoculated negative control, which had had so much texture deterioration, caused by the growth of P. expansum, that the work required to rupture both skin and flesh was significantly less than the original work at time zero.  4.6 Summary of findings for the hexanal project REV  produced  (UBC  model  VMD900W),  pectin-βCD,  hexanal-containing  microcapsules had low hexanal yields (15-36%). Scaling up to the larger 3.6 kW REV model machine increased the yield of the chosen formulation ”B” to 47%. The 3.6 kW REV model allowed for drying of larger batches in a more energetically efficient manner. When compared to freeze-drying, the process was considerably shorter (45 min vs. 48 hours). Hexanal containing REV processed microcapsules were just as effective as the direct addition of hexanal into the headspace at inhibiting P. expansum spores on PDA at 25°C. They were slightly less effective than the direct hexanal addition at 12°C. Treating both P.expansum spore inoculated and un-inoculated fresh-cut apple slices with  hexanal  containing  microcapsules  resulted  in  several  benefits.  Firstly,  microencapsulation made high doses of hexanal feasable. The slow hexanal release prevented browning deterioration and phytotoxic symptoms shown by the direct injection of the high concentrations of the compound. Microencapsulation also allowed for ease of handling and dosing due to hexanal stability and dilution when in microencapsulated form. A controlled release due to microencapsulation (by both drying methods) was achieved. Freeze-dried microcapsules showed a significantly higher release. 168  Un-inoculated apple slices stored at 5°C and treated with hexanal containing microcapsules (both drying methods) were still “sellable” by day 15 even when the negative control was nearing the “poor” quality rating. Texture analysis suggested that microcapsule-containing apple slices were capable of retaining their original texture when stored at 5°C for 15 days, in some cases even when the negative control could not.  169  5 Conclusions Vanillin had significantly higher encapsulation efficiency in the REV processed CDpectin matrix, compared to the freeze-dried form. The rate of vanillin release in an aqueous medium was rapid. Full release was obtained in 2 hours for the freeze-dried microcapsules, and a maximum release (80%) was achieved in 24 h for the REV microcapsules. Temperature of the medium had no significant effect on the rate of release. Cross-linking the matrix with calcium ions prior to processing was done in an attempt to decrease the rate of controlled release in an aqueous medium, however the vanillin yield obtained was unacceptable and the microcapsules exhibited reduced solubility in water. The latter would preclude application on the hydrophilic surface of apple slices, since leftover debris would make the slices unacceptable to consumers. Application of pectin in the amounts needed to achieve vanillin concentrations required for inhibition resulted in unappealing apple slices, both by “breading” and “gel application” methods. Increasing the vanillin to pectin ratio to double the amount in the formulations increased the encapsulation efficiency. However further increases in vanillin input showed that the encapsulation efficiency was inversely proportional to vanillin concentration. In addition, vanillin was found to be insufficiently volatile for gas phase antimicrobial applications. Given the limitations inherent to vanillin, an alternative antimicrobial with higher potency and volatility was considered for the remainder of this work. Hexanal was reported to exert strong antifungal activity and the comparatively higher volatility of the compound suggested it may be a better antimicrobial candidate for increasing the shelf-life of fresh-cut fruit. Although the yields were low, ranging from 15-36%, REV produced (UBC model VMD900W), pectin-βCD, hexanal-containing microcapsules could pose a viable method for a gradual incorporation of hexanal into the headspace of fresh-cut fruit, such as apple slices. In addition, using the larger 3.6 kW REV model machine increased the yield of the chosen formulation ”B” to 47%, as well as produced larger batches and was more 170  energetically efficient than the 900 W drier. This result suggests that there is potential for process improvement of  REV encapsulation of pectin-βCD, hexanal containing  microcapsules. REV processing has a considerable time reduction benefit when compared to freeze-drying (45 min vs 48 hours) as well as an energy reduction benefit (eliminating the need for freezing). Hexanal microencapsulated using REV proved to be just as effective as the direct addition of hexanal into the headspace when targeting P. expansum spores in laboratory media (PDA) at 25°C, and slightly less effective at 12°C, probably due to the fact that release was not as efficient at lower temperatures due to the lower hexanal vapour pressure. When treating both P. expansum spore inoculated and un-inoculated fresh-cut apple slices, microencapsulation showed a considerable advantage since the ability to dose hexanal at high amounts was feasible. Its slow release prevented the browning deterioration and phytotoxic symptoms shown by the direct headspace injection of the high concentrations. The direct injection of such amounts resulted in poor visual quality and considerable texture deterioration in the first 24 hours, while the same dosage in microencapsulated form did not have these effects with the added benefit of diluting the hexanal, so that it could be dosed with ease. The headspace hexanal concentration results showed (indirectly) that a large fraction of the hexanal released into the headspace seemed to became absorbed by the tissue. A controlled release due to microencapsulation (by both drying methods) was achieved. The freeze-dried microcapsules showed a significantly higher release. Visual assessment determined that un-inoculated apple slices stored at 5°C and treated with hexanal containing microcapsules (produced by both drying methods) were still “sellable” by day 15 even when the negative control was nearing the “poor” quality rating. Texture analysis shows that microcapsule-containing treatments were capable of retaining their original texture when stored at 5°C for 15 days, when the negative control did not (e.g. TPA “Chewiness” Fig. 57). 171  Finally, microencapsulation could also solve the issue of handling losses, due to hexanal’s high volatility, by making it more stable. Hexanal’s ability to significantly increase natural aroma production in apples as well as increasing colour stability is an added benefit as previously reported. Although not measured instrumentally, an increased aroma intensity was anecdotally observed in the apple slices that had been stored with the microencapsulated treatments of both drying methods. Taking into consideration the benefits and limitations of microencapsulating hexanal in a pectin-CD matrix, and the efficacy with which it is able to increase visual and textural shelf-life for fresh-cut apple slices at 5°C storage, hexanal REV and freeze-dried microcapsules seem like viable options for preservation. Future shelf-life studies are needed to define the maximum shelf-life extension possible for fresh-cut apple slice using REV encapsulated hexanal in the CD-pectin matrix employing optimal conditions. Optimal conditions would include: utilizing Natureseal®, fully eliminating carpal tissue, packaging in oxygen semipermeable films/containers and maintaining a 0-3°C storage temperature. Other interesting studies could involve the REVCD-pectin encapsulation of other effective antimicrobial C6 volatiles like trans-2hexenal. Before hexanal containing microcapsules are ready to be used in the food industry, other important tests, such as challenge tests for human pathogens (food safety) and sensory evaluation analysis (consumer acceptance) need to be performed. Although further studies yet need to be done, the REV and FD microencapsulation of hexanal in a βCD-pectin matrix seems to pose another step forward to finding a solution for the shelf-life problems associated with fresh-cut products.  172  References Abbasi, S. Rahimi, S. (2008). Microwave-assisted encapsulation of citric acid using hydrocolloids. International Journal of Food Science and Technology, 43, 1226-1232 Acosta E. (2008). Chapter 3: Testing the effectiveness of nutrient delivery systems. In Garti N (Ed.) Delivery and Controlled Release of Bioactives in Foods and Nutraceuticals, Cambridge: Woodhead Publishing Limited. Ahvenainen, R. (1996). New approaches in improving the shelf-life of minimally processed fruit and vegetables. Trends in Food Science & Technology, 7, 179-187 Al-Ati, T., and Hotchkiss, J.H. (2002) Chapter 10: Application of packaging and modified atmosphere to fresh-cut fruits and vegetables. In O. Lamikanra (Ed.) Fresh-Cut Fruits and Vegetables: Science, Technology, and Market, USA: CRC Press Alegre, I., Abadias, M., Anguera, M., Oliveira, M., Viñas, I. (2010). Factors affecting growth of foodborne pathogens on minimally processed apples. Food Microbiology, 27, 7076 Almenar, E., Auras, R., Rubino, M., & Harte, B. (2007). A new technique to prevent the main post harvest diseases in berries during storage: Inclusion complexes βcyclodextrin-hexanal. International Journal of Food Microbiology, 118(2), 164-172. Anonymous (2011). “Italy: La Veneta presents its new fresh-cut fruit line”. Retrieved on Dec 1, 2011 from http://www.mbg.com.my/MBG/news-a-updates/weekly-info/522-italy-laveneta-presents-its-new-fresh-cut-fruit-line.html Aprea, E. Biasioli, F., Carlin, S. Endrizzi, I., Gasperi, F. (2009). Investigation of volatile compound in two raspberry cultivars by two headspace techniques: solid phase microextraction/gas chromatography-mass spectrometry (SPME/GC-MS) and protontransfer reaction-mass spetrometry (PTR.MS). Journal of Agricultural and Food Chemistry, 57, 4011-4018 Arimura, G., Matsui, K., & Takabayashi, J. (2009). Chemical and molecular ecology of herbivore-induced plant volatiles: Proximate factors and their ultimate functions. Plant and Cell Physiology, 50(5), 911-923. Ártes, F., Gómez, P.A., Artés-Hernández, F. (2007). Physical, physiological, and microbial deterioration of minimally fresh processed fruits and vegetables. Food Science and Technology International, 13(3),177–188 Arthur, C.L., Killam, L.M., Buchholz, K.D., Pawliszyn, J. (1992). Automation and optimization of solid-phase microextraction. Analytical Chemistry, 64, 1960-1966 Astray, G., González-Barreiro, C. Mejuto, J.C., Rial-Otero, R., Simal-Gándara, J. (2009). A review on the use of cyclodextrins in foods. Food Hydrocolloids, 23, 1631-1640 Ayala-Zavala, J., Del-Toro,-Sánchez, L., Álvarez-Parrilla, E., González-Aguilar, G.A. (2008). High relative humidity in-package of fresh-cut fruits and vegetables: 173  Advantage or disadvantage considering microbiological problems and antimicrobial delivering systems? Journal of Food Science, 73(4), R41-47 Baert, K., Devlieghere, F., Bo, L., Debevere, J., De Meulenaer, B. (2008). The effect of inoculum size on the growth of Penicillium expansum in apples. Food Microbiology, 25, 212-217. Baert, K., Devlieghere, F., Flyps, H., Oosterlinck, M., Ahmed, M. M., Rajković, A., (2007a). Influence of storage conditions of apples on growth and patulin production by Penicillium expansum. International Journal of Food Microbiology, 119(3), 170-181. Baert, K., Valero, A., De Meulenaer, B., Samapundo, S., Ahmed, M.M., Bo, L., Debevere, J., Devlieghere, F. (2007b). Modelling the effect of temperatura on the growth rate and lag phase of Penicillium expansum in apples. International Journal of Food Microbiology, 118, 139-150. Baines, Z. V., & Morris, E. R. (1987). Flavour/taste perception in thickened systems: The effect of guar gum above and below c*. Food Hydrocolloids, 1(3), 197–205. Bakker, J. (1995). Flavour interactions with the food matrix and their effects on perception. In A. G. Gaonkar (Ed.), Ingredients interactions: Effects on Food Quality (pp. 411– 439). New York: Dekker. Baldwin, I.T. (2010). Plant volatiles. Current Biology, 20(9),R392-397. Bärlocher, F. Graca, M.A.S. (2005). Chapter 14: Total phenolics. In Graca, Bärlocher, & Gessner (Eds). Methods to Study Litter Decomposition: A Practical Guide. Netherlands: Springer. Barreiro, P., Ortiz, C., Ruiz-Altisent, M., De Smedt, V., Shotte, S., Andani, Z., Wakeling, I., Beyts, P.K. (1998) Comparison between sensory and instrumental measurements for mealiness assessment in apples. A collaborative test. Journal of Texture Studies. 29, 509-525. Beauchat, L., Ryu, J. (1997). Produce handling and processing practices. Emerging infectious diseases, 3, 459. Beltrán, D., Marín S., Gil, M.I. (2005). Ozonated water extends the shelf-life of fresh-cut lettuce. Journal of Agricultural and Food Chemistry, 53, 5654. Bennett, J.W., Klich, M. (2003). Mycotoxins. Clinical Microbiology Reviews, 16(3), 497-516 Bliss, R.M. (2003). Beating Back Blue Mold. Agricultural Research.,EBSCO Publishing. 51(8),17. Brijwani, K., Rigdon, A., Vadlani, P. V. (2010). Fungal Laccases: Production, Function, and Applications in Food Processing. Enzyme Research, Volume 2010. doi:10.4061/2010/149748 Calvo, J., Calvente, V., de Orellano, M. E., Benuzzi, D., & Sanz de Tosetti, M. I. (2007). Biological control of postharvest spoilage caused by Penicillium expansum and 174  Botrytis cinerea in apple by using the bacterium Rahnella aquatilis. International Journal of Food Microbiology, 113(3), 251-257. Cartwright, L.C. (1953). Solubility and volatility of propenyl guaethol, bourbonal, vanillin, and coumarin. Journal of Agricultural and Food Chemistry, 1(4), 312-314 Castro, R. Natera, R., Durán, E. (2008). Application of solid phase extraction techniques to analyse volatile compounds in wine and other enological products. European Food Research and Technology Journal, 228, 1-18 CDC. 2008. Guideline for Disinfection and Sterilization in Healthcare Facilities. Retrieved on Aug 15, 2011 from http://www.cdc.gov/hicpac/Disinfection_Sterilization/13_10otherSterilizationMethods.h tml Cerrutti, P., & Alzamora, S. M. (1996). Inhibitory effects of vanillin on some food spoilage yeasts in laboratory media and fruit purées. International Journal of Food Microbiology, 29(2-3), 379-386. Chang, Y. Y., & Cronan, J. E. Jr. (1999). Membrane cyclopropane fatty acid content as a major factor in acid resistance of Escherichia coli. Molecular Microbiology, 33, 249– 259. Chatjigakis, A.K., Donzé, C., Colemen, A.W. (1992).Solubility behavior of -cyclodextrin in water/cosolvent mixtures. Analytical Chemistry, 64, 1632-1634 Chattopadhyaya, S., Singhal, R. S., & Kulkarni, P. R. (1998). Oxidised starch as gum arabic substitute for encapsulation of flavours. Carbohydrate Polymers, 37(2), 143144. Chen, S.D., Ofoli, R.Y., Scott, E.P., Asmussen, J. (1993). Volatile retention in microwave freeze-dried model foods. Journal of Food Science, 58(5), 1157-1161 Chirife, J., Karel, M., Flink, J. (1973). Studies on mechanisms of retention of volatile in freeze-dried food models: the system PVP-n-propanol. Journal of Food Science, 38, 671-674 Chung, H.S., Toivonen, P.M.A., Moon, K,D. (2009). Effect of high vanillin treatment on storage quality of fresh-cut apples. Food Science and Biotechnology, 18(3), 636-649 Cole, M., Wood, R.K.S. (1961). Pectic enzymes and phenolic substances in apples rotted by fungi. Annals of Botany, 25(100), 435-452. Corbo M.R., Lanciotti, R., Gardini, F., Sinigaglia, M., Guerzoni, M.E. (2000). Effects of hexanal, trans-2-hexenal, and storage temperature on shelf sife of fresh sliced apples. Journal of Agricultural and Food Chemistry, 48(6), 2401-2408 Corte, F. V., Fabrizio, S. V., Salvatori, D. M., & Alzamora, S. M. (2004). Survival of Listeria innocua in apple juice as affected by vanillin or potassium sorbate. Journal of Food Safety, 24(1), 1-15.  175  Curme and Johnston. (1952). Glycols. Reinhold Publishing Corp. New York. Retrieved Oct 5, 2012 from Technical Data: Propylene Glycol, Boiling Point of Aqueous Propylene Glycol solutions. http://www.lyondellbasell.com/techlit/techlit/2519.pdf Del Valle, E. M. M. (2004). Cyclodextrins and their uses: A review. Process Biochemistry, 39(9), 1033-1046. De Souza Sant’Ana, A., Rosenthal, A., Rodriguez de Massaguer, P. (2008). The fate of patulin in apple juice processing: A review. Food Research International, 41, 441-453. Desai, K.G.H., Park H.J. (2005). Recent developments in microencapsulation of food ingredients. Drying Technology, 23(7), 1361-1394 Divakar, S. (1990). Structure of a -cyclodextrin-vanillin inclusion complex. Journal of Agriculture and Food Chemistry, 38(4), 940-944 Dixon, J., and Hewett, E.W. (2000). Factors affecting apple aroma/flavour volatile concentration: a review. New Zealand Journal of Crop and Horticultural Science, 28, 155-173. Durance, T. Sundaram, J. Yaghmaee, P., Burt, H., Wang, R. (2006). Technique for generation of porous biomaterials by application of radiant microwave energy under vacuum. Biomaterials Conference. 04/11/2006. Durance, T. & Yaghmaee, P. 2011. Microwave dehydration of food and food ingredients. In: Murray Moo-Young (Ed.), Comprehensive Biotechnology, 2nd Edition, Vol 4, pp. 617-628. Elsevier. Dyer, R.H., Martin, G.E. (1980). A comparison of the AOAC spectrophotometric method for vanillin and a GLC method for vanillin and ethyl vanillin in acoholic beverages. American Journal of Enology and Viticulture. 31(1), 37-39 Fan, L., Song, J., Beaudry, R.M., Hildebrand, P.D. (2006). Effect of hexanal vapor on spore viability of Penicillium expansum, lesion development on whole apples and fruit volatile biosynthesis. Journal of Food Science. 71, M105-109 FDA: US Food and Drug Administration. “EAFUS: Everything added to Food in the US”. Retrieved on May 11, 2010 from http://www.accessdata.fda.gov/scripts/fcn/fcnNavigation.cfm?rpt=eafusListing&display All=false&page=17 Fisher Scientific. (2012). Material Safety Data Sheet: Hexanal,96%. Retrieved on Oct 5, 2012 from http://fscimage.fishersci.com/msds/57717.htm Fitzgerald D.J. (2005). Structure-function analysis of the vanillin molecule and its antifungal properties. Journal of Agricultural and Food Chemistry, 53,1769-1775 Fitzgerald, D. J., Stratford, M., Gasson, M. J., Ueckert, J., Bos, A., & Narbad, A. (2004). Mode of antimicrobial action of vanillin against Escherichia coli, Lactobacillus plantarum and Listeria innocua. Journal of Applied Microbiology, 97(1), 104-113. 176  Flink, J., Karel, M. (1970). Effects of process variables on retention of volatiles in freezedrying. Journal of Food Science, 36, 444-447. Fonseca, S.C., Oliveira, F.A.R., Brecht, J.K. (2002). Modelling respiration rate of fresh fruits and vegetables for modified atmosphere packages: a review. Journal of Food Engineering, 52, 99-119 Gardini, F., Lanciotti, R., Caccioni, D.R.L., Guerzoni, M.E. (1997). Antifungal activity of hexanal as dependent on its vapour pressure. Journal of Agricultural and Food Chemistry, 45, 4297-4302. Gharsallaoui, A., Roudaut, G., Chambin, O., Voilley, A., & Saurel, R. (2007). Applications of spray-drying in microencapsulation of food ingredients: An overview. Food Research International, 40(9), 1107-1121. Gibbs, B.F., Kermasha, S., Alli, I., Mulligan, C.N. (1999). Encapsulation in the food industry: a review. International Journal of Food Sciences and Nutrition, 50, 213-224. Gil, M.I. , Gorny, J.R., Kader, A.A.(1998). Responses of “Fuji” apple slices to ascorbic acid treatments and low-oxygen atmospheres. HortScience, 33(2), 305-309 Giuffrida, F. Golay, P.A., Destaillats, F., Hug, B., Dionisi, F. (2005). Accurate determination of hexanal in beef buillons by headspace solid-phase microextraction gaschromatography mass-spectrometry. European Journal of Lipid Science and Technology, 107, 792-798. Godshall, M. A. (1997). How carbohydrates influence food flavour. Food Technology, 51(1), 63–67 Goff, S. A., & Klee, H. J. (2006). Plant volatile compounds: Sensory cues for health and nutritional value? Science, 311(5762), 815-819. Goubet, I. Le Quere, J.L., Voilley, A.J. (1998). Retention of aroma compounds by carbohydrates: influence of their physicochemical characteristics and their physical state: A review. Journal of Agricultural and Food Chemistry, 46, 1981-1990 Gouin, S. (2004). Microencapsulation: Industrial appraisal of existing technologies and trends. Trends in Food Science & Technology, 15(7-8), 330-347. Gumí, T., Gascón, S., Torras, C., & Garcia-Valls, R. (2009). Vanillin release from macrocapsules. Desalination,245(1-3), 769-775. Hatanaka, A. (1993). The biogeneration of green odour in green leaves. Phytochemistry, 34(5), 1201-1218 Hertel, M.O., Scheuren, H., Sommer, K. (2007). Solubilities of hexanal, benzaldehyde, 2furfural, 2- phenylethanol, phenylethanal, and -nonalactone in water at temperatures between (50-100) °C. Journal of Chemical and Engineering Data, 52, 2143-2145. [IRCHE] Iranian Chemical Engineers (2011). Antoine Coefficients for Vapor Pressure. Retrieved on Nov 20, 2011 from 177  http://www.irche.com/special/ANTOINE_COEFFICIENTS_FOR_VAPOR_PRESSUR E.pdf Jaya, S., Durance, T.D., Wang, R. (2009). Effect of alginate-pectin composition on drug release characteristics of microcapsules. Journal of Microencapsulation, 26(2), 143153 Kader, A. and Cantwell,M. 2005 “Rating Scale for Overall Visual Quality of Fresh-cut Produce” UC Davis Produce Quality Rating Scales and Color Charts. http://postharvest.ucdavis.edu/files/93651.pdf Kant, M.R., Bleeker, P.M., Van Wijk, M., Shuurink, R.C., Haring, M.A. (2009). Plant volatiles in defence. Advances in Botanical Research, 51, 613-666 Karathanos, V. T., Mourtzinos, I., Yannakopoulou, K., & Andrikopoulos, N. K. (2007). Study of the solubility, antioxidant activity and structure of inclusion complex of vanillin with β-cyclodextrin. Food Chemistry, 101(2), 652-658. Lanciotti, R., Gianotti, A., Patrignani, F., Belletti, N., Guerzoni, M. E., & Gardini, F. (2004). Use of natural aroma compounds to improve shelf-life and safety of minimally processed fruits. Trends in Food Science & Technology, 15(3-4), 201-208. Lanciotti, R., Belletti, N., Patrignani, F., Gianotti, A., Gardini, F., Guerzoni, M.E. (2003). Application of hexanal, (E)-2-hexenal, and hexyl acetate to improve the safety of fresh-sliced apples. Journal of Agricultural and Food Chemistry, 51, 2958-2963 Leepipattanawit R., Beaudry R.M., Hernandez R.J. (1997). Control of decay in modifiedatmosphere packages of sliced apples using 2-nonanone vapor. Journal of Food Science, 62, 1043–1047. Liu, L. (2007). Pectin in controlled drug delivery - a review. Cellulose, 14(1), 15. Lu, Z., Zhang, L., Lu, F. Bie, X. Yu, Z. (2007). Model of microbial growth on fresh-cut lettuce treated with chlorinated water during storage under different temperatures. Journal of Food Process Engineering, 30(1), 106-118. Madene, A. (2006). Flavour encapsulation and controlled release - a review. International Journal of Food Science and Technology, 41(1), 1-21 Matsui, K. (2006). Green leaf volátiles: hydroperoxide lyase pathway of oxylipin metabolism. Current Opinion in Plant Biology, 9, 274-280 McHugh, T.H., Senesi, E. (2000). Apple wraps: A novel method to improve the quality and extend the shelf-life of fresh-cut apples. Journal of Food Science, 65, 480-485. Mehinagic, E. , Royer, G., Symoneaux, R. , Bertrand, D., Jourjon, F., (2004). Prediction of the sensory quality of apples by physical measurements. Postharvest Biology and Technology, 34, 257-269 Meyer, S. (2011). “Just How Fresh Are McDonald’s New ‘Fresh-cut’ Happy Meal Apples?” Washington City Paper, Sept 19, 2011. Retrieved on Aug 1, 2012 from 178  http://www.washingtoncitypaper.com/blogs/youngandhungry/2011/09/19/just-howfresh-are-mcdonalds-new-fresh-cut-happy-meal-apples/ Miedes, E., Lorences, E. (2004). Apple (Malus domestica) and tomato (Lycopersicum esculentum) fruits cell-wall hemicelluloses and xyloglucan degradation during Penicillium expansum infection. Journal of Agricultural and Food Chemistry, 52, 79577963 Moon, K., Delaquis, P., Toivonen, P., & Stanich, K. (2006). Effect of vanillin on the fate of Listeria monocytogenes and Escherichia coli O157:H7 in a model apple juice medium and in apple juice. Food Microbiology,, 23(2), 169-174. Morales, H., Barros, G., Marín, S., Chulze, S., Ramos, A.J. Sanchis, V. (2008a). Effects of apple and pear varieties and pH on patulin accumulation by P. expansum. Journal of Science of Food and Agriculture, 88, 2738-2743. Morales, H., Sanchis, V., Coromines, J., Ramos, A.J., Marín, S. (2008b). Inoculum size and intraspecific interactions affects Penicillium expansum growth and patulin accumulation in apples. Food Microbiology, 25, 378-385 Morales, H., Marín, S., Ramos, A.J., Sanchis, V.(2010). Influence of post-harvest technologies applied during cold storage of apples in Penicillium expansum growth and patulin accumulation: A review. Food Control, 21, 953-962. Nakamura, S. Hatanaka, A. (2002). Green-leaf derived C6-aroma compounds with potent antibacterial action that act on both gram-negative and gram-positive bacteria. Journal of Agricultural and Food Chemistry, 50, 7639-7644 Neri ,F., Mari, M. and Brigati, S. (2006) Control of Penicillium expansum by plant volatile compounds. Plant Pathology, 55, 100–105 Nerín, C. Salafranca, J. Aznar, M. Batle, R. (2009). Critical review on recent developments in solventless techniques for extraction of analytes. Analytical and Bioanalytical Chemistry, 393, 809-833. Ngarmsak, M. (2007). Antifungal activity of vanillin on fresh-cut tropical fruits. Acta Horticulturae, 746, 409-416. North, H. (1949). Colorimetric determination of capsaicin in oleoresin of capsicum. Analytical Chemistry, 21, 934–936 Okull, D.O. and LaBorde, L.F. (2004). Activity of electrolyzed oxidizing water against Penicillium expansum in suspension and on wounded apples. Journal of Food Science, 69(1), 23-27 Olivas, G.I., Barbosa-Canovas, G.V. (2005). Edible Coatings for Fresh-cut fruits. Critical Reviews in Food Science and Nutrition, 45, 657-670 Partanen, R., Ahro, M., Hakala, M., Kallio, H., Forssell, P. (2002). Microencapsulation of caraway extract in beta-cyclodextrin and modified starches. European Food Research and Technology, 214(3), 242-247. 179  Peña, B., Casals, M., Torras, C., Gumi, T., García-Valls, R. (2009). Vanillin release from polysulfone macrocapsules. Industrial and Engineering Chemical Research, 48(3), 1562-1565 Poling, B.E. Prausnitz, J.M. O’Connell, J.P. (2001). The properties of gases and liquids. 5th Edition. App. A, New York: McGraw Hill. Retrieved Nov 20, 2011 from http://www.scribd.com/doc/41541675/Antoine-Constants Pothakamury, U.R., Barbosa-Cánovas, G.V. (1995) Fundamental aspects of controlled release in foods. Trends in Food Science and Technology, 6, 397-406. Prosser, J.I. (1993). Growth kinetics of mycelia colonies and aggregates of ascomycetes. Mycological Research, 97(5), 513-528 Queiroz, C., Mendes-Lopes, M.L., Fialho, E., Valente-Mesquita, V.L. (2008). Polyphenol oxidase: characteristics and mechanisms of browning control. Food Reviews International, 24, 361-375. Ragaert, P., Jacxsens, L., Vanderkinderen, I., Baert, L., Devlieghere, F., (2011) “Microbiological and safety aspects of fresh-cut fruits and vegetables. In MartínBelloso,O.M. and Soliva-Fortuny, R. (Eds.) Advances in Fresh-Cut Fruits and Vegetables Processing, Florida:CRC Press. Rahman, Shafiur. (1995). Food Properties Handbook. CRC Press Inc. Pg 203 Raybaudi-Massilia, R. M., Mosqueda-Melgar, J., Sobrino-López, A., Soliva-Fortuny, R., & Martín-Belloso, O. (2007). Shelf-life extension of fresh-cut “Fuji” apples at different ripeness stages using natural substances. Postharvest Biology and Technology, 45(2), 265-275. Reineccius, G.A. (2003) 29. Gas chromatography. In Nielsen, S.S. (Ed.) Food Analysis. Third Edition. USA: Springer Science and Media Inc. Risticevic, S., Niri, V.H., Vuckovic, D., Pawliszyn, J. (2009). Recent developments in solidphase microextraction. Analytical and Bioanalytical Chemistry, 393, 781-795 Rojas-Graü, M., Raybaudi-Massilia, R.M., Soliva-Fortuny, R.C., Avena-Bustillos, R.J, McHugh, T.H., Martín-Belloso, O. (2006). Apple puree-alginate edible coating as carrier of antimicrobial agents to prolong shelf-life of fresh-cut apples. Postharvest Biology and Technology, 45(2), 254. Rojas-Graü M., Grasa-Guillem R., Martin-Belloso, O. (2007). Quality changes in fresh-cut Fuji apple as affected by ripeness stage, antibrowning agents and storage atmosphere. Journal of Food Science, 72(1),S36-S43 Rojas-Graü M.A., Garner, E., Martín-Belloso, O. (2011). “The fresh-cut fruit and vegetables industry: current situation and market trends” In Martin-Belloso,O.M. and SolivaFortuny, R. (Eds.) Advances in Fresh-Cut Fruits and Vegetables Processing. Florida:CRC Press. Rolin, C., De Vries J. (1990). Chapter 10: Pectin. In Harris P. (Ed). Food Gels. Great Britain: Elsevier Science Publishers LTD 180  Rosenberg, M., Kopelman, I.J., Talmon, Y. (1985). A scanning electron microscopy study of microencapsulation. Journal of Food Science, 50, 139-144 Rounds, M.A. and Nielsen, S.S. (2003) Part V. Chromatography. IN Nielsen, S.S. (Ed.) Food Analysis. Third Edition. USA: Springer Science and Media Inc. Rupasinghe, H. P. V., Boulter-Bitzer, J., Ahn, T., & Odumeru, J. A. (2006). Vanillin inhibits pathogenic and spoilage microorganisms in vitro and aerobic microbial growth in fresh-cut apples. Food Research International, 39(5), 575-580. Sánchez-Ventura, S.E., Martínez Peniche, R.A., Castillo Tovar, J., Fdz Escartín, E. (2008). Antagonismo de las Levaduras Nativas contra la Pudrición Azul (Penicillium expansum link) en frutos de manzana. Revista Fitotecnia Mexicana. Sociedad Mexicana de Fitogenética, A.C, 31(004), 359-366. Sander, R. (1999). Compilation of Henry’s law constants for inorganic and organic species of potential importance in environmental chemistry. Max-Planck Institute of Chemistry. Version 3. Retrieved on Aug 5, 2012 from http://www.rolfsander.net/henry/henry.pdf Sanzani, S.M., De Girolamo, A., Schena, L., Solfrizzo, M., Ippolito, A., Visconti, A. (2009). Control of Penicillium expansum and patulin accumulation on apples by quercetin and umbelliferone. European Food Research and Technology, 228, 381-389. Scaman, C.H. and Durance, T.D. 2005. Combined microwave vacuum drying. Chapter 19 IN "Emerging Technologies for Food Processing." p.p. 507 - 530. Elsevier: London Secouard, S., Malhiac, C., Grisel, M., & Decroix, B. (2003). Release of limonene from polysaccharide matrices: Viscosity and synergy effect. Food Chemistry, 82(2), 227– 234. Shirey, R. E. (2012). 4 - SPME commercial devices and fibre coatings. In Janusz Pawliszyn (Ed.), Handbook of solid phase microextraction (pp. 99-133). Oxford: Elsevier. Shirey, Robert E., & Mindrup, Raymond F. (1999). SUPELCO: SPME Adsorption vs Absorption: Which Fiber is Best for your Application. Retrieved on Aug 20, 2011 from http://www.youngwha.com/tech/upload/T400011_fiber%20selection%20guide_ppt.pdf Sholberg, P.L. & Randall, P. (2007).Fumigation of Stored Pome Fruit with Hexanal Reduces Blue and Gray Mold Decay. HortScience, 42(3), 611-616. Singleton, V.I. & Rossi, J.A Jr. (1965) Colorimetry of total phenolics with phosphomolybdicphosphotungstic acid reagents. Phenolics determination. Journal of Enology and Viticulture, 16(3),144-158 Soliva-Fortuny, R. C., Elez-Martínez, P., & Martín-Belloso, O. (2004). Microbiological and biochemical stability of fresh-cut apples preserved by modified atmosphere packaging. Innovative Food Science & Emerging Technologies,,5(2), 215-224.  181  Soliva-Fortuny, R. C., & Martín-Belloso, O. (2003a). New advances in extending the shelflife of fresh-cut fruits: A review. Trends in Food Science & Technology, 14(9), 341353. Soliva-Fortuny, R.C., Lluch, M.A. Quiles, A., Grigelmo-Miguel, N. Martín-Belloso, O. (2003b) Evaluation of textural properties and microstructure during storage of minimally processed apples. Journal of Food Science, 68(1) 312-317. Song, J., Leepipattwit, R., Deng, W., Beaudry, R. (1998). Hexanal vapor acts as residueless antifungal agent that enhances aroma biosynthesis in apple fruit. Acta Horitculturae, 464, 219-224. Song, J., Gardner, B.D., Holland, J.F., Beaudry, R.M. (1997). Rapid analysis of volatile flavor compounds in apple fruit using SPME and GC/time-of-flight mass spectrometry. Journal of Agricultural and Food Chemistry, 1997, 1801-1807. Spafiu, F., Mischie, A., Ionita, P., Beteringhe, A., Constantinescu, T., Balaban, A.T., (2009). New alternatives for estimating the octanol/wáter partition coefficient and wáter solubility for volatile organic compounds using GLC data (Kovats retention índices). Arkivoc. 174-194. Retrieved on Aug 5, 2012 from http://www.arkat-usa.org/getfile/28920/ Spotts RA, Cervantes LA (1993). Filtration to remove spores of Penicillium expansum from water in pome fruit packinghouses. Tree Fruit Postharvest Journal, 4, 16–18 Stalin, T., Rajendiran, N. (2006). A study on the spectroscopy and photophysics of 4hydroxy-3-methoxybenzoic acid in different solvents, pH and -cyclodextrin. Journal of Molecular Structure, 794, 35-45 Sun Rich Fresh Foods Inc. (2012). “Products: Fresh-cut Fruit” Retrieved on Aug 14, 2012 from http://www.sun-rich.com/products/ Supelco (2012) “Tech tip: Did you know… Solid Phase Microextraction”. Retrieved on Aug 5, 2012 from http://www.supelco.com.tw/C-1%20%20%20349-359.pdf Szejtli, J. (1998). Introduction and general overview of cyclodextrin chemistry. Chemical Reviews, 98, 1743-1753. Szente, L., & Szejtli, J. (2004). Cyclodextrins as food ingredients. Trends in Food Science & Technology, 15(3-4), 137-142. Tari, T. A., Annapure, U. S., Singhal, R. S., & Kulkarni, P. R. (2003). Starch-based spherical aggregates: Screening of small granule sized starches for entrapment of a model flavouring compound, vanillin. Carbohydrate Polymers, 53(1), 45-51. Tari, T. A., & Singhal, R. S. (2002). Starch-based spherical aggregates: Stability of a model flavouring compound, vanillin entrapped therein. Carbohydrate Polymers, 50(4), 417421. Terta, M., Blekas, G., & Paraskevopoulou, A. (2006). Retention of selected aroma compounds by polysaccharide solutions: A thermodynamic and kinetic approach. Food Hydrocolloids, 20(6), 863-871 182  Thavong, P., Archbold, D.D., Pankasemsuk, T., Koslanund, R. (2010). Effect of hexanal vapour on longan fruit decay, quality, and phenolic metabolism during cold storage. International Journal of Food Science and Technology, 45, 2313-2320 Thies, C. (1996). A survey of microencapsulation processes. In Benita, S. (Ed). Microencapsulation: Methods and Industrial Applications. New York: Marcel Dekker Inc. Thostenson, E.T., Chou, T.W. (1999). Microwave processing: fundamentals and applications. Composites: Part A, 30, 1055-1071 Toivonen, P.M.A, Wiersma, P.A., Hampson, C., Lannard, B. (2010). Effect of short-term air storage after removal from controlled-atmosphere storage on apple and fresh-cut apple quality. Journal of the Science of Food and Agriculture, 90(4), 580-585. Toivonen, P, M, A., Brummel, D.A. (2008) Biochemical bases of appearance and texture changes in fresh-cut fruit and vegetables. Postharvest Biology and Technology. 48, 114 Tsami, E. Krokida, M.K., Drouzas, A.E. (1999). Effect of drying method on the sorption characteristics of model fruit powders. Journal of Food Engineering, 38, 381-392. Valentines, M.C., Vilaplana, R., Torres, R., Usall, J. , Larrigaudiere, C. (2005). Specific roles of enzymatic browning and lignification in apple disease resistance. Postharvest Biology and Technology, 36, 227-234. Vargas, M., Pastor, C., Chiralt, A., McClements, D. J. and González-Martínez, C. (2008) Recent advances in edible coatings for fresh and minimally processed fruits. Critical Reviews in Food Science and Nutrition, 48(6),496-511 Walton, N.J., Mayer, M.J., Narbad, A. (2003). Vanillin. Phytochemistry, 63, 505-515 Willaert, G.A., Dirinck, P. J., De Pooter, H.L., Schamp, N.N. (1983). Objective measurement of aroma quality of golden delicious apples as a function of controlled atmospheric storage time. Journal of Agricultural and Food Chemistry, 31(4), 809-813 Zhang, M., Tang, J., Mujumdar, A.S., Wang, S. (2006). Trends in microwave-related drying of fruits and vegetables. Trends in Food Science and Technology, 17, 524-534 Zhang, Z., Pawliszyn, J. (1993). Headspace solid-phase microextraction. Analytical Chemistry, 65, 1843-1852  183  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0073556/manifest

Comment

Related Items