@prefix vivo: . @prefix edm: . @prefix ns0: . @prefix dcterms: . @prefix skos: . vivo:departmentOrSchool "Applied Science, Faculty of"@en, "Chemical and Biological Engineering, Department of"@en ; edm:dataProvider "DSpace"@en ; ns0:degreeCampus "UBCV"@en ; dcterms:creator "Yuan, Zhaoyang"@en ; dcterms:issued "2017-04-05T22:28:36Z"@en, "2017"@en ; vivo:relatedDegree "Doctor of Philosophy - PhD"@en ; ns0:degreeGrantor "University of British Columbia"@en ; dcterms:description """In this work, the hydrothermal pretreatment under both acidic and alkaline conditions were conducted to study hemicellulose and silica removal from bamboo. In the first part of this work, evolution of proton concentration was examined during both auto- and dilute-acid hydrolysis of hemicellulose from green bamboo. An approximate mathematical model (toy model) to describe the proton concentration based upon conservation of mass and charge during deacetylation and ash neutralization coupled with a number of competing equilibria, was derived. This model was qualitatively compared to experiments where pH was measured as a function of time, temperature, and initial acid level. The toy model predicts the existence of a steady state proton concentration dictated by equilibrium constants, initial acetyl groups, and initial added acid. At room temperature, it was found that pH remains essentially constant both at low initial pH and autohydrolysis conditions. At elevated temperatures, one case of non-monotonic behaviour in which the pH initially increased, and then decreased at longer times, was found. As silica in bamboo creates processing problems, in the second part of this work, alkaline pretreatment of pure amorphous silica particles, bamboo powder and bamboo chips was carried out to study the underlying mechanism for silica and hemicellulose extraction. Response surface methodology was also used to optimize the treatment conditions that could completely extract silica and partially extract hemicellulose from bamboo chips prior to processing. Alkaline pretreatment resulted in significant improvement in the delignification of treated bamboo chips during subsequent kraft pulping, offering an option to reduce the effective alkali charge or the H-factor. The pre-extracted silica and hemicellulose in the liquor were recovered through a sequential procedure of CO2 and ethanol precipitation. Moreover, the feasibility of adopting alkaline pretreatment to the typical kraft pulping process was tested. Results demonstrated that all silica, and up to 50% of hemicellulose in raw feedstocks, could be extracted without degrading cellulose and lignin. Approximately 96% of extracted silica in the APEL could be recovered as a high purity (>99.8%) silica nanoparticles. These results demonstrated that the proposed modification may benefit kraft pulping and fit well into the proposed biorefinery concept."""@en ; edm:aggregatedCHO "https://circle.library.ubc.ca/rest/handle/2429/61124?expand=metadata"@en ; skos:note "Understanding Hemicellulose andSilica Removal from BamboobyZhaoyang YuanM.Sc., Tianjin University of Science and Technology, 2012A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinThe Faculty of Graduate and Postdoctoral Studies(Chemical and Biological Engineering)THE UNIVERSITY OF BRITISH COLUMBIA(Vancouver)April 2017c© Zhaoyang Yuan 2017AbstractIn this work, the hydrothermal pretreatment under both acidic and alkaline conditionswere conducted to study hemicellulose and silica removal from bamboo.In the first part of this work, evolution of proton concentration was examinedduring both auto- and dilute-acid hydrolysis of hemicellulose from green bamboo. Anapproximate mathematical model (toy model) to describe the proton concentrationbased upon conservation of mass and charge during deacetylation and ash neutral-ization coupled with a number of competing equilibria, was derived. This modelwas qualitatively compared to experiments where pH was measured as a function oftime, temperature, and initial acid level. The toy model predicts the existence ofa steady state proton concentration dictated by equilibrium constants, initial acetylgroups, and initial added acid. At room temperature, it was found that pH remainsessentially constant both at low initial pH and autohydrolysis conditions. At ele-vated temperatures, one case of non-monotonic behaviour in which the pH initiallyincreased, and then decreased at longer times, was found.As silica in bamboo creates processing problems, in the second part of this work,alkaline pretreatment of pure amorphous silica particles, bamboo powder and bamboochips was carried out to study the underlying mechanism for silica and hemicelluloseextraction. Response surface methodology was also used to optimize the treatmentiiAbstractconditions that could completely extract silica and partially extract hemicellulosefrom bamboo chips prior to processing. Alkaline pretreatment resulted in significantimprovement in the delignification of treated bamboo chips during subsequent kraftpulping, offering an option to reduce the effective alkali charge or the H-factor. Thepre-extracted silica and hemicellulose in the liquor were recovered through a sequen-tial procedure of CO2 and ethanol precipitation. Moreover, the feasibility of adoptingalkaline pretreatment to the typical kraft pulping process was tested. Results demon-strated that all silica, and up to 50% of hemicellulose in raw feedstocks, could beextracted without degrading cellulose and lignin. Approximately 96% of extractedsilica in the APEL could be recovered as a high purity (>99.8%) silica nanoparticles.These results demonstrated that the proposed modification may benefit kraft pulpingand fit well into the proposed biorefinery concept.iiiPrefaceIn this work, I was responsible for experimental design, experimental procedures anddata analysis. Dr. Martinez and Dr. Beatson supervised the research and providedfeedback and reviewed the manuscript. The work for the proton evolution duringauto- and dilute acid pre-hydrolysis of bamboo chips was co-performed with Dr.Nuwan Sella Kapu, a research associate in our laboratory. A list of journal andconference contributions is given below.1. Journal Papers(a) Zhaoyang Yuan, Nuwan S. Kapu, Rodger Beatson, Xue Feng Chang, D.Mark Martinez. Effect of alkaline pre-extraction of hemicellulose and silicaon kraft pulping of bamboo (Neosinocalamus affinis Keng). IndustrialCrops and Products, 2016, 91, 66-75.This publication presents a version of Chapters 3 and 4.(b) Nuwan S. Kapu, Zhaoyang Yuan, Xue Feng Chang, Rodger Beatson, D.Mark Martinez, Heather Trajano. Insight into the evolution of the pro-ton concentration during auto- and dilute acid hydrolysis of hemciellulose.Biotechnology for Biofuels, 2016, 9 (224), 1-10.This publication presents a version of Chapter 2.ivPreface(c) Lingfeng Zhao, Zhaoyang Yuan, Rodger Beatson, Xue Feng Chang,Nuwan S. Kapu, Heather L. Trajano, D. Mark Martinez. Increasing ef-ficiency of enzymatic hemicellulose removal from bamboo for productionof high-grade dissolving pulp. Bioresource Technology, 2017, 223, 40-46.The experiments in the paper were conducted by Lingfeng Zhao and mewith advisement from other co-authors.(d) Zhaoyang Yuan, Rodger Beatson, Xue Feng Chang, Nuwan S. Kapu,D. Mark Martinez. An eco-friendly scheme to eliminate silica problemsduring bamboo biomass fractionation. Nordic Pulp & Paper ResearchJournal, 2017, 32 (1), 4-13.This publication presents a version of Chapter 4.(e) Zhaoyang Yuan, Yangbing Wen, Nuwan S. Kapu, Rodger Beatson, D.Mark Martinez. A biorefinery scheme to fractionate bamboo into highgrade dissolving pulp and ethanol. Biotechnology for Biofuels, 2017, 10,38.The experiments in this paper were conducted by Dr. Wen and me withadvisement from other co-authors. This publication presents a version ofChapter 3 and further studies.(f) Zhaoyang Yuan, Yangbing Wen. Evaluation of an integrated process tofully utilize bamboo biomass during the production of bioethanol. Biore-source Technology (Accepted).This publication presents further investigations according to the experi-mental results obtained in this work.vPreface(g) Zhaoyang Yuan, Nuwan S. Kapu, Rodger Beatson, Xue Feng Chang,Heather L. Trajano, D. Mark Martinez. Insight into the understanding ofthe mechanism of alkaline pretreatment of bamboo biomass. This paper isin preparation and to be submitted to the peer reviewed journal.This publication presents the data shown in Chapters 3 and 4.2. Book Chapter(a) Jingqian Chen, Zhaoyang Yuan, Elisa Zanuso, Heather L. Trajano. Hy-drothermal Processing in Biorefineries-Production of Bioethanol and HighAdded-Value Compounds of Second and Third Generation Biomass. Springer(In press).In the preparation of this book chapter, I was responsible for writing thepart of hydrothermal pretreatment of bamboo and agricultural residuesand helped to integrate different parts. Part of this book chapter is pre-sented in Chapters 2 and 3.3. Conference Presentations(a) Zhaoyang Yuan, Nuwan S. Kapu, Xue Feng Chang, Rodger Beatson,D. Mark Martinez. Effect of alkaline pre-extraction of silica and hemi-celluloses on the kraft pulping of bamboo. PACwest Conference − 2016,Jasper, Canada, June, 2016. (Presentation).(b) Zhaoyang Yuan, Nuwan S. Kapu, Xue Feng Chang, Rodger Beatson,D. Mark Martinez. Removal of silica from bamboo for biorefinery appli-cations. 5th International Conference of Biorefinery-Towards Bioenergy,viPrefaceVancouver, Canada, August, 2015. (Presentation).(c) Zhaoyang Yuan, Nuwan S. Kapu, Rodger Beatson, D. Mark Martinez.Developing bamboo as an alternative feedstock for biorefinery applications:Solving the silica problem. PaperWeek Canada 2015-Biorefinery Session,Montreal, Canada, February, 2015. (Poster).(d) Nuwan S. Kapu, Zhaoyang Yuan, Xue Feng Chang, Lingfeng Zhao,Colby Song, Joel Kumlin, David Houghton, James Olson, D. Mark Mar-tinez, Rodger Beatson. Developing bamboo as an alternative feedstock forbio-products. Research Day, Department of Chemical & Biological Engi-neering, UBC, 2013. (Poster).viiTable of ContentsAbstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iiPreface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ivTable of Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viiiList of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiiList of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xivNomenclature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxAcknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxiii1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Evolution of the Proton Concentration During Auto- and DiluteAcid Hydrolysis of Hemicellulose . . . . . . . . . . . . . . . . . . . . . 52.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52.2 Model Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132.3 Materials and Methods . . . . . . . . . . . . . . . . . . . . . . . . . . 202.4 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . 25viiiTable of Contents2.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333 Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo 363.1 Dissolution of Pure Amorphous Silica in Sodium Hydroxide Solution . 373.1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 373.1.2 Materials and Methods . . . . . . . . . . . . . . . . . . . . . . 423.1.3 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . 443.1.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543.2 Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo Powder 543.2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543.2.2 Materials and Methods . . . . . . . . . . . . . . . . . . . . . . 613.2.3 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . 633.2.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 843.3 Alkaline Pre-extraction of silica and Hemicellulose from Bamboo Chips 843.3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 843.3.2 Materials and Methods . . . . . . . . . . . . . . . . . . . . . . 853.3.3 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . 863.3.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 933.4 Response Surface Experimental Design on Alkaline Pre-extraction ofBamboo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 943.4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 943.4.2 Materials and Methods . . . . . . . . . . . . . . . . . . . . . . 953.4.3 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . 983.4.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108ixTable of Contents3.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1094 Feasibility of Using Alkaline Pre-extraction in Kraft Pulping . . . 1114.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1114.2 Materials and Methods . . . . . . . . . . . . . . . . . . . . . . . . . . 1144.2.1 Raw Material . . . . . . . . . . . . . . . . . . . . . . . . . . . 1144.2.2 Kraft Pulping . . . . . . . . . . . . . . . . . . . . . . . . . . . 1154.2.3 Evaluation of Pulps . . . . . . . . . . . . . . . . . . . . . . . . 1164.2.4 Preparation of Silica Particles from the APEL . . . . . . . . . 1164.2.5 Isolation of Hemicellulose and Preparation of Hemicellulose-based Polymeric Film . . . . . . . . . . . . . . . . . . . . . . . 1184.2.6 Analytical Methods . . . . . . . . . . . . . . . . . . . . . . . . 1194.3 Results and Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . 1214.3.1 Kraft Pulping . . . . . . . . . . . . . . . . . . . . . . . . . . . 1214.3.2 Pulp Physical Properties . . . . . . . . . . . . . . . . . . . . . 1254.3.3 APEL Properties . . . . . . . . . . . . . . . . . . . . . . . . . 1304.3.4 Isolation of Silica from APEL with Carbon Dioxide . . . . . . 1314.3.5 Capture of CO2 from the CO2/N2 Mixture . . . . . . . . . . . 1354.3.6 Compositional and FTIR Analysis of Silica Powders . . . . . . 1384.3.7 Precipitation, Characterization and Utilization of Hemicellulosefrom the Treated APEL . . . . . . . . . . . . . . . . . . . . . 1404.3.8 Proposed Modification to Kraft Pulping Process and Mass Bal-ance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1424.4 Process Engineering . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147xTable of Contents4.4.1 Process Description . . . . . . . . . . . . . . . . . . . . . . . . 1474.4.2 Mass Balance of Pulp Line and Chemical Recovery Process . . 1504.5 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1525 Summary of Thesis and Recommendations for Future Research . 1545.1 Summary of Contributions . . . . . . . . . . . . . . . . . . . . . . . . 1545.2 Recommendations for Future Work . . . . . . . . . . . . . . . . . . . 159References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160Appendices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188A1 Bamboo Raw Material Characterization . . . . . . . . . . . . . . . . 188A2 Chip Washing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195A3 Mass Balance of the Pulping and Chemical Recovery Process . . . . . 200A3.1 Mass Balance of the Fibre Line . . . . . . . . . . . . . . . . . 200A3.2 Mass Balance of the Chemical Recovery Process . . . . . . . . 213xiList of Tables2.1 Chemical composition of bamboo, hardwood and softwood (Huang andRagauskas (2013); Saka (2004); Dence (1992); Li et al. (2012); Songet al. (2013); Torelli and Čufar (1995)). . . . . . . . . . . . . . . . . . 72.2 Composition of the bamboo chips used in this study. The values re-ported in this table are based upon the total mass of the bamboo chips. 222.3 A summary of the experimental conditions tested. . . . . . . . . . . . 243.1 A summary of the experimental conditions tested for the dissolutionof pure amorphous silica. . . . . . . . . . . . . . . . . . . . . . . . . . 433.2 Reaction rate constant (k2) calculated by experimental data fitting . 513.3 A summary of experimental conditions investigated for alkaline treat-ment of bamboo powder. . . . . . . . . . . . . . . . . . . . . . . . . . 623.4 Chemical composition of bamboo powder after pre-extraction withNaOH at 100 oC. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 773.5 Composition of hydrolysates (based on initial o.d. bamboo powder)obtained from alkaline treatment at 100 oC with NaOH concentrationof 0.45 mol/L. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83xiiList of Tables3.6 Experimental range of pre-extraction variables and coded levels ac-cording to response surface methodology. . . . . . . . . . . . . . . . . 963.7 Experimental design and observed responses of the dependent variables. 993.8 Values of regression coefficients of the fitted second order polynomials. 1053.9 Experimental design and observed responses of the dependent variables.1073.10 Confirmation runs of the alkaline pre-extraction of bamboo chips ac-cording to RSM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1084.1 Effect of alkaline pre-extraction and effective alkali charge on kraftpulping of bamboo. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1234.2 Chemical composition of alkaline pre-extraction liquor (APEL). . . . 1324.3 Gas phase CO2 measurements during silica precipitation. . . . . . . . 1364.4 Mineral composition of silica powder prepared from the APEL. . . . . 1384.5 Mass balance of the pulp line. . . . . . . . . . . . . . . . . . . . . . . 1504.6 Mass balance of the chemical recovery process. . . . . . . . . . . . . . 151A.1 Chemical composition of 5 year old original bamboo stem (Neosinocala-mus Affinis Keng). . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194xiiiList of Figures2.1 A simplified, hypothertical model of bamboo/grass secondary cell wallillustrating molecular interactions of hemicellulose with cellulose andlignin. This figure is reproduced from Sella Kapu and Trajano (2014). 92.2 Hydrolysis of hemicellulose during acidic pre-hydrolysis process. . . . 102.3 A schematic of the idealized hemicellulose - lignocellulose (LC) sub-strate considered in this work. Although xylan is hypothesized to becomprised of fast and a slow reacting fractions, we do not distinguishthese in this figure. The species Ac and Ar, which represent the acetyland arabinose groups, are initially bound to the xylan chain but arereleased through acid hydrolysis. The ash (MO) is not shown in thisfigure but is considered to be physically embedded in the LC portionof the matrix. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152.4 A schematic of the idealized reaction scheme. The chemical reactionsshown form the basis of the toy mathematical model of the protonconcentration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16xivList of Figures2.5 Temperature evolution for two representative cases. Both thermocou-ple signals are presented and replicate runs are shown but the differencebetween them is not perceptible on this scale. The thermocouples areplaced at the same elevation in the reactor but at two different radialpositions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262.6 The temperature evolution during the heat up period for Series 1-10.The results have been scaled using the form advanced in Equation 2.18with h/c set to be 0.15 min−1. . . . . . . . . . . . . . . . . . . . . . . 272.7 The effect of initial pH on the steady state pH measured after longterm pre-hydrolysis. The dashed line represents the toy model withthe equilibrium constant given in the text previously. The value ofKm is not stated in the text and is taken to be small. For practicalpurposes we set Km = 10−17M for this calculation. The two remainingparameters are set to be [XOAc]o = 0.025M and [MO]o = 0.001M anddetermined through regression. Although defined previously, Series 11represents the reaction at 160 oC and Series 12 at 180 oC. . . . . . . 292.8 Examination of the pH when the reaction proceeds at room temper-ature. To evaluate the toy model, the same values for the constantgiven in the caption of the previous Figure are used. In addition, weset k2 = 10M−1min−1 as determined by regression. . . . . . . . . . . 302.9 A evolution of pH under prehydrolysis conditions. To help understandthis we related the lowest initial pH experiments to the added acid:Series 17 0.25% (w/w) and Series 18 0.5 % (w/w). . . . . . . . . . . . 32xvList of Figures3.1 Basic structural unit of silica . . . . . . . . . . . . . . . . . . . . . . . 393.2 Effect of particle size on the dissolution rate of silica (temperature =70 oC and NaOH = 0.45 mol/L). . . . . . . . . . . . . . . . . . . . . 453.3 Effect of NaOH concentration on the dissolution rate of pure amor-phous silica (temperature = 70 oC and silica particle size = 250-400µm). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 473.4 Plot of 1 − α1/3 versus reaction time (t) with different silica particlesizes (temperature=70 oC; [OH−]o = 0.45 mol/L). . . . . . . . . . . . 493.5 Plot of 1 − α 13 versus reaction time (t) with different initial NaOHconcentrations (0.15-0.45 mol/L). Experiments were carried out withinitial silica particle size of 250-400 mum and at 70 oC. . . . . . . . . 503.6 Comparison between the experimental data and the predictions byEquation 3.15 for OH– concentration. The parameter is set m = 1 forthe calculation. Data points are from experimental results. . . . . . 523.7 Particle size distribution of the solid particles at different reaction times(initial silica particle size = 250-400 µm, temperature = 70 oC, initial[OH– ] = 0.15 mol/L). . . . . . . . . . . . . . . . . . . . . . . . . . . 533.8 Effect of NaOH concentration on the extraction of silica from bamboopowder at 70 oC. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 653.9 Application of shrinking core model on silica removal from bamboopowder (temperature = 70 oC). . . . . . . . . . . . . . . . . . . . . . 66xviList of Figures3.10 A schematic of the idealized hemicellulose - lignocellulose (LC) sub-strate considered in this work. X-X-X-X represents the backbone ofthe xylan structure. The species Ac, Ar and G, which represent theacetyl, arabinose and glucuronic acid groups, are initially bound tothe xylan chain but are released through alkaline treatment. Silicaand other ash components (MO) are not shown in this figure but areconsidered to be physically embedded in the LC portion of the matrix. 673.11 A schematic of the idealized reaction scheme. The chemical reactionsshown present the dominant reaction mechanisms during alkaline treat-ment of bamboo biomass. . . . . . . . . . . . . . . . . . . . . . . . . 733.12 Experimental yields of xylan remaining in the extracted milled bamboopowder at alkaline pre-extraction temperatures of 70-100 oC (initialNaOH concentration: a = 0.15 mol/L; b = 0.45 mol/L). . . . . . . . 793.13 Experimental yields of silica remaining in the extracted milled bamboopowder at alkaline pre-extraction temperatures of 70-100 oC (initialNaOH concentration: a = 0.15 mol/L; b = 0.45 mol/L). . . . . . . . 803.14 NaOH concentration during alkaline pre-extraction of bamboo powderat temperatures of 70-100 oC (initial NaOH concentration: a = 0.15mol/L; b = 0.45 mol/L). . . . . . . . . . . . . . . . . . . . . . . . . . 813.15 Experimental yields of xylan remaining in the extracted bamboo chipsat alkaline pre-extraction temperatures of 70-100 oC (initial NaOHconcentration: a = 0.15 mol/L; b = 0.45 mol/L). . . . . . . . . . . . 87xviiList of Figures3.16 Experimental yields of silica remaining in the extracted bamboo chipsat alkaline pre-extraction temperatures of 70-100 oC (initial NaOHconcentration: a = 0.15 mol/L; b = 0.45 mol/L). . . . . . . . . . . . 893.17 The effect of mass transfer on consumed NaOH during alkaline pre-extraction of bamboo (alkaline pre-extraction temperature = 100 oC). 913.18 The effect of mass transfer on xylose yields during alkaline pre-extractionof bamboo (alkaline pre-extraction was carried out at 100 oC with ini-tial NaOH concentration of 0.45 mol/L). . . . . . . . . . . . . . . . . 923.19 Effect of temperature and NaOH charge on the extraction of xylan ata fixed reaction time of 80 min. . . . . . . . . . . . . . . . . . . . . . 1013.20 Effect of temperature and NaOH charge on the extraction of silica ata fixed reaction time of 80 min. . . . . . . . . . . . . . . . . . . . . . 1033.21 Effect of temperature and NaOH charge on chip yield at a fixed reactiontime of 80 min. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1064.1 A schematic of experimental set-up for recovering silica and hemicel-lulose and reducing CO2 emissions. . . . . . . . . . . . . . . . . . . . 1184.2 Plot of pulp freeness (CSF) versus PFI mill revolution for pulps ob-tained from kraft pulping of extracted and non-extracted chips with19% EA. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1264.3 Effect of pre-extraction and effective alkali charge on the tensile indexof pulp. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1274.4 Effect of pre-extraction and effective alkali charge on the tensile indexof pulp. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128xviiiList of Figures4.5 Effect of temperature and pH on the precipitation of silica from theAPEL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1344.6 CO2 adsorption measurement curve for the precipitation of silica fromthe APEL at 60 oC. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1374.7 Fourier transform infrared (FTIR) spectra of silica produced from theAPEL of bamboo (a: before burning; b: after burning at 700 oC). . . 1404.8 Fourier transform infrared (FTIR) spectra of hemicellulose producedfrom the APEL. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1414.9 Proposed process of kraft pulping with extraction and recovery unitsof silica and hemicellulose from bamboo chips. . . . . . . . . . . . . . 1444.10 Combination of the new stages in a typical kraft bamboo pulp mill. . 148A.1 Ash and silica content in different height locations of the bamboo stem(a: ash content, b: silica content). The measurements were triplicated. 191A.2 Silica removal from bamboo chips by washing with water at two tem-peratures (a: continuous washing process, b: batch washing process).Note: batch process means wash water was replaced with fresh deion-ized water at 10 min intervals. . . . . . . . . . . . . . . . . . . . . . . 197A.3 Pilot-scale experiments on silica removal by washing with water. . . . 198xixNomenclatureA Reactant surface area, m2Ac Acetyl groupAcOH Acetic acidAPEL Alkaline pre-extraction liquoraq Aqueous phaseAr ArabinoseBL Black liquorC Cellulose polymerCa(OH2) Calcium hydroxideCCRD Central composite rotatable designCO2 Carbon dioxideCSF Canadian Standard Freeness, mLCTH Constant temperature (23 ± 0.5oC and humidity (50%)DCOOH D-gtucopyranosyluronic acidEA Effective alkaliECOOH Gluconic acidFTIR Fourier transform infrared spectraG Glucuronic acidxxNomenclatureGC Gas chromatographyH+ ProtonHCl Hydrochloric acidHMF HydroxymethylfurfuralH2O2 Hydrogen peroxideHPLC High performance liquid chromatographyH2SiO3 Silicic acidIFBR Integrated forest biorefineryISO Brightness unitKOH Potassium hydroxideLC LignocellulosicL:W Liquid-to-wood ratio (L/kg)Mn Number average molecular weight, g/molMO Ash in bambooMw Weight average molecular weight, g/molNaOH Sodium hydroxideNa2S Sodium sulfideNa2SiO3 Sodium silicateND Not detectedNH4OH Ammonia hydroxideNREL National Renewable Energy LaboratoryOH– Hydroxide iono.d. Oven driedxxiNomenclaturePDI Polydispersity indexr Reaction rater2 Regression coefficientR Radius of silica particle, mRSM Response surface methodologys Solid phaseT Temperature, oCt Time, minw/w Weight by weightWVP Water vapour permeabilityWVTR Water vapour transfer rateX Xylose unitki Reaction rate constantKi Equilibrium constant, unitlessα Fraction of residual silica, unitlessβ Reaction order the hydroxide ion concentration, unitlessρ Density of silica, kg/m3xxiiAcknowledgementsIn the first place, I would like to express my sincerest gratitude to my supervisors,Prof. D. Mark Martinez and Prof. Rodger P. Beatson for their invaluable guidanceand support throughout my Ph.D. study. Above all and the most needed, theyprovided me support and friendly help in various ways. I deeply grateful to Dr.Nuwan Sella Kapu for his guidance, suggestions and valuable discussions throughoutthe course of this research.In addition, I am deeply thankful to Xue Feng Chang for his helpful guidance,discussions and critical comments. My special thanks go to George Soong for hisassistance with all types of technical problems, Lingfeng Zhao for friendship andenlarging my vision about pulp and paper.I am grateful to my committee members, Dr. Heather L. Trajano, Dr. RichardChandra and Dr. Valdeir Arantes for their great comments and input about myresearch. I offer my heartfelt thanks to my departed committee member, Prof. CarlDouglas, for his comments and suggestions through the initial phase of my research.Specially thanks are given to my family and friends for their constant love andsupport. At finally, I would like to specially thank Chunyu for her constant supportand always being by my side.xxiiiChapter 1IntroductionConcerns over climate change have driven the development of a bio-based economyin which energy and numerous consumer products are manufactured from renewablelignocellulosic feedstocks. Among the feedstocks, non-wood materials are attractingattention for use in conventional pulp and papermaking, and as a feedstock for thebiorefinery applications (Farrell et al. (2000); Machmud et al. (2013); Salmela et al.(2008)). Of the non-woods, bamboo, a perennial woody grass, has received substantialattention due to its chemical composition (similar to wood), fast-growth that becomesharvest-ready in 3-5 years, and high abundance in many Asian countries (Luo et al.(2013); Okubo et al. (2004)). While bamboo is already used for the commercialmanufacture of numerous products, there is interest in expanding its use for producingdissolving pulp (Ma et al. (2011)), transportation fuels (Leenakul and Tippayawong(2010)) and various chemicals (Liu et al. (2011); Littlewood et al. (2013)).Production of dissolving grade pulp (i.e. pulp with > 92 w/w% α-cellulose withvery low quantities of \"impurities\" such as hemicellulose, ash and lignin) generallyrequires a pre-hydrolysis step in which the lignocellulosic material is subjected tohydrothermal (water/steam) or dilute acid treatment to remove hemicellulose andincrease the accessibility of substrate to chemicals used in subsequent processes (Sixta(2006); Agbor et al. (2011); Öhgren et al. (2007)). Pre-hydrolysis is also referred to1Chapter 1. Introduction’pretreatment’ in the lignocellulosic fuels literature (Chandra et al. (2016); Chandraet al. (2015); Trajano and Wyman (2013); Fengel and Wegener (1983)). One of theopen remaining questions in the literature is the incomplete understanding of thekinetics of this process. This could facilitate more efficient process optimization andscale-up in these areas.Indeed, there are difficulties to adopt bamboo in traditional processes. Comparedto wood, bamboo contains a much higher level of silica (0.5-5%). Silica creates variousdownstream challenges during the industrial utilization of bamboo. For example,(a) The high silica level creates problems such as scaling of evaporators and de-creasing the causticizing efficiency in the causticizer in the recovery cycle of theconventional kraft pulping process.(b) During the production of dissolving grade pulp, high amounts of residual sil-ica causes poor filterability and interferes with the downstream conversion ofdissolving pulps into other products (Liese (1987); Salmela et al. (2008); Tsujiet al. (1965)).(c) During the industrial applications of lignin obtained from bamboo pulping orethanol fermentation residues, the silica stays in the residual lignin hinder-ing thermal motions (softening, decomposition or degradation) and preventingthermal processing of lignin into value-added bio-products such as emulsionstabilizers and dispersing agents (Kadla et al. (2002); Pye (2008))(d) Silica in the water stream also causes several complications such as membranefouling and pipeline scaling (Le et al. (2015); Negro et al. (2001)).2Chapter 1. IntroductionThus, the ideal method to solve the silica issues encountered in processing bamboofor industrial processes is that silica should be separated completely from the rawmaterial prior to processing. The extracted silica could then be recovered as a by-product for further utilization, for example as a filler in cement or for the productionof catalysts and absorption agents. Therefore, understanding the pre-extraction ofhemicellulose and silica from bamboo and their subsequent recovery is critical inexpanding bamboo usage.In this work, hydrothermal pretreatment, using water as the media with/withoutthe addition of chemicals (acid or alkali), was used to extract hemicellulose and silicafrom bamboo biomass. The thesis is presented in two distinct yet complimentaryparts. In the first study, a mathematical model was developed to describe the protonconcentration evolution during auto- and dilute-acid pretreatment of bamboo chips.The proposed model provides guidelines to evaluate the effect of temperature, acidaddition and time on the H+ generation during pre-hydrolysis.In the second study, we proposed a novel methodology to pre-condition bambooto transform it to a useful feedstock for dissolving pulp or kraft pulp. This proposedmethod of solving the silica associated problems will be a guide to improving chemicalrecovery of kraft mills using non-wood materials as well as optimizing present dis-solving pulp or kraft pulping processes in order to obtain high quality pulp productsat high yield.The hypotheses of this work are as follows:1. A proton evolution model that covers auto- and dilute acid pre-hydrolysis ofbamboo can be generated to guide the acidic pre-treatment of lignocellulosic3Chapter 1. Introductionbiomass.2. NaOH can selectively remove silica and hemicellulose from bamboo.3. Extracted silica and hemicellulose can be recovered in an economical and envi-ronmentally friendly way for further applications.4. The process of combination of pre-extraction and recovery of silica and hemi-cellulose is suitable for the production of high-grade kraft pulp from bamboowhile resolving the silica associated challenges.This thesis is organized into 5 chapters. The motivation of this work is given inChapter 1. Chapter 2 presents the evolution of proton concentration during auto-and dilute-acid pretreatment of bamboo biomass. Chapter 3 presents alkaline pre-extraction of silica and hemicellulose from bamboo prior to pulping. The study onthe the effect of alkaline pretreatment on kraft pulping of bamboo and the recoveryof extracted silica and hemicellulose from the spent liquor of alkaline pre-extractionas byproducts are presented in Chapter 4 and a mass balance for the proposed mod-ification to a typical kraft pulping process is also given. The highlights of this workare summarized in Chapter 5 and recommendations for future research are given.4Chapter 2Evolution of the ProtonConcentration During Auto- andDilute Acid Hydrolysis ofHemicellulose2.1 IntroductionIn this chapter the kinetics of proton generation during pre-hydrolysis of bamboochips in a batch reactor is examined. Bamboo is a perennial species belonging to theGraminease family and Bambuseae subfamily. It is highly abundant worldwide andencompasses over 1250 species within 75 genera (Scurlock et al. (2000)). In Chinaalone, there are approximately 300 species in 44 genera, occupying 33000 km2 ofthe country’s total forest area (Littlewood et al. (2013)). Bamboo plantations alsohave several advantages such as limiting soil erosion in cropping systems, improvingwater quality, and requiring relatively low chemical and nutrients during growing(García-Aparicio et al. (2011)). Moreover, compared to woods, most bamboo speciesneed less time (3-5 years) to mature (Gratani et al. (2008); Krzesińska et al. (2009)).52.1. IntroductionMoreover, bamboo is considered a promising species for cultivation on marginal landfor biofuels and bio-products (Littlewood et al. (2013)). Belonging to the family ofgrass, bamboo has a stem structure with numerous vascular bundles scattered in amatrix of parenchyma storing cells, all of which are surrounded by a strong and denseepidermis. Vascular bundles are commonly known as bamboo fibres (Amada (1995)).The three main components of bamboo are cellulose, hemicellulose and lignin,and are found at levels comparable to those of hardwoods and softwoods (Table 2.1)(Luo et al. (2013); Peng et al. (2009); Jun et al. (2012); Huang and Ragauskas(2013)). Cellulose is a linear polymer of repeating sugar units of glucose linkedby β − 1, 4 glycosidic bonds and has cyrstalline and paracrystalline regions (Hal-lac and Ragauskas (2011)). Hemicelluloses are branched, amorphous polysaccharidescomposed of short lateral chains consisting of several different monosaccharides andfunctional groups such as xylose, mannose, galactose, arabinose, acetyl groups andglucuronic acid groups (Scheller and Ulvskov (2010)). Like agricultural residuals,such as wheat-straw and rice-straw, the hemicellulose of bamboo is primarily glu-curonoarabinoxylan which has a xylose backbone with arabinose, glucuronic acid andacetyl side-groups (Pauly et al. (1999); Scheller and Ulvskov (2010)). Lignin, whichis hydrophobic in nature, is an amorphous polyphenol of the primary monolignols:coumaryl, coniferyl, and sinapyl alcohol (Saake and Lehnen (2007)). Within the cellwall, lignin is tightly bound to cellulose and hemicellulose through hydrogen bonds,lignocellulosic complexes and covalent bonds. Biomass also contains some compoundsknown collectively as extractives (soluble in water or organic solvent). In bamboo,the extractives are mainly composed of resins, fats, nonstructural sugars, nitrogenousmaterial, chlorophyll, and waxes (He and Yue (2008)). The extractives often have62.1. IntroductionTable 2.1: Chemical composition of bamboo, hardwood and softwood (Huang and Ra-gauskas (2013); Saka (2004); Dence (1992); Li et al. (2012); Song et al. (2013); Torelli andČufar (1995)).Cellulose Hemicellulose Lignin Silica Pectin, starch,(%) (%) (%) (%) etc. (%)Bamboo 38-51 24-28 21-31 0.5-5 1-5.5Hardwood 42-51 23-35 19-26 0-0.05 1-3Softwood 41-47 23-31 27-33 0-1.2 1-2protective biological and anti-microbial activities and aid in the chemotaxonomic divi-sion of plant species by their specific biosynthetic pathways (Torssell (1997)). Othercommon extractives include phenolics, terpenes, aliphatic acids and alcohols (Fen-gel and Wegener (1983)). Besides these organics, bamboo also contains inorganicminerals, which is commonly referred to as ash, and includes both plant structuralcomponents and inorganic materials such as in soil picked up in harvesting opera-tions. The ash content is composed of salts and oxides containing elements such aspotassium, calcium, magnesium, sodium and silicon. As a non-wood, bamboo ashcontains high level of silica content of up to 70% of total ash (Liese (1987)).Cellulose is the main component of dissolving grade pulp and both cellulose andhemicellulose can be readily hydrolyzed into fermentable sugars for the productionof biofuels and chemicals. Therefore, from the pulping or biorefinery perspective, thetotal content of cellulose and hemicellulose of bamboo is about 65-75% of the rawbiomass, which indicates that it is a suitable candidate as an alternative feedstock forpulp and paper industry and biorefineries.With regards to the structure of the biomass, the hemicellulose-lignin matrix andstructural proteins are thought to serve as a physical barrier and adheres to, and teth-ers, cellulose macrofibrils through hydrogen bonds and van der Waal’s interactions,72.1. Introductionsee Fig. 2.1 (Altaner and Jarvis (2008); Dammström et al. (2008); Cosgrove et al.(2012)). The strong barrier has high stability and hinders degradation of biomasscomponents. Altering the biomass structure and increasing cell wall accessibility isessential during the utilization of lignocellulosic biomass for the production of fu-els, chemicals, pulps and other bio-based materials. Different precondition methods,which can be categorized as physical, chemical, biological and hydrothermal, havebeen developed to solubilize and remove hemicellulose and/or a fraction of lignin,and increase pore volume and surface area of the biomass (Xu and Huang (2014)).Among the practical precondition technologies, acidic pretreatment is regarded tobe an essential unit operation and has been commercially applied in the productionof dissolving pulp and biofuels from lignocellulosic biomasses. This is because thistreatment method takes advantage of the high moisture content within the biomass.It can also efficiently convert the polysaccharides into monomeric sugars and theircorresponding degradation products in a low energy and environmentally friendlymanner. During the production of dissolving grade pulp, acidic pretreatment is alsoreferred to pre-hydrolysis, which includes both auto-hydrolysis (using water/steam asthe media) and dilute acid hydrolysis.Pre-hydrolysis refers to the reaction pathway to remove hemicellulose from lig-nocellulosic materials prior to subsequent chemical or enzymatic treatment (Little-wood et al. (2013); Sixta (2006); Trajano and Wyman (2013)). During acidic pre-hydrolysis, an acid catalyzes the breakdown of the long hemicellulose chains to formshorter chian oligomers and sugar monomers in the presence of water or steam. Theproduction of dissolving pulp and nanocrystalline cellulose (NCC) (a product madefrom dissolving pulp) requires the removal of more than 70% of hemicellulose during82.1. IntroductionFigure 2.1: A simplified, hypothertical model of bamboo/grass secondary cell wall il-lustrating molecular interactions of hemicellulose with cellulose and lignin. This figure isreproduced from Sella Kapu and Trajano (2014).pre-hydrolysis. Lignocellulosic ethanol production uses acid hydrolysis to generatean enzymatically digestible cellulose-fraction. The hydrolyzed hemicellulose-derivedoligomers and monomers can be used in food applications or in the textile, paper, ex-plosives, cosmetic, petroleum, and mining industries (González-Muñoz et al. (2012)).In addition, hemicellulose oligomers can also be used to produce gels, films, coatings,and adhesives (González-Muñoz et al. (2012); Sella Kapu and Trajano (2014)).Pre-hydrolysis is different from torrefaction wherein biomass is treated at 200-300 oC in an inert gas environment (Neupane et al. (2015)). During hemicellulosetorrefication, it is degraded into volatile organic compounds including CO2 and CO,and char (Neupane et al. (2015)). Pre-hydrolysis of lignocellulosic biomass is nor-mally conducted between 120-220 oC in the presence of water or steam with/without92.1. IntroductionFigure 2.2: Hydrolysis of hemicellulose during acidic pre-hydrolysis process.the addition of acid (Sanchez and Cardona (2008); Borrega et al. (2013); Rissanenet al. (2014a); Rissanen et al. (2014b); Yan et al. (2014)). The breaking of glycosidiclinkages between monomers in the polymeric chains of hemicellulose and cellulose isthe fundamental hydrolysis mechanism (Fengel and Wegener (1983)). Hydrolysis ofthe glycosidic linkage is initiated by the protonation of either glycosidic oxygen orring oxygen to form a carbonium cation and an end chain product, i.e. glucose inthe depicted case. Further, a water molecule is added to the carbonium cation, thereaction teminates, resulting in the formation of two monomeric sugars and a proton(Grénman et al. (2011); SriBala and Vinu (2014)). The protons, released by the disso-ciation of water at high temperature, the added acid and the cleavage of acetyl groupsfrom hemicellulose, subsequently catalyze the hydrolysis of hemicellulose (Grénmanet al. (2011)). Fig. 2.2 shows the hemicellulose hydrolysis reaction process.In acidic pre-hydrolysis process, monomeric pentoses and hexoses released duringthe hydrolysis of hemicellulose and cellulose may undergo subsequent dehydration re-actions to form furfural and hydroxymethylfurfural (HMF), respectively (Rogalinskiet al. (2008)). Increasing the hydrolysis temperature or prolonging the residence timewill increase the generation of hemicellulose and cellulose degraded byproducts suchas furfural, HMF and other light organics (Pu et al. (2013)). Furfural and HMF mayfurther react to generate formic or levulinic acid (Pu et al. (2013)). Using birch wood,Borrega et al. (2011) reported a 7-10% yield of furfural during auto-hydrolysis between102.1. Introduction200-240 oC using a batch reactor. Bamboo exhibits the same trend as wood duringacidic hydrolysis: an increase in monomeric or polymeric sugar yield with increasingtemperature/time followed by a decrease due to increased degradaded byproductsproduction. For example, when the temperature of steam treatment of bamboo wasincreased from 186 to 200 oC, the xylose yield in the liquid phase increased by morethan 50% (García-Aparicio et al. (2011)). However, above 200 oC, the xylose contentin the liquid phase decreased by about 25% (García-Aparicio et al. (2011)). In ad-dition, during acidic pretreatment, the addition of acid accelerates the hemicellulosehydrolysis through the availability of proton catalyst (Larsson et al. (1999); Gütschet al. (2012)).Both auto-hydrolysis and dilute acid hydrolysis are considered viable pretreatmentoptions in the production of dissolving pulp and lignocellulosic ethanol. However,kinetic modelling still remains at the forefront and the evolution of concentration ofthe acid catalyst [H+] is one of the longstanding unanswered questions (Sella Kapuand Trajano (2014)). One major reason might be the chemical complexity of biomass.The literature on the kinetics of the removal of hemicellulose is substantial. Themodelling approach was built upon the approach used for dilute acid hydrolysis ofcellulose (Saeman (1945)). For hemicellulose, complex behavior is evident and numer-ous groups consider that two fractions of hemicellulose are distributed spatially overtwo separate domains in the solid matrix to help simplify the analysis (Kobayashiand Sakai (1956)). Each fraction reacts with available protons at differing rates dueto differences in reaction activation energy. This model has been adopted widely andis commonly referred to as the \"biphasic model\". It consists of two solid species,fast and slow hemicellulose, denoted as Xi(s), which hydrolyze following first-order112.1. IntroductionkineticsXi(s)ki−−→H+X(aq) ri = ki[Xi] (2.1)where ri and ki are defined as the rate of reaction and the rate constant, respectively,to form a set of soluble products, X(aq), which are susceptible to further hydrolysisor decomposition reactions. The subscript i represents either fast or slow. Theinitial values for Xi are considered to be intrinsic for the biomass (Esteghlalian et al.(1997); Ma et al. (2011)). Variations on this approach are available in the literatureto describe subtle effects such as the formation of oligomeric intermediates or masstransfer rates (Cahela et al. (1983); Conner and Lorenz (1986); Tillman et al. (1990);Carrasco and Roy (1992); Garrote et al. (2001); Kim and Lee (2002); Brennan andWyman (2004); Hosseini and Shah (2009); Morinelly et al. (2009); Mittal et al. (2009);Shen and Wyman (2011); Visuri et al. (2012); Liu et al. (2012); Luo et al. (2013);Zhao et al. (2012); Aguilar et al. (2002)). However, no physical or chemical attributeshave been identified to differentiate fast and slow hemicellulose.One of the open remaining questions in this literature is an understanding of theevolution of the concentration of the acid catalyst. What makes this problem particu-larly challenging is that there are competing pathways governing proton evolution andneutralization. Although difficult to substantiate, a number of authors have advancedrate constants ki of the formki = koi exp(−EaiRT)f(t, [H+]) (2.2)122.2. Model Developmentwhere koi is the pre-exponential factor; and Eai is the activation energy. The functionf(t, [H+]) is determined empirically and is found to vary greatly in the literature. Thisfunction is included to allow for different reaction rates with different acid levels. Inone extreme we find that this function varies linearly in time while in the otherextreme it is considered as a constant and set to its initial value. We summarizethese forms asf(t, [H+]) =a+ bt autohydrolysis[H+]no dilute acid(2.3)depending upon if the experiment is conducted under dilute-acid or autohydrolysisconditions. Here a, b and n are empirical constants and [H+]no is the initial con-centration of the acid catalyst. n is typically found to be between 0.8-1.3 and wenote that Shen and Wyman (2011) set n = 1 for corn stover. The utility of thisfunctional form has been questioned and it is evident that there is no theoreticalbasis for the form of the assumed functions (Esteghlalian et al. (1997); Conner andLorenz (1986); Morinelly et al. (2009); Maloney et al. (1985); Malester et al. (1992);Lloyd and Wyman (2004); Lloyd and Wyman (2005); Hong et al. (2013)). In thischapter we attempt to gain insight into the assumed form of Equation 2.3. We doso by examining the evolution of the proton concentration during reaction throughexperiment and mathematical modelling.2.2 Model DevelopmentThe analysis presented in this section is aimed at understanding the evolution of[H+] during the reaction. The goal is to develop a qualitative understanding of this132.2. Model Developmentform by posing a hypothetical reaction scheme which, at some level of approximation,represents the true reaction scheme. It is done at a level in which the analysis is math-ematically transparent and of sufficient detail to capture the dominant mechanisms.As a result the approach is referred to as a \"toy model\".One of the many complicating factors hindering the modelling process is thatthere is a large number of chemical species such as cellulose, hemicellulose, ligninand different ash components, which are distributed throughout the cell wall in acomplex manner. To simplify, classes of species which behave similarly are groupedtogether and represented as one hypothetical species. For example, we representthe ash constituents as a lumped parameter MO, that is, the ash is an oxide of thehypothetical species M with a valence state of 2+; this hypothetical species servesto neutralize the available protons. This can be reposed at another valence state orwith secondary effects, such as precipitation from solution, included. In a similarmanner the hemicellulose constituents have been reduced to a linear xylose polymer,denoted by X, fast and slow, having arabinose (Ar) side chains (Fig. 2.3). Protonsare represented by H+ and the hydroxyl groups by OH– ; both of these species areconsidered to be in the aqueous phase and the (aq) notation has been dropped. Wehave included the potential of an acid being added to the system and denote thisspecies as H2A because sulfuric acid is most commonly used in the literature. Theacetyl group Ac is defined as H3C−C(−O)−. Mass transfer effects are neglected.We consider four primary reactive pathways in Fig. 2.4 and each individual reac-tion is assumed to follow elementary kinetics. In the first of these, shown on the farleft of Fig. 2.4, we consider deacetylation where Ac is cleaved from the hemicellulose142.2. Model Development X XO AcX XArX XLCX XO AcX XLCXnFigure 2.3: A schematic of the idealized hemicellulose - lignocellulose (LC) substrateconsidered in this work. Although xylan is hypothesized to be comprised of fast and a slowreacting fractions, we do not distinguish these in this figure. The species Ac and Ar, whichrepresent the acetyl and arabinose groups, are initially bound to the xylan chain but arereleased through acid hydrolysis. The ash (MO) is not shown in this figure but is consideredto be physically embedded in the LC portion of the matrix.backbone through an acid hydrolysis of the ester.XOAc + H2Ok1−−→H+XOH(s) + AcOH(aq) r1 = k1[XOAc][H+] (2.4)This reaction may occur with acetyl groups which are attached to either solubleor solid phases of the hemicellulose. For simplicity, any differences in rate betweenthe deacetylation reaction occurring in the solid or liquid phases are ignored. Asthe product AcOH(aq), acetic acid, behaves as a weak acid, it adopts the followingequilibrium in solutionAcOH(aq)KAcOH−−−−⇀↽ − AcO− + H+ KAcOH = [AcO−][H+][AcOH]= 1.8× 10−5 (2.5)where Ki, from this point forward is defined as the equilibrium constant and thevalue quoted is at room temperature. Both Garrote et al. (2001) and Aguilar et al.(2002) have used similar modelling approaches to describe deacetylation. Aguilaret al. (2002), for example, explicitly indicates that this reaction follows first order152.2. Model DevelopmentNeutralizationAcid AdditionDeacetylationHydrolysisAcid CycleFigure 2.4: A schematic of the idealized reaction scheme. The chemical reactions shownform the basis of the toy mathematical model of the proton concentration.162.2. Model Developmentkinetics (Aguilar et al. (2002)). We build upon these studies by including the effectsof the weak-acid behavior of acetic acid (see Equation 2.5).To continue, water disassociation demandsH2OKw−−⇀↽− OH− + H+ Kw = [OH−][H+] = 1× 10−14 (2.6)and this serves as an additional source of H+. Because of these equilibria, H+ isavailable for both the neutralization and hydrolysis reactions.In addition to this, protons may also be available if acid is added to the system.We capture the reaction scheme as if the added acid is sulfuric acid, as this is themost common addition in the literature:H2A(aq)fast−−→ HA− + H+ (2.7)HA−Ka−−⇀↽− A2− + H+ Ka = [A2−][H+][HA−]= 1× 10−2 (2.8)Like others in the literature, we consider the disassociation give in Equation 2.7 tobe instantaneous. The final aspect to consider is the neutralization of the protons bythe ash. As mentioned above the reaction scheme depends upon the species involved.As discussed, we consider that MO reacts according to the following schemeMO(s) + 2 H+k2−−→ M2+(aq) + H2O r2 = k2[MO(s)][H+] (2.9)M2+ + 2 OH−Km−−⇀↽− M(OH)2(aq) Km = [M2+(aq)][OH−]2[M(OH)2]→ 0 (2.10)As the equilibrium constant Km is unknown, we simply assign this value to be a172.2. Model Developmentvery small number to reduce the number of free parameters. It should be noted thatwe do not characterize a number of potential secondary reactions in solutions, eventhough they may affect the proton levels to small degree. For example, we ignore thepotential reaction between M2+ and A2– for mathematical transparency as these donot affect the proton concentration.Having established the chemistry of the toy model, we now construct the math-ematical model. We build the model upon two conservation laws: conservation ofmass of each of the species found in solution and an overall charge neutrality of thesolution. Conservation of mass expresses that the initial moles of a certain speciesmust sum to total moles of the species in the reaction products. For example, theinitial moles of M in [MO]o, must balance the number of moles of M, in the species[MO], [M2+], and [M(OH)2] at any time throughout the course of the reaction. Thiscan be expressed as[MO]o = [MO] + [M2+] + [M(OH)2] = [MO] + [M2+](1 +[OH−]2Km)(2.11)through use of the equilibrium relationship given in Equation 2.10. In a similarmanner, conservation of mass for the species Ac can be expressed as[XOAc]o = [XOAc] + [AcOH] + [AcO−] = [XOAc] + [AcO−](1 +[H+]KAcOH)(2.12)and A as[H2A]o = [HA−] + [A2−] = [A2−](1 +[H+]Ka)(2.13)182.2. Model Developmentwith use of Equations 2.5 and 2.8, respectively. To continue, charge neutralization isinvoked, i.e.[AcO−] + [OH−] + [HA−] + 2[A2−] = [H+] + 2[M2+] (2.14)which can be expressed as [XOAc]o − [XOAc](1 + [H+]KAcOH)+ Kw[H+]+ [H2A]o([H+] + 2Ka[H+] +Ka)= [H+] + 2 [MO]o − [MO](1 + Kw2Km[H+]2)(2.15)through use of Equations 2.11-2.13. This equation indicates that the proton concen-tration in the solution is governed by charge neutralization and is related to molesof acetic acid formed (first term on LHS of equation), the amount of ash neutral-ized (second term on RHS of equation), three different equilibria found in solution(Km, Ka, KAcOH), and the amount of acid initially added [H2A]o. To complete thisdescription, we use the rate expressions given in Equations 2.4 and 2.9ddt[XOAc] = −k1[XOAc][H+] [XOAc(0) ] = [XOAc]o (2.16)ddt[MO] = −k2[MO][H+] [MO(0) ] = [MO]o (2.17)where the subscript o represents the initial concentration of the species. Equations2.15-2.17 represent the toy model to describe the proton concentration during reac-tion. The utility of this set of equations will be tested experimentally in three limitingcases, i.e.192.3. Materials and Methods(a) at long reaction times where the reactions with XOAc and MO are nearly com-plete.(b) with the reaction occurring at room temperature in order to examine the pro-posed ash neutralization scheme.(c) at typical reaction temperatures found for pre-hydrolysis.as a function of initial pH.2.3 Materials and MethodsBamboo chips, prepared from 3− 7 year old trees, were provided by the Lee & ManPaper Manufacturing Ltd. China. The obtained chips were stored at 4 oC until usedfor experimentation. The chips were air dried for approximately 24 h and re-chippedusing a Wiley mill (Thomas Scientific, NJ, USA) and screened with a 45-16-9.5 mmstacked sieve system. Chips retained on the 9.5 mm were designated as accepts forexperimentation. The accepts were then washed with deinoized water at 35 oC for10 min at a liquid-to-wood ratio of 20 L/kg using a laboratory mixer to removeimpurities, such as soil and sand (see Appendix A1 and A2). The washed chips wereair dried for approximately 24 h and then stored at 4 oC until used for subsequentexperiments. The chips were analyzed with respect to lignin content, carbohydratecomposition, extractives, ash and silica content. All chemicals used in this study werereagent grade and purchased from Fisher Scientific, Canada.The moisture content of solid samples was measured by drying at 105 ± 2 oC toconstant weight. The contents of water and solvent extractives of bamboo chips were202.3. Materials and Methodsdetermined using a Soxhlet extractor according to TAPPI T 204 cm-97.Carbohydrates and lignin contents of the solids were determined after air dry-ing according to National Renewable Energy Laboratory (NREL) standard protocols(Sluiter et al. (2012)). Briefly, the chips were air-dried and ground to pass througha 40-mesh screen of a Wiley mill. The samples were then subjected to a two-stepH2SO4 hydrolysis protocol to digest the polysaccharides into monomeric sugars. Af-ter hydrolysis, Klason lignin was separated through filtration and measured gravi-metrically. Acid soluble lignin in the hydrolysate (after removing Klason lignin) wasdetermined at wavelength 205 nm using a UV-vis spectrophotometer (Dence (1992)).Acid hydrolysates were then filtered using 0.2 µm syringe filters (ChromatographicSpecialties, Inc. ON, Canada) and analyzed for monomeric sugars using a DionexICS 5000+ HPLC (high performance liquid chromatography) system equipped withan AS-AP autosampler and an electrochemical detector (Thermo Fisher Scientific,MA, USA). The monomeric sugars were separated on a Dionex CarboPac SA10 ana-lytical column (Thermo Fisher Scientific, MA, USA) at 45 oC using 1 mM NaOH asthe mobile phase, and the sugars were quantified using electrochemical detection andChromeleon software (Thermo Fisher Scientific, MA, USA). High purity monomericsugar standards, arabinose, galactose, glucose, xylose and mannose were purchasedfrom Sigma-Aldrich (ON, Canada). Fucose was used as the internal standard.Total ash content of the raw bamboo chips was determined according to TAPPIT211 om-02. Silica content of bamboo chips and pulp was measured gravimetrically,using a method modified from (Ding et al. (2008)). Briefly, about 5 g of dried andpowdered bamboo sample was completely ashed at 550 oC. After cooling, 10 mL ofHCl (6 mol/L) was added to the ash to precipitate silica and dissolve acid-soluble212.3. Materials and MethodsTable 2.2: Composition of the bamboo chips used in this study. The values reported inthis table are based upon the total mass of the bamboo chips.Composition % od, BambooHemicellulose as: 21.8Xylan 20.3Arabinan 0.8Galactan 0.7Ash as: 2.1SiO2 1.12CaO 0.37K2O 0.28Al2O3 0.15Cellulose as: 48.7α-cellulose 47.3β-cellulose 1.4Lignin as: 25.1Acid Soluble 0.9Acid Insoluble 24.2Extractives as: 4.6Water 3.4Solvent 1.2Acid groups as: 3.6Acetyl group 2.7Uronic acid 0.9ash. The resultant solution was gently boiled to near dryness in a boiling water bath.HCl treatment was repeated three times in about 30 min, after which another 15mL of HCl (6 mol/L) was added to the solution. After 2 more minutes, the solutionwas filtered off through No. 42 ashless filter paper (Fisher Scientific, Canada). Theprecipitate was rinsed 5-6 times with 1 mol/L HCl solution and 5-6 times with hotdeionized water (≈ 50 oC). Both the filter paper with the precipitate was ashed at700 oC and calcined at 1000 oC in a muffle furnace to reach a constant weight. Theresultant silica residue was weighed to determine silica and ash content. All mea-surements were run at least in triplicate. Detailed analysis of the metal composition222.3. Materials and Methodsof ash was done using inductively coupled plasma time of flight mass spectrometry(ICP-TOFMS) (Benkhedda et al. (2000)). During the ash composition analysis, highpurity (> 99.9%) nitric acid (HNO3) was used as the dissolution agent. All chemicalsused were analytical grade. A summary of the compositional analysis is shown inTable 2.2.Four separate studies were conducted in this work, as summarized in Table 2.3.In all cases bamboo chips and water were mixed at defined liquor to wood ratios(see Table 2.3) and placed in a 300 mL stainless steel reactor. The total mass ofthe chips and water for all liquid-to-wood ratios was kept constant at 217 g; thisslurry filled about 80% of the available volume of the reactor. The purpose of thefirst study (Series 1-10) was to characterize the reactor temperature response overtime. The reactor was immersed in an oil bath set at a defined temperature, Tb.The temperature of the mixture was continually monitored with two thermocouplesmounted in the middle of the reactor, on the central plane but at two different radialpositions. Upon completion of a run, the reactor was cooled by immersion in anice-water bath.In the second study (Series 11 and 12), conducted to investigate the equilibriumproton concentration after a long period of time, the pH of bamboo chips-liquidmixture having a liquor-to-wood ratio of 6.5 L/kg was measured as a function initialacid content after a minimum of 315 min (in some cases 10 h). In the third study,the second study was repeated but at room temperature (Series 13-16). In the fourthseries, the time evolution of the proton concentration was measured as a function oftime, temperature and initial acid addition (Series 17-21). In this case the liquid-to-wood ratio was 6.5 L/kg.232.3. Materials and MethodsTable 2.3: A summary of the experimental conditions tested.Experiment Series L : W [H+]o Tb t(L/kg) (pH) (oC) (min)Temp. Measurement 1 water only 7.1 120 0 < t < 452 water only 7.1 150 0 < t < 453 6.5 7.1 140 0 < t < 454 6.5 7.1 150 0 < t < 455 6.5 7.1 160 0 < t < 456 6.5 7.1 180 0 < t < 457 8 7.1 120 0 < t < 458 8 7.1 150 0 < t < 459 10 7.1 120 0 < t < 4510 10 7.1 150 0 < t < 45Long time behavior 11 6.5 1.3-6.8 160 t > 31512 6.5 1.5-7.1 180 t > 315Room Temperature 13 6.5 1.5 23 0 < t < 115514 6.5 2.9 23 0 < t < 115515 6.5 5.0 23 0 < t < 115516 6.5 6.0 23 0 < t < 1155Elevated Temperature 17 6.5 1.7 160 0 < t < 36018 6.5 1.5 160 0 < t < 39019 6.5 7.2 160 0 < t < 39020 6.5 7.2 180 0 < t < 39021 6.5 3.5 160 0 < t < 360Footnotes: In Series 1 through 10, the temperature was sampled at a frequency of 1Hz. In Series 11-12 a total of 18 samples were measured at times ranging from 315to 390 min. In Series 13-14, four different initial pH were tested and approximately11 samples, obtained at different times, were measured. In this case Tb is defined asthe oil bath temperature. \"L : W\" and \"t\" refer to liquid-to-wood ratio and time.242.4. Results and Discussion2.4 Results and DiscussionBefore proceeding to the main findings, it is instructive to first examine the temper-ature profile in the reactor after immersion in the oil bath. For each experimentalcondition, the temperature of the reaction mixture (chips and liquid phases) wasrecorded using two temperature transducers located at different radial positions inthe reactor. For all conditions (Series 1-21) there was no significant difference be-tween the two transducer signals, and the reactor seemingly behaved as if there wereno spatial gradients in the system. Two results representative of all the runs areshown in Fig. 2.4. For each experimental condition, the signals from both temper-ature transducers, located at different radial positions in the reactor are reported.What is evident in this figure is that there is no significant difference in the signalsand the reactor seemingly behaves as if there are no spatial gradients in the system,i.e. it is at a uniform temperature. The trend with all the data sets is that the heatup period is about ∼ 15− 20min i.e. the heat-up rate was essentially the same. Thecool-down rate is approximately ∼ 25 oC/min.It is curious that there are no strong radial temperature gradients in the system.This result is evident in both the pure water case (Series 1-2) and cases with liquid-to-wood ratio as low as 6.5 L/kg (Series 3-6). Two speculative arguments are proposedto explain this. In the first case we argue that the thermal mass of the steel reactor,i.e. the product of its mass and heat capacity, to be significantly larger than thereactants. As a result, the temperature response of the reactants is dictated by theheating or cooling of the reactor. The second argument is somewhat more delicate.It is also possible that convection occurs due to difference in density of the fluid near252.4. Results and Discussion0 10 20 30 40 50 60t (min)20406080100120140160T(oC)Series 4Series 7Figure 2.5: Temperature evolution for two representative cases. Both thermocouple signalsare presented and replicate runs are shown but the difference between them is not perceptibleon this scale. The thermocouples are placed at the same elevation in the reactor but at twodifferent radial positions.the outer wall in comparison to the bulk. Convective currents in the reactor wouldtend to diminish the radial gradients.With the notion of uniform spatial temperature gradients, we examine the temper-ature evolution throughout the reaction. We propose the temperature profile followsan equation of the formcdTdt= h(Tb − T ) ⇒ T¯ = Tb − TTb − To = exp(−hct) (2.18)262.4. Results and Discussion0 10 20 30 40 50t (min)00.10.20.30.40.50.60.70.80.91T¯Exponential decaySeries 1-10Figure 2.6: The temperature evolution during the heat up period for Series 1-10. Theresults have been scaled using the form advanced in Equation 2.18 with h/c set to be 0.15min−1.where c is the product of the effective mass and heat capacity of the reactor andreactants; h is the overall heat transfer coefficient; and T and Tb are temperature andoil bath temperature, respectively. The utility of this equation is tested by plottingSeries 1-10, shown in Table 2.3, in Fig. 2.6 using the scalings indicated in Equation2.18. What is evident in this figure is that the system displays nearly exponentialbehavior as the experimental data (the red dotted lines) somewhat follow Equation2.18, shown as the thick black line. However, we were unable to achieve a similarscaling during the cool-down period.272.4. Results and DiscussionAt this point we begin to explore the utility of the toy model (Equations 2.15-2.17). The first aspect of the model that we will explore is the long-time behavior andexamine if a steady state proton concentration is possible. Experimentally the pHwas measured at long-time by simply allowing the reaction to proceed for at least 315minutes at an elevated temperature. From the toy model, we see that a steady stateconcentration for [H+] exists and can only be achieved when both the deaceytlyationand neutralization reactions are complete, i.e.d[XOAc]dt=d[MO]dt= 0, thus, [XOAc] = [MO] = 0 (2.19)Indeed, at steady state the proton concentration may be estimated directly fromEquation 2.15, i.e.[XOAc]o(1 + [H+]KAcOH) + Kw[H+]+ [H2A]o([H+] + 2Ka[H+] +Ka)= [H+] + 2[MO]o(1 + Kw2Km[H+]2) (2.20)which is a sixth-order polynomial in [H+]. The steady state concentration is givenby the roots of this polynomial and the behavior of this function is given in Fig.2.7. This equation was solved for [H+] in MATLAB using the built-in root findingprocedure. Superimposed on this is the experimental data given as Series 11 and 12.Two observations are clearly evident. The first observation that can be made is thatwe find a remarkably similar trend with the toy model. The second observation isthat there two distinct regions. Under autohydrolysis conditions, i.e. the right handportion of the graph, the steady state (or long-time) pH is independent of the initialpH. Here, the steady state pH is governed by the weak-acid equilibrium and by ash282.4. Results and Discussion1 2 3 4 5 6 7 8Initial pH11.522.533.544.5Long time pHToy ModelSeries 11Series 12Figure 2.7: The effect of initial pH on the steady state pH measured after long term pre-hydrolysis. The dashed line represents the toy model with the equilibrium constant given inthe text previously. The value of Km is not stated in the text and is taken to be small. Forpractical purposes we setKm = 10−17M for this calculation. The two remaining parametersare set to be [XOAc]o = 0.025M and [MO]o = 0.001M and determined through regression.Although defined previously, Series 11 represents the reaction at 160 oC and Series 12 at180 oC.neutralization or buffering. With increasing levels of added acid, we find that thelong time pH approximately equals the initial pH. This is shown on the left handportion of Fig. 2.7.These results support the kinetic modelling for xylan removal under dilute acidconditions. As discussed in Section 2.1, a number of authors have assigned the protonconcentration to be constant and equal to its initial value ( see Kobayashi and Sakai(1956); Esteghlalian et al. (1997); Wyman et al. (2005); Shen and Wyman (2011) for292.4. Results and Discussion0 200 400 600 800 1000 1200t (min)1234567pHSeries 13Series 14Series 15Series 16Toy ModelFigure 2.8: Examination of the pH when the reaction proceeds at room temperature. Toevaluate the toy model, the same values for the constant given in the caption of the previousFigure are used. In addition, we set k2 = 10M−1min−1 as determined by regression.example). However, under autohydrolysis conditions, this does not occur. There is avast difference between the initial and steady state pH of the system.We continue the discussion of the toy model and examine a second limiting casewhen the reaction proceeds at room temperature, see Fig. 2.8. Here, four caseswere examined in which the amount of acid added initially was varied. At roomtemperature it can be assumed that the deacetylation reaction proceeds at a muchslower rate in comparison to the ash neutralization scheme. Under this assumption,302.4. Results and Discussionthe toy model reduces toddt[MO] = −k2[MO][H+] (2.21)Kw[H+]+ [H2A]o([H+] + 2Ka[H+] +Ka)= [H+] + 2 [MO]o − [MO](1 + Kw2Km[H+]2) (2.22)which has been solved numerically in MATLAB using a Runge-Kutta scheme (ODE23s)coupled with root finding procedure for the proton concentration. The equations aresolved simultaneously. As shown in Fig. 2.8, at low initial pH (Series 13), pH isconstant as the concentration of added acid is in excess of the neutralization poten-tial of the ash. With decreasing initial added acid (Series 14-15), the neutralizationreaction proceeds until all the ash is reacted. With Series 16, no added acid, the pHvaries weakly with time. We interpret this result through the toy model, and advancethe argument that the neutralization reaction proceeds but the kinetics are extremelyslow due to the low proton concentration.We now move to perhaps the main findings in this chapter and examine theevolution of pH during pre-hydrolysis treatment. In our final set of experiments, weexamined the evolution of the proton concentration at elevated temperatures. Herewe must include the effect of deacetylation and as a result, the full toy model mustbe solved numerically using MATLAB. We treat Equations 2.16 and 2.17 as a systemof equations and solve this initial value problem in conjuction with a root findingprocedure to estimate [H+] from Equation 2.15. The results are shown in Fig. 2.9.Again at low initial pH, proton concentration varies weakly with time (Series 17and 18) and remains essentially at its initial value. This was the expected result as312.4. Results and Discussion0 100 200 300 400 500 600t (min)12345678pHSeries 17Series 18Series 19Series 20Series 21Toy ModelFigure 2.9: A evolution of pH under prehydrolysis conditions. To help understand thiswe related the lowest initial pH experiments to the added acid: Series 17 0.25% (w/w) andSeries 18 0.5 % (w/w).322.5. Summarydemonstrated earlier through steady state analysis under excess acid conditions, thepH should remain essentially constant. Below this limit complex behavior is observed.Under autohydrolysis conditions (Series 19 and 20), there is a rapid initial drop inpH followed by a diminished rate at longer times. However, the most curious result isgiven by Series 21 where non-monotonic behavior, i.e. the pH initially rises and thenfalls, is observed. We base our interpretation on the toy model which indicates thatthe neutralization reaction is initially proceeding faster than deacetylation. At longertimes, the ash is completely reacted and the pH diminishes from deacetylation.These results can now be used to interpret the form of the rate constant used forxylan removal. As seen in Equation 2.3, the rate constant under dilute acid conditionsis related to the initial pH. This is quite reasonable as we have shown that the pHshould be essentially constant during the course of the reaction. However, we cannotmake any comment on the value of n in this equation. Below this limit, the behaviorof [H+] is very complex. Simple linear functions may indeed apply for a particularsystems of interest. However, this can not be generalized as the pH response dependsstrongly on the rate of ash neutralization in comparison to the rate of deacetylation.2.5 SummaryIn this chapter the evolution of the proton concentration was examined during thehydrolysis of bamboo chips. At issue was the seemingly disparate model descriptionsin the literature which treat dilute acid differently than autohydrolysis conditions.We have attempted to address this issue by posing a \"toy model\" in which we haveincluded a number of chemical components to help describe the reaction. We advance332.5. Summarythat the proton concentration is governed by a charge neutrality in the solution andinfluenced by the:(a) weak acid equilibrium formed from the deaceytlation of the acetyl group fromthe xylan(b) equilibrium created by water dissociation(c) ash neutralization and the associated equilibrium in solution(d) added acidThere are a number of outcomes from the toy model which have been tested ex-perimentally. The first, and perhaps most significant outcome, is that the toy modelpredicts the existence of a steady state solution. The steady state value is dictatedby the equilibrium constants, and the initial added acid and acetyl group levels. Themodel qualitatively follows the trend given by experiment. It is difficult to performa quantitative comparison as auxiliary relationships, such the variation of the equi-librium constant with temperature are not known. The model was tested at roomtemperature to examine the changes in pH when ash neutralization is the dominantmechanism. Under these conditions we find, surprisingly, that the pH remains es-sentially constant both at low initial pH and under autohydrolysis conditions. Acidis likely in excess of the neutralization potential of the ash, in the former case, andthe kinetics of neutralization become exceedingly small in the latter case due to thelow proton concentration. Finally, when the hydrolysis reaction proceeded at elevatedtemperatures, we found one case of non-monotonic behaviour in which the pH initiallyincreased, and then decreased at longer times. This is attributed to the difference in342.5. Summaryrates between the neutralization and deacetylation reactions.As described in the introduction the evolution of the proton concentration duringpre-hydrolysis is poorly modeled using empirical functions (Equation 2.3) which arenot rooted in a proper chemical reaction scheme. With our toy model we proposea chemical reaction pathway that satisfactorily describes experimentally determinedproton concentration under both auto- and dilute-acid hydrolysis conditions. Accu-rate modeling of the proton concentration would significantly improve the existingkinetic models of hemicellulose hydrolysis and facilitate more efficient process opti-mization and scale-up in pulping and biorefinery areas.35Chapter 3Alkaline Pre-extraction of Silica andHemicellulose from BambooAs discussed in Chapter 1, silica creates downstream processing problems for pulpingand bioconversion processes. Considerable effort has been expended in trying to solvethis silica issue. These efforts have been mainly focused on retaining silica in the finalpulp or desilication of black liquor (Jahan et al. (2006); Kopfmann and Hudeczek(1988); Pan et al. (1999); Tsuji et al. (1965)). Unfortunately, until now, none ofthis work has led to a commercial process. Recently, the proposed integrated forestbiorefinery concept (IFBR) has been advanced as a means of addressing concerns overenergy security and climate change. According to the IFBR concept, hemicellulose ispartially pre-extracted prior to the pulping stages for the generation of value addedproducts such as bioethanol, furfural, acetone, or to be used as papermaking addi-tives (Bai et al. (2012); Hamzeh et al. (2013); Liu et al. (2013); Mao et al. (2008)).Moreover, pre-extraction of lignocellulosic biomass has been shown to be an efficientway of extracting biomass components while also improving the digestibility of theresidual biomass and preserving pulp quality (van Heiningen (2006)). Thus, a novelway to solve the silica problems encountered when pulping bamboo could involvepre-extraction of silica along with hemicellulose prior to pulping processes.363.1. Dissolution of Pure Amorphous Silica in Sodium Hydroxide SolutionIn this chapter, alkali was used as the reagent for the dissolution of silica from bam-boo biomass. To understand the reaction mechanism during alkaline pre-extractionof bamboo, four separate experiments were conducted and presented, i.e.(a) dissolution of pure amorphous silica particles with different sizes to investigatethe mechanism of silica reaction.(b) alkaline treatment of bamboo powder to study the chemical reactions occurringduring alkaline treatment.(c) alkaline treatment of bamboo chips to understand the mass transfer effectsduring the removal of silica and other bamboo components .(d) with response surface methodology in order to examine the mechanisms of al-kaline treatment of bamboo chips and implement the proposed technology intoan industrial scale process.3.1 Dissolution of Pure Amorphous Silica inSodium Hydroxide Solution3.1.1 IntroductionSilicon (Si) is the second-most abundant element after oxygen in the Earth’s crust andis critical to many geochemical and biochemical processes. Plants of Equisetaceae,Cyperaceae and Poaceae are rich in silica compared to hardwoods and softwoods. Be-longing to the family Poaceae, bamboo takes up the monomeric silicic acid, Si(OH)4,373.1. Dissolution of Pure Amorphous Silica in Sodium Hydroxide Solutionfrom the soil during the growth (Sangster and Parry (1981)). Most of silica is ac-cumulated as amorphous hydrated silica (SiO2 ·nH2O) in the plant body with littlecrystalline phases (Motomura et al. (2006); Sterling (1967)). Once deposited, the sil-ica in the tissue is immobile and the process appears to be irreversible. Three typesof silica deposits have been recognized in vascular plants (bamboo included), whichare silica incrustations of cell walls, silica infilling of the interior of cells (cell lumen)and silica extracellular deposits such as the silica layer associated with outer cuticle ofleaf or stem of plant (Fauteux et al. (2005); Le et al. (2015); Richmond and Sussman(2003)). These deposits have been suggested to play a role in mechanical strengthand stability of tissues, protection against fungi, insects and herbivores, resistanceto drought, improvement of light interception, alleviation of problems caused by nu-trient deficiency and excess, and increase of photosynthetic efficiency (Raven (1983);Prychid et al. (2003); Ma and Yamaji (2006)).One of the key steps to efficiently implement bamboo in the traditional pulp andpaper industry is to remove silica. In order to understand the extraction mechanismof silica from bamboo, understanding the reaction kinetics between pure amorphoussilica and NaOH is of great importance.The term silica refers to the chemical compound silicon dioxide, with the chemicalformula SiO2, but as a monomer it is never found in this form. In nature, silica consistsof three different forms viz., crystalline silica, vitreous silica , and amorphous silica.In most forms of silica and silicate minerals, the basic chemical structural unit isthe tetrahedral arrangement of four oxygen atoms surrounding a central silicon (Si)atom, shown as Fig. 3.1 (Bergna and Roberts (2005)). Silica found in nature ismade up of three-dimensional branched chains of alternating silicon and oxygen atoms383.1. Dissolution of Pure Amorphous Silica in Sodium Hydroxide SolutionFigure 3.1: Basic structural unit of silica(tetrahedrons or octahedrons), in which the chains terminate in hydroxyl groupslinked to the silicon atom (Cox (1993)). Different arrangements of tetrahedrons oroctahedrons in silica yield different structures and thereby properties. For example,the random packing of tetrahedrons in amorphous silica makes it less stable than theordered dense packing of tetrahedrons in quartz (Bergna and Roberts (2005); Le et al.(2015)).At ordinary temperatures silica is chemically stable and resistant to many commonreagents, especially acids such as sulfuric acid (H2SO4) and hydrochloric acid (HCl)(Cox (1993)). However, it undergoes a wide variety of chemical transformationsunder harsh conditions such as high temperature and high pH (Brückner (1970)).Moreover, the reactivity of silica is greatly dependent on the crystalline structure,treatment conditions, and state of subdivision of the particular sample investigated.For example, finely divided amorphous silica is considerable more reactive than otherforms of silica (Sierka and Sauer (1997)).With regards to the solubilization of silica in water (solubility), it is reported to393.1. Dissolution of Pure Amorphous Silica in Sodium Hydroxide Solutionbe in accordance to the reversible reaction (Niibori et al. (2000)), asSiO2(s) + 2 H2O −−⇀↽− Si(OH)4(aq) (3.1)where s and aq represent solid and aqueous phases, respectively. The solubility ofsoluble Si(OH)4 in water is affected by various factors such as silica structure, temper-ature and pH (Wirth and Gieskes (1979)). Quartz is the thermodynamically stableform of silica below 870 oC (Bergna and Roberts (2005)) and is essentially water-insolubile with a solubility of around 10 ppm at 25 oC (Lier et al. (1960)). On thecontrary, amorphous silica has a higher solubility, which is about 100 ppm at roomtemperature (Iler (1979)). The solubility also increases with increasing the temper-ature, which in the case of amorphous silica increases to about 900 ppm at 200 oC(Gunnarsson and Arnórsson (2000)). Moreover, the presence of impurities such assodium sulfate (Na2SO4) and sodium chloride (NaCl) in the solution increases thesolubility of silica (Chen and Marshall (1982)). In addition, both low and high pHvalues increase the solubility of silica (Fleming and Crerar (1982)). At very low pH(< 2) the solubility increases by the reaction of Si(OH)4 with H+. The solubilityappears relatively constant from pH 2 to pH 9 at room temperature, but increasesconsiderably above pH 9 with more than a six-fold increase in solubility realized atpH 11 due to the formation of silicate ion in addition to Si(OH)4 (Alexander et al.(1954); Iler (1979)).Ordinary acids do not react with silica except for hydrofluroic acid (HF) and phos-phoric acid (H3PO4) (at elevated temperatures) (Talvitie (1951); Blumberg (1959)).In contrast, silica shows an acidic character by its reaction with a large number of403.1. Dissolution of Pure Amorphous Silica in Sodium Hydroxide Solutionbasic oxides to form silicates (Zaitsev et al. (2000)). For example,2 Na2O + SiO2 −−→ Na4SiO4 (3.2)Na2O + SiO2 −−→ Na2SiO3 (3.3)Na2O + 2 SiO2 −−→ Na2Si2O5 (3.4)Silica is known to dissolve in both sodium hydroxide (NaOH) and sodium car-bonate (Na2CO3) solutions. The reaction rates increase with increasing the reactiontemperature. The reactions are occurring according to following schemesSiO2 + 2 NaOH −−→ Na2SiO3 + H2O (3.5)SiO2 + Na2CO3 −−→ Na2SiO3 + CO2 ↑ (3.6)The formed sodium silicate is soluble in aqueous solution. However, when thepH of the alkali solution is reduced from above pH 11 to between pH 9-10, sodiumsilicate will be converted into insoluble silicic acid. This pH adjustment step can beachieved by the use of mineral acids such as sulfuric acid (H2SO4) or carbon dioxide(CO2) gas. As an example, with using CO2 treatment, the reaction can be expressedasNa2SiO3 + 2 CO2 + 2 H2O −−→ H2SiO3 ↓+ 2 NaHCO3 (3.7)In addition, during the dissolution of silica in alkaline solutions, the presence of413.1. Dissolution of Pure Amorphous Silica in Sodium Hydroxide Solutionsome polyvalent cations such as Ca2+, Mg2+, Fe3+, Al3+ and Cu2+ will interact withsilica causing co-precipitation and forming insoluble deposits (Le et al. (2015); Eastet al. (2013)), in turn reducing the dissolution of silica particles.Since the silica in bamboo is mainly in amorphous phase (Motomura et al. (2006)),we review and study the chemistry of amorphous silica in this work. Several researchgroups have studied the kinetics and mechanisms of the dissolution of pure amorphoussilica in aqueous solutions of alkali. They assumed the dissolution of silica to followequilibrium reaction kinetics and assumed that the dissolution rate is as a function ofthe reactive surface area (Anatskii and Ratinov (1969); Greenberg and Price (1957);Iler (1979); Okunev et al. (1999); Thornton et al. (1988)). The dissolution rate wasmodeled as dependent on temperature, time, and hydroxide ion (OH– ) and silicaconcentrations according the spherical shrinking core model. The reaction order withrespect to [OH– ] in the proposed models remains an open question. Therefore, theobjective of this part of work was to examine whether the dissolution of silica in NaOHsolution follows the spherical shrinking core model under the studied conditions.3.1.2 Materials and MethodsAmorphous silica particles with three diameters (0.2-0.3, 250-400, and 850-1000 µm)were purchased from Sigma-Aldrich (Saint Louis, MO, USA) with a purity higherthan 99%. All chemicals used in this study were reagent grade and solutions wereprepared with purified deionized water.The experiments on the dissolution of amorphous silica in NaOH solution were car-ried out in a round bottomed three-necked flask, equipped with a stirrer, thermometer423.1. Dissolution of Pure Amorphous Silica in Sodium Hydroxide SolutionTable 3.1: A summary of the experimental conditions tested for the dissolution of pureamorphous silica.Silica particle size (µm) Temperature (oC) [OH−]o (mol/L) time (min)0.2-0.3 70 0.45 0 < t < 90250-400 70 0.15 0 < t < 100250-400 70 0.30 0 < t < 100250-400 70 0.45 0 < t < 100850-1000 70 0.45 0 < t < 130Note: [OH– ] and t refer to NaOH concentration and reaction time.and pH meter, immersed in a laboratory-scale heated oil bath. The experimental con-ditions used for the dissolution of pure amorphous silica particles are summarized inTable 3.1. In all cases, the liquid to silica ratio was kept constant at 100:1.12 (L : kg).A flask with 0.5 g pure amorphous silica particles and the calculated volume of wa-ter was placed in the bath for 15 min to heat up to 70 oC. Then, the calculatedvolume of NaOH solution (stock concentration of 100 g/L) was added into the mix-ture to maintain the target NaOH concentrations (0.15-0.45 mol/L) and the desiredliquid-to-silica ratio. During the reaction process, he stirring rate was fixed at 100rpm. At the end of the experimental run, the samples in the reactor were filteredwith the membrane filter of pore size of 0.02 µm, the dissolved silica in the filtratewas determined photometrically by the yellow silicomolybdate method (Tong et al.(2005)). Undissolved silica particles on the membrane were collected and oven driedat 105 ± 2 oC to constant weight. The weight of the solids was determined withan analytical balance. The particle size distribution of the silica particles of varioustimes (0-130 min) during the reaction was measured by a Mastersizer 2000 (Malvern,United Kingdom). The residual alkali concentration of the filtrate was measured. Allexperiments were completed in triplicate.433.1. Dissolution of Pure Amorphous Silica in Sodium Hydroxide SolutionAfter the NaOH treatment, the residual fraction of silica particles was calculatedasα =the amount of residual silicathe amount of initial silica introduced into the alkaline reaction system(3.8)where α is the fraction of residual silica during the dissolution process.3.1.3 Results and DiscussionIn this section, the goal of alkaline dissolution of pure amorphous silica is to determinethe reaction rate between silica and NaOH in the absence of competing reactions andmass transfer effects created by bamboo biomass. It is well known that the basicchemical structural unit of silica is the tetrahedral arrangement of four oxygen atomssurrounding a central silicon (Si) atom. When silica is put into a NaOH solution,several reactions take place simultaneously (Wirth and Gieskes (1979)), i.e.SiO2(s) + H2O −−⇀↽− H4SiO4(aq) Ksp = 5.32× 10−4M (3.9)mSiO2(s) + 2 NaOH → Na2O ·mSiO2(aq) + H2O m = 1− 4 (3.10)Fig. 3.2 shows the mass fraction of residual silica (α) against the reaction time.As shown, under the same NaOH concentration (0.45 mol/L) decrease the particlesize of initial silica powders resulted in the faster dissolution of silica particles. Forexample, the times needed for the dissolution of 98% of initial silica particles with thesize ranges of 0.2-0.3, 250-400 and 850-100 µm were 30, 80 and 110 min, respectively.This could be attributed to the fact that silica particles with smaller particle size443.1. Dissolution of Pure Amorphous Silica in Sodium Hydroxide Solution0 20 40 60 80 100 120 1400.00.20.40.60.81.0 Fraction of residual silica ()Time (min) 0.2-0.3 m 250-400 m 850-1000 mFigure 3.2: Effect of particle size on the dissolution rate of silica (temperature = 70 oCand NaOH = 0.45 mol/L).453.1. Dissolution of Pure Amorphous Silica in Sodium Hydroxide Solutionhave a larger total reaction surface area than that of silica with larger particle size atthe same mass in the reaction system (Mgaidi et al. (2004)).To study the effect of NaOH concentration on the dissolution of pure amorphoussilica. Silica powder with particle size of 250-400 µm was used (Fig. 3.3). Increasingthe NaOH concentration resulted in faster dissolution of the silica. For example,when treating 250-400 µm silica powder with NaOH solution having a concentrationof 0.15 mol/L, 89% of initial silica was dissolved in 60 min, whereas, increasing theNaOH concentration to 0.45 mol/L, 95% of silica was dissolved in 60 min. One likelyreason is that the contact probability of silica and hydroxide ion is increased withmore hydroxide ion in the reaction system, resulting in faster depolymerization ofsilica.To interpret the experimental data and describe the silica dissolution in NaOHsolution, the classic shrinking core model was applied, where the reaction rate isconsidered to be proportional to the surface area of the unreacted silica particles(Niibori et al. (2000); Mgaidi et al. (2004)). In applying the spherical shrinking coremodel, the surface of silica particle is assumed to have the equal reactivity and asmooth reaction interface. Thus, the rate of reaction was assumed asr1 = k1A[OH−]β (3.11)where r1 is the reaction rate of silica (mol/min), k1 is the reaction rate constant perunit surface area [mol/(m2 · min)], A is the reactant surface area for the reaction(m2), [OH– ] is the concentration of hydroxide ion (mol/L), and β is the reactionorder with respect to hydroxide ion concentration ([OH– ].463.1. Dissolution of Pure Amorphous Silica in Sodium Hydroxide Solution0 20 40 60 80 1000.00.20.40.60.81.0Fraction of residual silica ()Time (min) 0.15 mol/L 0.30 mol/L 0.45 mol/LFigure 3.3: Effect of NaOH concentration on the dissolution rate of pure amorphous silica(temperature = 70 oC and silica particle size = 250-400 µm).473.1. Dissolution of Pure Amorphous Silica in Sodium Hydroxide SolutionBased on Equation 3.11, the dissolution rate can be obtained as a function ofundissolved silica (α)d(1− α)dt=3MRoρk1[OH−]βα2/3 (3.12)where M is the molecular weight of SiO2 (kg/mol), Ro is the radius of the initial silicaparticles (m), respectively, and ρ is the density of silica (kg/m3).Since the NaOH charge used for the silica dissolution is in stoichiometrically largelyexcessive (≥ 5:1), [OH– ] could be considered to be constant during the reactionprocess. The time required for the conversion (1− α) is given by1− α1/3 = k2t k2 = Mk1Roρ[OH−]β (3.13)Experimental data shown in Figs. 3.2 and 3.3 were tested with Equation 3.13.The utility of Equation 3.13 was shown in Figs. 3.4 and 3.5. Straight lines passingthrough the origin were obtained for each particle size and NaOH concentration,indicating that Equation 3.13 could describe the silica dissolution process under thestudied conditions. The results of the nature of the rate-controlling mechanism wasa function of the extent of dissolution with the reaction order of 23with respect tosilica concentration, which is in agreement with the reported results of 0.6-0.8 onthe dissolution of silica aerogel into NaOH solution (Okunev et al. (1999)). Table3.2 shows the rate constants k2 obtained from Figs. 3.4 and 3.5. As a reaction rateconstant, k1 is independent of silica particle size. Based on the values of k2, it canbe deduced that the value of k1 ranging from 5.8-6.5 (mol0.36L0.64m−2min−1) wasobtained.483.1. Dissolution of Pure Amorphous Silica in Sodium Hydroxide Solution0 20 40 60 80 100 120 1400.00.20.40.60.81.0r2 = 0.98r2 = 0.97 0.2-0.3 m 250-400 m 850-1000 m 1 - Time (min)r2 = 0.98Figure 3.4: Plot of 1 − α1/3 versus reaction time (t) with different silica particle sizes(temperature=70 oC; [OH−]o = 0.45 mol/L).493.1. Dissolution of Pure Amorphous Silica in Sodium Hydroxide Solution0 20 40 60 80 1000.00.20.40.60.81.0r 2 = 0.97r 2 = 0.980.15 mol/L0.3 mol/L0.45 mol/L1-Time (min)r 2 = 0.96Figure 3.5: Plot of 1 − α 13 versus reaction time (t) with different initial NaOH concen-trations (0.15-0.45 mol/L). Experiments were carried out with initial silica particle size of250-400 mum and at 70 oC.503.1. Dissolution of Pure Amorphous Silica in Sodium Hydroxide SolutionTable 3.2: Reaction rate constant (k2) calculated by experimental data fittingSilica particle size (µm) [NaOH] (mol/L) k2 (min−1))0.2-0.3 0.45 0.024250-400 0.15 0.0091250-400 0.3 0.014250-400 0.45 0.018850-1000 0.45 0.0071For silica with the same particle size, the value of k1 should be constant. Thus,the relationship between k2 and k1k1 =k2Roρ3M [OH−]β(3.14)gives an approach to estimate the value of β. With the values of k2 of experimentalresults with the same silica particle size (250-400 µm) (table 3.2), we could estimatethe value of β = 0.64 by assuming the [OH– ] is constant during the reaction process.From above results (Figs. 3.4, 3.5 and Table 3.2), it can be concluded that theshrinking core model is able to describe silica dissolution in NaOH solution.Moreover, during the dissolution of pure amorphous silica in NaOH solution, theresidual alkali concentration is given by the following equation[OH−] = [OH−]o −2m([SiO2]o − [SiO2]) +Kw[OH−]Kw = 10−14 (3.15)which is a second-order polynomial in [OH– ]. Clearly, the residual OH– concentra-tion is given by the root of the polynomial. This equation was solved numerically for[OH– ] in MATLAB using the built-in root finding procedure. Fig. 3.6 presents thecomparison of calculated [OH– ] values against experimental data for the dissolution513.1. Dissolution of Pure Amorphous Silica in Sodium Hydroxide Solution0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.450.100.150.200.250.300.350.400.45 [OH] calculated (mol/L)[OH ] experimental (mol/L)Figure 3.6: Comparison between the experimental data and the predictions by Equation3.15 for OH– concentration. The parameter is set m = 1 for the calculation. Data pointsare from experimental results.of amorphous silica in NaOH solutions. By using a linear regression, it was foundthat the model prediction is very good, with a regression coefficient (r2) of 0.982. Ad-ditionally, m = 1 was obtained from the plot of predicted [OH– ] against experimental[OH– ].One of the limitations of the classic shrinking model is that it does not considerparticle size distribution during the dissolution process. Taking the silica particlesrange 250-400 µm as an example, the particle size distribution during the reaction atan initial NaOH concentration of 0.15 mol/L was determined (Fig. 3.7). After heat-523.1. Dissolution of Pure Amorphous Silica in Sodium Hydroxide Solution0 100 200 300 400 500 600 7000510152025 Raw silica After heat-up t=30 minDifferential Volume (%)Mean particle diameter ( m)Figure 3.7: Particle size distribution of the solid particles at different reaction times (initialsilica particle size = 250-400 µm, temperature = 70 oC, initial [OH– ] = 0.15 mol/L).up period, about 15% of the volume of raw silica materials with particle diameterless than 60 µm disappeared and the volume diameter was decreased to about 300µm. As the chemical reaction proceeded, the volume diameter of silica particles wasfurther reduced. The volume diameter was about 150 µm after 30 min. These resultsconfirmed that silica with smaller particle size dissolves faster compared to larger size.Thus, from the results obtained, it was reasonable to conclude that the dissolutionrate of amorphous silica in NaOH solution is controlled by the reactant surface area.533.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo Powder3.1.4 SummaryExperimental results on the dissolution of pure amorphous silica particles in NaOHsolution demonstrated that the classic shrinking core model could be used to describethe silica dissolution rate under the alkali concentrations studied in this work. Withthe decrease of initial silica particle size, the dissolution of silica particles into NaOHsolution was much faster compared to that of silica with large larger particle sizes.The reaction order with respect to OH– concentration of 0.64 was obtained under thestudied conditions. This is helpful to understand the removal of silica from bamboobiomass or other high-silica biomasses following alkaline extraction to resolve thesilica associated challenges during biorefinery applications.3.2 Alkaline Pre-extraction of Silica andHemicellulose from Bamboo Powder3.2.1 IntroductionThe degree of silica accumulation is quite different among various tissues of bambootree; distribution can vary even within the same tissue. The silica content in bambooincreases from the stem, through the branches, to the leaves (Lux et al. (2003); Dinget al. (2008)). In the leaf, silica is mainly accumulated in epidermal cells, with thehighest levels in specialized idioblasts (silica cells) and lower levels in the mesophyllcells and vascular bundles; silica in the bamboo root is accumulated exclusively inthe endodermal cells (Bennett and Sangster (1981); Motomura et al. (2000); Luxet al. (2003)). In the stem of bamboo, silification mainly takes place as infillings of543.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo Powderthe interior epidermis cells, which are located as the outmost layer of plant tissues(Piperno (2014)). Other tissues in the bamboo culm, such as hypodermal and vasculartissue, may also be silicified, but to a lesser degree than the epidermal tissue (Le et al.(2015); Motomura et al. (2000)).Silica content of the bamboo stem ranges from 0.5-5% (w/w) (Liese (1992)). Incontrast, the silica content of wood is usually less than 0.01% (w/w), with few species,however, having higher silica content (up to 1%) (Song et al. (2013); Torelli and Čufar(1995)). Thus, bamboo silica takes up a considerable part of the biomass.During the utilization of bamboo biomass in biorefineries, rather than taking silicaas a complication, silica in the raw material can be used as a sustainable feedstockfor various high-value products such as for the production of catalysts, thixotropicagents, pharmaceuticals, film substrates, or used as mescoporous structured silicafor adsorption processes and as fillers in cement (Kalapathy et al. (2002); Liou andYang (2011); Klapiszewski et al. (2015)). Accordingly, similar to the pre-extractionof hemicellulose and lignin from biomass, the pre-extraction of silica along with lig-nocellulosic components from raw materials prior to pulping and biofuels productioncould be a promising method to resolve the silica problems.Alkali can be used to dissolve silica and transfer the generated soluble silicatesinto the bulk liquor; this provides a means of extracting silica prior to subsequentprocessing steps (Section 3.1). Alkali dissolution of silica from lignocellulosic biomasscan also be referred as alkaline pretreatment of biomass. Alkaline pretreatment is an-other chemical/hydrothermal pretreatment technology that has been used to extracthemicellulose from biomass. Several alkaline reagents, including sodium hydroxide(NaOH), calcium hydroxide (Ca(OH)2), potassium hydroxide (KOH), aqueous ammo-553.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo Powdernia, ammonia hydroxide (NH4OH), and NaOH incombination with hydrogen peroxide(H2O2), have been investigated for the pretreatment of lignocellulosic feedstocks (Ku-mar et al. (2009); Jun et al. (2012); Walton et al. (2010); Alvarez-Vasco and Zhang(2013)). Although the type of alkali had little effect on the extraction efficiency ofhemicellulose from lignocellulosic feedstocks, it did affect the extent of lignin removal(Jin et al. (2010)). For example, NaOH and KOH have been reported to be more effec-tive in lignin removal while NH4OH has little effect (Huang et al. (2008)). Moreover,different from the pretreatment under acidic conditions, the alkaline pretreatmentcan be carried out at relative low temperatures (room temperature) and wide rangeof times (from seconds to days) (Mosier et al. (2005)).Hydrolysis under alkaline conditions also causes the cleavage of lignin bonds andglycosidic hemicellulose bonds as well as disruption of ester bonds crosslinking ligninand hemicellulose, resulting in the removal of hemicellulose and lignin. The most im-portant alkali-catalyzed reactions include polysaccharides dissolution, deacetylationof hemicellulose, peeling reactions of carbohydrates, and random hydrolysis followedby secondary peeling reactions (Jin et al. (2010); Lehto and Alén (2013)). The alka-line peeling reaction removes terminal anhydro-sugar units to generate new reducingend groups until a competitive stopping reaction begins and forms a stable saccharideacid end group. At the same time, dissolution or/and degradation of lignin, removalof extractives, and saponification of esters (fats and waxes) are occurring (Lehto andAlén (2013)). Alkaline pretreatment also removes acetyl and uronic substitutents onhemicellulose by alkaline saponification (Zhang and Lynd (2004)).Alkaline pretreatment of biomass results in the reduction in the degree of poly-merization and crystallinity of cellulose and swelling of fibres, which increases the563.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo Powdersurface area and accessibility of treated solids to enzymes and chemicals used duringsubsequent processing. In addition, compared to the pretreatments with other alkalis,NaOH pretreatment can also cleave the ester bonds between lignin and/or hemicel-lulose and hydroxycinnamic acids, such as p-coumaric and ferulic acids (Spencer andAkin (1980)), thereby enhancing the removal of hemicellulose and lignin.Compared with acid and autohydrolysis, alkaline treatment is more effective methodat breaking ester bonds between lignin, hemicellulose and cellulose, and limiting frag-mentation of hemicellulosic polymers; many of the sodium salts of organic acids canbe recovered and/or regenerated (Gáspár et al. (2007); Xu and Huang (2014)). Un-der alkaline conditions, softwood glucomannan is rapidly degraded by the peelingreaction while most of xylan in hardwood, bamboo and cereal straw is solubilizedin the oligomer form (van Heiningen (2006); Jin et al. (2010)). Moreover, it hasbeen reported that alkaline pretreatment is regarded to be more effective at remov-ing hemicellulose from xylan-rich lignocellulosic biomasses (hardwoods, bamboo andagricultural residues) than softwoods (van Heiningen (2006); Huang et al. (2010);Yoon and van Heiningen (2010)).Increasing severity by increasing temperature, time, or alkali charge facilitatedsolids dissolution from lignocellulosic biomass (Yoon and van Heiningen (2010); Venaet al. (2013)). Vena et al. (2013) investigated the alkaline pre-treatment of hardwoodsprior to pulping. They reported that the maximum xylan recovery yield of 16% (basedon dry wood mass) could be obtained by increasing the temperature, time and NaOHconcentration to 90 oC, 240 min and 2 mol/L, respectively (Vena et al. (2013)).However, high temperature, long reaction time and high alkaline charge facilitategeneration of carboxylic acids. For example, Lehto and Alén (2015) investigated573.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo PowderNaOH treatment of softwood chips at 130 and 150 oC with 1-8% alkali charge for30-120 min and these treatments resulted in 2.0-13.6% removal of the original drywood material with the main constituents in the dissolved organic fraction beingvarious carboxylic acids (volatile formic and acetic acids and non-volatile hydroxymonocarboxylic and hydroxy dicarboxylic acids). The lower alkali charges were foundto favor the formation of the carboxylic acids.Yoon and van Heiningen (2010) investigated the green liquor hydrolysis of Loblollypine chips at temperature of 170-190 oC with green liquor charge of 2-6% (w/w) (asNa2O) and compared to pure water extraction with final pH of about 4 at 170-190oC (Yoon and van Heiningen (2008); Yoon and van Heiningen (2010)). The resultsdemonstrated that the sugar yields of xylose and mannose were both lower with alkaliaddition; the mannose yield after green liquor hydrolysis was approximately 92.7%lower than that of the autohydrolysis yield. A similar green liquor pretreatmentof mixed hardwood chips dissolved approximately 60% of initial mannan using analkali charge of 8% (w/w) at 160 oC (Jin et al. (2010)). Xylan dissolution increasedwith alkali charge, however, it was difficult to remove more than 25% of the initialxylan even at an alkali charge of 20% (w/w). Nearly 100% cellulose could be preservedduring alkaline pretreatment. The low extent of polysaccharides removal could be dueto the pH being too low to start random hydrolysis and secondary peeling reactions(Jin et al. (2010)).The monomer concentration in the alkaline hydrolysate is lower than that of acidichydrolysate (Yoon and van Heiningen (2010); Lehto and Alén (2013)). Low alkalicharge (< 6% of biomass) favors the monomer production (Lehto and Alén (2013)).Walton et al. (2010) compared hydrolysis of mixed southern hardwood chips with hot583.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo Powderwater and alkali (alkali charge of 2-8%) at the same conditions of 160 oC for 1-2 h.Xylose concentration in alkaline hydrolysate did not show as strong correlation withthe severity of treatment as that was found for water hydrolysis (Walton et al. (2010)).During alkaline pretreatment, glucose units are liberated by alkaline hydrolysis orpeeling reaction from cellulose chains. Due to the high crystallinity and degree ofpolymerization (DP), cellulose is more resistant towards alkaline media and suffers lessdegradation than hemicellulose. For example, less than 2% of cellulose was removedduring alkaline treatment of wheat straw with the utilization of 1.5% NaOH at 20 oCfor up to 144 h while up to 90% of hemicellulose was extracted (Sun et al. (1995)).Several studies have also been conducted with bamboo. Yamashita et al. (2010)investigated alkaline peroxide pretreatment to improve the enzymatic saccharifica-tion of treated bamboo chips. They found that the combination of 1% NaOH with1% (v/v) H2O2 could yield 399 mg/g (initial dry sample) of glucose without usingsevere conditions such as high NaOH concentration or high temperature. Li et al.(2014) subjected bamboo powder (2.0 mm) from different bamboo layers (bamboogreen, bamboo timber and bamboo yellow) to alkaline pretreatment with 6-12% NaOH(w/w) at 180 oC for 30 min. These treatments removed 54.5% of xylose and 60.2%of lignin from bamboo when treating with 12% NaOH (w/w) at 180 oC for 30 min(Li et al. (2014)). This study showed that the dissolution of lignin and hemicellulosewas mainly dependent on the loading of NaOH (Li et al. (2014)). During alkalinepretreatment, the separated and fully exposed micro-fibrils increased the external sur-face area and the porosity, thereby facilitating subsequent enzymatic and chemicalprocessing for the production of dissolving grade pulp and lignocellulosic fuels.However, previous studies on the alkaline pretreatment of bamboo chips prior to593.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo Powderpulping were mainly focused on the extraction of hemicellulose/lignin for the pro-duction of kraft pulp, high-grade dissolving pulp or fermentable sugar (Leenakul andTippayawong (2010); Luo et al. (2013); Yamashita et al. (2010); Li et al. (2014);Sathitsuksanoh et al. (2010)). Limited investigation has been conducted in extract-ing silica from bamboo. Additionally, the extracted silica can also be an excellentresource for silica-derived products (Zhang et al. (2013)). Thus, it would be usefulto fabricate nanosilica from renewable silica-containing biomass material in a cost-effective way.In addition, alkaline pretreatment of lignocellulosic materials can be well-integratedwith an existing alkaline process such as kraft pulping, since it can lower the alkalicharge required in subsequent cooking or bleaching step and hence preserve the pulpquality (Jun et al. (2012); Huang et al. (2008); Huang et al. (2010); Helmerius et al.(2010)). Moreover, the hemicellulose extracted during the alkaline pre-extractionprocess can be used for the generation of value-added products such as bioethanol,furfural, acetone or papermaking additives (Bai et al. (2012); Liu et al. (2013); Maoet al. (2008); Hamzeh et al. (2013)). In addition, for the production of dissolvinggrade pulp, near-complete removal of hemicellulose is required (Sixta (2006)). Al-kaline pre-extraction of hemicellulose also promises an alternative way to producedissolving pulp from high silica biomasses by pre-extracting silica and hemicelluloseprior to pulping.Among different alkali reagents investigated for the removal of silica from bam-boo or other high silica biomasses, NaOH is considered to be a better choice as itis readily available in the form of white liquor in kraft pulping operations and theco-precipitation caused by the interaction of silica with some polyvalent alkali cations603.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo Powder(Ca2+, Mg2+, Fe3+, Al3+ and Cu2+) can be significantly alleviated (Le et al. (2015)).Within this context, the pre-extraction of silica prior to subsequent commercial pulp-ing may not only be able to solve the silica problems of using bamboo in kraft pulpingbut also would add value and increase revenue to the mill.Moreover, it has been suggested that chip size play an important role in the ex-traction of hemicellulose or other products from wood chips (Brennan and Wyman(2004); Rissanen et al. (2014a)). Thus, to minimize the mass transfer effects on theextraction of silica from bamboo, alkaline pretreatment of bamboo powder was eval-uated. In the work presented in this section, alkaline pretreatment was carried out tocompletely extract silica and partially separate hemicellulose from bamboo powder.The effects of pretreatment conditions (NaOH concentration, temperature and time)on silica and hemicellulose pre-extraction were investigated. The comparison of thedissolution of silica from bamboo powder and pure amorphous silica was made to un-derstand the mechanism for silica extraction. Moreover, chemical reactions involvedin alkaline treatment of bamboo biomass are proposed and a toy model to describethe evolution of OH– concentration is given.3.2.2 Materials and MethodsThe washed commercial bamboo chips used in Chapter 2 were taken as the rawfeedstock. Some dried bamboo chips were ground using a Wiley Mill and sieved to aparticle size of 40-60 mesh. The bamboo powder was collected in glass jars for furtheruse. All chemicals used were reagent grade and solutions were prepared with purifieddeionized water.613.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo PowderTable 3.3: A summary of experimental conditions investigated for alkaline treatment ofbamboo powder.Temperature (oC) time (min) [OH−]o (mol/L)70 0 < t < 180 0.15-0.52580 0 < t < 180 0.15, 0.4590 0 < t < 180 0.15, 0.45100 0 < t < 180 0.15, 0.45[OH−]o and t refer to initial NaOH concentration and time.Alkaline pre-extraction experiments on bamboo powder (40-60 mesh) were car-ried out in 4 silicate glass bottles of 500 mL capacity immersed in a laboratory-scaleheated oil bath. A series of isothermal pre-extraction experiments were conductedover various temperatures (70-100 oC), times (5-180 min) and initial NaOH concen-trations (0.15-0.45 mol/L) at constant liquid-to-wood ratio (10 L/kg) (Table 3.3).For an alkaline pre-extraction run, bamboo powder of 20 g oven dried (o.d.) and thecalculated volume of deionized water and NaOH solution (stock concentration of 100g/L) were mixed and placed in a reactor. Subsequently, the reactor was placed in theoil bath pre-heated to the target temperature. After the treatment, the vessels wererapidly cooled down in an ice/water bath. The pre-treated bamboo powder was sep-arated from the liquor through filtration. The liquor was collected and then stored at4 oC for the compositional analysis. The treated powder was washed thoroughly withdeionized water to remove dissolved substances and collected for component analysis.All experiments were performed in triplicate.The chemical composition was determined according to the NREL standard pro-tocol described in Chapter 2. The ash and silica content of the solid samples weremeasured based on the methods described in Chapter 2. The silica content of theliquor was measured by using the silicon molybdenum blue photometric method (Tong623.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo Powderet al. (2005)). Briefly, 1 mL APEL was dissolved in 10 mL HNO3 solution. Aftershaking, 10 mL ammonium molybdate ((NH4)6Mo7O24) was added into the solution.The solution was gently heated at 30 oC for 12 min in a shaking water bath (75rpm). After cooling, 40 mL ammonium ferrous sulfate ((NH4)2Fe(SO4)2 ·6H2O) wasadded to the solution. The resultant solution was made up to 250 mL with a volu-metric flask. The silica content was measured at wavelength 813 nm with a UV-visspectrophotometer. The residual alkali concentration in the liquor was determinedby titration with hydrochloric acid (HCl) according to TAPPI T 625 cm-85.3.2.3 Results and DiscussionApplication of Shrinking Core Model in Silica Removal from BambooPowderThe composition of the bamboo feedstock used in this study has been determined andshown in Table 2.2 in Chapter 2. In this section, the main purpose of alkaline pre-extraction of bamboo powder is to understand the reaction kinetics on alkaline pre-extraction of bamboo biomass by minizing the mass transfer effects. The other goal isto preserve lignin and cellulose in the pretreated biomass. This is because the presenceof lignin in the spent pre-extraction liquor hampers its utilization in bioethanol orxylitol production as lignin degradation products inhibit the growth and metabolicactivity of micro-organisms used in bioconversion processes. Moreover, silica recoverythrough lowering the pH of alkaline pre-extraction liquor is also negatively affecteddue to the co-precipitation of lignin (Minu et al. (2012); Shi et al. (2011)). Sincepre-extraction processes using high temperature and low alkali concentration have633.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo Powderseveral drawbacks such as high capital investment cost and low molecular-mass ofthe extracted hemicellulose (Jun et al. (2012); Yoon and van Heiningen (2010)), highalkali concentration and relatively lower reaction temperatures were investigated inthis work.To minimize the mass transfer limitations, bamboo powder (250-400 µm) was usedfor the alkaline treatment. To verify the utility of shrinking core model in modellingthe extraction of silica from bamboo powder, alkaline treatment was carried out atthe same temperature (70 oC) as used for the dissolution of pure amorphous silicaunder various NaOH concentrations (Fig. 3.8). As shown, the extraction of silicaincreased with the increase of NaOH concentration.Fig. 3.9 shows the results of the application of Equation 3.13 to describe the dis-solution of silica from bamboo powder. Here, it can be observed that large deviationswere obtained when plotting (1 − α 13 ) against reaction time t (min) at lower NaOHconcentrations (< 0.45 mol/L), indicating that the removal of silica from bamboopowder does not follow the fitting of standard shrinking core model. This might bedue to competing reaction for the NaOH by wood components such as such as acetyland uronic acid groups, xylan, and other extractives. When increasing the NaOHconcentration, straight lines with r2 > 0.95 passing through the origin were obtained,which means that the shrinking core model works at the high OH– concentration.One explanation for this is that when using high OH– concentration, the stoichio-metric ratio of OH– to silica is excessively large throughout the reaction process.Thus, it can be concluded that the increase of NaOH concentration could increasethe silica dissolution rate from bamboo powder.643.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo Powder0 20 40 60 80 100 120 140 160 1800.00.20.40.60.81.0Fraction of residual silica ()Time (min) 0.15 mol/L 0.225 mol/L 0.30 mol/L 0.375 mol/L 0.45 mol/L 0.525 mol/LFigure 3.8: Effect of NaOH concentration on the extraction of silica from bamboo powderat 70 oC.653.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo Powder0 20 40 60 80 100 120 140 160 1800.00.10.20.30.40.50.60.70.80.9r2 = 0.95 0.15 mol/L 0.225 mol/L 0.30 mol/L 0.375 mol/L 0.45 mol/L 0.525 mol/L1 - Time (min)r2 = 0.98Figure 3.9: Application of shrinking core model on silica removal from bamboo powder(temperature = 70 oC).663.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo Powder X XO AcX XO GX XLCX XO AcX XO GXnFigure 3.10: A schematic of the idealized hemicellulose - lignocellulose (LC) substrateconsidered in this work. X-X-X-X represents the backbone of the xylan structure. Thespecies Ac, Ar and G, which represent the acetyl, arabinose and glucuronic acid groups, areinitially bound to the xylan chain but are released through alkaline treatment. Silica andother ash components (MO) are not shown in this figure but are considered to be physicallyembedded in the LC portion of the matrix.Toy ModelDuring alkaline pretreatment of bamboo biomass, the behavior of NaOH is complex.There are a number of pathways governing OH– consumption and generation. Basedon the discussion and definition in Chapter 2, an idealized hemicellulose structure foruse in modelling the alkaline pretreatment is proposed and shown in Fig. 3.10.For establishing the chemistry of the toy model, hydroxyl groups are representedby OH– and protons by H+; both of these species are considered to be in aqueousphase and the aq notation has been dropped. Sodium hydroxide (NaOH) is the alkalibeing added to the system. The acetyl group Ac and glucuronic acid G are defined asH3C−C(−O)− and C5H9O5COOH. During the alkaline pre-extraction of silica frombamboo, several reactions take place simultaneously.As disscussed in Section 3.1, when silica is put into the NaOH solution, severalreactions take place simutaneously (Wirth and Gieskes (1979)). The dissolution rateof silica in NaOH is considered to be a function of total reactant surface area of silicapartciels (Niibori et al. (2000)). For simplicity, we consider the dissolution of silica673.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo Powderfollows Equation 3.10 and the dissolution rate is given with Equation 3.11, in whichri and ki are defined as the chemical reaction rate and rate constant, respectively,A is the total surface area of silica particles, β is the reaction order with respect to[OH– ].We consider alkali consumption by Ac and G, which are cleaved from the hemi-cellulose backbone through an alkaline saponification of the ester bonds. Based onthe stoichiometry, the reactions between NaOH and acetic and glucuronic acid followfirst order kinetics.XOAc + NaOHk2−−→ Ac−O−Na+(aq) + XOH(s)r2 = k2[XOAc][OH−](3.16)XOG + NaOHk3−−→ G−O−Na+(aq) + XOH(s)r3 = k3[XOG][OH−](3.17)where ri and ki are defined as the chemical reaction rate and rate constant, respec-tively.These two reactions may occur with acetyl and glucuronic groups which are at-tached to either soluble or solid phases of the hemicellulose. For simplicity, anydifferences in rate between the alkaline saponification reaction occurring in the solidor liquid phases are ignored. As the products AcONa(aq) and GONa(aq), sodiumacetate and sodium glucuronate, behave as salts with strong alkaline base; they adopt683.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo Powderthe following equilibria in solutionAc−O−Na+(aq) + H2O KAcONa−−−−⇀↽ − Ac−OH + OH− + Na+KAcONa =[AcOH][OH−][AcONa]= 5.56× 10−10(3.18)G−O−Na+(aq) + H2O KGONa−−−−⇀↽ − G−OH + OH− + Na+KGONa =[GOH][OH−][GONa]= 8.51× 10−12(3.19)where Ki, from this point is defined as the equilibrium constant and the value quotedis at room temperature (Avdeef et al. (1993); Wang et al. (1991) ).We consider xylan dissolution under mild alkaline conditions (< 140 oC) is mainlyderived from the peeling reaction (Sixta (2006)). Although the stopping reaction alsooccurs, but compared to peeling reaction, the stopping reaction rate is much slower,which is ignored in this modelling process. The degradation of xylan is assumed tofollow the mechanismXn(s)kp1−−−→OH−Xn−1(s) + DCOOH(aq) n = 2, 3...rp1 = kp1[Xn][OH−](3.20)DCOOH + NaOHk4−−→ DCOO−Na+ + H2Or4 = k4[DCOOH][NaOH](3.21)where DCOOH is the degradation product of xylan (xyloisosaccharinic acid). Theneutralization of the formed degradation acid (Equation 3.21) consumes the alkali inthe pretreatment liquor. Since the rate of neutralization reaction is fast, the degra-dation rate of xylan is mainly controlled by Equation 3.20. The disassociation of693.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo PowderDCOO–Na+ in solution (Perrin et al. (1981))DCOO−Na+ + H2OKD−−⇀↽− DCOOH + OH− + Na+KD =[DCOOH][OH−][DCOO−Na+]= 1× 10−9(3.22)serves as a source of OH– .In a similar manner, cellulose, a linear polymer of repeating sugar units of glucose,also undergoes degradation to some extent by the peeling reactionCn(s)kp2−−−→OH−Cn−1(s) + ECOOH(aq) n = 2, 3...rp2 = kp2[Cn][OH−](3.23)where ECOOH is the degradation product of cellulose (glucoisosaccharinic acid). Theneutralization of ECOOHECOOH + NaOHk5−−→ ECOO−Na+ + H2Or5 = k5[ECOOH][NaOH](3.24)ECOO–Na+ adopts the following equilibrium in solution (Käkölä and Alén (2006))ECOO−Na+ + H2OKE−−⇀↽− ECOOH + OH− + Na+KE =[ECOOH][OH−][ECOO−Na+]= 1.58× 10−11(3.25)Since the neutralization reaction takes place fast, we consider the reactions givenin Equations 3.21 and 3.24 to be instantaneous. To continue, lignin might also be703.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo Powderremoved to some extentLignin(s) + b[OH−] kL−−→ Lignin(s) + P (aq)rL = kL[Lignin][OH−]b(3.26)where b is the stoichiometry parameter with respect to NaOH and P is the degradationproduct of lignin. Since alkaline delignification of biomass starts at temperaturehigher than 100 oC (Sixta (2006)), the dissolved lignin could be mainly mono- oroligo-lignols (Arato et al. (2005)).In the reaction system, water disassociation demandsH2OKw−−⇀↽− OH− + H+ Kw = [OH−][H+] = 1× 10−14 (3.27)and this serves as an additional source of OH– .In addition to these equilibria, the disassociation of NaOH is the main source ofOH– in the reaction systemNaOHfast−−→ Na+ + OH− (3.28)As NaOH is a strong alkali, we consider the disassociation given in Equation 3.28 tobe instantaneous. The final aspect to consider is the dissolution of ash in water. Asmentioned above the reaction scheme depends upon the species involved. Here, weconsider a hypothetical oxide MO which reacts according to the following schemeMO(s) + H2O −−⇀↽− M(OH)2 (aq) (3.29)713.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo PowderThe dissolution of ash (MO) in the solution is assumed to be instantaneous.M(OH)2(aq)Km−−⇀↽− M2+ + 2 OH−Km =[M2+(aq)][OH−]2[M(OH)2]→ 0(3.30)As the equilibrium constant Km is unknown, we simply assign this value to be a verysmall number to reduce the number of free parameters. It should be noted that wedo not characterize a number of the potential secondary reactions in solutions, eventhough they may affect the OH– levels to a small degree. For example, we ignore thestopping reactions of xylan and glucan with NaOH for mathematical transparency asthese have much slower reaction rate compared to that of the peeling reaction. Theprecipitation of metal hydroxides was also ignored for mathematical transparency.Since the alkaline treatment temperature used in this study was low (≤ 100 oC),the degradation of cellulose and lignin can be ignored. Thus, reactions shown in Fig.3.11 can be used to describe the dominant mechanisms during alkaline treatment ofbamboo.Having established the chemistry of the toy model, we now construct the math-ematical model. We build the model upon two conservation laws: conservation ofmass of each of the species found in solution and an overall charge neutrality of thesolution. Conservation of mass expresses that the initial moles of a certain speciesmust sum to the total moles of the species in the reaction products. For example, theinitial moles of M is [MO]o, must balance the number of moles of M, in the species of[MO], [M2+], and [M(OH)2] at any time throughout the courses of the reaction. This723.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo PowderFigure 3.11: A schematic of the idealized reaction scheme. The chemical reactions shownpresent the dominant reaction mechanisms during alkaline treatment of bamboo biomass.can be expressed as[MO]o = [MO] + [M2+] + [M(OH)2]= [MO] + [M2+](1 +[OH−]2Km) (3.31)through use of the equilibrium relationship given in Equation 3.30. In a similarmanner, conservation of mass for the species Si, Ac, G, Na+ can be expressed as[SiO2]o = [SiO2] + [H4SiO4] +m[Na2O ·mSiO2] (3.32)733.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo Powder[XOAc]o = [XOAc] + [AcO−Na+] + [AcOH]= [XOAc] + [AcONa](1 +KAcONa[OH−]) (3.33)[XOG]o = [XOG] + [GO−Na+] + [GOH] = [GOAc] + [GONa](1 +KGONa[OH−])(3.34)[NaOH]o = 2[Na2O ·mSiO2] + [AcONa] + [GONa]+ [DCOONa] + [Na+]= 2[Na2O ·mSiO2] +([AcOH][OH−]KAcONa)+([GOH][OH−]KGONa)+([DCOOH][OH−]KD)+ [Na+] (3.35)with use of Equations 3.10, 3.11, 3.18, 3.19, and 3.22. To continue, the charge neu-tralization conservation equation is invoked, i.e.[Na+] + [H+] + 2 [M2+] = [OH−] (3.36)Thus, Equation 3.36 can be expressed as[OH−] = [OH−]o −2m([SiO2]o − [SiO2])− [XOAc]o − [XOAc](1 + KAcONa[OH−])− [XOG]o − [XOG](1 + KGONa[OH−])− [X]o − [X](1 + KD[OH−])+ Kw[OH−]+ 2([MO]o − [MO]1 + [OH−]2Km)(3.37)743.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo Powderthrough use of Equations 3.31-3.35. This equation (Equation 3.37) gives the residualOH– concentration in the reaction system during the process of alkaline treatment ofbamboo powder. This equation also indicates that the residual alkali concentrationin the solution is governed by silica dissolution, alkaline saponification of acetyl andglucuronic acid groups, the amount of neutralization by the formed degradation prod-ucts from xylan and cellulose, five different equilibria found in solution (Km, KAcONa,KGONa, KD), and the amount of alkali initially added ([NaOH]0).Removal of Silica and Hemicellulose from Bamboo PowderA series of experiments were carried out over a wide range of reaction conditions (seeMaterial and Methods Section). The chemical composition of the treated sampleswas analyzed. As an example, the composition of bamboo powder obtained from theextraction at 100 oC is shown in Table 3.4. The yields are reported in percentage andrelated to original oven-dried bamboo mass. As can be observed from Table 3.4, thetreatment yield decreased with increasing the extraction severity. For example, at 100oC, the use of 0.15 mol/L NaOH resulted in biomass yields between 89% and 99%while the yield decreased to 82%-97% with initial NaOH concentration of 0.45 mol/L.Results in Table 3.4 also showed that alkaline extraction under low temperatures (<100 oC) resulted in little loss of cellulose and lignin while significantly reducing thehemicellulose (mainly xylan) and silica contents. On the basis of the powder yield andcompositional analysis, the actual loss of the different bamboo components during thealkaline extraction was calculated (calculation not shown). At 100 oC, by increasingthe initial NaOH concentration from 0.15 to 0.45 mol/L and reaction time from 5 to180 min, the calculated cellulose and lignin mass fraction losses were 0.4-5.1% and753.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo Powder0.3-6.4% (based on initial starting raw material), respectively. The high crystallinityand limited accessibility towards chemicals make cellulose very recalcitrant towardsdegradation under mild conditions such as those utilized in this study (Engström et al.(2006)). Low amounts of extracted lignin would benefit the recovery of silica fromalkaline pre-extraction liquor (APEL) by reducing lignin co-precipitation, resultingin high purity silica particles. The loss of galactan and arabinan, did not contributemuch to yield loss as the content in the starting material was low (total mass fractionless than 1.5% in raw chips); the majority of the extracted hemicellulose was xylan.Fig. 3.12 shows the fraction of residual xylan remaining in the extracted bamboopowder as a function of reaction time. The error bars are calculated from the averagevalues from triplicate experiments at 100 oC with the standard deviation (95% confi-dence interval). As shown in Fig. 3.12, the amount of residual xylan decreased withincreasing reaction time, treatment temperature and NaOH concentration. Duringinitial alkaline treatment, xylan solubilization is rapid and solubilization slows downwith increasing time, especially at low initial NaOH loading. This is probably be-cause the consumption of NaOH by bamboo components such as silica, acetyl andglucuronic acid groups lower the hydroxyl ions (OH– ) available for the attack ofthe hemicellulose structure. For example, at the initial NaOH concentration of 0.15mol/L and the treatment time of 180 min, increasing the temperature from 70 to 100oC only resulted in the removal of about 7-18% of initial xylan (Fig. 3.12a). Approx-imately 20% of the original xylan mass was extracted by treating bamboo powder at70 oC with initial NaOH concentration of 0.45 mol/L for 180 min. In contrast, at100 oC with the same initial NaOH concentration (0.45 mol/L), up to 56% of xylanwas extracted in 180 min (Fig. 3.12b).763.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo PowderTable 3.4: Chemical composition of bamboo powder after pre-extraction with NaOH at 100 oC.NaOH Time Yield Glucose Xylose Galactose Arabinose Lignin Silica(mol/L) (min) (%)a (%)b (%)b (%)b (%)b (%)b (%)b0.15 5 98.9 47.63±0.78 20.05±0.15 0.65±0.05 0.71±0.04 24.38±0.14 1.01±0.0215 97.4 47.78±1.26 19.56±0.24 0.68±0.03 0.54±0.10 24.34±0.18 0.85±0.0230 95.6 48.79±0.45 18.86±0.29 0.58±0.06 0.63±0.07 24.55±0.25 0.68±0.0160 92.8 50.07±0.96 18.09±0.32 0.62±0.04 0.56±0.06 25.04±0.37 0.45±0.0390 91.6 50.36±0.85 17.63±0.18 0.65±0.06 0.39±0.12 25.21±0.49 0.32±0.02120 90.3 50.62±1.12 17.16±0.36 0.48±0.10 0.55±0.08 25.36±0.66 0.26±0.02150 89.2 50.71±0.72 16.85±0.27 0.36±0.12 0.62±0.04 25.69±0.68 0.21±0.03180 88.7 50.83±0.66 16.65±0.31 0.43±0.08 0.52±0.09 25.81±0.32 0.19±0.010.45 5 97.2 48.43±0.88 19.84±0.25 0.68±0.04 0.74±0.05 24.82±0.14 0.89±0.0315 95.6 48.94±1.02 18.47±0.32 0.73±0.02 0.70±0.04 24.98±0.25 0.48±0.0230 92.2 50.14±0.57 17.39±0.27 0.69±0.06 0.64±0.10 25.56±0.47 0.21±0.0360 87.1 52.17±1.34 15.62±0.44 0.64±0.04 0.57±0.08 26.7±0.86 0.05±0.0290 85.6 52.96±1.86 14.70±0.24 0.52±0.10 0.48±0.12 26.97±0.74 0.01±0.006120 84.5 52.57±1.67 13.45±0.38 0.36±0.12 0.53±0.06 27.23±0.95 ND150 83.4 53.48±0.94 13.63±0.25 0.57±0.22 0.64±0.09 27.44±0.83 ND180 82.2 53.93±0.72 13.33±0.43 0.49±0.14 0.42±0.14 27.56±0.72 NDa Calculations were based on initial o.d. chip mass. b Values were calculated based on treated o.d. chip mass. ND-not detected.773.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo PowderThe other goal of the alkaline pre-extraction was to extract silica prior to pulping.With alkaline extraction, the amount of silica removal from bamboo powder increasedwith increasing NaOH charge and temperature. Fig. 3.13 shows the fraction ofresidual silica remaining in the extracted solids as a function of time. The error barsare calculated from the average values from triplicate experiments at 100 oC withthe standard deviation (95% confidence interval). Results shown in Fig. 3.13 clearlyillustrate that faster removal of significant amounts of silica requires either increasingthe initial NaOH concentration or treating at higher temperature for longer times. Forexample, alkaline treatment of bamboo powder for 180 min with the initial NaOHconcentration of 0.15 mol/L at 70 and 100 oC resulted in the extraction of about65% and 84% of initial silica, respectively. One reason has been discussed in Section3.1, which is that increasing the treatment temperature increases the reaction ratebetween silica and NaOH. The other likely reason for the less effectiveness of silicaremoval at lower initial NaOH concentration such as 0.15 mol/L might be that somealkali is consumed or neutralized by bamboo components such as acetyl and uronicacid groups (Gossett et al. (1982); Pavlostathis and Gossett (1985)), resulting inmuch reduced [OH– ] in later stages for the reaction. In contrast, using initial NaOHconcentration of 0.45 mol/L at 100 oC, nearly 95% of silica could be extracted frombamboo powder in 45 min. Moreover, after the removal of 96% of silica mass, thesilica content of treated material was about 0.04% or even less (based on treated o.d.solid mass), which means that even the silica impact on the production of high puritydissolving-grade pulp can be eliminated (Sixta (2006)). With such low amount ofsilica (6 0.04%, w/w) in the treated bamboo sample, the adverse effect of silica onthe chemical recovery process of kraft pulping can be significantly reduced.783.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo Powder0 20 40 60 80 100 120 140 160 1800.750.800.850.900.951.0070 C80 C90 C100 CFraction of residual xylan (x)Time (min)a0 20 40 60 80 100 120 140 160 1800.50.60.70.80.91.0b 70 C 80 C 90 C 100 CFraction of residual xylan (x)Time (min)Figure 3.12: Experimental yields of xylan remaining in the extracted milled bamboopowder at alkaline pre-extraction temperatures of 70-100 oC (initial NaOH concentration:a = 0.15 mol/L; b = 0.45 mol/L). 793.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo Powder0 20 40 60 80 100 120 140 160 1800.00.20.40.60.81.070 C80 C90 C100 CFraction of residual silica ()Time (min)a0 20 40 60 80 100 120 140 160 1800.00.20.40.60.81.070 C80 C90 C100 CFraction of residual silica ()Time (min)bFigure 3.13: Experimental yields of silica remaining in the extracted milled bamboo powderat alkaline pre-extraction temperatures of 70-100 oC (initial NaOH concentration: a = 0.15mol/L; b = 0.45 mol/L). 803.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo Powder0 20 40 60 80 100 120 140 160 1800.000.030.060.090.120.15 NaOH concentration (mol/L)Time (min) 70 C 80 C 90 C 100 C a0 20 40 60 80 100 120 140 160 1800.000.150.300.45 NaOH concentration (mol/L)Time (min) 70 C 80 C 90 C 100 CbFigure 3.14: NaOH concentration during alkaline pre-extraction of bamboo powder attemperatures of 70-100 oC (initial NaOH concentration: a = 0.15 mol/L; b = 0.45 mol/L).813.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo PowderTo further assess the effects of alkaline pre-extraction on the dissolution of bamboocomponents, the sugars, soluble lignin and silica in the extract liquors were analyzed.Table 3.5 shows the chemical composition of liquors obtained from the alkaline pre-extraction at 100 oC with initial NaOH concentration of 0.45 mol/L. The yields arereported in weight percentage of original o.d. wood mass. An interesting observa-tion is that the sum of powder yield (Table 3.4) and solid contents of the alkalinepre-extraction liquor (APEL) (Table 3.5) was larger than 100% in the treatment ex-periments. The excess values of the mass balance might be partly due to the sodiumions bound to dissolved components such as acetic or glucuronic acid. On the otherhand, the weight of extraneous inorganic compounds dissolved in the alkaline extractssuch as NaOH also contributed to the total solid content. As shown in Table 3.5, vary-ing the severity of the pre-treatment had little effect on the extraction of glucan andlignin from bamboo powder into the APEL. In contrast, a large portion of silica andhemicellulose in the raw material were extracted. In addition, the silica content atthe maximum silica removal (120-180 min) was calculated to be in the range of 1.10-1.13% (based on original o.d. mass of bamboo biomass), indicating almost all silicain bamboo was extracted during alkaline pre-treatment (Table 3.5). At extractionconditions that resulted in the removal of more than 97% of initial silica, the xylancontent in the APEL reached up to 7% (based on the original o.d. bamboo powder).Moreover, data in Tables 3.4 and 3.5 also show that the sum of xylan content in bothbiomass residuals and APELs was 88.3-96.6% of the initial xylan in raw material,revealing a reasonable mass balance. The 4-12% xylan not accounted for might belost during sample washing after the pre-extraction or through xylan degradationinto products undetected by the methodology used. Moreover, as shown in Table823.2. Alkaline Pre-extraction of Silica and Hemicellulose from Bamboo PowderTable 3.5: Composition of hydrolysates (based on initial o.d. bamboo powder) obtainedfrom alkaline treatment at 100 oC with NaOH concentration of 0.45 mol/L.time (min)Products 30 60 90 120 150 180Xylose (%) 3.36±0.24 5.54±0.31 6.25±0.27 7.12±0.51 7.07±0.46 6.97±0.34Galactose (%) 0.05±0.2 0.02±0.02 0.16±0.2 0.14±0.04 0.17±0.05 0.10±0.02Arabinose (%) 0.15±0.02 0.15±0.03 0.17±0.04 0.13±0.06 0.21±0.04 0.27±0.02Glucose (%) 0.98±0.12 1.23±0.73 1.19±0.65 1.36±0.52 1.26±0.81 1.55±0.79Lignin (%) 0.45±0.14 0.51±0.61 0.53±0.22 0.68±0.17 0.70±0.56 0.84±0.67Silica (%) 0.90±0.03 1.04±0.02 1.07±0.02 1.10±0.03 1.10±0.03 1.10±0.03Total solids (%) 26.54 29.77 30.67 31.08 32.39 32.61The values are expressed as percentage of the initial o.d. chip mass.3.5, the content of total xylose (both monomers and oligomers) initially increasedwith treatment severity, thereafter it decreased a little. This is probably due to thepeeling reaction of the reducing end groups of xylooligomers into xyloisosaccharinicacid via α and β benzilic acid rearrangement, which confirmed the assumption thatxylan could be degraded during alkaline pre-treatment of bamboo biomass.Equation 3.37 illustrates the consumption of NaOH by different bamboo compo-nents during alkaline pre-extraction of bamboo. To further confirm this, alkali con-centration of the APEL was determined and shown in Fig. 3.14. It can be observedthat at lower initial NaOH concentration (0.15 mol/L), the residual alkali concen-tration after treatment for 180 min ranged from 0.014-0.03 mol/L, corresponding tothe fact that more than 80% of initial NaOH was consumed during the reaction (Fig.3.14a). In contrast, when using initial NaOH concentration of 0.45 mol/L and treat-ment temperature of 100 oC, the lowest NaOH concentration after alkaline treatmentwas 0.26 mol/L (Fig. 3.14b), which still had high concentration of OH– available forthe dissolution of silica. These results to some extent explain the deviations obtained833.3. Alkaline Pre-extraction of silica and Hemicellulose from Bamboo Chipsduring the fitting of shrinking core model in the removal of silica from bamboo powder(Fig. 3.9).3.2.4 SummaryNaOH concentration plays an essential role in the dissolution of silica from bamboopowder. During alkaline pretreatment of bamboo powder, several reactions betweenbamboo components and NaOH take place in parallel and consuming a significantamount of OH– , resulting in a silica dissolution rate in bamboo powder slower thanthat found for pure amorphous silica particles under the same alkaline dissolutionconditions. The treatment parameters such as temperature, time and NaOH concen-tration had a positive effect on the removal of silica and xylan during the alkalinepre-extraction of bamboo powder. All silica and up to 60% of xylan in bamboo pow-der could be extracted under the conditions investigated in this work. A toy modelthat describes the evolution of OH– concentration and the extraction of silica andxylan in the reaction process was proposed.3.3 Alkaline Pre-extraction of silica andHemicellulose from Bamboo Chips3.3.1 IntroductionDuring the pre-extraction of biomass for the removal of hemicellulose or other com-ponents, particle size of the raw biomass also plays an important role. It has beensuggested that the reduction of the particle size accelerates the removal of wood com-843.3. Alkaline Pre-extraction of silica and Hemicellulose from Bamboo Chipsponents from lignocellulosic feedstocks (Chundawat et al. (2007); Zhao et al. (2008)).In the previous section, we described the work on alkaline pre-extraction of bamboopowder (250-400 mesh). Results demonstrated that almost all silica and up to 60%of hemicellulose could be extracted under the studied conditions.In the pulp and paper industry, commercial bamboo chips, which have a largersize (≈ 1×2×3 cm) than bamboo powder, are used as the raw material. Thus,mass transfer effects might affect the transportation of treatment chemicals and thediffusion of solubilized degradation products. Thus, the work presented in the sectionwas to evaluate the mass transfer effects on the removal of silica and hemicellulosefrom bamboo chips during alkaline pre-extraction. Understanding of the underlyingmechanisms of alkaline treatment of commercial bamboo chips also favors the processscale-up and optimization.3.3.2 Materials and MethodsBamboo chips used for alkaline pre-extraction were the same as that used in Chapter 2.All chemicals were reagent grade and solutions were prepared with purified deionizedwater.Alkaline pre-extraction of bamboo chips were carried out in 4 silicate glass reactorsof 2 L capacity immersed in a laboratory-scale oil bath. Before using for the experi-ments, bamboo chips were placed in a bucket with deionized water and soaked for 72h to allow uniform penetration of treatment chemicals into the chip during alkalinetreatment process. The alkaline pre-extraction temperature and NaOH concentrationwere the same as that used in the treatment of bamboo powder. To extract more853.3. Alkaline Pre-extraction of silica and Hemicellulose from Bamboo Chipssilica and hemicellulose from bamboo chips, longer reaction time was investigatedcompared to that used for the treatment of bamboo powder. For each experiment,100 g o.d. bamboo chips and the calculated NaOH solution (stock concentration of100 g/L) and deionized water were mixed and placed in a reactor. The treatmentprocess and the sample collection were the same as previous described (see Section3.2). All experiments were performed in triplicate.Chemical composition of insoluble solids and liquid samples were analyzed accord-ing to the NREL standard protocol (Chapter 2). Silica in the liquor was measuredaccording to Tong et al. (2005) (Section 3.2).3.3.3 Results and DiscussionTo understand mass transfer effects on the extraction of silica and hemicellulose frombamboo, alkaline pre-treatment of commercial bamboo chips was carried out underthe same conditions used for the treatment of bamboo powder. Figs. 3.15 and 3.16show the experimental data for alkaline pre-extraction of bamboo chips obtained inthe laboratory at 70-100 oC with initial NaOH concentrations of 0.15 and 0.45 mol/Lfor up to 300 min. Similar to the trends obtained from the alkaline pre-extractionof bamboo powder, the extraction of xylan and silica increased with enhancing thetreatment intensity, such as higher temperature, increased NaOH loading and longertime. However, when comparing the data obtained from bamboo chips (Figs. 3.15and 3.16) to that of bamboo powder (Figs. 3.12 and 3.13), it can be observed that theamounts of solubilized xylan and silica with bamboo chips were generally lower thanbamboo powder under the same alkaline treatment conditions. For example, alkaline863.3. Alkaline Pre-extraction of silica and Hemicellulose from Bamboo Chips0 50 100 150 200 250 3000.750.800.850.900.951.00 Fraction of residual xylan (x)Time (min) 70 C 80 C 90 C 100 Ca0 50 100 150 200 250 3000.50.60.70.80.91.0Fraction of residual xylan (x)Time (min) 70 C 80 C 90 C 100 CbFigure 3.15: Experimental yields of xylan remaining in the extracted bamboo chips atalkaline pre-extraction temperatures of 70-100 oC (initial NaOH concentration: a = 0.15mol/L; b = 0.45 mol/L).873.3. Alkaline Pre-extraction of silica and Hemicellulose from Bamboo Chipstreatment of bamboo biomass at 90 oC and NaOH concentration of 0.45mol/L for 100min, about 34% of xylan and 97% of silica were extracted from bamboo powder whilethe extracted xylan and silica from bamboo chips were 18% and 92%, respectively(Figs . 3.12, 3.13, 3.15 and 3.16). This might be attributed to the mass transfer effectsplay an important role in the solubilization of silica and xylan from bamboo chips.During alkaline pretreatment of bamboo biomass, NaOH molecules penetrate into theinside of bamboo mainly through the lumen. Therefore, the decrease of the particlesize can improve the transportation of chemicals into the inside of the bamboo chip.Moreover, since bamboo powder has a smaller particle size and a larger specific surfacearea than bamboo chips, more silica molecules and ester linkages are accessible fordissolution and hydrolysis by OH– , resulting in higher silica and xylan solubilizationunder the same extraction conditions (Mittal et al. (2009)). In contrast, when treatingbamboo chips, not only hydroxide ions (OH– ) have to penetrate into inside bamboochip pores to dissolve silica and hydrolyze xylan, but also the dissolved silica andsolubilized xylan products also have to transport from the inside of chip pores to thebulk liquor. The limited accessibility of bamboo components to chemicals and therecalcitrant nature of bamboo biomass negatively affect the extraction of silica andxylan from bamboo chips compared to bamboo powder.Since the goal of alkaline pre-extraction of bamboo chips was to extract the mostsilica and hemicellulose while minimizing the loss of cellulose and lignin to preservethe final pulp yield and the heating value of the generated black liquor. Longertreatment time (up to 300 min) was used than bamboo powder (180 min). Figs.3.15b and 3.16b illustrate that at 80-100 oC using initial NaOH concentration of 0.45mol/L, about 30-50% of xylan and all silica in bamboo chips could be removed within883.3. Alkaline Pre-extraction of silica and Hemicellulose from Bamboo Chips0 50 100 150 200 250 3000.00.20.40.60.81.0 Fraction of residual silica ()Time (min) 70 C 80 C 90 C 100 C a0 50 100 150 200 250 3000.00.20.40.60.81.0Fraction of residual silica ()Time (min) 70 C 80 C 90 C 100 CbFigure 3.16: Experimental yields of silica remaining in the extracted bamboo chips atalkaline pre-extraction temperatures of 70-100 oC (initial NaOH concentration: a = 0.15mol/L; b = 0.45 mol/L). 893.3. Alkaline Pre-extraction of silica and Hemicellulose from Bamboo Chips240 min. These results indicate that alkaline pretreatment of bamboo chips could bea promising method to resolve the associated silica challenges encountered in pulpingwith bamboo.To further investigate the mass transfer effects during alkaline treatment of bam-boo chips versus bamboo powder, NaOH concentration of the alkaline pre-extractionliquors was measured. Taking the experiments carried out at 100 oC as an example,the consumed OH– was calculated and shown in Fig. 3.17. As shown, the consumedNaOH during alkaline treatment of bamboo powder was generally higher than thatof bamboo chips. This could be attributed to the slower diffusion of NaOH from thebulk liquor to the inside of bamboo chips and the degraded products such as sodiumacetate from the inside of bamboo chips to the bulk liquor. These results confirmedthat mass transfer plays an important role in the diffusion of chemicals into and outof the bamboo chips.The total xylose content (xylose and xylooligomers) in the hydrolysates obtainedfrom the treatments of bamboo chips and bamboo powder was also determined andshown in Fig. 3.18. As shown in Fig. 3.18, more xylose was present in the bulkliquors in bamboo-powder experiments compared to that of liquors obtained from thetreatment of bamboo chips. These results also revealed the importance of accessibilityof bamboo components to NaOH and transportation of formed products inside thebamboo chips to the bulk liquor during the alkaline pre-extraction process. Moreover,compared the data shown in Fig. 3.18 to Figs. 3.12 and 3.15, it can be found that withthe increase of treatment time from 140 to 180 min, the xylan removal from bamboobiomass was increased, whereas the xylan content in the hydrolysates was almostconstant or even decreased a little. This might be explained by the fact that some903.3. Alkaline Pre-extraction of silica and Hemicellulose from Bamboo Chips0 50 100 150 200 250 3000.000.050.100.150.20 [OH-]o - [OH-] (mol/L)Time (min)NaOH: 0.15 mol/L NaOH: 0.45 mol/L Bamboo powder Bamboo powder Bamboo chips Bamboo chipsFigure 3.17: The effect of mass transfer on consumed NaOH during alkaline pre-extractionof bamboo (alkaline pre-extraction temperature = 100 oC).913.3. Alkaline Pre-extraction of silica and Hemicellulose from Bamboo Chips0 20 40 60 80 100 120 140 160 180012345678Xylose (g/100 g of o.d. bamboo)Time (min) Bamboo powder Bamboo chipsFigure 3.18: The effect of mass transfer on xylose yields during alkaline pre-extraction ofbamboo (alkaline pre-extraction was carried out at 100 oC with initial NaOH concentrationof 0.45 mol/L).xylose units were degraded to undetected products when increasing the treatmentseverity.On the basis of above results and discussion, the reaction mechanism for thealkaline pre-treatment of bamboo chips can be generalized as follows:1) Transportation of hydroxide ions from the bulk liquor to the exterior surface ofchips;2) Penetration of hydroxide ions to the inside of chips;3) Alkaline solubilization of silica and xylan inside bamboo chips;923.3. Alkaline Pre-extraction of silica and Hemicellulose from Bamboo Chips4) Diffusion of dissolved xylooligomers, xylose and silicates to the chip exterior;5) Transportation of the soluble silicates and xylan degradation products into thebulk liquor.It is apparent that mass-transfer effects play a significant role in the reactionkinetics of alkaline pre-extraction of silica and hemicellulose from bamboo chips.3.3.4 SummaryLow temperature alkaline pre-extraction of bamboo chips prior to subsequent pulpingprocesses was demonstrated to be an effective way to selectively extract silica andhemicellulose from bamboo biomass without degrading cellulose and lignin. Thecomparison of the extraction of silica and xylan between bamboo powder and bamboochips revealed the importance of mass-transfer effects during the process of alkalinepre-extraction of bamboo chips. Under the studied extraction conditions, almost100% of silica and up to 50% of hemicellulose in bamboo chips were extracted. Takingaccount into the consideration for efficient dissolution of silica and hemicellulose,the removal was more efficient at a pre-extraction temperature of 100 oC comparedto 80 oC. Moreover, during alkaline pre-extraction silica and hemicellulose frombamboo chips, the increase of treatment severity led to the degradation of extractedxylan, resulting in the decrease of xylan content detected in the extraction liquor. Inaddition, based on the comparison of responses of bamboo powder and bamboo chipsto the alkaline pre-extraction process, the mechanism of the extraction of the bamboocomponents has been proposed, mainly including the transportation of chemicalsfrom bulk liquor to the inside of bamboo chips and diffusion of dissolved bamboo933.4. Response Surface Experimental Design on Alkaline Pre-extraction of Bamboocomponents from the inside of bamboo chips to the bulk liquor.3.4 Response Surface Experimental Design onAlkaline Pre-extraction of Bamboo3.4.1 IntroductionIn previous sections, we presented the work with pure amorphous silica, bamboopowder and bamboo chips to understand the reaction mechanisms during alkalinepre-extraction of silica and hemicellulose from the bamboo biomass. To investigatethe feasibility of industrial application of the proposed technology, it is necessary toexpand the study and implement the extraction process at an industrial scale. Forthe process scale-up and optimization, the establishment of the relationship betweenresponses of bamboo components and treatment conditions is of great importance.To optimize the alkaline pre-extraction conditions, the response surface methodology(RSM), a widely practiced approach for the optimization of various industrially impor-tant processes (Chang et al. (2002); Poojary and Mugeraya (2012)), was investigatedin this section. In the work reported below, by means of central composite rotatabledesign (CCRD) and RSM, the optimal pre-extraction process parameters were de-termined for the extraction of silica and hemicelluloes while preserving cellulose andlignin in the residual solids (maintaining yield of chip and pulp and minimizing effectof lignin co-precipitation). Moreover, a lower liquid to wood ratio (4 L/kg versus 10L/kg) for alkaline pre-extraction of commercial bamboo chips was investigated in thissection to reduce the fresh water consumption.943.4. Response Surface Experimental Design on Alkaline Pre-extraction of Bamboo3.4.2 Materials and MethodsWashed commercial bamboo chips used in this section were the same as that used inChapter 2. Chemicals were reagent grade and solutions were prepared with purifieddeionized water.Response surface methodology (RSM) was assembled for the alkaline pre-extractionof bamboo with the objective of achieving the most hemicellulose and silica extractionand the least loss of cellulose and lignin thus residual hemicellulose and silica in thetreated chips and chip yield were considered as the responses. To achieve the goal ofthis study, a three-factor central composite rotatable design (CCRD) was used in thecurrent work.To improve the utility of this proposed process, NaOH charge instead of NaOHconcentration was investigated in the current work. Based on the concentration ofNaOH used in the previous sections conduted on the treatment of bamboo powder andbamboo chips, the range for the NaOH charge was chosen to be 6-18% (based on theo.d. weight of the original chip mass). The three variables i.e. reaction temperature(T ), NaOH charge ([OH−]o) and extraction time (t) were studied at five levels (-1.682,-1, 0, +1, +1.682), where the desired ranges of values of the variables were coded tolie at ±1 for the factorial points, 0 for centre points and ±1.682 for the axial points.The experimental domain was defined by previous preliminary experiments, whichshowed the potential of increasing hemicellulose and silica extraction efficiency whilemaintaining cellulose and lignin in the residual solids. The ranges of the independentvariables and experimental design levels are listed in Table 3.6. According to CCRD,a set of 20 experimental runs was carried out (Cochran and Cox (1968)), shown in953.4. Response Surface Experimental Design on Alkaline Pre-extraction of BambooTable 3.6: Experimental range of pre-extraction variables and coded levels according toresponse surface methodology.Variable Symbol Coded variable levelLowest Low Centre High Highest-1.682 -1 0 +1 +1.682Temperature (oC) T 59.8 70 85 100 110.2NaOH charge (%) [OH−]o 1.91 6 12 18 22.09Extraction time (min) t 13 30 55 80 98Table 3.7. The order of the experiments was randomized. Runs 1-14 were performedin duplicate.Data were analyzed by multiple regressions through the least-square method withthe Design Expert Version 6.0.6 software (Stat-Ease, Inc., Minneapolis, MN, USA).A second-order polynomial equation was used to express the responses as a functionof the three independent variables (Equation 3.38).Y = b0 +k∑i=1bixi +k∑i=1biix2i +k∑i=1k∑j=i+1bijxixj k = 3 (3.38)where Y is the predicted response, b0 is the constant coefficient, bi, bii and bij are thecoefficients of linear, quadratic and second-order interaction regression terms, respec-tively, k is the number of independent variables, xi and xj are the coded independentvariables (temperature, NaOH charge, and time). The estimation of regression co-efficient parameters, three dimensional (3-D) response surface graphs and responseoptimization were performed using MATLAB Version 8.5.0 software.The alkaline pre-extraction experiments were carried out in a rotating reactorsystem (Aurora Products, Savona, BC, Canada) which consists of 4 stainless steeldigesters of 2 L each placed in a single rotating frame. The reactors were rotated963.4. Response Surface Experimental Design on Alkaline Pre-extraction of Bambooat 50 rpm with 60-s clockwise rotations followed by 60-s counter clockwise rotationsthroughout the reaction process. The maximum rated pressure and temperature ofthe digesters are 15 bar and 205 oC, respectively, which are monitored wirelesslyby Honeywell XYR 5000 transducers. In all alkaline pre-extraction experiments, theliquid-to-wood ratio was kept constant at 4 L/kg. For the alkaline pre-extraction run,bamboo chips of 100 g o.d. and the calculated volume of deionized water and NaOHsolution (stock concentration of 100 g/L) were mixed and placed in a digester. Sub-sequently, the reactor was placed in the digester system for alkaline pre-extraction.The temperature ramp-up time was kept constant at 20 min. Before using for exper-iments, bamboo chips were pre-soaked in deionized water for 72 h to expel air insidebamboo chips, aid the impregnation of NaOH from the bulk liquor to the inside ofbamboo chips and ensure a uniform penetration during the alkaline pre-extractionprocess. Upon completion of a run, the vessel was rapidly cooled in a cold waterbath and the pre-extracted chips were recovered from the liquor through filtration.The chips were washed thoroughly with deionized water until the filtrate was neutral.The the wet washed bamboo chips were stored at 4 oC for further analysis and kraftcooking. The alkaline pre-extraction liquor (APEL) was collected and stored at 4 oCfor the chemical composition analysis and further experimentation.The chemical composition (carbohydrates and lignin) of both solid and liquid frac-tions was determined according to the NREL standard protocol described in Chapter2. The ash and silica content were also determined according to the methods de-scribed in Chapter 2. The silica content of the APEL was measured by using thesilicon molybdenum blue photometric method (Tong et al. (2005)). All experimentswere performed at least in triplicate.973.4. Response Surface Experimental Design on Alkaline Pre-extraction of Bamboo3.4.3 Results and DiscussionTo extend the application of the proposed process and optimize the NaOH pre-treatment parameters, response surface methodology (RSM) with a central compos-ite rotatable design (CCRD) was employed. The experimental conditions and corre-sponding responses of the dependent variables (percentages of extracted hemicelluloseand silica, chip yield) are listed in Table 3.7. The experimental data were used tocalculate the coefficients of the second-order polynomial equation. According to theanalysis of variance (ANOVA), the statically significant model terms with p-valueless than 0.05 were obtained and used for the response regression and the model ex-pression. The mathematical models were expressed in terms of coded variables. Thenon-significant model terms were removed.As discussed in Section 3.2 and data shown in Table 3.7, over 93% of extractedhemicellulose was in the form of xylan. The second-order polynomial functions rep-resenting xylan extraction (Y1), silica extraction (Y2) and chip yield (Y3) can be ex-pressed as a function of three operating parameters of alkaline pre-extraction, namelyextraction temperature (T ), NaOH charge ([OH−]o) and reaction time (t). The re-sultant models adequately represented the experimental data as r2 > 0.95.As shown in Table 3.7, the amount of xylan removal ranged from 8.1 to 39.4% ofthe initial xylan in the starting material. Equation 3.39 shows the relationship be-tween the percentage of extracted xylan and pre-treatment variables of NaOH charge,983.4. Response Surface Experimental Design on Alkaline Pre-extraction of BambooTable 3.7: Experimental design and observed responses of the dependent variables.Run no. Independent variables Dependent variablestemperature (oC) NaOH (%) time (min) Extracted Extacted Chipxylan (%) silica(%) yield (%)1 110.2 12 55 34.6 92.1 86.62 59.8 12 55 10.2 42.2 95.63 85 1.91 55 8.4 22.3 97.14 85 22.09 55 36.5 90.2 89.35 85 12 97 21.4 85.6 90.36 85 12 13 13.6 45.7 95.77 70 18 80 21.5 83.7 91.48 70 18 30 20.4 58.7 93.89 70 6 30 8.1 28.2 96.610 100 18 30 26.1 91.6 90.411 70 6 80 8.2 60.9 95.512 100 18 80 39.4 99.7 86.313 100 6 80 21.3 70.9 91.214 100 6 30 12.3 62.3 93.515 85 12 55 19.3 78.3 92.116 85 12 55 18.3 77.8 91.517 85 12 55 19.2 76.9 91.218 85 12 55 18.7 78.7 91.419 85 12 55 18.2 80.2 91.720 85 12 55 18.5 79.5 91.2993.4. Response Surface Experimental Design on Alkaline Pre-extraction of Bambootemperature and time.Y1 = 18.20 + 5.93T + 7.74[OH−]o + 2.61t+ 1.14T 2 + 1.15[OH−]o2+ 2.76Tt(r2 = 96.32%, radj.2 = 94.09%)(3.39)where Y1 is fraction of extracted xylan (%, based on the starting xylan mass), T ,[OH−]o and t are reaction temperature, NaOH charge and reaction time, respectively.Independent variables are standardized.The studied alkaline pre-extraction variables showed positive effects on the extrac-tion of xylan from bamboo chips. The results of ANOVA for the CCRD are presentedin Table 3.8; the tests of F -value and P -value were used to determine the significanceof the regression coefficients of each parameter. The larger the value of F and smallervalue of P , the more significant of the corresponding coefficient term (Atkinson et al.(1992)), which also means that the larger coefficient the stronger effect of the cor-responding term. Accordingly, during the alkaline extraction of hemicellulose frombamboo, the NaOH charge had the strongest effect on the extraction of xylan whilereaction time had the weakest effect (Table 3.8).Fig. 3.19 shows the 3-D surface plots of the effect of the interaction betweenNaOH charge (%) and reaction temperature (oC) on the fraction of extracted xylanat the fixed reaction time of 80 min. As shown, up to 50% of original xylan in rawbamboo chips could be extracted during alkaline pre-extraction and that the amountof extracted xylan increased with increasing NaOH charge and reaction temperature(Fig. 3.19). For example, about 6% of the original xylan mass was extracted bytreating bamboo chips at 60 oC with 12% NaOH charge for 80 min. In contrast, at1003.4. Response Surface Experimental Design on Alkaline Pre-extraction of BambooFigure 3.19: Effect of temperature and NaOH charge on the extraction of xylan at a fixedreaction time of 80 min.110 oC with the same NaOH charge (12% of o.d. chip mass), about 40% of xylancould be extracted (Fig. 3.19). The maximum xylan extraction predicted by responsesurface was 59.9% when treating bamboo chips at 110 oC with 22.1% NaOH chargefor 97 min.During alkaline pre-treatment, the amount of extracted silica is significantly de-pended on the severity of the pre-extraction process. As shown in Table 3.7, thefraction of extracted silica ranged from 22.3 to 99.7% of the initial silica mass. Basedon the experimental data, a quadratic regression model (second-order) was obtained,after exclusion of the non-significant terms according to the analysis of variance. The1013.4. Response Surface Experimental Design on Alkaline Pre-extraction of Bambooequation for the fraction of extracted silica is shown in Equation 3.40.Y2 = 72.91 + 12.55T + 16.15[OH−]o + 10.24t− 5.27[OH−]o2 − 5.75Tt(r2 = 93.97%, radj.2 = 91.81%)(3.40)where Y2 is fraction of extracted silica (%, based on the starting silica mass), T ,[OH−]o and t are reaction temperature, NaOH charge and reaction time, respectively.Independent variables are standardized.Equations 3.40 revealed that the amount of silica extraction from bamboo chipswas positively affected by the all three studied factors, namely extraction temperature,NaOH charge and reaction time. As illustrated in Table 3.8, NaOH charge had thestrongest effect on the extraction of during alkaline pre-extraction of bamboo chipsfollowed by extraction temperature and time. This is not in agreement with theliterature observation obtained from sodium carbonate (Na2CO3) pretreatment ofwheat straw (Pekarovic et al. (2005)). One likely reason for the difference is that thebamboo chips used in the present research had been pre-soaked in deionized waterfor 72 h to expel the air inside the chips. In contrast to the dry wheat straw usedby Pekarovic et al. (2005), pre-soaked chips are expected to experience more uniformpenetration of NaOH solution and relatively reduce the mass transfer effects. Inaddition, the range of conditions tested as well as the alkali reagent used (NaOHversus Na2CO3) in this study are different from those used by Pekarovic et al. (2005).Based on the industrial aspect, NaOH might be a better choice due to its abundanceas white liquor in a typical kraft pulp mill.Fig. 3.20 shows the 3-D surface plots of the effect of the interaction between NaOH1023.4. Response Surface Experimental Design on Alkaline Pre-extraction of BambooFigure 3.20: Effect of temperature and NaOH charge on the extraction of silica at a fixedreaction time of 80 min.charge and temperature on the extraction of silica at the fixed time of 80 min. Thefraction of extracted silica increased with increasing NaOH charge and temperature.For example, treating bamboo chips for 80 min with 6% NaOH charge at 70 oCand 100 oC, about 61% and 71% of initial silica in bamboo chips could be removed,respectively. In contrast, at 100 oC with 18% NaOH charge for 80 min, all silica inuntreated bamboo chips was removed. The dissolution rate of silica in NaOH solutionincreased with increasing NaOH concentration and reaction temperature (Fig. 3.20);this is in agreement with previous studies on the dissolution of amorphous silica inNaOH solutions (Niibori et al. (2000)). Moreover, with the increase of treatmenttemperature, more xylan was extracted, resulting in more pores of the treated chips;this improves the penetration of NaOH into chips and the diffusion out of solublesilicates, thus leading to faster extraction of silica.Results in Table 3 indicated that the chip yield depended on all three factors,1033.4. Response Surface Experimental Design on Alkaline Pre-extraction of BambooNaOH loading, extraction temperature and time; the chip yield from the pretreatedsamples ranged from 84% to 97% of the initial dry chip mass. Based on the experi-mental data, the coefficients of the second-order response function on the coded levelsrepresenting the chip yield was calculated, shown as Equation 3.41.Y3 = 91.67− 2.29T − 2.08[OH−]o − 1.38t+ 0.57[OH−]o2(r2 = 94.86%, radj.2 = 93.48%)(3.41)where Y3 is chip yield (%, based on initial chip mass), T , [OH−]o and t are reactiontemperature, NaOH charge and reaction time, respectively. Independent variablesare standardized.According to Equation 3.41, the chip yield was inversely affected by increasingtemperature, NaOH charge and extraction time. The chip yield loss was mainlyaffected by NaOH charge and in minor extent by temperature and time (Table 3.8).As shown in Fig. 3.21, increasing the treatment intensity decreased the chip yield.For example, the use of 6% NaOH resulted in chip yield between 95% and 97% at 80oC while the yield decreased to 91-94% at 100 oC.To further assess the effects of alkaline pretreatment on the dissolution of bamboocomponents, the composition (sugars, lignin and silica) of the solid fraction remain-ing after pretreatment was determined. Table 3.9 shows the results of extractionin terms of chemical composition of pre-treated chips. As shown in Table 3.9, theglucan content in the treated bamboo chips increased due to the removal of hemicel-lulose. The increase of cellulose content is favorable for the subsequent treatmentsdesigned to produce dissolving grade pulp or fermentable glucan for lignocellulosic1043.4. Response Surface Experimental Design on Alkaline Pre-extraction of BambooTable 3.8: Values of regression coefficients of the fitted second order polynomials.Factor Regression coefficient F -value P -valuePercentage of extracted xylan - - -Model - 78.73 < 0.0001 (significant)b0 18.20 - < 0.0001Linear - - -b1 5.93 152.3 < 0.0001b2 7.74 260.5 < 0.0001b3 2.61 29.50 0.0001Quadratic - - -b11 1.14 5.97 0.295b22 1.15 6.16 0.0275Interaction - - -b13 2.76 19.39 0.0007Percentage of extracted silica - - -Model 43.60 < 0.0001 (significant)b0 72.91 - < 0.0001Linear - - -b1 12.55 59.99 < 0.0001b2 16.15 99.34 < 0.0001b3 10.24 39.94 < 0.0001Quadratic - - -b22 -5.27 11.37 0.0046Interaction - - -b13 -5.75 7.37 0.0167Chip yield - - -Model 69.14 < 0.0001 (significant)b0 91.67 - < 0.0001Linear - - -b1 -2.29 122.45 < 0.0001b2 -2.08 101.35 < 0.0001b3 -1.38 44.51 < 0.0001Quadratic - - -b22 0.57 8.24 0.0117Note: the response surface regression and models were expressed in terms of codedvariables, without taking into account the statistically non-significant terms.b1, b2 and b3 are values of the coefficients of temperature, NaOH charge, and time,respectively.1053.4. Response Surface Experimental Design on Alkaline Pre-extraction of Bamboo6070Temperature (°C)80901001102015NaOH charge (%)1059692888480100Chip yield (%)8486889092949698Figure 3.21: Effect of temperature and NaOH charge on chip yield at a fixed reaction timeof 80 min.fuels production. However, for the production of regular kraft pulp, the removal ofhemicellulose might negatively affect the physical properties of final paper due to theloss of hydrogen bonding (Bai et al. (2012)). These results showed that the treatmenttime, temperature and NaOH charge used in this study did not significantly degradethe lignin in bamboo chips. These results confirmed the obtained conclusions thatalkaline pre-extraction under studied conditions had little effect on the extractionof cellulose and lignin from bamboo chips and alkaline pre-extraction is a promisingmethod for solving the silica problems during bamboo biomass fractionation.Equations 3.39, 3.40 and 3.41 were used to identify suitable NaOH treatment con-ditions to the reach the target of extracting more than 97% of silica from bamboochips while persevering cellulose and lignin in the treated solids. Moreover, the se-lected conditions should also be able to lower capital costs and heat losses, and reducethe pretreatment time. Using the software (Design Expert Version 6.0.6), the opti-1063.4. Response Surface Experimental Design on Alkaline Pre-extraction of BambooTable 3.9: Experimental design and observed responses of the dependent variables.Run Composition (%)ano. Glucose Galactose Arabinose Lignin Xylose Silica1 53.5(0.8) 0.4(0.05) 0.8(0.1) 27.8(0.2) 15.3(0.3) 0.10(0.02)2 49.8(0.3) 0.6(0.02) 0.7(0.05) 26.0(0.1) 19.1(0.6) 0.68(0.4)3 49.9(0.05) 0.3(0.1) 0.7(0.07) 25.7(0.2) 19.2(0.4) 0.9(0.05)4 52.4(0.1) 0.5(0.02) 0.6(0.1) 27.3(0.6) 14.4(0.3) 0.12(0.03)5 52.1(0.8) 0.6(0.06) 0.6(0.05) 27.0(0.1) 17.7(0.7) 0.15(0.04)6 50.1(0.4) 0.7(0.05) 0.7(0.07) 25.7(0.4) 18.3(0.3) 0.64(0.04)7 52.1(0.5) 0.4(0.08) 0.6(0.05) 26.7(0.3) 17.4(0.2) 0.20(0.02)8 50.8(0.2) 0.8(0.07) 0.8(0.1) 26.1(0.4) 17.2(0.2) 0.49(0.03)9 49.7(0.4) 0.6(0.1) 0.4(0.06) 25.6(0.2) 19.3(0.8) 0.83(0.02)10 52.3(0.8) 0.6(0.1) 0.8(0.05) 27.0(0.2) 16.6(0.3) 0.10(0.03)11 50.4(0.6) 0.4(0.04) 0.6(0.06) 25.8(0.1) 19.5(0.5) 0.46(0.02)12 53.8(0.2) 0.8(0.05) 0.5(0.1) 27.8(0.2) 14.2(0.4) 0.01(0.01)13 51.5(0.6) 0.4(0.06) 0.6(0.09) 26.7(0.3) 17.5(0.7) 0.36(0.04)14 50.9(0.8) 0.3(0.07) 0.7(0.04) 26.2(0.2) 19.0(0.6) 0.45(0.03)15 51.5 0.9 0.8 26.5 17.8 0.2616 51.6 0.7 0.5 26.6 18.1 0.2717 52.1 0.6 0.7 26.8 18.0 0.2818 52.0 0.7 0.8 26.2 17.7 0.2619 51.7 0.4 0.6 26.6 18.1 0.2420 52.1 0.8 0.4 26.1 18.2 0.25Note: values of range between duplicate measurements are given within bracket.aComposition of the solid fractions resulting from the pre-treatments were expressedas the percentage of dry matter of treated bamboo chips.1073.4. Response Surface Experimental Design on Alkaline Pre-extraction of BambooTable 3.10: Confirmation runs of the alkaline pre-extraction of bamboo chips according toRSM.Alkaline Extracted xylan (%) Extracted silica (%) Chip yield (%)pre-treatment Actual Predicted Actual Predicted Actual Predicted18% NaOH, 37.8 39.5 99.8 100 87.3 86.5100 oC, 90 min17% NaOH, 32.4 35.7 94.8 96.2 89.4 87.797 oC, 70 minmized alkaline pre-extraction conditions were determined to be at a temperature of95-100 oC with NaOH charge of 16-18% (based on the o.d. chip mass) for 70-90 min.To verify the model and validate the proposed methodology, pre-treatment experi-ments were conducted under two different conditions which were using 18% NaOH at100 oC for 90 min and 17% NaOH charge at 97 oC for 70 min, respectively. Table3.10 shows the experimental results on the alkaline pre-extraction of bamboo chips.The actual silica and xylan extraction data were close to the predicted values. For ex-ample, treating bamboo chips with 18% NaOH at 100 oC for 90min resulted in 97.6%and 34.3% of initial silica and xylan removal, which were close to predicted valuesof 99.6% and 36.8%, respectively. The results clearly demonstrated the noteworthyextraction of silica and hemicellulose from bamboo chips.3.4.4 SummaryThe application of response surface methodology (RSM) and central composite ro-tatable design (CCRD) for modelling the influence of the three operating variables(temperature, NaOH charge and time) on the treatment efficiency (extraction of sil-ica and hemicellulose from bamboo chips) has been discussed. RSM has helped tolocate the conditions for best silica removal with minimal loss of cellulose and lignin.1083.5. SummaryAccording to model equations and confirmatory experiments, 97% desilication at achip yield of about 82% was reached by treating bamboo chips at temperatures of95-100 oC with NaOH charge of 16-18% (based on the original o.d. chip mass) for70-90 min.3.5 SummaryIn this chapter the dissolution of pure amorphous silica in NaOH solution and NaOHtreatment of bamboo biomass were investigated.Under the alkali concentrations studied, the shrinking core model could be usedto describe the silica dissolution rate of pure amorphous silica. During alkaline treat-ment of bamboo biomass, several reactions between bamboo components and NaOHtake place in parallel and consuming a significant amount of OH– . To describe thereactions, toy chemistries, representing the dominant mechanism of alkaline pretreat-ment, were posed. A toy model that illustrates the change of OH– concentrationduring the reaction process was established. The main objective of this part of workis to remove the most silica and hemicellulose from bamboo chips and solve the silicaassociated challenges during pulping bamboo, so the experimental data on the disso-lution of silica and xylan were presented. The test of the utility of the proposed toymodel was presented in the published paper.The consumption of OH– by other bamboo components resulted in lower silicadissolution rate than that found for pure amorphous silica particles. With the increaseof the OH– concentration (≥ 0.45 mol/L) used for alkaline pretreatment of bamboopowder, the silica removal rate could be described by the classic shrinking core model.1093.5. SummaryThe comparison of the extraction of the extraction of silica and xylan betweenbamboo powder and bamboo chips revealed the importance of mass-transfer effectsduring alkaline pre-extraction of bamboo chips. Based on the comparison of responsesof bamboo powder and bamboo chips to the alkaline pre-extraction process, the mech-anism of the extraction of the bamboo components has been proposed.The application of response surface methodology (RSM) and central compositerotatable design (CCRD) for modelling the influence of the three operating variables(temperature, NaOH charge and time) on the treatment efficiency (extraction ofsilica and hemicellulose from bamboo chips) has been discussed. RSM has helped tolocate the conditions for best silica removal with minimal loss of cellulose and lignin.According to model equations and confirmatory experiments, 97% desilication at achip yield of about 82% was reached by treating bamboo chips at temperatures of95-100 oC with NaOH charge of 16-18% (based on the o.d. chip mass) for 70-90 min.In this chapterïijŇ the removal of silica and hemicellulose from bamboo biomasswas investigated during the alkaline pre-extraction of bamboo powder and bamboochips. Results demonstrate that alkaline pre-extraction of bamboo prior to pulp-ing and biorefinery processes is an effective way to extract silica and hemicelluloesselectively without degrading cellulose and lignin.110Chapter 4Feasibility of Using AlkalinePre-extraction in Kraft Pulping4.1 IntroductionThe kraft process has long been the dominant method for the production of chem-ical pulp due to its versatility in dealing with different raw materials coupled withhigh quality of resultant pulp and the efficient recovery of energy and cooking chem-icals (Garrote et al. (2003)). Moreover, for bamboo, kraft pulping is generally pre-ferred to soda pulping when preparing chemical pulp or dissolving grade pulp (Ryd-holm (1965)). Therefore, the technologies used for bamboo delignification are similarto those generally applied to wood pulping. The kraft pulping treatment involvesthe heating of lignocellulosic materials in an aqueous solution of sodium hydroxide(NaOH) and sodium sulfide (Na2S) from approximately 70 oC to cooking tempera-ture (up to 180 oC), followed by 1-2 h cooking period. During this treatment, alkalinedelignification reactions break both phenolic and nonphenoplic β−O−4 ether bonds,and remove the methoxyl groups (–O–CH3), leading to the formation of phenolateions, which are soluble. Na2S in the reaction system accelerates the lignin breakdown, in turn reducing the cellulose degradation caused by NaOH.1114.1. IntroductionTo resolve the silica challenges encountered with bamboo pulping, alkaline pre-extraction of silica and hemicellulose from bamboo chips have been extensively studiedin Chapter 3. The treated chips , with silica content less than 0.03% (w/w), can beused as the ideal starting material for the production of pulp and paper or evendissolving-grade pulp. Moreover, alkaline pre-extraction has been reported to be ableto increase pore volume and surface area of wood chips, which might benefit for thesubsequent pulping processes. Thus, it is of great importance to investigate the effectof alkaline pre-extraction on kraft pulping of bamboo.In addition, the liquor generated during alkaline treatment of bamboo chips cannot be directly mixed with black liquor and sent to chemical recovery process. Becauseif we do so, the silica associated challenges will come back to the kraft pulping process.The alkaline pre-extraction liquor (APEL) obtained from alkaline pretreatment con-tains a large amount of silica and hemicellulose, which could be a potential sustainablesource for various high-value products if recovered. Silica is an essential starting mate-rial for the consumer products such as catalysts, thixotropic agents, pharmaceuticals,film substrates, electric and thermal insulators composite filler, etc. (Kalapathy et al.(2002); Liou and Yang (2011)). On the other hand, currently, industrial production ofnanosilica is mainly from relative expensive sources of tetraethoxysilane and tetram-ethoxysilane under high temperature treatment (> 1300 oC) (Affandi et al. (2009)).Likewise, hemicellulose can also be used for various applications such as production ofethanol, xylitol, bio-polymeric films, or as papermaking additives (Jun et al. (2012);Huang et al. (2010); Ren et al. (2009); Schild et al. (2010)). Therefore, the recoveryof these valuable dissolved materials, silica and hemicellulose, from the APEL in aneconomical and eco-friendly manner is critical to their downstream processing and1124.1. Introductionutilization for value-added products.Among the chemicals used for the precipitation of silica from alkali aqueous me-dia (Minu et al. (2012); Zhang et al. (2013)), carbon dioxide (CO2), present in thewaste flue gas from the pulp mill recovery circle, provides a convenient option. Intypical chemical pulp mills, a mixture of combustible materials and inorganic chemi-cals known as black liquor (spent cooking liquor) is a by-product of fibre extractionfrom wood. Black liquor is normally burned in the recovery boiler to recover pulpingchemicals and energy for the mill operation. The burning of black liquor in the millrecovery cycle and the calcination of lime mud generated in the lime kiln generate alarge amount of waste flue gas, containing mainly CO2 (5-40%) and water vapor withsmall amounts of sulfur dioxide (SO2) and nitric oxides (NOX) (Berglin and Bernts-son (1998); Hektor and Berntsson (2007)), which can be used to precipitate silicafrom the APEL. Moreover, the utilization of waste gas can, to some extent, alleviateenvironmental problems by reducing pulp mill greenhouse gas emissions. Thus, theAPEL treatment process can also be regarded as a CO2-capture system. Moreover,different from the sodium-based salts formed by applying other acids like sulfuric acid,nitric acid or hydrochloric acid in silica precipitation, the sodium carbonate (Na2CO3)formed when using CO2 can be easily recovered as sodium hydroxide (NaOH) by thereaction with low-price calcium oxide (CaO), which is called the causticizing reaction.Moreover, the generated calcium carbonate (CaCO3) can be recovered as CaO againby the calcining process in the lime kiln. Causticizing and calcining are standardchemical recovery practices in commercial kraft or soda pulping. Compared to theblack liquor obtained from alkaline pulping of high silica content non-woody mate-rials such as cereal straw, bamboo or switchgrass (Schild et al. (2010)), the lignin1134.2. Materials and Methodscontent of the APEL obtained from bamboo was very low (see Chapter 3). Thus,lignin co-precipitation, which normally occurs during the desilication of black liquorwith CO2 treatment (Kopfmann and Hudeczek (1988); Schild et al. (2010); Minuet al. (2012); Zhang et al. (2013)), will be significantly less and as a result, pure silicacan be separated and the calorific value of the black liquor obtained from subsequentkraft pulping will be preserved.In this chapter, a novel and green concept is proposed for the production of kraftpulp, pure amorphous silica particles and polymeric hemicellulose from bamboo chips.In the proposed scheme, alkaline pretreatment was initially carried out to remove themost of silica from bamboo chips. The treated chips were served as raw materials forthe production of pulp and paper and the liquid phase was sequentially treated withcarbon dioxide (CO2) and ethanol for the recovery of dissolved silica and hemicellu-lose.4.2 Materials and Methods4.2.1 Raw MaterialBamboo chips were the same as that used in Chapter 2. Alkaline treated chipswere obtained from the experimental runs carried out in Chapter 3. The three pre-treatment runs were those using 18% NaOH (based on o.d. weight of chips) at 100oC for 1 and 5 h and 18% NaOH at 80 oC for 3 h. All chemicals used in this chapterwere analytical grade and purchased from Fisher Scientific, Canada.Alkaline pre-extraction liquor (APEL) was generated from the sodium hydroxide1144.2. Materials and Methods(NaOH) pretreatment of bamboo chips with alkali charge of 12% (w/w) (based onoven dried chips) and the liquid-to-wood ratio of 8 L/kg in a 5 L flask for 3 h at90 oC heated using a laboratory-scale water bath (see Chapter 3). Gases of CO2and nitrogen (N2) with 99.99% purity were purchased from Praxair Technology Inc.,Canada in cylinders. Distilled and deionized water was applied for all treatmentprocesses.4.2.2 Kraft PulpingPulping of pre-extracted and untreated (without alkaline pre-extraction) bamboochips by the kraft process was conducted in four 300 mL stainless steel reactorsin an oil bath under conditions covering the range of practical interest. For all cooks,the temperature was raised to 165 oC in 85 min and held at 165 oC for 75 min. Theliquid-to-wood ratio and sulfidity were fixed at 4 L/kg and 25% (percentage of Na2S,expressed as Na2O), respectively. The effective alkali (EA) (expressed as Na2O) wasvaried in the range of 13-19% (based on oven dried wood mass). For each kraft cook,45 g o.d. extracted or non-extracted bamboo chips and the calculated volume ofcooking chemicals and deionized water were placed in the reactor and mixed for 10min. Afterwards, the cooking process was carried out according to the conditionsbeing investigated. Upon completion of a cook, the reactor was rapidly cooled andkraft pulp was recovered using vacuum filtration. The kraft pulp was thoroughlywashed with deionized water until the pH of the filtrate reached neutral. Then, thepulp was disintegrated, screened, and filtered to measure total yield, screened yield,and rejects of the kraft cooking process. All experiments were performed in triplicate.1154.2. Materials and Methods4.2.3 Evaluation of PulpsThe ash content of pulp was determined according to TAPPI T211 om-02. Silicacontent of the pulp was measured according to the methods shown in Chapter 3.The kappa number of screened pulps was determined according to TAPPI T236 om-99. The fines content of the pulps was measured with a Fibre Quality Analyzer (OpTest Equipment Inc., ON, Canada) based on TAPPI T271 om-07. The pre-extractedand untreated bamboo pulps were beaten in a laboratory disc refiner (PFI mill) atdifferent revolutions according to TAPPI T248 sp-00. The freeness (drainability)of the pulps was determined according to TAPPI T227 om-99 (Canadian StandardMethod). Standard handsheets of about 60 g/m2 were made by TAPPI T 205 sp-02.The handsheets were tested for tensile and tear strength properties using TAPPI T220 sp-01.4.2.4 Preparation of Silica Particles from the APELThe mechanisms for the liquor neutralization and silica precipitation with carbondioxide (CO2) are shown in Equations 4.1 and 4.2. A schematic diagram of theexperimental apparatus used in this work is illustrated in Figure 4.1. The apparatuswas constructed to measure the effects of CO2 treatment time and temperature on thesilica precipitation from the APEL and analyze the compositions of all coexisting gasesvia an online gas chromatograph. Glass reactors with the ratio of length to diameterof 6 were used (No. 7 apparatus in Fig. 4.1). The composition of residual gas wasanalyzed on a CX-3400 gas chromatograph (GC) (Varian Canada Inc., Missinssauga,Ontario) equipped with a thermal conductivity detector (TCD) and flame ionization1164.2. Materials and Methodsdetector along with a CP-PoraPLOT U capillary column. Ultra high purity Helium(He) was used as carrier gas. The gas sample was collected with a stainless steelsampling valve (Agilent Technologies, Inc., Model 8134). The sampling valve wasflushed out three times with He before collecting a sample for analysis.2 NaOH + CO2 −−→ Na2CO3 + H2O (4.1)Na2SiO3 + 2 CO2 + 2 H2O −−→ H2SiO3↓+ 2 NaHCO3 (4.2)Two separate studies were conducted in this work. In the first study, a series ofCO2 treatment experiments were carried out over various temperatures (20-100 oC)and times (5-120 min). The temperature of the single experiment was maintainedin an electrically heated water bath. For a silica precipitation run, 200 mL APELwas added into the reactor before placing in the water bath for 15 min to reach thedesired temperature. Then, the CO2 was introduced into the bottom of the reactorat the flow rate of 0.1 L/min. Upon completion of an experimental run, the reactorwas taken out from the water bath and let for precipitation for 24 h. The precipitatewas washed several times with deionized water and carefully collected. The collectedprecipitate was dried at 120 oC in the oven for 24 h. The moisture free samples wereused for the fourier transform infrared spectroscopy (FTIR) analysis. Then, the drysample was incinerated in a muffle furnace at 700 oC for 4 h to completely removeany adsorbed organics.In the second study, the mixture gas of CO2 and nitrogen (N2) was used to simulatefor the waste flue gas under conditions (such as temperature) optimized by the firststudy. The flow rates of CO2 and N2 were 0.1 L/min and 0.2 L/min, respectively. The1174.2. Materials and MethodsFigure 4.1: A schematic of experimental set-up for recovering silica and hemicellulose andreducing CO2 emissions.treatment time was set to 5-120 min at the fixed temperature. After each treatment,gas samples (300 µL) were taken for analysis with the Gas Chromatography (GC).The precipitate was collected as previously described.4.2.5 Isolation of Hemicellulose and Preparation ofHemicellulose-based Polymeric FilmThe separation of hemicellulose from the CO2-treated APEL was achieved by ethanolprecipitation and filtration. In the precipitation process, 95% ethanol was slowlyadded to the filtrate obtained from the silica isolation with a volumetric ratio of 3while stirring. The stirring was stopped after another 5 min after final addition. Themixture was kept at 20 oC for 24 h to allow the formed precipitate to settle. Then,the hemicellulose were recovered by centrifugation (3000 rpm, 10 min) and washed1184.2. Materials and Methodswith 95% ethanol until the supernatant became colorless. Finally, the obtained hemi-cellulose particles were vacuum dried at 45 oC for 36 h.The hemicellulose-based polymeric films were made with adding 30% plasticizer(glycerol) without adding any other agents according to Mikkonen and Tenkanen(2012). Briefly, the precipitated hemicellulose were dissolved in deionized water at85 oC. Two hemicellulose concentrations, 10 and 20 g/L, were used for the testingof mechanical properties including tensile strength and elongation at break, watervapour transfer rate (WVTR), and water vapour and oxygen permeability, respec-tively. After cooling for 5 min and adding glycerol, the solutions were sonicated for 5min, cast into Teflon-coated Petri dishes, and dried in a constant temperature (23 ±0.5 oC) and humidity (50%) (CTH) room for 7 days. The films were conditioned inthe CTH room before analysis, with exception of the samples for the measurementsof water sorption which were stored in vacuum desiccators.4.2.6 Analytical MethodsThe residual alkali concentration in the black liquor obtained from kraft pulping wasdetermined by the titration with hydrochloric acid (HCl) according to TAPPI T 625cm-85.For the determination of the chemical composition of the APEL, the liquor wasneutralized using dilute sulfuric acid. Then the samples were autoclaved with 4%(w/w) H2SO4 for 60 min. After hydrolysis, the compositional analysis was followingthe NREL standard protocols used in previous chapter (see Chapter 2). Total solidcontent of the APEL was determined by vacuum drying at 45 oC for 48 h. Silica1194.2. Materials and Methodscontent of liquors was measured by using the silicon molybdenum blue photometricmethod described in Chapter 3 (Tong et al. (2005)). Lignin content ofthe black liquorswas determined gravimetrically by acid precipita-tion and centrifugation (Rocha et al.(2012)).Mineral composition of the silica powders obtained after incineration at 700 oCwas determined by atomic emission measurements using an ICP spectrometer (iCAP6000 Series, Thermo Scientific, MA, USA) with nitric acid (HNO3) to prepare testingsolutions. Fourier transform infrared spectroscopy experiments of silica particles andhemicellulose were carried out on a Cary 630 FTIR Spectrometer (Agilent Technolo-gies, ON, Canada) using the ATR model. The absorption spectra were recorded inthe absorption band mode in the range of 4000 to 500 cm−1. The number-averagemolecular weight (Mn) and weight-average molecular weight (Mw) of hemicellulosewere determined by size exclusion chromatography (SEC) with a DAWNEOS-OptilabrEX (Wyatt Technology Inc., USA) equipped with a high performance liquid chro-matography (HPLC) pump (Waters Corp., USA) and two columns, TSK-GELG-4000PWx1 (7.8×300 mm) and TSK-GEL G-2500 PWx1 (7.8×300 mm). The eluent phaseused was 0.02 mol/L potassium dihydrogen phosphate (KH2PO4) containing 0.2 NNaCl. The flow rate was 0.5 mL/min, and the operating temperature was 35 oC.The solution was filtered with 0.45 µm membrane and injected in the SEC system foranalysis, the glucan was used calibration standards. The PDI (polydispersity index)is the ratio of weight-average molecular weight (Mw) to number-average molecularweight (Mn).The tensile strength and elongation at break of the conditioned films were de-termined according to ASTM D882-12 (2005). Water vapor transfer rate (WVTR),1204.3. Results and Discussionwater vapor permeability (WVP) and oxygen transmission were measured accord-ing to ASTM E96/E96M-10, ASTM E96/E96M-05 and ASTM D3985-81 standards,respectively.4.3 Results and Discussion4.3.1 Kraft PulpingChips from three representative alkaline pre-treatment runs with residual silica con-tent less than 0.04% (w/w) were selected for subsequent kraft pulping. To improvethe efficiency when applied in a mill process, higher NaOH charge and shorter reac-tion times were used for the pre-extraction. These conditions have the potential toextract higher levels of hemicellulose and silica with less cellulose degradation (seeprevious discussion in Chapter 3).To evaluate the effect of alkaline pretreatment on kraft pulping of bamboo chips,effective alkali (EA) charge was chosen as the variable while reaction temperature andtime were kept constant. In the kraft cooking process, four levels of effective alkali(EA) charges were studied to investigate the effectiveness of alkaline pre-extractionon the kraft pulping of bamboo. To compare the kraft pulping process of extractedbamboo chips and original chips on a uniform basis, the alkali charged to the treatedchips was adjusted according to the analysis of the residual alkali in the extract liquor.Table 4.1 shows the effect of alkaline pre-extraction on the kraft pulping of bamboochips. It should be noted that the pulping yield was expressed as the overall pulpyield (measured based on the initial o.d. mass of bamboo chips). As shown in1214.3. Results and DiscussionTable 4.1, at all EA charges, kraft pulps from alkaline pre-treated chips had lowerkappa numbers (lower residual lignin content in the kraft pulp) than the non pre-extracted bamboo chips, showing that the alkaline pre-treatment had a positive effecton delignification during kraft cooking. This might be due to the fact that the removalof hemicellulose/lignin during alkaline pre-extraction process resulted in chips havinga more open structure thus improving the accessibility of the cooking chemicals tolignin in chips and improving the rate of diffusion of degraded lignin into the blackliquor. For example, a kappa number of 16.0 could be achieved by pulping extractedchips at 100 oC for 1 h with 17% EA while 19% EA was needed when pulping withcontrol chips (without pre-treatment) under the same conditions. Increasing the pre-treatment severity (longer time and higher temperature) also decreased the kappanumber of pulp obtained under the same kraft cooking conditions (results of chipstreated at 100 oC for 1 and 5 h, respectively (Table 4.1). Compared to the normalkappa number (30-50) of brownstock in a modern kraft pulp mill (Salmela et al.(2008); Mân Vu et al. (2004)), lower kappa numbers (lower lignin contents) of pulpsobtained from alkaline pretreated chips were achieved in this study; this translatesinto lower demand for bleaching chemicals and, hence lower bleaching costs.1224.3. Results and DiscussionTable 4.1: Effect of alkaline pre-extraction and effective alkali charge on kraft pulping of bamboo.Pre-extracted EA Rejectsa Total yielda Kappa Residual silica Silica CSF Finesb Hemicellu-bamboo chips (%) (%) (%) number in pulpa (%) in BLa (%) (mL) (%) loseb (%)Non-extracted 13 2.3 ±0.2 55.1 ±1.4 30.1 ±0.5 0.19 ±0.03 0.91 ±0.03 549 ±10 17.4 25.515 1.8 ±0.4 53.9 ±1.2 22.6 ±0.2 0.16 ±0.01 0.95 ±0.04 538 ±8 18.6 23.817 1.5 ±0.5 53.6 ±1.7 18.8±0.6 0.14±0.02 0.95 ±0.03 544 ±12 19.1 20.419 1.3 ±0.3 53.1±1.2 15.9±0.5 0.13±0.03 0.97±0.04 531±7 19.8 15.4180 oC 3 h 13 0.5 ±0.2 54.3 ±1.8 27.2 ±0.6 ≈ 0.01 ≈ 0.01 629 ±12 14.2 20.1115 0.4 ±0.3 53.1 ±1.2 20.7 ±0.7 ≈ 0.01 ≈ 0.01 624 ±9 14.4 16.8517 0.2 ±0.1 52.2 ±1.5 16.5 ±0.4 ND ≈ 0.04 608 ±13 14.4 13.7719 ≈ 0.01 51.0 ±0.8 13.9 ±0.2 ND ≈ 0.02 618 ±11 14.7 11.68100 oC 1 h 13 0.3 ±0.1 52.3 ±2.3 27.6 ±1.2 ≈ 0.01 ≈ 0.01 635 ±8 13.8 19.3815 ≈ 0.1 52.7 ±1.3 19.8 ±0.9 ND 0.04 ±0.02 624 ±8 14.0 16.7617 ≈ 0.1 52.1 ±1.5 16.4 ±0.3 ND ≈ 0.02 619 ±12 14.4 13.5519 ND 51.4 ±1.2 12.9 ±0.8 ND 0.04 ±0.01 620 ±11 14.6 11.25100 oC 5 h 13 ≈ 0.1 52.2 ±1.6 22.4 ±1.0 ND ND 649 ±5 13.1 15.2215 ND 51.5 ±1.9 16.1 ±0.7 ND ND 640 ±16 13.6 14.6617 ND 50.9 ±1.0 13.8 ±0.6 ND ND 620 ±9 14.0 12.3619 ND 50.3 ±1.0 11.1 ±0.4 ND ND 612 ±12 14.2 10.17a Calculations were based on original oven dried chip mass. b Calculations were based on oven dried pulp mass.EA-effective alkali. ND-not detected. BL-black liquor. CSF-Canadian standard freeness.1234.3. Results and DiscussionThe total pulp yield of extracted chips was generally slightly lower than that of thecontrol under the same kraft cooking conditions (Table 4.1). It should be noted thatpulps from extracted chips had lower kappa numbers (lower lignin content), whichaccounted for 0.3-0.6% of the pulp yield. In addition, a lower rejects content in thepulp could be obtained with alkaline pre-extracted chips (below 0.5%) compared tothe 1.3-1.5% of the controls; this could be related to the fact that extracted chips havea more open structure enabling better penetration of cooking chemicals resulting in amore even cook. The screened pulp yield obtained from the extracted chips was similarto that of the control or even higher. Total pulp yield decreased with increasing EAcharge for both the pre-treated chips and control. Similar results have been obtainedby kraft pulping of alkaline pre-extracted aspen chips (Jun et al. (2012)).The initial kraft pulp (brownstock), drainage resistance is an important param-eter as it strongly affects the downstream operations such as pulp washing. In thisstudy, the drainage resistance of the brownstock was determined as Canadian Stan-dard Freeness (CSF). As shown in Table 4.1, the CSF of pre-extracted pulps werein the range of 600-700 mL whereas the freeness of the non-extracted pulps rangedfrom 520-560 mL. Additionally, the measured amount of fines, determined as fibrousmaterials with sizes between 0.07 and 0.2 mm, was lower in pre-extracted pulps thanin control pulps (Table 4.1); these results were in agreement with studies on kraftpulping of hemicellulose extracted sugar maple (Duarte et al. (2011)). This might beattributed to the pulping process was not strong enough to degrade fibres themselves(Duarte et al. (2011)). Therefore, the higher CSF values of extracted pulps were inaccordance with the decrease in the fines content. Higher CSF means faster rates ofwater drainage during brownstock washing, which improves the mill efficiency.1244.3. Results and DiscussionThe most significant observation was the very low residual silica content in kraftpulp from pre-extracted chips. As can be seen in Table 4.1, the residual silica contentof the pulp from pre-extracted chips was below 0.02% while it was 0.15-0.19% (basedon original o.d. chip mass) in the control. One reason for the high residual silicacontent of the pulp from non-extracted chips might be that the silicates dissolvedduring pulping adhere onto the fibre surface and are not removed during subsequentpulp washing. High silica contents in kraft pulp make it unsuitable for use in highgrade products such as ashless filter paper or facial tissue. In addition to the challengesof silica in the kraft pulp, high silica content in black liquor also causes problems inthe chemical recovery process such as scaling of evaporators, decrease in causticizingefficiency, and the generation of large amount of solid waste (calcium silicate mixedwith calcium carbonate). Moreover, silica in the black liquor is difficult to removebecause of the high lignin content. Therefore, alkaline pre-extraction is a promisingapproach to solve the silica problems when pulping bamboo.4.3.2 Pulp Physical PropertiesIt is important to assess the impact of the extraction process on the physical propertiesof resulting pulps. As indicated in Fig. 4.2 the initial freeness of the control pulpwas much lower than the initial freeness of the pulps from chip pre-extracted with19% EA. The rate of freeness drop with PFI refining was similar for all the pulps.These results indicate that pulps from alkaline pre-extracted chips need more refiningenergy to attain the same level of freeness as the control. This agrees well withthe generalized experience that pulps with low content of hemicellulose and fines are1254.3. Results and Discussion0 1000 2000 3000 4000300350400450500550600650 Control 80 C 3 h 100 C 1 h 100 C 5 h CSF (mL)Beating energy (Revolutions)Figure 4.2: Plot of pulp freeness (CSF) versus PFI mill revolution for pulps obtained fromkraft pulping of extracted and non-extracted chips with 19% EA.difficult to beat to a target freeness due to the small degree of internal fibrillation withincreased refining (Walton et al. (2010)). Similar refining results were also obtainedin kraft pulping with extracted and non-extracted chips with 15%, 17% and 19% EA(data not shown).The strength properties of all pulps were determined at the CSF of 425 mL.Plots of tensile and tear indices against EA charge of handsheets of pulps from pre-extracted chips and the control pulp are shown Figs. 4.3 and 4.4, respectively. Tensilestrength index of pulps from alkaline pretreated bamboo chips initially increased withincreasing EA charge, thereafter it decreased. For the control samples, the tensile1264.3. Results and Discussion13 14 15 16 17 18 19303540455055606570 Tensile index (Nm/g)Effective alkali charge (%) Control 80 C 3 h 100 C 1 h 100 C 5 hFigure 4.3: Effect of pre-extraction and effective alkali charge on the tensile index of pulp.index increased with increasing EA charge. These results confirmed the result thatthe delignification rate of pre-extracted chips was faster than the control. Withthe higher removal of lignin from treated chips at lower EA charge, more bondingcould be formed among cellulosic fibres than in the case of the control, resulting inthe improvement of physical properties of handsheets. However, with continuallyincreasing the EA charge, compared to the control pulps, the handsheet strength ofpulps from extracted bamboo chips decreased significantly. One likely reason is thelower hemicellulose content in extracted pulps, resulting in fewer bonding.It can be seen in Fig. 4.3 that the tensile strength index of pulps from chipsextracted using milder pre-treatment conditions (80 oC for 3 h and 100 oC for 1 h)1274.3. Results and Discussion13 14 15 16 17 18 1945678910Tear index (mN m2/g)Effective alkali charge (%) Control 80 C 3 h 100 C 1 h 100 C 5 hFigure 4.4: Effect of pre-extraction and effective alkali charge on the tensile index of pulp.1284.3. Results and Discussiondecreased at EA charges higher than 17%. This decrease began at 15% EA in thepulps from chips extracted at high severity (100 oC for 5 h). This is probably dueto the loss of different amounts of hemicellulose during the alkaline pre-extractionprocesses. With the increase in the intensity of alkaline pre-extraction, more hemi-cellulose were removed, resulting in higher cellulose/hemicellulose ratio, which willform larger macrofibrils during handsheet making (Molin and Teder (2002), Waltonet al. (2010)), resulting in lower bonding. In addition, at the highest tensile indexof handsheets, hemicellulose contents of the pulps were 15.41%, 11.68%, 11.25% and10.17% (based on o.d. pulp mass) for non-treated chips and chips pre-treated underconditions of 80 oC for 3 h, 100 oC for 1 h, and 100 oC for 5 h, respectively. Thus, itconfirmed the assumption that hemicellulose content plays an important role in thestrength properties of paper.Fig. 4.4 shows that the tear index of extracted pulp was better than that of non-extracted pulp. The overall improvement of tear strength with alkaline pre-extractioncan be explained by the removal of lignin and hemicellulose making the resultant kraftpulps to contain more cellulose per gram handsheet. On the other hand, the removalof fines might also improve the tear index of the extracted pulps.Results of the kraft pulping with extracted and non-extracted chips show that thealkaline pre-extraction is suitable for the production of papermaking fibre. Moreover,it also might be an alternative approach for the production of dissolving grade pulps(high purity cellulose) from bamboo. Moreover, the pre-extraction process removeshemicellulose and silica from the black liquor cycle which will improve the efficiencyof the chemical recovery process for the recovery of energy and inorganic chemicals(NaOH and Na2S). The hemicellulose and silica removed during the process can1294.3. Results and Discussionpotentially be converted into other value-added products.4.3.3 APEL PropertiesBased on our previous study on alkaline pretreatment of bamboo chips to extractthe maximum silica and hemicellulose while minimizing the degradation of celluloseand lignin (Chapter 3), the APEL was produced from the alkaline extraction with aNaOH charge of 10-14% (based on o.d. chip mass) at 90 oC for 3 h with a liquid-to-wood ratio of 8 L/kg in the laboratory. Table 4.2 shows the chemical compositionand other properties of the generated APELs.The concentration of silica, released as soluble sodium silicate (Na2SiO3), was1.45, 1.73 and 1.70 g/L in the APELs obtained from 10, 12 and 14% NaOH charges,respectively. Based on the silica concentration in the APEL, the liquid-to-wood ratio(8 L/kg) used for the extraction and the moisture content of treated chips (60-65%),it was calculated that up to 99% of silica in raw chips was removed with 12-14%NaOH, whereas about 86% of silica was removed with 10% NaOH charge (based ono.d. chip mass). This was further confirmed by the measurement of the silica contentof extracted chips, which was 0.17% (w/w) in the case of 10% NaOH charge and lessthan 0.02% (w/w) with 12% and 14% NaOH charge.The maximum concentration of lignin (mainly monomeric and oligomeric lignols)of the liquors was 1.99 g/L; this means that only about 5.1% of initial lignin wasremoved during alkaline pre-extraction under the studied conditions. The low ligninconcentration is favorable for the silica precipitation from the APEL. The extractedglucan in the liquors ranged from 1.24-1.46 g/L with increasing NaOH charge from1304.3. Results and Discussion10-14% (based on o.d. chip mass), corresponding to 1.7-2.0% of the initial glucan. Inthe liquors, xylan had a concentration of 8.78-9.97 g/L depending on NaOH charge,corresponding to about 30% of xylan being removed during alkaline pretreatment.Concentrations of galactan and arabinan in the APELs were 0.2-0.5 g/L, respectively,while acetic acid concentration was about 2.75 g/L. According to the measurements,the sum of biomass yield and dissolved materials in the APELs were calculated to be98.5% for the treatments under the three NaOH charges (10-14% w/w). This 1.5%biomass loss might be due to other undetermined dissolved inorganics and degradationproducts. The pH of generated APELs were 12.97, 13.12 and 13.28 for 10%, 12% and14% NaOH charges, respectively.Based on the above experimental data, utilization of 12% NaOH charge (on o.d.chip mass) not only had the potential of removing all silica from bamboo chips butalso could reduce the chemical consumption compared to that of 14% NaOH charge.Consequently, the APEL obtained from the pretreatment with 12% NaOH charge(based on o.d. chip mass) was selected and subjected for the recovery of silica andhemicellulose.4.3.4 Isolation of Silica from APEL with Carbon DioxideDuring CO2 treatment of the APEL, soluble sodium silicates (Na2SiO3) are convertedto the insoluble colloid silicic acid (H2SiO3), which can be removed from the liquor(see Equation 4.2). In this work, silica was recovered from the APEL by lowering itspH with CO2 at different temperatures (20-100 oC). Fig. 4.5 shows the effect of tem-perature and pH on the removal of silica from the APEL. Evidently, the temperature1314.3. Results and DiscussionTable 4.2: Chemical composition of alkaline pre-extraction liquor (APEL).Component NaOH charge10 12 14Biomass yield (%) 88.8 87.5 86.6Solid content (%) 3.26 3.31 3.32Lignin (g/L) 1.42 1.74 1.99Glucan (g/L) 1.24 1.29 1.46Xylan (g/L) 8.78 9.26 9.97Galactan (g/L) 0.21 0.22 0.34Arabinan (g/L) 0.32 0.41 0.45Silica (g/L) 1.45 1.73 1.70Sodium acetate (g/L) 2.76 2.77 2.74pH 12.97 13.12 13.28Values are expressed as averages of the three replicate measurements.had a minor effect on the amount of silica precipitated from the APEL. At the fivestudied temperatures, the percentage of removed silica varied 93-97% at a final liquorpH of about 9.0. The amount of silica removed was increased slightly by increasingthe temperature from 20 to 60 oC. Increasing in temperature to 100 oC led to a smallreduction in silica removal. The maximum silica removal was achieved at 60-80 oC.Moreover, compared to low temperature (20-60 oC), higher temperature (80-100 oC)required more CO2 to reach the same final pH. For example, it required about 120min to decrease the pH of the APEL to 9.0 at 100 oC at a 0.1 L/min CO2 flow rate;in contrast, at 20 oC with the same flow rate, pH 9.0 was reached in about 25 min.This is due to the decreasing solubility of CO2 with increasing temperature of thesolution, which means that the utilization of flue gas would be less efficient at highertemperatures. Therefore, it can be concluded that the optimum temperature for theprecipitation and recovery of silica from the APEL was around 60 oC. Additionally,since the temperature of the obtained APEL after pre-treatment (conducted at 90oC) was around 65 oC, the proposed process would not require extra heat in the1324.3. Results and Discussionprecipitation step.In contrast to temperature, the liquor pH plays a much more important role insilicic acid precipitation from the APEL. During the bubbling process, the pH of theAPEL gradually decreased over time towards a plateau at all studied temperatures.As shown in Fig. 4.5, in the pH range of 13.12 to 11.0, only about 4.8-8.7% of silicawas precipitated from the APEL at all studied temperatures. However, when thepH of the liquor was decreased from 11.0 to 9.0 by bubbling CO2 for a longer time(25-120 min), a sharp increase in the amount of precipitated silicic acid occurred; upto 93-97% of raw silica in the APEL was removed at the five treatment temperatures.For example, at 60 oC, about 96% of silica was removed from the APEL by loweringthe pH to 9.0. With continuous bubbling of CO2 into the liquor, more carbonicacid is formed, and the pH drops as the acid concentration increases resulting in therapid formation of the insoluble silicic acid at pH 11.0 to 9.0. About 2-5% of silicastill remained dissolved in the APEL even when the pH was decreased to 8.2. Thiscan be attributed to the solubility of silica even at low pH values, which means notall silica can be precipitated by lowering the pH of the solution (Hegde and Rao(2006); Liu et al. (2008)). In addition, it should be noted that it was very difficult toreduce the pH of the APEL below 8.0 even at the lowest temperature (20 oC) testedin this study, which might be due to the bicarbonate/carbonate buffering systemformed in the APEL. In contrast, lowering the pH to 8.2 was much easier to achieve.Accordingly, the pH of silica precipitation from the APEL should be limited to about8.2 in order to maintain mill efficiency.The APEL after silica removal can be used as a raw feedstock for the recovery ofhemicellulose to produce ethanol, xylitol or polymeric films; so it is of great impor-1334.3. Results and Discussion13 12 11 10 9 80.00.20.40.60.81.01.21.41.61.8Residual silica in APEL (g/L)pH 20 C 40 C 60 C 80 C 100 CFigure 4.5: Effect of temperature and pH on the precipitation of silica from the APEL.1344.3. Results and Discussiontance to minimize the loss of hemicellulose during the CO2 treatment stage. Analysisof the chemical composition of APEL recovered after maximum silica removal at 60oC showed that only about 4% of hemicellulose was removed during the silica re-moval. The likely reasons for hemicellulose loss include the potential co-precipitationof hemicellulose with silica as well as wash-off during the silica separation process. Itcan be concluded that the precipitation and filtration of silica from the APEL doesnot significantly influence the hemicellulose concentration in the supernatant, andwill not have a large negative effect on subsequent hemicellulose recovery.4.3.5 Capture of CO2 from the CO2/N2 MixtureWith the fundamental information obtained from the silica precipitation with CO2only, an attempt for developing the utilization of flue gas for recovering silica wasinitiated. Thus, in this study, a mixture gas of CO2 (0.1 L/min) and N2 (0.2 L/min)was used to simulate for the flue gas from the chemical recovery circle of the kraftpulp mills (Fan et al. (2009)). During the treatment process, the unused gas wasreleased from the reactor to make the pressure in the reactor constant at 1 atm.The released gas phase composition was analyzed during the silica precipitation atdifferent times, the CO2 composition is given in Table 4.3. The CO2 concentration inthe gas phase was found to gradually increase with increasing treatment time. This isbecause initially more CO2 was dissolved in the liquor and consumed by the reactionswith chemicals such as NaOH and Na2SiO3 in the APEL; with prolonged treatment,the chemical reactions were completed and the content of CO2 in the APEL reachedsaturation.1354.3. Results and DiscussionTable 4.3: Gas phase CO2 measurements during silica precipitation.Time (min) CO2 (%) (± 0.3)0 33.3310 20.3420 20.7830 21.4340 22.5350 23.5960 25.2570 26.4380 27.4290 28.19100 28.73110 29.15120 29.52Based on the compositional analysis of the residual gas phase, the actual adsorp-tion or consumption of CO2 for a period of 5-120 min was calculated. As shown inFig. 4.6, the adsorption/uptake of CO2 by the APEL initially increased with increas-ing time, thereafter it reached a plateau. At 60 and 120 min, the pH of the treatedAPEL was determined to be 8.16 and 8.08, respectively. These results confirmedthat the maximum consumption of CO2 was completed at 60 min and it was verydifficult to reduce the pH of the APEL below 8.0 with CO2. As illustrated in Fig.4.5, during the process of reducing the pH of APEL from 13.12 to 8.16, 1.94 L CO2(at a density of 1.842 g/L) was consumed per 200 mL APEL. Thus, the adsorptionof CO2 is calculated to be 7.15 gCO2/LAPEL (based on the 12% NaOH charge used inthis study). Moreover, since the real flue gas might contain small amounts of otheracidic compounds such as SO2, NOX or H2S (Douskova et al. (2009)), which will alsobe removed by the APEL, resulting in a relatively clean gas for emissions.1364.3. Results and Discussion0 20 40 60 80 100 120012345678Adsorbed CO2 (g/L)Treatment time (min)Figure 4.6: CO2 adsorption measurement curve for the precipitation of silica from theAPEL at 60 oC.1374.3. Results and DiscussionTable 4.4: Mineral composition of silica powder prepared from the APEL.Element Weight percentage (%)Calcium (Ca) 0.12Manganese (Mn) 0.0026Iron(Fe) 0.0071Aluminum (Al) Not detectedCopper (Cu) 0.0097Magnesium (Mg) 0.0075Potassium (K) 0.0057Sodium (Na) 0.0034Sum 0.1564.3.6 Compositional and FTIR Analysis of Silica PowdersTo remove organics (lignin and sugars) and prepare highly purified silica powder,oven dried silica samples were burnt at 700 oC for 4 h. The weight percentages of themineral ingredients present in the silica powders after burning were measured using aniCP spectrometer (Table 4.4). As can be seen in Table 4.4, the impurities containedin the silica powders were calcium, manganese, iron, copper, magnesium, potassiumand sodium; among which the concentration of calcium was the highest (0.12% w/w)while the total impurities accounted for about 0.16% (w/w). These results showedthat the produced powders contained more than 99.8% (w/w) of silica, confirming thehypothesis that high purity silica could be produced from the APEL of bamboo. Suchhigh purity of the precipitated porous silica particles makes it an excellent startingmaterial for various high-value products such as pharmaceuticals that require silicapurity higher than 99.7% (Morpurgo et al. (2010)).Fig. 4.7 shows the FTIR spectra of the silica powders before (a) and after (b)high temperature (700 oC) burning. As shown in Fig. 4.7a, the two absorption peaksat 3408 cm−1 and 1596 cm−1 are principally attributed to stretching of -OH groups1384.3. Results and Discussionand bending modes of absorbed water in precepitated silica particles, respectively(Kačuráková et al. (1998)). The predominant absorption band at 1074 cm−1 can beassigned to Si-O-Si asymmetric stretching and, absorbance peaks at 794 cm−1 is dueto the symmetric Si-O-Si bond (An et al. (2010); Knauss and Wolery (1988)). Peaksof Fig. 4.7a (before high temperature burning) confirmed that very little hemicelluloseco-precipitated during the silicic acid precipitation process. Comparison of the FTIRspectra 4.7b (after burning at 700 oC) and 4.7a shows the disappearance of the twoabsorption peaks at 3408 cm−1 and 1596 cm−1 upon burning, indicating the completeremoval of organics and water producing high purity silica powders. Other absorbancebands at 1084 cm−1 and 792 cm−1 (Fig. 4.7b) are also assigned to Si-O-Si asymmetricstretching and symmetric Si-O-Si bond, respectively (Hegde and Rao (2006)). Theshift in the 1074 cm−1 peak in Fig. 4.7a to 1084 cm−1 in Fig. 4.7b is likely dueto the elimination of the hydrogen bond (O-H) and the weakening of Si-O-Si afterhigh temperature treatment. The IR bands of silica powder investigated in this studywere similar to those reported in other investigations (Barik et al. (2008); Kalapathyet al. (2002)). Also, the absence of peaks at around 1510 cm−1 and 2930 cm−1 inthe FTIR analysis confirmed that lignin did not co-precipitate with silica duringAPEL processing (Liu et al. (2008)). Therefore, alkaline pre-extraction effectivelydecouples desilication from the major delignification reactions of kraft pulping. Thisis a significant benefit in kraft pulping because, with a pre-extraction step, silicadoes not fractionate into the black liquor during the kraft cook. Scaling and otherproblems that occur during kraft chemical recovery are hence eliminated. Moreover, ifnecessary, cleaner lignin fractions with lower ash, can be recovered from black liquor.1394.3. Results and DiscussionFigure 4.7: Fourier transform infrared (FTIR) spectra of silica produced from the APELof bamboo (a: before burning; b: after burning at 700 oC).4.3.7 Precipitation, Characterization and Utilization ofHemicellulose from the Treated APELThe hemicellulose recovery was conducted on the APEL after silica removal at opti-mum conditions (CO2 treatment at 60 oC to pH 8.6). Chemical composition analysisof the isolated hemicellulose showed that the total sugar content was 92.2%, in whichxylan, galactan, arabinan and glucan contents were 78.1%, 2.1%, 1.8%, and 10.2%,respectively. The measured lignin and ash contents were 2.7% and 1.1%, respectively.The uronic acid content was determined to be 4.0%, which was in good accordancewith reports that about 4-5% of uronic acid appeared as 4-O-methyl-D-glucuronicacid residues in bamboo feedstocks (Billa et al. (1996)). In the literature, simi-lar composition results have been reported for alkaline extracted hemicellulose fromhardwood and sugar cane bagasse (Peng et al. (2009); Jun et al. (2012)). Due to theco-precipitation of lignin and possibly extractives, the isolated hemicellulose had a1404.3. Results and DiscussionFigure 4.8: Fourier transform infrared (FTIR) spectra of hemicellulose produced from theAPEL.brightness of 32.6% ISO. After hemicellulose precipitation, the residual liquor can besent to the traditional kraft chemical recovery process to recover inorganics such assodium hydroxide for alkaline pre-extraction or kraft cooking.Fig. 4.8 shows the FTIR spectra of the isolated hemicellulosic fraction. Thestrong broad band at 3419 cm−1 is attributed to the stretching of O-H bond. Theabsorbance at 2920 cm−1 was due to the C-H stretching vibration. The absorbanceat 1611 cm−1 can be assigned to absorbed water in hemicellulose (Wen et al. (2011)).The prominent absorption peak of 1044 cm−1 is due to C-O stretching in C-O-Clinkages. The absorption bands of 1385, 1329, and 1246 cm−1 are characteristic ofhemicellulose obtained by alkali extraction (Xu et al. (2007)). The absorbance at897 cm−1 can be assigned to β − glycosidic linkages between the sugar units (Liuet al. (2011)). The small absorption peak of 1513 cm−1 is characterized by aromaticskeleton vibrations belonging to lignin (Billa et al. (1996)).To further characterize the isolated hemicellulose, size exclusion chromatography1414.3. Results and Discussion(SEC) was used to determine the average molecular weight and the polydispersityindex (PDI). The measured weight-average (Mw), number-average (Mn) molecularweights and the PDI were 26,770, 18,920 g/mol and 1.41 respectively. The small PDIvalue indicates that the obtained hemicellulose have a good chemical and structuralhomogeneity, which is favorable for downstream applications such as production ofhemicellulose-based bio-polymeric films or other applications.The obtained hemicellulose were used to prepare biopolymeric films by the ad-dition of 30% glycerol as the plasticizer. The tensile strength, elongation at break,WVTR and WVP were determined to be 30 MPa, 3.0%, 310 g/(m2 · d), and 0.8gmm/(m2 · d · kPa), respectively. At 50% relative humidity (RH), the oxygen perme-ability of the films formed with hemicellulose was 0.23 cm3 ·m/(m2 · d · kPa), whichwas substantially lower compared to the films formed with amylose and ethylene vinylalcohol at the same pH (Wen et al. (2011); Xu et al. (2007)). These results showedthat the hemicellulose recovered from APEL could be used as a good alternative,sustainable resource for the production of high-value added films (Liu et al. (2011);Stading et al. (1998)). As the property standards vary depending on end-use, poly-meric film produced from bamboo xylan (obtained in this study) needs to be furtherinvestigated to increase the tensile strength and the brightness.4.3.8 Proposed Modification to Kraft Pulping Process andMass BalanceBased on our findings, shown in Chapters 3 and 4, we propose a process for theintegration of low temperature alkaline pre-extraction of silica and hemicellulose from1424.3. Results and Discussionbamboo chips into a commercial kraft pulping as illustrated in Fig. 4.9. In a typicalkraft pulp mill, alkali is readily abundant as kraft white liquor, so the integrationof this pre-extraction stage could be achieved without major capital investment ofextensive process changes. In the proposed process scheme, washed bamboo chipsare treated with alkali solution under atmospheric conditions prior to kraft pulping.The treated chips with very low silica content are processed through kraft pulping orpre-hydrolysis kraft pulping to produce high-grade kraft pulp or dissolving pulp. Theextraction liquor (APEL) is treated for the recovery of dissolved materials includingsilica and hemicellulose, which is critical to maximize revenue to the kraft pulp millthrough the production of value added products. To precipitate silica from the APEL,CO2, a waste gas readily available as flue gas in pulp mills, was used to convert thesoluble silicates into insoluble silicic acid, which is isolated via filtering. Then, thefiltered samples were dried and burnt at 700 oC to remove any residual organicsand produce silica particles. The obtained amorphous silica particles can be easilypulverized to nanosilica for various applications (Liou (2004)). The separation ofthe maximum amounts of silica from the APEL significantly improves the recoveryof residual alkali and lime mud in the chemical recovery cycle of a typical kraftpulping process. Ethanol was then added to the CO2 treated APEL to precipitatethe hemicellulose. The recovered hemicellulose can be used as the starting materialfor bio-polymeric films production, biofuels fermentation or other applications. It isexpected that the alkaline pre-extraction will lower the alkali and other chemicalscharges required during subsequent kraft pulping and bleaching stages (Helmeriuset al. (2010)). Moreover, kraft white liquor, generated during the chemical recoverycircuit in the kraft pulp mill, can be readily used as the alkali solution for silica and1434.3. Results and DiscussionFigure 4.9: Proposed process of kraft pulping with extraction and recovery units of silicaand hemicellulose from bamboo chips.hemicellulose pre-extraction (Jun et al. (2012)). After the sequential steps of silicaand hemicellulose recovery, the double-treated APEL can be mixed with black liquorfrom kraft pulping and sent to the traditional chemical recovery circuit to recoverenergy and inorganic chemicals for mill operations.Taking the 3 h pre-extraction at 100 oC with 18% NaOH as an example, themass balance of the main components of bamboo around the whole system from pre-treatment to kraft cooking can be determined (Fig. 4.9). Mass balance of each frac-tion is expressed in terms of dried material mass. With regards to the pre-treatmentstage, the sum of recovered organics (cellulose, hemicellulose, lignin) and inorganic(silica) in both biomass residual and APEL corresponded to 97.4% of those in the rawbamboo chips. The 2.6% material loss is due to the degradation of lignin and/or hemi-cellulose into unidentified products. The analysis of the APEL showed that about35% of hemicellulose (mainly xylan) and 99% of silica contained in raw bamboo chipswere extracted, showing a good revenue source for the mill and a novel way for solving1444.3. Results and Discussionthe silica problems.The generated APEL was subjected to CO2 treatment to recover silica particles;about 1.06 g silica (after combustion at 700 oC) was produced. The CO2 treated liquorwas used for the hemicellulose recovery, in which the precipitated hemicellulose wasdetermined to be 7.1 g. Based on 100 g o.d. mass of chips entering the pre-extractionprocess, the mass of obtained silica (1.06 g) and hemicellulose (7.1 g) correspondedto 95.5% and 92.2% of the two components in the APEL, respectively. The lossesof silica and hemicellulose during the recovery process might be due to the solubilityof silica even at low pH and hemicellulose separation during CO2 treatment process,respectively.The pretreated bamboo chips (without washing) were subjected to kraft cooking,performed at 15% EA charge and 25% sulfidity. Based on 100 g o.d. mass of chipsentering the pre-extraction-kraft pulping process, the mass of obtained pulp was de-termined to be 52.5 g, in which masses of cellulose, hemicellulose, and lignin were 44.1g, 4.9 g, and 2.7 g, respectively. The recovered masses of cellulose and hemicelluloesin the brownstock corresponded to 94.7% and 36.5% of the two components in thepre-treated chips, respectively, showing reasonable agreement with previous studies(Pinto et al. (2005); Mân Vu et al. (2004)). The losses of cellulose and hemicelluloseduring delignification might be due to peeling reactions and alkaline degradation,respectively (Rocha et al. (2012)). For the material balance of lignin during kraftpulping, as shown, the total amount of precipitated lignin in the black liquor andlignin in the final pulp was 19.9 g, which was equivalent to 82.8% of the lignin in pre-treated chips. The 17.2% loss of lignin in raw chips during kraft cooking might be dueto water washing of the kraft pulp and incomplete lignin precipitation during black1454.3. Results and Discussionliquor acidification. Finally, with regards to the main components (cellulose, hemi-cellulose, and lignin), the overall recovery of the proposed system from pre-extractionto kraft pulping and byproducts recovery showed a good mass balance of 80.2%.Moreover, one important aspect of the proposed scheme is the adsorption of CO2in a cost-effective way, which transfers the industrial waste (flue gas) to a valuablestarting material; this has the potential for lowering capital and operating costs andreducing pulp mill greenhouse gas emissions. In using silica-rich biomasses for biore-fineries, this proposed process has several advantages, for example, it creates a silica-free substrate with high digestibility. During the subsequent processing of alkalineextracted chips, lower chemical or enzyme charge could be expected (Kumar et al.(2009); Jun et al. (2012); Huang et al. (2008)). Additionally, a relatively clean ligninfraction could be obtained from the spent liquor after cooking or enzymatic hydrolysisof silica- and extractive-free bamboo chips.For a biomass-to-pulp plant that has a capacity of 600 tons bleached kraft pulp perday, about 1500 o.d. tons of bamboo chips are required. With the studied bamboo,about 16.8 tons/day of silica is processed into the pulping scheme, which causes seri-ous problems in chemical recovery circuit and quality of final pulp. With the proposedtechnology, alkaline pretreatment of bamboo chips carried out with a liquid-to-woodratio of 8 L/kg at 90 oC and 12% NaOH charge (based on o.d. chip mass) for 3 h,about 9500 m3/day APEL will be generated. To precipitate the dissolved silica inthe APEL, about 67.9 tons/day of CO2 is required, which will might improve thehandling of acidic compounds in the mill waste gas. Moreover, the separation of themaximum amounts of silica from the APEL significantly improves the recovery ofresidual alkali and lime mud in the chemical recovery cycle of a typical kraft pulp-1464.4. Process Engineeringing process. The extraction of silica prior to pulping also increases the evaporationefficiency by lowering the viscosity of black liquor obtained from kraft pulping andeliminating the scale buildup in the evaporators (Cardoso et al. (2009)). In addition,the reduction of silica content of the lime mud generated during the causticizing stageof kraft chemical recovery makes the lime mud recyclable, necessitating disposal oflarge amount of solid waste which is detrimental to the environment (Pekarovic et al.(2005)). In addition, hemicellulosic sugars in the APEL can be separated by ultrafil-tration or nanofiltration (Ahsan et al. (2014)), which greatly reduces the amount ofethanol required.One of the most advantageous aspects of our proposed approach is that the methoddecouples silica from organics-rich black liquor enabling silica recovery without sub-stantial loss of lignin and hemicellulose. Moreover, the proposed approach may workfor a bioethanol plant as well with using high-silica biomasses as raw material.4.4 Process Engineering4.4.1 Process DescriptionTo evaluate the feasibility of the proposed process in an industrial mill application, amass balance of each operation stage was calculated. Taking into account the existingequipment and the pulping process of the typical kraft pulp mill (Chongqing Lee &Man mill, China as an example), a process combination was designed to resolve thesilica problems and recover byproducts, shown in Fig. 4.10.In the proposed process, before kraft cooking, bamboo chips are pre-treated with1474.4. Process EngineeringFigure 4.10: Combination of the new stages in a typical kraft bamboo pulp mill.1484.4. Process Engineeringa NaOH solution (white liquor) in the digester, which converts insoluble silica intosoluble sodium silicate. The solution of dissolved sodium silicate is separated from thebamboo chips by displacement with fresh white liquor. Then carbon dioxide (flue gas)is added into the pre-extraction liquor (APEL) to lower the pH and convert sodiumsilicate into insoluble silicic acid, which can be precipitated and removed from theliquor. After this, the resulting liquor is mixed with black liquor from Kraft cookingto be sent to the chemical recovery process. Because the silica in the mixed liquorhas been removed, current problems encountered during chemical recovery processes,including scaling of evaporators, low evaporation efficiency, and the inability to recyclelime mud, are substantially resolved.1494.4. Process EngineeringTable 4.5: Mass balance of the pulp line.Program Liquid phase (L) Solid phaseinput output input rejects dissolvedmaterialChips 470 - 530 - -Screening system - 16.45 - 18.55 -Chip washing system (5 L/kg) 2650 2638.52 511.45 7.67 -Silica/hemicelluloseextraction (8 L/kg) 3565.21 3316.78 503.78 5.04 70.53Kraft cooking (4.4 L/kg) 1241.80 607.32 428.21 - -Washing of brown stock 3637.25 2566.29 8.35 141.52O2 delignification stage 8082.83 8323.13 269.99 2.70 27.0D0 stage 3641.69 3653.08 242.99 1.21 1.21EOP stage 3608.40 3631.29 240.56 1.20 1.68D1 stage 3565.05 3576.24 237.67 0.71 0.48Overall 30458.23 28329.10 236.48 293.52Final pulp storage Lin − Lout = 2129.13 Qin −Qout = 236.48Real pulp consistency inthe storage tank 236.48/(2129.13 + 236.48) = 10%4.4.2 Mass Balance of Pulp Line and Chemical RecoveryProcessAccording to the calculations of the mass balance of each operation step (AppendixA3), a summary of the pulping process with the silica and hemicellulose extractionstage was made, shown in Table 4.5. The APEL obtained from the silica/hemicellulosestage is 3312.35 L, which will consume about 23.7 kg CO2 for the recovery of silicafrom the APEL. Moreover, according to the calculation for illustrating an industrialprocess, a cost estimate on the proposed stages, alkaline pretreatment and recoveryof dissolved biomass, can be made.Table 4.6 shows the mass balance of the chemical recovery process. Due to thelack of data with stages of recovery boiler, causticizing system and calcination, mass1504.4. Process EngineeringTable 4.6: Mass balance of the chemical recovery process.Program Input Outputliquid (L) solid (kg) liquid (L) solids weight (kg)Silica recovery 3450.89 75.57 3379.79 71.1Hemicellulose recovery 3379.79 64.44 3043.09 336.7Evaporation 5575.98 170.52 284.2 -Recovered silica 71.1× 10% = 7.11 kgRecovered hemicellulose 336.7× 10% = 33.67 kgbalance of this part was not made.This proposed modification to the typical kraft pulping process not only solvesthe silica problems in the kraft pulping of bamboo, but also provides a kraft pulpmill a green pathway to manufacture high-value products from the pre-extracted andprecipitated silica and hemicellulose. This proposed process has several advantages,for example, it increases the revenue to the kraft pulp mill by recovering silica andhemicellulose for various consumer products. It also converts waste industrial flue gascontaining CO2 to a valuable starting material for silica precipitation. Moreover, theextraction of silica prior to pulping increases the evaporation efficiency by loweringthe viscosity of black liquor obtained from kraft pulping and eliminating the scalebuildup of evaporators. In addition, the reduction of silica content of the lime mudgenerated during the causticizing stage of chemical recovery circuit of kraft pulpingprocess makes the lime mud recyclable, as opposed to its disposal as a solid waste.Another significant aspect of the process is the ease in which it can be implementedin existing kraft-based pulping mills.1514.5. Summary4.5 SummaryAlkaline pre-extraction of the bamboo chips improved the delignification during kraftpulping even at lower effective alkali charges. Pulp from pre-extracted bamboo chipsshowed similar screened yield to that of non-extracted chips while initial drainageresistance (CSF) improved slightly. The tensile strength index of the pulp actuallybenefited from the alkaline pre-treatment at low EA charges. The tear strength indexof the kraft pulp was improved by the alkaline pre-extraction. Silica was not detectedin the kraft pulp and black liquor obtained with pre-extracted bamboo chips indicatingthat the process provides a solution to silica problems encountered in bamboo pulping.All the silica and a part of the hemicellulose in bamboo chips were fractionated intoan alkaline pre-extraction liquor (APEL) during alkali pre-treatment of bamboo chips.To recover the dissolved silica and hemicellulose from the APEL, a sequential routeinvolving carbon dioxide (CO2) precipitation and ethanol treatment was developed.Up to 97% of silica dissolved in the APEL was recovered by treating the APEL at60 oC with CO2 (flue gas could be used in pulp mills) to a final pH of 8.2. TheCO2 adsorption capacity of the APEL was determined to be 7.15 gCO2/LAPEL. Theprecipitated silica nanoparticles had a purity of 99.8%, indicating that the isolatedsilica can be used as the starting material for various high-value products. Thehemicellulose isolated from the CO2-treated APEL had a uniform molecular weightdistribution, showing their potential for downstream conversion into high-value bio-products. Moreover, polymeric film fabricated from the isolated hemicellulose showedgood properties such as high tensile strength and elongation at break and, relativelylow WVTR, WVR and oxygen permeability.1524.5. SummaryThrough this study, an environmentally friendly way to produce pure silica nanopar-ticles and polymeric hemicellulose from bamboo has been developed. In addition, theremoval of silica prior to biomass processing in pulp mills or biorefineries solves theproblems created by silica in the recovery cycle of kraft pulping. The proposed useof pulp mill waste gas containing carbon dioxide would recycle pulp mill greenhousegas emissions and contribute to reduce global warming.A mass balance of the proposed alkaline pretreatament-kraft pulping process wasperformed. The silica extracted during pretreatment can be recovered from the APELby lowering the pH prior to pulping the raw material. Moreover, CO2 present inthe waste flue gas can be used to lower the pH of the APEL avoiding increasedCO2 emissions from the lime kiln. In addition, the removal of silica also makes thegenerated lime mud recyclable during the subsequent calcination process producinggood quality CaO that can be used in causticizing thus avoiding solid waste disposalissues and the necessity of purchasing fresh lime. Both the utilization of flue gas andthe recovery of CaO improve the sustainability of the conventional pulping processand make the bamboo pulping process more environmentally friendly.153Chapter 5Summary of Thesis andRecommendations for FutureResearch5.1 Summary of ContributionsIn this thesis, a series of experimental and numerical studies were conducted to studyhemicellulose hydrolysis during acidic pre-hydrolysis and the removal of silica andhemicellulose during alkaline pretreatment. The effect of alkaline pretreatment onsubsequent kraft pulping was also investigated by evaluating the resultant pulp prop-erties. To investigate if additional value could be obtained from the pulping process(a biorefinery concept) the feasibility of recovering dissolved silica and hemicellulosicsugars in the alkaline pre-extraction liquor was carried out. To further confirm thisfeasibility, a statistical mass balance of the whole proposed process was performed.Specifically, the conclusions of this thesis are:1. The evolution of proton concentration was examined during the auto- anddilute-acid hydrolysis of bamboo chips. A \"toy model\" was successfully pro-posed that described the reaction. The toy model predicts the existence of a1545.1. Summary of Contributionssteady state solution, which is dictated by the equilibrium constants, and theinitial added acid and acetyl group levels. The model qualitatively follows thetrend given by experiment. The model was tested at room temperature to ex-amine the changes in pH when ash neutralization is the dominant mechanism.Under these conditions, the pH remains essentially constant both at low initialpH and under autohydrolysis conditions. When the hydrolysis reaction pro-ceeded at elevated temperatures, the pH initially increased, then decreased atlonger times due to the difference in the rates of the neutralization and deacety-lation reactions. With our toy model we propose a chemical reaction pathwaythat satisfactorily describes experimentally determined proton concentrationunder both autohydrolytic and dilute acid hydrolysis conditions. This accuratemodeling of the proton concentration should significantly improve the existingkinetic models of hemicellulose hydrolysis and facilitate more efficient processoptimization and scale-up.2. Alkaline pretreatment with NaOH was demonstrated to be an effective wayto selectively extract silica and hemicellulose from bamboo biomass withoutdegrading cellulose and lignin.(a) The dissolution of pure amorphous silica with different particle sizes inNaOH solution was experimentally studied in the lab. Results showed thatsurface area was the controlling parameter of the reaction rate betweensilica and NaOH. The standard shrinking core model could be used todescribe the silica dissolution rate in the alkali solution under the NaOHconcentrations studied in this work.1555.1. Summary of Contributions(b) Large deviations were obtained when using the shrinking core to describethe removal of silica from bamboo powder at low NaOH concentration (<0.45 mol/L).(c) During alkaline pretreatment of bamboo biomass, several reactions, suchas degradation of xylan, dissolution of silica and neutralization of acetyland uronic acid groups, occur in parallel. Due to the consumption of OH–by several bamboo chemical components, the dissolution rate of silica inbamboo powder was lower than that of pure amorphous silica particles.(d) The extraction of xylan and silica from bamboo powder increased with theincrease of the treatment severity such as higher temperature, higher NaOHcharge and longer reaction time. All silica and up to 55% of xylan couldbe extracted by treating bamboo powder with initial NaOH concentrationof 0.45 mol/L at 100 oC for 180 min.(e) To understand the effect of chip size on the extraction rates of silica andxylan from bamboo biomass, alkaline pretreatment of bamboo chips wascarried out under the same conditions used for bamboo powder. Resultsdemonstrated that mass-transfer effects play an important role in the pre-extraction of both silica and xylan from bamboo chips. A 5-step extractionmechanism was proposed based on the experimental results.(f) The response surface methodology (RSM) and central composite rotatabledesign (CCRD) were used to determine the most significant variables formaximum removal of silica and hemicellulose while maintaining celluloseand lignin content. Polynomial equations describing extraction of xylan1565.1. Summary of Contributionsand silica, and chip yield were obtained as a function of the treatment vari-ables of temperature, NaOH charge and time. According to the equationsand confirmatory experiments, 97% desilication could be reached by treat-ing bamboo chips for 70 -90 min at temperatures of 95-100 oC with NaOHcharges of 16-18% (based on the o.d. chip mass). Under these conditionsup to 40% of the xylan could be extracted.(g) All silica and up to 45% of hemicellulose were removed from bamboo chipsunder the studied conditions. The alkaline pretreatment of bamboo chipsprovides a novel way to resolve the silica issue encountered when pulpingwith bamboo.3. Alkaline pre-extraction of the bamboo chips improved delignification duringsubsequent kraft pulping even at lower effective alkali (EA) charge. Pulp frompre-extracted bamboo chips showed similar screened pulp yield to that of non-extracted chips while initial pulp freeness (CSF) improved slightly. The tensilestrength index of the pulp actually benefited from the alkaline pre-treatmentat low EA charges. The tear strength index of the kraft pulp was improved bythe alkaline pre-extraction. Silica was not detected in the kraft pulp and blackliquor obtained using pre-extracted bamboo chips, indicating that the processprovides a solution to silica problems encountered in bamboo pulping.4. Recovery of dissolved silica and hemicellulose from the spent liquor is of greatimportance in converting the kraft pulp mill into a biorefinery unit.(a) To recover the dissolved silica and hemicellulose from the alkaline pre-extraction liquor (APEL), a sequential route involving carbon dioxide1575.1. Summary of Contributions(CO2) and ethanol precipitation was developed. Up to 97% of silica dis-solved in the APEL was recovered by treating the APEL at 60 oC withCO2 to a final pH of 8.2. The CO2 adsorption capacity of the APELwas determined to be 7.15 gCO2/LAPEL with the 12% NaOH charge usedfor pretreatment of bamboo chips. The precipitated silica had a purityof 99.8%, indicating it can be an excellent starting material for varioushigh-value products. For a typical bamboo-to-pulp mill with a capacity of600-700 tons pulp per day, about 67.9 tons/day CO2 is required and 16.8tons/day silica is generated.(b) The hemicellulose isolated from the CO2-treated APEL had a uniformmolecular weight distribution, showing their potential for downstream con-version into high-value bio-products. Moreover, polymeric film fabricatedfrom the isolated hemicellulose showed good properties such as high ten-sile strength and elongation at break and, relatively low WVTR, WVPand oxygen permeability.5. According to our findings, we proposed a scheme that integrates alkaline extrac-tion and the recovery of silica and hemicellulose from bamboo into a commercialkraft pulping process. In the proposed process, extracted bamboo chips serveas the starting material for a kraft-based pulping processes to produce conven-tional kraft pulp, while the extraction liquor is used for the recovery of dissolvedmaterials such as silica and hemicellulose. Moreover, the pre-extraction of silicafrom bamboo chips also makes the generated lime mud (CaCO3) from causti-cizing recyclable in the lime kiln, avoiding disposal of large amounts of solid1585.2. Recommendations for Future Workwaste and the requirement for purchase of fresh lime.6. A statistical mass balance for the proposed process was also presented. Resultsdemonstrated the feasibility of the proposed process for the production of kraftpulp and the resolution of silica challenges.5.2 Recommendations for Future WorkThe results of this thesis provide a strong foundation for future work. The focus ofthe future work could be in several areas, these are:1. Further development of the kinetic model for the hydrolysis of xylan duringauto- and dilute-acid hydrolysis of bamboo chips.2. Establishing the mathematical model to describe the reactions during alkalinepretreatment of bamboo powder and bamboo chips.3. To develop a more efficient and costly-effective way to recover hemicellulosefrom the silica-free alkaline pre-extraction liquor (APEL).4. An optimization study can be conducted to produce high grade dissolving pulpby integrating silica pre-extraction into the pulping process.5. Reactor design. Design a liquor tank for the recovery and separation of silicicacid from the alkaline pre-extraction liquor (APEL) with flue gas treatment.6. A techno-economical analysis is generally required to assess the feasibility ofthe proposed process for the utilization of bamboo in biorefineries.159ReferencesAffandi, S., Setyawan, H., Winardi, S., Purwanto, A., and Balgis, R. (2009). A facilemethod for production of high-purity silica xerogels from bagasse ash. Adv. PowderTechnol., 20(5):468–472.Agbor, V. B., Cicek, N., Sparling, R., Berlin, A., and Levin, D. B. (2011). Biomasspretreatment: fundamentals toward application. Biotechnol. Adv., 29(6):675–685.Aguilar, R., Ramırez, J., Garrote, G., and Vázquez, M. (2002). Kinetic study of theacid hydrolysis of sugar cane bagasse. J. Food Eng., 55(4):309–318.Ahsan, L., Jahan, M. S., and Ni, Y. (2014). Recovering/concentrating of hemicellu-losic sugars and acetic acid by nanofiltration and reverse osmosis from prehydrol-ysis liquor of kraft based hardwood dissolving pulp process. Bioresour. Technol.,155:111–115.Alexander, G. B., Heston, W., and Iler, R. K. (1954). The solubility of amorphoussilica in water. J. Physical Chem., 58(6):453–455.Altaner, C. and Jarvis, M. (2008). Modelling polymer interactions of the ’molecularvelcro’ type in wood under mechanical stress. J. Theoretical Biology, 253(3):434–445.160ReferencesAlvarez-Vasco, C. and Zhang, X. (2013). Alkaline hydrogen peroxide pretreatment ofsoftwood: hemicellulose degradation pathways. Bioresour. Technol., 150:321–327.Amada, S. (1995). Hierarchical functionally gradient structures of bamboo, barley,and corn. Mrs Bulletin, 20(01):35–36.An, D., Guo, Y., Zhu, Y., and Wang, Z. (2010). A green route to preparation of silicapowders with rice husk ash and waste gas. Chem. Eng. J., 162(2):509–514.Anatskii, F. and Ratinov, V. (1969). Mechanism and kinetics of reaction between ac-tive silicon and alkali-metal hydroxides. Doklady Akademi Nauk Sssr, 186(6):1341.Arato, C., Pye, E. K., and Gjennestad, G. (2005). The lignol approach to biorefiningof woody biomass to produce ethanol and chemicals. App. Biochem. Biotechnol.,123(1-3):871–882.Atkinson, A., Donev, A., and Tobias, R. (1992). Optimum experimental designs.Oxford University Press, New York, NY.Avdeef, A., Comer, J. E., and Thomson, S. J. (1993). ph-metric log p. 3. glasselectrode calibration in methanol-water, applied to pka determination of water-insoluble substances. Analytical Chem., 65(1):42–49.Bai, L., Hu, H., and Xu, J. (2012). Influences of configuration and molecular weight ofhemicelluloses on their paper-strengthening effects. Carbohydr. Polym., 88(4):1258–1263.Barik, T., Sahu, B., and Swain, V. (2008). NanosilicaâĂŤfrom medicine to pestcontrol. Parasitology Res., 103(2):253–258.161ReferencesBenkhedda, K., Infante, H. G., Ivanova, E., and Adams, F. C. (2000). Trace metalanalysis of natural waters and biological samples by axial inductively coupledplasma time of flight mass spectrometry (icp-tofms) with flow injection on-line ad-sorption preconcentration using a knotted reactor. J. Analy. Atomic Spectrometry,15(10):1349–1356.Bennett, D. M. and Sangster, A. (1981). The distribution of silicon in the adventitiousroots of the bamboo sasa palmata. Canadian J. Botany, 59(9):1680–1684.Berglin, N. and Berntsson, T. (1998). Chp in the pulp industry using black liquorgasification: thermodynamic analysis. App. Thermal Eng., 18(11):947–961.Bergna, H. E. and Roberts, W. O. (2005). Colloidal silica: fundamentals and appli-cations, volume 131. CRC Press.Billa, E., Tollier, M.-T., and Monties, B. (1996). Characterisation of the monomericcomposition ofin situwheat straw lignins by alkaline nitrobenzene oxidation: Effectof temperature and reaction time. J. Sci. Food Agricul., 72(2):250–256.Blumberg, A. A. (1959). Differential thermal analysis and heterogeneous kinetics thereaction of vitreous silica with hydrofluoric acid. J. Physical Chem., 63(7):1129–1132.Borrega, M., Nieminen, K., and Sixta, H. (2011). Effects of hot water extraction in abatch reactor on the delignification of birch wood. BioResources, 6(2):1890–1903.Borrega, M., Tolonen, L. K., Bardot, F., Testova, L., and Sixta, H. (2013). Potential162Referencesof hot water extraction of birch wood to produce high-purity dissolving pulp afteralkaline pulping. Bioresour. Technol., 135:665–671.Brennan, M. A. and Wyman, C. E. (2004). Initial evaluation of simple mass transfermodels to describe hemicellulose hydrolysis in corn stover. App. Biochem. Biotech-nol., 115(1-3):965–976.Brückner, R. (1970). Properties and structure of vitreous silica i. J. Non-crystallineSolids, 5(2):123–175.Cahela, D. R., Lee, Y., and Chambers, R. (1983). Modeling of percolation process inhemicellulose hydrolysis. Biotechnol. Bioeng., 25(1):3–17.Cardoso, M., de Oliveira, É. D., and Passos, M. L. (2009). Chemical composition andphysical properties of black liquors and their effects on liquor recovery operation inbrazilian pulp mills. Fuel, 88(4):756–763.Carrasco, F. and Roy, C. (1992). Kinetic study of dilute-acid prehydrolysis of xylan-containing biomass. Wood Sci. Technol., 26(3):189–208.Chand, N., Jain, D., and Nigrawal, A. (2006). Investigation on gradient dielectriccharacteristics of bamboo (dentroclamus strictus). J. App. Poly. Sci., 102(1):380–386.Chandra, R. P., Arantes, V., and Saddler, J. (2015). Steam pretreatment of agricul-tural residues facilitates hemicellulose recovery while enhancing enzyme accessibil-ity to cellulose. Bioresour. Technol., 185:302–307.163ReferencesChandra, R. P., Chu, Q., Hu, J., Zhong, N., Lin, M., Lee, J.-S., and Saddler, J.(2016). The influence of lignin on steam pretreatment and mechanical pulping ofpoplar to achieve high sugar recovery and ease of enzymatic hydrolysis. Bioresour.Technol., 199:135–141.Chang, Y.-N., Huang, J.-C., Lee, C.-C., Shih, L., and Tzeng, Y.-M. (2002). Useof response surface methodology to optimize culture medium for production oflovastatin by monascus ruber. Enzyme Microbial Technol., 30(7):889–894.Chen, C.-T. A. and Marshall, W. L. (1982). Amorphous silica solubilities iv. behaviorin pure water and aqueous sodium chloride, sodium sulfate, magnesium chloride,and magnesium sulfate solutions up to 350 oC. Geochimica et Cosmochimica Acta,46(2):279–287.Chundawat, S. P., Venkatesh, B., and Dale, B. E. (2007). Effect of particle size basedseparation of milled corn stover on afex pretreatment and enzymatic digestibility.Biotechnol. Bioeng., 96(2):219–231.Cochran, W. and Cox, G. (1968). Experimental designs, john wiley, new york.Collin, B., Doelsch, E., Keller, C., Panfili, F., and Meunier, J.-D. (2012). Distributionand variability of silicon, copper and zinc in different bamboo species. Plant Soil,351(1-2):377–387.Conner, A. H. and Lorenz, L. F. (1986). Kinetic modeling of hardwood prehydrolysis.part iii. water and dilute acetic acid prehydrolysis of southern red oak. Wood FiberSci., 18(2):248–263.164ReferencesCosgrove, D. C. et al. (2012). Comparative structure and biomechanics of plantprimary and secondary cell walls. Frontiers Plant Sci., 3:1–6.Cox, G. B. (1993). The influence of silica structure on reversed-phase retention. J.Chromatography A, 656(1):353–367.Dammström, S., Salmén, L., and Gatenholm, P. (2008). On the interactions betweencellulose and xylan, a biomimetic simulation of the hardwood cell wall. BioRe-sources, 4(1):3–14.De Lopez, S., Tissot, M., and Delmas, M. (1996). Integrated cereal straw valorizationby an alkaline pre-extraction of hemicellulose prior to soda-anthraquinone pulping.case study of barley straw. Biomass Bioenergy, 10(4):201–211.Dence, C. W. (1992). The determination of lignin. In :Methods in Lignin Chemistry,pages 33–61. Springer.Ding, T., Zhou, J., Wan, D., Chen, Z., Wang, C., and Zhang, F. (2008). Siliconisotope fractionation in bamboo and its significance to the biogeochemical cycle ofsilicon. Geochimica et Cosmochimica Acta, 72(5):1381–1395.Douskova, I., Doucha, J., Livansky, K., Machat, J., Novak, P., Umysova, D., Zach-leder, V., and Vitova, M. (2009). Simultaneous flue gas bioremediation and re-duction of microalgal biomass production costs. App. Microbiology Biotechnol.,82(1):179–185.Duarte, G. V., Ramarao, B. V., Amidon, T. E., and Ferreira, P. T. (2011). Effect of165Referenceshot water extraction on hardwood kraft pulp fibers (acer saccharum, sugar maple).Ind. Eng. Chem. Res., 50(17):9949–9959.East, C. P., Fellows, C. M., and Doherty, W. O. (2013). Aspects of the kinetics andsolubility of silica and calcium oxalate composites in sugar solutions. J. Food Eng.,117(3):291–298.Engström, A.-C., Ek, M., and Henriksson, G. (2006). Improved accessibility and reac-tivity of dissolving pulp for the viscose process: pretreatment with monocomponentendoglucanase. Biomacromolecules, 7(6):2027–2031.Esteghlalian, A., Hashimoto, A. G., Fenske, J. J., and Penner, M. H. (1997). Modelingand optimization of the dilute-sulfuric-acid pretreatment of corn stover, poplar andswitchgrass. Bioresour. Technol., 59(2):129–136.Fan, S., Li, S., Wang, J., Lang, X., and Wang, Y. (2009). Efficient capture of co2from simulated flue gas by formation of tbab or tbaf semiclathrate hydrates. EnergyFuels, 23(8):4202–4208.Farrell, E. P., Führer, E., Ryan, D., Andersson, F., Hüttl, R., and Piussi, P. (2000).European forest ecosystems: building the future on the legacy of the past. ForestEcology Management, 132(1):5–20.Fauteux, F., Rémus-Borel, W., Menzies, J. G., and Bélanger, R. R. (2005). Silicon andplant disease resistance against pathogenic fungi. FEMS Microbiol. Lett., 249(1):1–6.166ReferencesFengel, D. andWegener, G. (1983). Wood: chemistry, ultrastructure, reactions. Walterde Gruyter, New York, USA.Fleming, B. and Crerar, D. (1982). Silicic acid ionization and calculation of silicasolubility at elevated temperature and ph application to geothermal fluid processingand reinjection. Geothermics, 11(1):15–29.García-Aparicio, M., Parawira, W., Van Rensburg, E., Diedericks, D., Galbe, M.,Rosslander, C., Zacchi, G., and Görgens, J. (2011). Evaluation of steam-treated gi-ant bamboo for production of fermentable sugars. Biotechnol. Progress, 27(3):641–649.Garrote, G., DomıÌĄnguez, H., and Parajó, J. C. (2001). Generation of xylose solu-tions from eucalyptus globulus wood by autohydrolysis–posthydrolysis processes:posthydrolysis kinetics. Bioresour. Technol., 79(2):155–164.Garrote, G., Eugenio, M. E., Díaz, M. J., Ariza, J., and López, F. (2003). Hydrother-mal and pulp processing of Eucalyptus. BioresouR. Technol., 88(1):61–68.Gáspár, M., Kálmán, G., and Réczey, K. (2007). Corn fiber as a raw material forhemicellulose and ethanol production. Process Biochem., 42(7):1135–1139.González-Muñoz, M., Alvarez, R., Santos, V., and Parajó, J. (2012). Productionof hemicellulosic sugars from pinus pinaster wood by sequential steps of aqueousextraction and acid hydrolysis. Wood Sci. Technol., 46(1-3):271–285.Gossett, J. M., Stuckey, D. C., Owen, W. F., and McCarty, P. L. (1982). Heattreatment and anaerobic digestion of refuse. J. Environ. Eng. Div., 108:437–454.167ReferencesGratani, L., Crescente, M. F., Varone, L., Fabrini, G., and Digiulio, E. (2008).Growth pattern and photosynthetic activity of different bamboo species growing inthe botanical garden of rome. Flora-Morphology, Distribution, Functional EcologyPlants, 203(1):77–84.Greenberg, S. and Price, E. (1957). The solubility of silica in solutions of electrolytes.J. Physical Chem., 61(11):1539–1541.Grénman, H., EraÌĹnen, K., Krogell, J., WillfoÌĹr, S., Salmi, T., and Murzin, D. Y.(2011). Kinetics of aqueous extraction of hemicelluloses from spruce in an intensifiedreactor system. Ind. Eng. Chem. Res., 50(7):3818–3828.Gunnarsson, I. and Arnórsson, S. (2000). Amorphous silica solubility and the ther-modynamic properties of H4SiO4 in the range of 0 to 350 oc at p sat. Geochimicaet Cosmochimica Acta, 64(13):2295–2307.Gütsch, J. S., Nousiainen, T., and Sixta, H. (2012). Comparative evaluation of auto-hydrolysis and acid-catalyzed hydrolysis of eucalyptus globulus wood. Bioresour.Technol., 109:77–85.Hallac, B. B. and Ragauskas, A. J. (2011). Analyzing cellulose degree of polymeriza-tion and its relevancy to cellulosic ethanol. Biofuels, Bioprod. Biorefining, 5(2):215–225.Hamzeh, Y., Ashori, A., Khorasani, Z., Abdulkhani, A., and Abyaz, A. (2013). Pre-extraction of hemicelluloses from bagasse fibers: Effects of dry-strength additiveson paper properties. Ind. Crops Prod., 43:365–371.168ReferencesHe, Y. and Yue, Y. (2008). A review of the effective component and applications ofextracts from bamboo leaves. Biomass Chem. Eng., 3:31–37.Hegde, N. D. and Rao, A. V. (2006). Organic modification of teos based silica aero-gels using hexadecyltrimethoxysilane as a hydrophobic reagent. App. Surface Sci.,253(3):1566–1572.Hektor, E. and Berntsson, T. (2007). Future co2 removal from pulp mills–processintegration consequences. Energy Conv. Management, 48(11):3025–3033.Helmerius, J., von Walter, J. V., Rova, U., Berglund, K. A., and Hodge, D. B.(2010). Impact of hemicellulose pre-extraction for bioconversion on birch kraftpulp properties. Bioresour. Technol., 101(15):5996–6005.Hong, B., Xue, G., Guo, X., and Weng, L. (2013). Kinetic study of oxalic acidpretreatment of moso bamboo for textile fiber. Cellulose, 20(2):645–653.Hosseini, S. A. and Shah, N. (2009). Multiscale modelling of hydrothermal biomasspretreatment for chip size optimization. Bioresour. Technol., 100(9):2621–2628.Huang, F. and Ragauskas, A. (2013). Extraction of hemicellulose from loblolly pinewoodchips and subsequent kraft pulping. Ind. Eng. Chem. Res., 52(4):1743–1749.Huang, H.-J., Ramaswamy, S., Al-Dajani, W. W., and Tschirner, U. (2010). Processmodeling and analysis of pulp mill-based integrated biorefinery with hemicellulosepre-extraction for ethanol production: A comparative study. Bioresour. Technol.,101(2):624–631.169ReferencesHuang, H.-J., Ramaswamy, S., Tschirner, U., and Ramarao, B. (2008). A reviewof separation technologies in current and future biorefineries. Separa. PurificationTechnol., 62(1):1–21.Iler, R. K. (1979). The chemistry of silica: solubility, polymerization, colloid andsurface pro perties, and biochemistry. Wiley.Jahan, M. S., Lee, Z. Z., and Jin, Y. (2006). Organic acid pulping of rice straw. i:cooking. Turkish J. Agricul. Forestry, 30(3):231–239.Jin, Y., Jameel, H., Chang, H.-m., and Phillips, R. (2010). Green liquor pretreatmentof mixed hardwood for ethanol production in a repurposed kraft pulp mill. J. WoodChem. Technol., 30(1):86–104.Jun, A., Tschirner, U. W., and Tauer, Z. (2012). Hemicellulose extraction fromaspen chips prior to kraft pulping utilizing kraft white liquor. Biomass Bioenergy,37:229–236.Kačuráková, M., Belton, P. S., Wilson, R. H., Hirsch, J., and Ebringerová, A. (1998).Hydration properties of xylan-type structures: an ftir study of xylooligosaccharides.J. Sci. Food Agricul., 77(1):38–44.Kadla, J., Kubo, S., Venditti, R., Gilbert, R., Compere, A., and Griffith, W. (2002).Lignin-based carbon fibers for composite fiber applications. Carbon, 40(15):2913–2920.Käkölä, J. and Alén, R. (2006). A fast method for determining low-molecular-mass170Referencesaliphatic carboxylic acids by high-performance liquid chromatography–atmosphericpressure chemical ionization mass spectrometry. J. Sep. Sci., 29(13):1996–2003.Kalapathy, U., Proctor, A., and Shultz, J. (2002). An improved method for productionof silica from rice hull ash. Bioresour. Technol., 85(3):285–289.Kim, S. B. and Lee, Y. (2002). Diffusion of sulfuric acid within lignocellulosicbiomass particles and its impact on dilute-acid pretreatment. Bioresour. Technol.,83(2):165–171.Klapiszewski, Ł., Bartczak, P., Wysokowski, M., Jankowska, M., Kabat, K., andJesionowski, T. (2015). Silica conjugated with kraft lignin and its use as a novelâĂŸgreenâĂŹsorbent for hazardous metal ions removal. Chem. Eng. J., 260:684–693.Knauss, K. G. and Wolery, T. J. (1988). The dissolution kinetics of quartz as afunction of ph and time at 70 c. Geochimica et Cosmochimica Acta, 52(1):43–53.Kobayashi, T. and Sakai, Y. (1956). Hydrolysis rate of pentosan of hardwood in dilutesulfuric acid. J. Agricul. Chem. Society of Japan, 20(1):1–7.Kopfmann, K. and Hudeczek, W. (1988). Desilication of black liquor: pilot planttests. Tappi J., 71(10):139–147.Krzesińska, M., Zachariasz, J., and Lachowski, A. (2009). Development of monolithiceco-composites from carbonized blocks of solid iron bamboo (dendrocalamus stric-tus) by impregnation with furfuryl alcohol. Bioresour. Technol., 100(3):1274–1278.171ReferencesKumar, R., Mago, G., Balan, V., and Wyman, C. E. (2009). Physical and chemicalcharacterizations of corn stover and poplar solids resulting from leading pretreat-ment technologies. Bioresour. Technol., 100(17):3948–3962.Larsson, S., Palmqvist, E., Hahn-Hägerdal, B., Tengborg, C., Stenberg, K., Zacchi,G., and Nilvebrant, N.-O. (1999). The generation of fermentation inhibitors duringdilute acid hydrolysis of softwood. Enzyme Microbial Technol., 24(3):151–159.Le, D. M., Sørensen, H. R., Knudsen, N. O., and Meyer, A. S. (2015). Implicationsof silica on biorefineries–interactions with organic material and mineral elements ingrasses. Biofuels, Bioproducts Biorefining, 9(1):109–121.Leenakul, W. and Tippayawong, N. (2010). Dilute acid pretreatment of bamboo forfermentable sugar production. J. Sustain. Energy Environment, 1:117–120.Lehto, J. and Alén, R. (2013). Alkaline Pre-treatment of Hardwood Chips Prior toDelignification. J. Wood Chem. Technol., 33(2):77–91.Lehto, J. and Alén, R. (2015). Alkaline pre-treatment of softwood chips prior todelignification. J. Wood Chem. Technol., 35(2):146–155.Li, X. (2004). Physical, chemical, and mechanical properties of bamboo and its utiliza-tion potential for fiberboard manufacturing. PhD thesis, Beijing Forestry University.Li, X., Shupe, T., Peter, G., Hse, C., and Eberhardt, T. (2007). Chemical changeswith maturation of the bamboo species phyllostachys pubescens. J. Tropical ForestSci., pages 6–12.172ReferencesLi, Z., Jiang, Z., Fei, B., Cai, Z., and Pan, X. (2014). Comparison of bamboo green,timber and yellow in sulfite, sulfuric acid and sodium hydroxide pretreatments forenzymatic saccharification. Bioresour. Technol., 151:91–99.Li, Z., Jiang, Z., Fei, B., Pan, X., Cai, Z., Yu, Y., et al. (2012). Ethanol organo-solv pretreatment of bamboo for efficient enzymatic saccharification. BioResources,7(3):3452–3462.Lier, J. v., Bruyn, P. d., and Overbeek, J. T. G. (1960). The solubility of quartz. J.Physical Chem., 64(11):1675–1682.Liese, W. (1987). Research on bamboo. Wood Sci. Technol., 21:189–209.Liese, W. (1992). The structure of bamboo in relation to its properties and utilization.In Zhu, S., Li, W., Zhang, X. Wang, Z. ed., Bamboo and its use. Proceedings ofthe International symposium on Industrial Use of Bamboo, Beijing, China, pages7–11.Liou, T.-H. (2004). Preparation and characterization of nano-structured silica fromrice husk. Materials Sci. Eng.: A, 364(1):313–323.Liou, T.-H. and Yang, C.-C. (2011). Synthesis and surface characteristics of nanosilicaproduced from alkali-extracted rice husk ash. Materials Sci. Eng.: B, 176(7):521–529.Littlewood, J., Wang, L., Turnbull, C., and Murphy, R. J. (2013). Techno-economicpotential of bioethanol from bamboo in china. Biotechnol. Biofuels, 6(1):1.173ReferencesLiu, H., Hu, H., Jahan, M. S., and Ni, Y. (2013). Furfural formation from the pre-hydrolysis liquor of a hardwood kraft-based dissolving pulp production process.Bioresour. Technol., 131:315–320.Liu, Q., Wang, S., Zheng, Y., Luo, Z., and Cen, K. (2008). Mechanism study of woodlignin pyrolysis by using tg–ftir analysis. J. Analy. App. Pyrolysis, 82(1):170–177.Liu, X., Lu, M., Ai, N., Yu, F., and Ji, J. (2012). Kinetic model analysis of dilutesulfuric acid-catalyzed hemicellulose hydrolysis in sweet sorghum bagasse for xyloseproduction. Ind. Crops Prod., 38:81–86.Liu, Z., Ni, Y., Fatehi, P., and Saeed, A. (2011). Isolation and cationization ofhemicelluloses from pre-hydrolysis liquor of kraft-based dissolving pulp productionprocess. Biomass Bioenergy, 35(5):1789–1796.Lloyd, T. A. and Wyman, C. E. (2004). Predicted effects of mineral neutralizationand bisulfate formation on hydrogen ion concentration for dilute sulfuric acid pre-treatment. In Proceedings of the Twenty-Fifth Symposium on Biotechnology forFuels and Chemicals Held May 4–7, 2003, in Breckenridge, CO, pages 1013–1022.Springer.Lloyd, T. A. and Wyman, C. E. (2005). Combined sugar yields for dilute sulfuricacid pretreatment of corn stover followed by enzymatic hydrolysis of the remainingsolids. Bioresour. technol., 96(18):1967–1977.Luo, X., Ma, X., Hu, H., Li, C., Cao, S., Huang, L., and Chen, L. (2013). Kinetic studyof pentosan solubility during heating and reacting processes of steam treatment ofgreen bamboo. Bioresour. Technol., 130:769–776.174ReferencesLux, A., Luxová, M., Abe, J., Morita, S., and Inanaga, S. (2003). Silicification ofbamboo (Phyllostachys heterocycla Mitf.) root and leaf. Springer.Ma, J. F. and Yamaji, N. (2006). Silicon uptake and accumulation in higher plants.Trends Plant Sci., 11(8):392–397.Ma, X., Huang, L., Chen, Y., and Chen, L. (2011). Preparation of bamboo dissolvingpulp for textile production; part 1. study on prehydrolysis of green bamboo forproducing dissolving pulp. BioResources, 6(2):1428–1439.Machmud, M. N., Fadi, F., Fuadi, Z., and Kokarkin, C. (2013). Alternative fibersources from gracilaria sp and eucheuma cottonii for papermaking. Inter. J. Sci.Eng., 6(1):1–10.Malester, I. A., Green, M., and Shelef, G. (1992). Kinetics of dilute acid hydrolysis ofcellulose originating from municipal solid wastes. Ind. Eng. Chem. Res., 31(8):1998–2003.Maloney, M. T., Chapman, T. W., and Baker, A. J. (1985). Dilute acid hydrolysisof paper birch: Kinetics studies of xylan and acetyl-group hydrolysis. Biotechnol.Bioeng., 27(3):355–361.Mân Vu, T. H., Pakkanen, H., and Alén, R. (2004). Delignification of bamboo (Bam-busa procera acher). Ind. Crops Prod., 19(1):49–57.Mao, H., Genco, J., Yoon, S.-H., van Heiningen, A., and Pendse, H. (2008). Technicaleconomic evaluation of a hardwood biorefinery using the \"near-neutral\" hemicellu-lose pre-extraction process. J. Biobased Materials Bioenergy, 2(2):177–185.175ReferencesMgaidi, A., Jendoubi, F., Oulahna, D., El Maaoui, M., and Dodds, J. (2004). Kineticsof the dissolution of sand into alkaline solutions: application of a modified shrinkingcore model. Hydrometallurgy, 71(3):435–446.Mikkonen, K. S. and Tenkanen, M. (2012). Sustainable food-packaging materialsbased on future biorefinery products: Xylans and mannans. Trends Food Sci.Technol., 28(2):90–102.Minu, K., Jiby, K. K., and Kishore, V. V. N. (2012). Isolation and purification oflignin and silica from the black liquor generated during the production of bioethanolfrom rice straw. Biomass Bioenergy, 39:210–217.Mittal, A., Chatterjee, S. G., Scott, G. M., and Amidon, T. E. (2009). Modeling xylansolubilization during autohydrolysis of sugar maple wood meal: Reaction kinetics.Holzforschung, 63(3):307–314.Molin, U. and Teder, A. (2002). Importance of cellulose/hemicellulose-ratio for pulpstrength. Nordic Pulp Paper Res. J., 17(1):14–19.Morinelly, J. E., Jensen, J. R., Browne, M., Co, T. B., and Shonnard, D. R. (2009).Kinetic characterization of xylose monomer and oligomer concentrations duringdilute acid pretreatment of lignocellulosic biomass from forests and switchgrass.Ind. Eng. Chem. Res., 48(22):9877–9884.Morpurgo, M., Teoli, D., Pignatto, M., Attrezzi, M., Spadaro, F., and Realdon, N.(2010). The effect of na2co3, naf and nh4oh on the stability and release behav-ior of sol–gel derived silica xerogels embedded with bioactive compounds. ActaBiomaterialia, 6(6):2246–2253.176ReferencesMosier, N., Wyman, C., Dale, B., Elander, R., Lee, Y., Holtzapple, M., and Ladisch,M. (2005). Features of promising technologies for pretreatment of lignocellulosicbiomass. Bioresour. Technol., 96(6):673–686.Motomura, H., Fujii, T., and Suzuki, M. (2000). Distribution of silicified cells inthe leaf blades of pleioblastus chino (franchet et savatier) makino (bambusoideae).Annals Botany, 85(6):751–757.Motomura, H., Fujii, T., and Suzuki, M. (2006). Silica deposition in abaxial epidermisbefore the opening of leaf blades of pleioblastus chino (poaceae, bambusoideae).Annals Botany, 97(4):513–519.Motomura, H., Mita, N., and Suzuki, M. (2002). Silica accumulation in long-livedleaves of sasa veitchii (carrière) rehder (poaceae–bambusoideae). Annals Botany,90(1):149–152.Negro, C., Blanco, M., Lopez-Mateos, F., DeJong, A., LaCalle, G., Van Erkel, J., andSchmal, D. (2001). Free acids and chemicals recovery from stainless steel picklingbaths. Separa. Sci. Technol., 36(7):1543–1556.Neupane, S., Adhikari, S., Wang, Z., Ragauskas, A., and Pu, Y. (2015). Effect oftorrefaction on biomass structure and hydrocarbon production from fast pyrolysis.Green Chem., 17(4):2406–2417.Niibori, Y., Kunita, M., Tochiyama, O., and Chida, T. (2000). Dissolution Rates ofAmorphous Silica in Highly Alkaline Solution. J. Nuclear Sci. Technol., 37(4):349–357.177ReferencesÖhgren, K., Bura, R., Saddler, J., and Zacchi, G. (2007). Effect of hemicellulose andlignin removal on enzymatic hydrolysis of steam pretreated corn stover. BioresouR.Technol., 98(13):2503–2510.Okubo, K., Fujii, T., and Yamamoto, Y. (2004). Development of bamboo-basedpolymer composites and their mechanical properties. Composites Part A: App.Sci. Manufacturing, 35(3):377–383.Okunev, A., Shaurman, S., Danilyuk, A., Aristov, Y. I., Bergeret, G., and Renouprez,A. (1999). Kinetics of the sio 2 aerogel dissolution in aqueous naoh solutions:experiment and model. J. Non-crystalline Solids, 260(1):21–30.Pan, X.-J., Sano, Y., and Ito, T. (1999). Atmospheric acetic acid pulping of rice strawii: behavior of ash and silica in rice straw during atmospheric acetic acid pulpingand bleaching. Holzforschung, 53(1):49–55.Pauly, M., Albersheim, P., Darvill, A., and York, W. S. (1999). Molecular domainsof the cellulose/xyloglucan network in the cell walls of higher plants. Plant J.,20(6):629–639.Pavlostathis, S. G. and Gossett, J. M. (1985). Alkaline treatment of wheat straw forincreasing anaerobic biodegradability. Biotechnol. Bioeng., 27(3):334–344.Pekarovic, J., Pekarovicova, A., Joyce, T. W., et al. (2005). Desilication of agriculturalresidues-the first step prior to pulping. Appita J., 58(2):130–134.Peng, F., Ren, J.-L., Xu, F., Bian, J., Peng, P., and Sun, R.-C. (2009). Fractional178Referencesstudy of alkali-soluble hemicelluloses obtained by graded ethanol precipitation fromsugar cane bagasse. J. Agricul. Food Chem., 58(3):1768–1776.Perrin, D. D., Dempsey, B., and Serjeant, E. P. (1981). pKa prediction for organicacids and bases, volume 1. Springer.Pinto, P., Evtuguin, D., and Neto, C. P. (2005). Structure of hardwood glucuronoxy-lans: modifications and impact on pulp retention during wood kraft pulping. Car-bohydrate Poly., 60(4):489–497.Piperno, D. R. (2014). Phytolyth analysis: an archaeological and geological perspective.Elsevier.Poojary, H. and Mugeraya, G. (2012). Laccase production by phellinus noxius hp f17:Optimization of submerged culture conditions by response surface methodology.Res. Biotechnol., 3(1):9–20.Prychid, C. J., Rudall, P. J., and Gregory, M. (2003). Systematics and biology ofsilica bodies in monocotyledons. Botanical Review, 69(4):377–440.Pu, Y., Treasure, T., Gonzalez, R., Venditti, R. A., and Jameel, H. (2013). Auto-hydrolysis pretreatment of mixed softwood to produce value prior to combustion.BioEnergy Res., 6(3):1094–1103.Pye, E. K. (2008). Industrial lignin production and applications. Biorefineries-Industrial processes and products: Status quo and future directions, pages 165–200.Raven, J. A. (1983). The transport and function of silicon in plants. BiologicalReviews, 58(2):179–207.179ReferencesRen, J., Peng, F., and Sun, R. (2009). The effect of hemicellulosic derivatives on thestrength properties of old corrugated container pulp fibres. J. Biobased MaterialsBioenergy, 3(1):62–68.Richmond, K. E. and Sussman, M. (2003). Got silicon? the non-essential beneficialplant nutrient. Current Opinion Plant Biology, 6(3):268–272.Rissanen, J. V., Grénman, H., WillfoÌĹr, S., Murzin, D. Y., and Salmi, T. (2014a).Spruce hemicellulose for chemicals using aqueous extraction: kinetics, mass trans-fer, and modeling. Ind. Eng. Chem. Res., 53(15):6341–6350.Rissanen, J. V., Grénman, H., Xu, C., Willför, S., Murzin, D. Y., and Salmi, T.(2014b). Obtaining spruce hemicelluloses of desired molar mass by using pressurizedhot water extraction. ChemSusChem, 7(10):2947–2953.Rocha, G. J., Martín, C., da Silva, V. F., Gómez, E. O., and Gonçalves, A. R. (2012).Mass balance of pilot-scale pretreatment of sugarcane bagasse by steam explosionfollowed by alkaline delignification. Bioresour. Technol., 111:447–452.Rogalinski, T., Ingram, T., and Brunner, G. (2008). Hydrolysis of lignocellulosicbiomass in water under elevated temperatures and pressures. J. Supercritical Fluids,47(1):54–63.Rousset, P., Aguiar, C., Labbé, N., and Commandré, J.-M. (2011). Enhanc-ing the combustible properties of bamboo by torrefaction. Bioresour. Technol.,102(17):8225–8231.180ReferencesRydholm, S. A. (1965). Pulping processes. Interscience publishers, Wiley, New York,USA, pages pp. 676–687.Saake, B. and Lehnen, R. (2007). Lignin. in: Ullmann’s Encyclopedia of IndustrialChemistry.Saeman, J. F. (1945). Kinetics of wood saccharification-hydrolysis of cellulose anddecomposition of sugars in dilute acid at high temperature. Ind. Eng. Chem. Res.,37(1):43–52.Saka, S. (2004). The raw materials of ca 2.1 wood as natural raw materials for celluloseacetate production. In Macromolecular Symposia, volume 208, pages 7–28. WileyOnline Library.Salmela, M., Alén, R., and Vu, M. T. H. (2008). Description of kraft cooking andoxygenâĂŞalkali delignification of bamboo by pulp and dissolving material analysis.Ind. Crops Prod., 28(1):47–55.Sanchez, O. J. and Cardona, C. A. (2008). Trends in biotechnological production offuel ethanol from different feedstocks. Bioresour. Technol., 99(13):5270–5295.Sangster, A. and Parry, D. (1981). Ultrastructure of silica deposits in higher plants.Springer.Sathitsuksanoh, N., Zhu, Z., Ho, T.-J., Bai, M.-D., and Zhang, Y.-H. P. (2010).Bamboo saccharification through cellulose solvent-based biomass pretreatment fol-lowed by enzymatic hydrolysis at ultra-low cellulase loadings. Bioresour. Technol.,101(13):4926–4929.181ReferencesScheller, H. V. and Ulvskov, P. (2010). Hemicelluloses. Plant Biology, 61(1):263.Schild, G., Sixta, H., and Testova, L. (2010). Multifunctional alkaline pulping. delig-nification and hemicellulose extraction. Cellulose Chem. Technol., 44(1):35–45.Scurlock, J., Dayton, D., and Hames, B. (2000). Bamboo: an overlooked biomassresource? Biomass Bioenergy, 19(2000):229–244.Sella Kapu, N. and Trajano, H. L. (2014). Review of hemicellulose hydrolysis insoftwoods and bamboo. Biofuels, Bioprod. Biorefining, 8(6):857–870.Shen, J. and Wyman, C. E. (2011). A novel mechanism and kinetic model to ex-plain enhanced xylose yields from dilute sulfuric acid compared to hydrothermalpretreatment of corn stover. Bioresour. Technol., 102(19):9111–9120.Shi, H., Fatehi, P., Xiao, H., and Ni, Y. (2011). A combined acidification/PEOflocculation process to improve the lignin removal from the pre-hydrolysis liquor ofkraft-based dissolving pulp production process. BioresouR. Technol., 102(8):5177–5182.Sierka, M. and Sauer, J. (1997). Structure and reactivity of silica and zeolite catalystsby a combined quantum mechanics [ndash] shell-model potential approach basedon dft. Faraday Discussions, 106:41–62.Sixta, H. (2006). Handbook of pulp, vol. 1, wiley-vch, weinheim.Sluiter, A., Hames, B., Ruiz, R., Scarlata, C., Sluiter, J., Templeton, D., and Crocker,D. (2012). Determination of structural carbohydrates and lignin in biomass. Lab-oratory Analytical Procedure, 1617.182ReferencesSong, Z., Liu, H., Li, B., and Yang, X. (2013). The production of phytolith-occludedcarbon in china’s forests: implications to biogeochemical carbon sequestration.Global Change Biology, 19(9):2907–2915.Spencer, R. R. and Akin, D. E. (1980). Rumen microbial degradation of potas-sium hydroxide-treated coastal bermudagrass leaf blades examined by electron mi-croscopy. J. Animal Sci., 51(5):1189–1196.SriBala, G. and Vinu, R. (2014). Unified kinetic model for cellulose deconstructionvia acid hydrolysis. Ind. Eng. Chem. Res., 53(21):8714–8725.Stading, M., Hermansson, A.-M., Gatenholm, P., et al. (1998). Structure, mechan-ical and barrier properties of amylose and amylopectin films. Carbohydr. Poly.,36(2):217–224.Sterling, C. (1967). Crystalline silica in plants. American J. Botany, pages 840–844.Sun, R., Lawther, J. M., and Banks, W. (1995). Influence of alkaline pre-treatmentson the cell wall components of wheat straw. Ind. Crops Prod., 4(2):127–145.Talvitie, N. (1951). Determination of quartz in presence of silicates using phosphoricacid. Analy. Chem., 23(4):623–626.Thornton, S., Radke, C., et al. (1988). Dissolution and condensation kinetics of silicain alkaline solution. SPE Reservoir Eng., 3(02):743–752.Tillman, L., Lee, Y., and Torget, R. (1990). Effect of transient acid diffusion onpretreatment/hydrolysis of hardwood hemicellulose. App. Biochem. Biotechnol.,24(1):103–113.183ReferencesTong, G.-l., Lu, Q., Wang, Y., and Zhou, C. (2005). Spectrophotometeric determi-nation of silicon in rice straw and black liquor after rice straw alkaline pulping bysilicomolybdate blue. China Pulp Paper Ind., 26:64–66.Torelli, N. and Čufar, K. (1995). Mexican tropical hardwoods. comparative study ofash and silica content. Holz als Roh-und werkstoff, 53(1):61–62.Torssell, K. B. (1997). Natural product chemistry: a mechanistic, biosynthetic andecological approach. 2, Apotekarsocieteten, Stockholm.Trajano, H. L. and Wyman, C. E. (2013). Fundamentals of biomass pretreatment atlow pH. Wiley Online Library, NJ, USA.Tsuji, H., Isono, Z., and Ono, K. (1965). Studies on dissolving bamboo pulp. Univ.Osaka Prefecture, Ser. B, Agric. Biology, 16:89–104.van Heiningen, A. (2006). Converting a kraft pulp mill into an integrated forestbiorefinery. Pulp Paper Can., 107(6):38–43.Vena, P. F., Brienzo, M., del Prado García-Aparicio, M., Görgens, J. F., and Ryp-stra, T. (2013). Hemicelluloses extraction from giant bamboo (bambusa balcooaroxburgh) prior to kraft or soda-aq pulping and its effect on pulp physical proper-ties. Holzforschung, 67(8):863–870.Visuri, J. A., Song, T., Kuitunen, S., and Alopaeus, V. (2012). Model for degradationof galactoglucomannan in hot water extraction conditions. Ind. Eng. Chem. Res.,51(31):10338–10344.184ReferencesWalton, S. L., Hutto, D., Genco, J. M., Van Walsum, G. P., and van Heiningen,A. R. P. (2010). Pre-extraction of hemicelluloses from hardwood chips using analkaline wood pulping solution followed by kraft pulping of the extracted woodchips. Ind. Eng. Chem. Res., 49:12638–12645.Wang, H., Loganathan, D., and Linhardt, R. J. (1991). Determination of the pkaof glucuronic acid and the carboxy groups of heparin by 13c-nuclear-magnetic-resonance spectroscopy. Biochemical J., 278(3):689–695.Wen, J.-L., Xiao, L.-P., Sun, Y.-C., Sun, S.-N., Xu, F., Sun, R.-C., and Zhang, X.-L.(2011). Comparative study of alkali-soluble hemicelluloses isolated from bamboo(bambusa rigida). Carbohydrate Res., 346(1):111–120.Wirth, G. and Gieskes, J. (1979). The initial kinetics of the dissolution of vitreoussilica in aqueous media. J. Colloid Interface Sci., 68(3):492–500.Wyman, C. E., Decker, S. R., Himmel, M. E., Brady, J. W., Skopec, C. E., and Viikari,L. (2005). Hydrolysis of cellulose and hemicellulose. Polysaccharides: StructuralDiversity Functional Versatility, 1:1023–1062.Xu, F., Sun, J., Geng, Z., Liu, C., Ren, J., Sun, R., Fowler, P., and Baird, M.(2007). Comparative study of water-soluble and alkali-soluble hemicelluloses fromperennial ryegrass leaves (lolium peree). Carbohydrate Poly., 67(1):56–65.Xu, Z. and Huang, F. (2014). Pretreatment methods for bioethanol production. App.Biochem. Biotechnol, 174(1):43–62.Yamashita, Y., Shono, M., Sasaki, C., and Nakamura, Y. (2010). Alkaline perox-185Referenceside pretreatment for efficient enzymatic saccharification of bamboo. CarbohydratePoly., 79(4):914–920.Yan, L., Zhang, L., and Yang, B. (2014). Enhancement of total sugar and ligninyields through dissolution of poplar wood by hot water and dilute acid flowthroughpretreatment. Biotechnol. Biofuels, 7(1):1.Yoon, S.-H. and van Heiningen, A. (2008). Kraft pulping and papermaking propertiesof hot-water pre-extracted loblolly pine in an integrated forest products biorefinery.Tappi J, 7(7):22–27.Yoon, S.-H. and van Heiningen, A. (2010). Green liquor extraction of hemicellulosesfrom southern pine in an integrated forest biorefinery. J. Ind. Eng. Chem., 16(1):74–80.Zaitsev, A., Shelkova, N., and Mogutnov, B. (2000). Thermodynamics of na2o-sio2melts. Inorganic Materials, 36(6):529–543.Zhang, X., Zhao, Z., Ran, G., Liu, Y., Liu, S., Zhou, B., and Wang, Z. (2013).Synthesis of lignin-modified silica nanoparticles from black liquor of rice strawpulping. Powder Technol., 246:664–668.Zhang, Y.-H. P. and Lynd, L. R. (2004). Toward an aggregated understanding ofenzymatic hydrolysis of cellulose: noncomplexed cellulase systems. Biotechnol.Bioeng., 88(7):797–824.Zhao, X., Zhang, L., and Liu, D. (2012). Biomass recalcitrance. part i: the chemical186Referencescompositions and physical structures affecting the enzymatic hydrolysis of ligno-cellulose. Biofuels, Bioprod. Biorefining, 6(4):465–482.Zhao, Y., Wang, Y., Zhu, J., Ragauskas, A., and Deng, Y. (2008). Enhanced enzy-matic hydrolysis of spruce by alkaline pretreatment at low temperature. Biotechnol.Bioeng., 99(6):1320–1328.187AppendicesA1 Bamboo Raw Material CharacterizationEven though bamboo belongs to the grass family, having a woody stem it is struc-turally quite different from other lignocellulosic materials such as softwoods, hard-woods and cereal straws. Bamboo stem is composed of three parts: epidermal (out-most cell layer of the stem), mid-cortex (the region between epidermal and innercortex), and inner cortex (the portion encircling the hollow center of the culm). Dueto the unique location in the bamboo stem and their functions, these three regions dif-fer substantially in chemical composition including extractives, ash and silica content(Chand et al. (2006)). These differences will affect the bamboo feedstock processingmethods in kraft-based pulping and biorefinery applications. It is thought that it maybe beneficial to remove one or two regions of the bamboo stem (bamboo epidermalor inner cortex) before pulping. However, the removal of bamboo parts as waste willreduce the useful biomass fraction and cause environmental problems (Li (2004)).Accordingly, before optimizing the pulping processes, the chemical composition andsilica mass distribution in the bamboo stem needs to be quantified.Fresh bamboo trees (1-5 years old) of Neosinocalamus affinis Keng were collectedin September, 2012 from a natural forest in Sichuan Province, China. Each bamboo188A1. Bamboo Raw Material Characterizationstem was evenly cut into three parts along the length of the bamboo stem: top-middle-bottom. All samples were washed thoroughly with deionized water to removethe dust and other impurities such as sand and soil from the surface of bamboo chips.This washing operation was performed 10 times and then samples were dried for 48h at room temperature. Subsequently, the samples were cut into small strips with arazor blade and ground using a Wiley Mill. The sample powder that passed througha 40-mesh sieve yet retained on a 60-mesh sieve was collected in glass jars for furtheruse.For the analysis of chemical composition in different layers along the radial di-rection, the stems were separated mannually with a plane block (Lee Valley ToolsLtd. Canada) to three layers: epidermal layer (≈ 0.1 mm), mid-cortex (≈ 3 mm)and inner cortex (≈ 0.2 mm). The three fractions of bamboo samples were ground,screened and stored as previously described.The organic components (carbohydrates and lignin) and inorganics (ash) compo-sition was measured according to the analytical methods (NREL standard protocol)described in Chapter 2.Fig. A.1 shows the ash and silica contents in different parts of the bamboo stemof different ages. Results showed that the ash content of one year old bamboo wassignificantly higher than that of three and five year old bamboo. It has been suggestedin the literature that the higher ash content observed in younger bamboo could bea result of the larger mass fraction of epidermal tissue present in younger plants (Liet al. (2007)). Analysis of ash along the length of the bamboo stem showed thatthe ash content was the highest at the top part, an observation also reported by Li(2004) and Rousset et al. (2011). The results indicate that a substantial portion of189A1. Bamboo Raw Material Characterizationthe ash in bamboo consists of silica. As shown in Fig. A.1b, the silica content inthe five year old bamboo stem was higher than that of one and three year old, andthe effect of age was more pronounced in the top regions of the stem. This is inagreement with previous studies that reported increasing silica content with bambooaging (Motomura et al. (2002)). Moreover, silica level in the bamboo stem decreasedgradually from the apical to basal portions of the stem (Fig. A.1b). This observationis similar to the findings by Collin et al. (2012). Analysis of silica distribution in thebamboo stem illustrated that the silica content of bamboo increases with increasingage. With consideration of biomass generated by each unit plantation area, bambootrees with age > 3 years old should be harvested for pulp and biorefinery applications.To study whether some parts of the bamboo stem should be removed before pro-cessing, take five year old bamboo as an example, the composition of different fractionsalong the length (top, middle, bottom) of the bamboo stem and layers (epidermal,mid-cortex, inner cortex) across the cross section of the original bamboo stem was de-termined, shown in Table A1. The composition of each component analyzed (glucan,xylan, galactan, arabinan, lignin, extractives, ash, silica), is expressed as the averagemass percentage of this component in the oven dry solids, determined at least threetests. With regards to the composition along the length of the bamboo stem, thecellulose and lignin content of the three fractions along the length of the stem (top,middle, bottom) were similar at around 46.5% and 23.5%, respectively, while thehemicellulose (mainly xylan) content of the bamboo at the top of the stem (24.8%)was the highest, followed by that in the middle (23.4%). Hemicellulose content wasthe lowest at the bottom of the stem (21.4%). Arabinan and galactan content in all190A1. Bamboo Raw Material Characterization0 1 2 3 4 5 60.00.51.01.52.02.53.0 Ash content (%)Bamboo age (years) Top Middle Bottoma0 1 2 3 4 5 60.00.20.40.60.81.01.2 Silica content (%)Bamboo age (years) Top Middle BottombFigure A.1: Ash and silica content in different height locations of the bamboo stem (a:ash content, b: silica content). The measurements were triplicated.191A1. Bamboo Raw Material Characterizationthree parts were less than 1% while mannan was undetectable by the HPLC method-ology used. The main difference in chemical composition among the three parts alongthe stem was found in the silica content. At the top of the stem the silica content was0.91% which decreased by almost 50% to a value of 0.47% at the bottom of the stem.This is in agreement with previous studies that reported decreasing silica levels fromthe apical to basal portions of the stem (Collin et al. (2012)). The most importantobservation was that the ratio of silica mass in the three parts was about 1:1:1 dueto the difference in the weight fraction of bamboo biomass in each region.The chemical composition and silica mass distribution of different layers (epi-dermal, mid-cortex, inner cortex) along the radial direction of the stem are shown inTable A1. It was found that the glucan content of the bamboo mid-cortex (49.4%) wasmuch higher than that of the epidermal layer (44.3%) and the inner cortex (43.3%).Lignin content was found to be highest in the bamboo epidermal portion (28.5%), fol-lowed by the inner cortex (24.1%) and mid-cortex (23.2%). The hemicellulose (xylan,arabinan and galactan) content in the epidermal region, mid-cortex, and inner cortexwere comparable at 22.3%, 23.5% and 24.4%, respectively. The highest extractivescontent was in the bamboo inner cortex (8.4%), followed by the mid-cortex (5.2%)and the epidermal layer (4.8%). The ash and silica contents of the bamboo epidermalpart were higher than those of the bamboo mid-cortex and inner cortex. The differ-ence in silica content (1.3%) was especially high with that in the epidermal regionbeing about seven-times higher than in the mid-cortex or the bamboo inner cortex.However, the majority of silica mass in the bamboo was located in bamboo mid-cortex(63%) because this region accounts for about 88% of the biomass while the epidermalportion and the inner cortex account for only 6% each. According to the composition192A1. Bamboo Raw Material Characterizationanalysis results, the removal of bamboo epidermal layer or inner cortex does not resultin a significant decrease of silica amount input into the pulping processes; in contrast,it would increase the capital cost and cause environmental problems by the increasingindustrial wastes. Moreover, compared with bamboo mid-cortex, bamboo epidermalpart and inner cortex have comparable concentrations of hemicellulose, which canbe pre-extracted for bioconversion. Consequently, rather than removing biomass ofepidermal region and inner cortex the whole bamboo stem should be used as the rawmaterial in pulping or biorefinery processes.193A1. Bamboo Raw Material CharacterizationTable A.1: Chemical composition of 5 year old original bamboo stem (Neosinocalamus Affinis Keng).Conponent Length direction Cross sectionTop Middle Bottom Epidermal Mid-cortex Inner cortexGlucan (%) 46.57±1.52 46.71±1.85 46.89±1.94 44.32±1.56 49.44±2.12 43.35±1.85Xylan(%) 24.82±1.02 23.43±0.87 21.41±0.98 21.53±1.12 22.61±1.26 23.37±1.55Galactan (%) 0.58±0.12 0.44±0.25 0.35±0.14 0.18±0.04 0.20±0.04 0.20±0.02Arabinan (%) 0.63±0.05 0.52±0.11 0.41±0.13 0.62±0.13 0.77±0.14 0.91±0.14Lignin (%) 23.14±0.21 23.75±0.32 23.56±0.20 28.52±0.22 23.21±0.31 24.08±0.26Extractives (%) 5.64±0.72 4.12±0.64 3.83±0.81 4.78±0.44 5.24±0.36 8.41±0.29Ash (%) 1.35±0.10 1.31±0.12 1.28±0.13 2.12±0.08 0.93±0.04 1.26±0.04Silica (%) 0.91±0.05 0.60±0.02 0.47±0.04 1.32±0.03 0.17±0.02 0.20±0.05Silica massfraction 0.36 0.34 0.30 0.32 0.63 0.05194A2. Chip WashingA2 Chip WashingAccording to the results of bamboo stem characterization, bamboo chips were pre-pared from 3 to 7 year old bamboo trees. Due to the commonly practiced methods ofharvest and open-pile chip storage, a substantial amount of soil can be detected onthe surface and structural pores of bamboo chips. As soil comprises 50-70% of silicaby mass, the removal of soil is necessary to minimize silica input into the bambooprocessing system. Indeed, the impact of chip cleanliness is underscored by the ob-servation that fresh bamboo chips prepared in the laboratory contained 1.1% (w/w)silica while chips provided by a functional mill measured 1.56% (w/w). Therefore,we examined the feasibility of chip screening and washing with water to remove soiland silica. In the past, several researchers have used screening and washing to re-move the soil or other impurities on the surface of bamboo and other non-wood rawmaterials (De Lopez et al. (1996)). However, the authors did not describe the wash-ing process in detail, especially the important process parameters and their impacton the efficiency of washing. Moreover, to our knowledge, pilot-scale studies on theimpact of bamboo washing have not been documented. In the present study, twotypes of washing methods, continuous washing and batch washing, were investigated(Fig. A.2). The chips used originated from approximately three-year old bamboo andwere provided by a mill operated by Lee & Man Paper Manufacturing Ltd., China.Deionized water and recycled white water from pulp line were used in the lab-scaleand pilot-scale studies, respectively.Fig. A.2a shows the effects of washing time and temperature of lab-scale contin-uous washing on the residual silica content of bamboo chips. At both 17 oC and 35195A2. Chip WashingoC the residual silica content of chips decreased rapidly in the first 10 min and thenleveled off. The rate of silica removal increased with increasing temperature; a 10minwash at 17 oC and 35 oC reduced the silica content of the chips by about 25% and28%, respectively. Silica is reported to be soluble in water, albeit to a minute extent,even at 25 oC. Although the solubility is low, the solubility is reported to increasewith increasing temperature (Chen and Marshall (1982)). Higher temperature can beexpected to increase the kinetic energy of silica molecules. These factors may play arole in the observed increase in silica removal at 35 oC compared to 17 oC. Prolongedwashing beyond 10 min (up to 180 min) did not result in any further silica removal.It suggests that the fraction of silica loosely entrapped in the material is removed bydeionized water washing for a short time. Moreover, during the continuous washingprocess, the silica content fluctuated, suggesting that silica and silica contained soilin the wash water can be regained by the chips. Bamboo stem has cells with largelumens and it is possible that a portion of the removed silica in the wash water isre-trapped in these pores.To test whether more silica could be removed by changing wash water at intervals,a series of experiments were carried out by replacing the wash water with fresh waterat 10 min intervals. As shown in Fig. A.2b, the residual silica content decreasedrapidly in the first 10 min and only very little silica was removed thereafter withprolonged washing. The degree of silica removal in 10 min was almost identical tocontinuous washing without changes of wash water, about 25% and 28% at 17 oCand 35 oC, respectively. This indicates that it is difficult to remove more than 25-28% silica just by washing with water, even with extended washing and wash waterchanges. Fig. A.2b also shows that the silica regaining process did not occur when196A2. Chip Washing0 30 60 90 120 150 1800.60.81.01.21.41.6 17 C 35 C Silica content of bamboo chips (%)Washing time (min)a0 10 20 30 400.60.81.01.21.41.6 17 C 35 C Silica content of bamboo chips (%)Washing time (min)bFigure A.2: Silica removal from bamboo chips by washing with water at two temperatures(a: continuous washing process, b: batch washing process). Note: batch process meanswash water was replaced with fresh deionized water at 10 min intervals. 197A2. Chip Washing0 2 4 6 8 100.60.81.01.21.41.61.561.21.050.931.281.21.12 Pilot-scale experiment Laboratory-scaleSilica content of bamboo chips (%)Washing time (min)Figure A.3: Pilot-scale experiments on silica removal by washing with water.wash water was changed every 10 min. This observation supports the idea that ifthe removed silica is still in the washing system, it might be regained by chips. Thisemphasizes the practical implications of wash water recycling in mill operations.Based on the results obtained in the laboratory process, the washing technologywas tested for silica removal from bamboo at pilot-scale during kraft pulping by theLee & Man Paper Manufacturing, Ltd., China. In this mill, white water at about35 oC, generated from a fourdrinier kraft bamboo pulp board machine, was used towash the raw chips. Following the laboratory protocol, the mill changed the usualpractice of a 2 min wash to a 10 min wash (Fig. A.3). The costs incurred by theprolonged washing time were negligible. Results of Fig. A.3 show that after a 10 min198A2. Chip Washingwash the residual silica content of pilot-scale washed chips was about 0.93% (basedon the dry weight of bamboo chips), which is lower than the value obtained in the lab(1.12%) (w/w). A likely reason for the higher effectiveness of the pilot-scale washingis the use of a screw press for dewatering which could make the chips abrade againsteach other and squeeze out liquid in chips. Therefore, it is possible that more silica-rich epidermal layers of the bamboo stem as well as some silica entrapped in the celllumens were removed. Results of this set of experiments demonstrated that a 10 minwash at 35 oC was an economical and effective method for silica removal from bamboochips.199A3. Mass Balance of the Pulping and Chemical Recovery ProcessA3 Mass Balance of the Pulping and ChemicalRecovery ProcessA3.1 Mass Balance of the Fibre LineThe calculation is based on 1 tonne of original bamboo chips, with the moisturecontent of 47%. Before carrying out the calculation, several definitions are made.Q : Dry fibre content of the materials (kg)D : Total weight of the material (kg)C : Solids consistency of the material (%)R : Reject ratio of each step (%)S : Solution or waterK : Yield (based on the previous step)Chip preparation process1. Known:• D1 = 1000 kg• C1 = 1− 47% = 53%• R3 = 3.5%200A3. Mass Balance of the Pulping and Chemical Recovery Process2. Calculation:C1 = C2 = C3 = 53% (5.1)Q1 = D1 × C1 = 1000× 53% = 530 kg (5.2)Q2 = Q1 × (1−R3) = 530× (1− 3.5%) = 511.45 kg (5.3)D2 = Q2/C2 = 511.45/53/Q3 = Q1 −Q2 = 530− 511.45 = 18.55 kg (5.4)D3 = Q3/C3 = 18.55/53% = 35 kg (5.5)Chip washing system1. Known:• Q2 = 511.45 kg• D2 = 965 kg• C2 = 53%• C4 = 1− 48% = 52% (moisture content of washed chips is 48%)• S5 = 2560 L (liquid-to-solid ratio=5 L/kg, based on initial dry chips)201A3. Mass Balance of the Pulping and Chemical Recovery Process• R6 = 1.5%2. Calculation:Q6 = Q2 ×R6 = 511.45× 1.5% = 7.67 kg (5.6)Q4 = Q2 −Q6 = 511.45− 7.67 = 503.78 kg (5.7)D4 = Q4/C4 = 503.78/52/D6 = S5 +D2 −D4 = 2650 + 965− 968.81 = 2646.19 kg(5.8)Silica and hemicellulose extraction process1. Known:• Q7 = Q4 = 503.78 kg• C7 = 52%• C8 = 1− 60% = 40% (moisture content of alkali treated chips is 60%)• D7 = D4 = 968.81 kg202A3. Mass Balance of the Pulping and Chemical Recovery Process• S9 = 8×Q7− (D7−Q7) = 8×503.78− (968.81−503.78) = 3565.21 L (L :W = 8 L/kg)• K8 = 85% (based on previous step)• R11 = 1%• C11 = C8 = 40%2. Calculation:Q8 = Q7 ×K8 = 503.78× 85% = 428.21 kg (5.9)D8 = Q8/C8 = 428.21/40% = 1070.53 kg (5.10)Q11 = Q7 ×R11 = 503.78× 1% = 5.04 kg (5.11)D11 = Q11/C11 = 5.04/40% = 12.60 kg (5.12)S10 = S9 +D7 −D8 −D11 = 3565.21 + 968.81− 1070.53− 12.6= 3450.89 kg(5.13)Q10 = Q7 −Q8 −Q11 = 503.78− 428.21− 5.04 = 70.53 kg (5.14)Kraft cooking process1. Known:• Q12 = Q8 = 428.21 kg• D12 = D8 = 1070.53 kg• S12 = 1070.53− 428.21 = 642.32 L• C12 = C8 = 40%203A3. Mass Balance of the Pulping and Chemical Recovery Process• C13.1 = 17%• L : W = 4.4 L/kg• S14 = 428.21× 4.4− 642.32 = 1241.8 L• K13.1 = 65% (based on previous step)204A3. Mass Balance of the Pulping and Chemical Recovery Process2. Calculation:Q13.1 = Q12 ×K13 = 428.21× 65% = 278.34 kg (5.15)D13.1 = Q13/C13 = 278.34/17% = 1637.29 kg (5.16)S13.2 = S14 +D12 −D13.1 = 1241.8 + 1070.53− 1637.29 = 675.04 kg (5.17)Washing and screening before bleaching1. Known:• D15 = D13.1 = 1637.29 kg• Q15 = Q13.1 = 278.34 kg• C15 = 17%• C16 = 10%• S18 = 9.1 m3/t pulp (assumed the washing cleanliness with Kappa No.=8-9)• R19 = 3%• C19 = 20% (water saturated)205A3. Mass Balance of the Pulping and Chemical Recovery Process2. Calculation:Q19 = Q15 ×R19 = 278.34× 3% = 8.35 kg (5.18)D19 = Q19/C19 = 8.35/20% = 41.75 kg (5.19)S18 = 9.1× 278.34 = 2532.89 kg (5.20)Q16 = Q15 −Q19 = 278.34− 8.35 = 269.99 kg (5.21)D16 = Q16/C16 = 269.99/10% = 2699.9 kg (5.22)S17 = S18 +D19 +D16 −D15 = 2532.89 + 41.75 + 2699.9− 1637.29 = 3637.25 L(5.23)Two-stage O2 delignification1. Known:• D20 = D16 = 2699.9 kg• Q20 = Q16 = 269.99 kg• C20 = 10%206A3. Mass Balance of the Pulping and Chemical Recovery Process• C21 = 10%• K21 = 90% (based on previous step)• S22 (determined by NaOH charge and conventration,which are 2.5% and 400 g/L, respectively)• S23 (determined by pulp consistency in the washing machine, C=4%)• R25 = 1%• C25 = 20% (water saturated)2. Calculation:207A3. Mass Balance of the Pulping and Chemical Recovery ProcessQ21 = Q20 × 90% = 269.99× 90% = 242.99 kg (5.24)D21 = Q21/C21 = 242.99/10% = 2429.9 kg (5.25)S22 = (Q20 × 2.5%)/0.4 = (269.99× 2.5%)/0.4 = 16.87 L (5.26)S23 = Q20/4%− S22 −D20 = 269.99/4%− 16.87− 2699.9 = 4032.98 L (5.27)Teo− stagewashing : 2S23 = 2× 4032.98 = 8065.96 L (5.28)Q25 = Q20 ×R25 = 269.99× 1% = 2.7 kg (5.29)D25 = Q25/C25 = 13.5 kg (5.30)S24 = S22 + 2× S23 +D20 −D21 −D25= 16.87 + 8065.96 + 2699.9− 2429.9− 13.5 = 8339.33 kg(5.31)D0 bleaching stage1. Known:• D26 = D21 = 2429.9 kg• Q26 = Q21 = 242.99 kg• S28 = 1.3%×Q26 (charge of ClO2 is 1.3%, based on o.d. pulp)• S29 (determined by pulp consistency in the washing machine, C=4%)• C27 = 10%• K27 = 99% (based on previous step)• R31 = 0.5%• C31 = 20% (water saturated)208A3. Mass Balance of the Pulping and Chemical Recovery Process2. Calculation:Q27 = Q26 ×K27 = 242.99× 99% = 240.56 kg (5.32)D27 = Q27/C27 = 240.56/10% = 2405.6 kg (5.33)S28 = (Q26 × 1.3%) = (242.99× 1.3% = 3.16 kg (5.34)S29 = Q26/4%−D22 − S28 = 6074.75− 2429.9− 3.16 = 3641.69 L (5.35)Q31 = Q26 ×R31 = 242.99× 0.5% = 1.21 kg (5.36)D31 = Q31/C31 = 1.21/20% = 6.07 kg (5.37)S30 = S28 + S29 +D26 −D27 −D31= 3.16 + 3641.69 + 2429.9− 2405.6− 6.07 = 3663.08 kg(5.38)EOP bleaching stage1. Known:• D32 = D27 = 2405.6 kg• Q32 = Q27 = 240.56 kg209A3. Mass Balance of the Pulping and Chemical Recovery Process• S34 = (1.3%×Q32)/0.4 + 6×Q32/1000(charges of NaOH andH2O2 are 1.5% and 6 kg/t, respectively based on pulp)• S35 (determined by pulp consistency in the washing machine, C=4%)• C33 = 10%• K33 = 98.8% (based on previous step)• R37 = 0.5%• C37 = 20% (water saturated)2. Calculation:Q33 = Q32 ×K33 = 240.56× 98.8% = 237.67 kg (5.39)D33 = Q33/C33 = 237.67/10% = 2376.7 kg (5.40)S34 = (Q32 × 1.5%)/0.4 + 6×Q32/1000= (240.56× 1.5%)/0.4 + 6× 240.56/1000 = 10.46 L(5.41)S35 = Q32/4%− S34 −D32 = 240.56/4%− 10.46− 2405.6 = 3597.94 L (5.42)Q37 = Q32 ×R37 = 227.47× 0.5% = 1.2 kg (5.43)D37 = Q37/C37 = 1.2/20% = 6.01 kg (5.44)S36 = S34 + S35 +D32 −D33 −D37= 10.46 + 3597.94 + 2405.6− 2376.7− 6.01 = 3631.29 L(5.45)210A3. Mass Balance of the Pulping and Chemical Recovery ProcessD1 bleaching stage1. Known:• D38 = D33 = 2376.7 kg• Q38 = Q33 = 237.67 kg• S40 = 0.05%×Q38 (charge of ClO2 is 0.5%, based on o.d. pulp)• S41 (determined by pulp consistency in the washing machine, C=4%)• C39 = 10%• K39 = 99.5% (based on previous step)• R43 = 0.3%• C43 = 20% (water saturated)211A3. Mass Balance of the Pulping and Chemical Recovery Process2. Calculation:Q39 = Q38 ×K39 = 237.67× 99.5% = 236.48 kg (5.46)D39 = Q39/C39 = 237.67/10% = 2376.7 kg (5.47)S40 = Q38 × 0.5% = 224.74× 0.5% = 1.12 kg (5.48)S41 = Q38/4%−D38 − S40 = 237.67/4%− 2376.7− 1.19 = 3563.86 L (5.49)Q43 = Q39 ×R43 = 236.48× 0.3% = 0.71 kg (5.50)D43 = Q43/C43 = 0.71/20% = 3.55 kg (5.51)S42 = D38 + S40 + S41 −D39 −D43= 2376.7 + 1.19 + 3563.86− 2364.8− 3.55 = 3573.4 L(5.52)212A3. Mass Balance of the Pulping and Chemical Recovery ProcessA3.2 Mass Balance of the Chemical Recovery ProcessAccording to the results in Chapters 3 and 4, 100% of silica in bamboo chips couldbe removed and 97% of the extracted silica could be recovered from the APEL withCO2 treatment. The silica content of raw bamboo chips is 1.12% (based on the o.d.chips). Before carrying out the calculation, several definitions are made.L : Liquor amount of each step (L)D : Total weight of the material (kg)M : Solids content of each step (kg)E : Efficiency of each step (%)W : Weight of silicon compounds in each step (kg)H : Weight of hemicellulose in each step (kg)Recovery of silica from the APEL2 NaOH + SiO2 −−→ Na2SiO3 + H2ONa2SiO3 + CO2 + 2 H2O −−⇀↽− H2SiO3 ↓+ NaHCO31. Known:• L44 = 3450.89 L213A3. Mass Balance of the Pulping and Chemical Recovery Process• M44 = 503.78× 15% = 75.57 kg (dissolved lignin, hemicellulose,cellulose, sodium silicates etc.)• W44 = 503.78× 1.12%× 122/60 = 11.47 kg (molecular weights ofNa2SiO3 and SiO2 are 122 and 60 g/mol, respectively)• C47 = 10%• E48 = 95%2. Calculation:W47 = W44 × 97%× 78/122 = 11.47× 97%× 78/122 = 7.11 kg(molecular weight of Na2SiO3 and H2SiO3 are 122 and 78 g/mol, respectively)(5.53)D47 = W47/C47 = 7.11/10% = 71.1 kg (5.54)W48 = W47 × 60/78× E48 = 7.11× 60/78× 95% = 5.19 kg (5.55)L45 = L44 −D47 = 3450.89− 71.1 = 3379.79 L (5.56)M45 = M44 −W44 × 97% = 75.57− 11.47× 97% = 64.44 kg (5.57)Residual alkali in the APEL was not included.214A3. Mass Balance of the Pulping and Chemical Recovery ProcessRecovery of hemicellulose from the silica removed APEL1. Know:• L45 = 3379.79 L• M45 = 64.44 kg• E50 = 95%• H45 = M45 × 55%• C50 = 10%2. Calculation:H45 = M45 × 55% = 64.44× 55% = 35.44 kg (5.58)H50 = H45 × E50 = 35.44× 95% = 33.67 kg (5.59)D50 = H50/C50 = 33.67/10% = 336.7 kg (5.60)L49 = L45 −D50 = 3379.79− 336.7 = 3043.09 L (5.61)M49 = M45 −H50 = 64.44− 35.44 = 29 kg (5.62)Assuming no other components was removed except for hemicellulose.215A3. Mass Balance of the Pulping and Chemical Recovery ProcessEvaporation system1. Known:• L52 = L49 + L51• L49 = 3043.09 L• L51 = 2532.89 L• M49 = 29 kg• M51 = 141.52 kg• M53/L53 = 60%2. Calculation:L52 = 3043.09 + 2532.89 = 5575.98 L (5.63)M52 = M49 +M51 = 29 + 141.52 = 170.52 kg (5.64)M53 = M52 = 170.52 kg (5.65)L53 = 175.37/60% = 284.2 L (5.66)216A3. Mass Balance of the Pulping and Chemical Recovery ProcessRecovery boiler, causticizing system and calcinationFor the calculation of these three stages, accurate data are required for the specificpulping system. So, we did not do the mass balance of this part.217"@en ; edm:hasType "Thesis/Dissertation"@en ; vivo:dateIssued "2017-05"@en ; edm:isShownAt "10.14288/1.0343481"@en ; dcterms:language "eng"@en ; ns0:degreeDiscipline "Chemical and Biological Engineering"@en ; edm:provider "Vancouver : University of British Columbia Library"@en ; dcterms:publisher "University of British Columbia"@en ; dcterms:rights "Attribution-NonCommercial-NoDerivatives 4.0 International"@* ; ns0:rightsURI "http://creativecommons.org/licenses/by-nc-nd/4.0/"@* ; ns0:scholarLevel "Graduate"@en ; dcterms:title "Understanding hemicellulose and silica removal from bamboo"@en ; dcterms:type "Text"@en ; ns0:identifierURI "http://hdl.handle.net/2429/61124"@en .