UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Fundamentals of strength loss in recycled paper Nazhad, Mousa M. 1994

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

Item Metadata


831-ubc_1995-983302.pdf [ 5.75MB ]
JSON: 831-1.0088893.json
JSON-LD: 831-1.0088893-ld.json
RDF/XML (Pretty): 831-1.0088893-rdf.xml
RDF/JSON: 831-1.0088893-rdf.json
Turtle: 831-1.0088893-turtle.txt
N-Triples: 831-1.0088893-rdf-ntriples.txt
Original Record: 831-1.0088893-source.json
Full Text

Full Text

FUNDAMENTALS OF STRENGTH LOSS IN RECYCLED PAPERbyMousa M. NazhadA THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THEREQUIREMENTS FOR THE DEGREE OFDoctor of PhilosophyInTHE FACULTY OF GRADUATE STUDIESDEPARTMENT OF FORESTRYWe accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIADecember, 1994Mousa M. Nazhad, 1994In presenting this thesis in partial fulfillment of the requirements for an advanced degreeat The University of British Columbia, I agree that the Library shall make it freelyavailable for reference and study. I further agree that permission for extensive copyingscholarly purposes may be granted by the Head of my department or by his or herrepresentative. It is understood that copying or publication of this thesis for fmancialgain shall not be allowed.Facultyt1itof ForestryThe University of British ColumbiaVancouver, Canada(signature)Date: December 22, 1994iiABSTRACTConsiderable work has been devoted to the upgrading of recycled chemical (lowyield) pulp fibers during the past decade. There is also disagreement on the effectivenessof an upgrading process regardless, whether of chemical or mechanical origin. Oneserious problem which restricts sustainable progress in the field of fine-paper recycling isthe lack of knowledge of the mechanism by which recycling affects the texture andarrangement of the cell wall which ultimately causes inferior properties of the recycledfibers.The deteriorative effect of recycling on fine-paper manifested itself on the loss inpotential bonding of recycled fibers. The loss in potential bonding of the recycled fiberstranslated into homification (i.e., loss in fiber wet-flexibility) and/or surface deactivationby recycling. The susceptibility of the fibers for hornification rather than surfacedeactivation during recycling is substantiated with different techniques. It is concludedthat the hornification is responsible for inferior properties of recycled fibers. Moreimportantly, observations in the present work suggest that refining/beating does notdevelop any new surface area. The effect of refining is restricted to a reduction in therigidity of the lamellae by mechanical fatigue and subsequently, increased swelling andplasticization of the fiber wall. Thus, drying of never-dried fibers (unbeaten or beaten)from water pulls the lamellae toward each other by surface tension forces and binds thelamellae rich in surface by crystallization forces. These forces lead to an increase in thecrystallization of the cell wall provided that the condition required for crystallization, ismet by the molecular orientation in the cell wall. When these fibers are re-wetted again,the delamination does not reverse completely, and the lamellae remain partially closed.This results in increased rigidity of unraveled lamellae and restricts the internal surfacesof fibers to access by water. The concomitant result is restricted swelling and thus, lossin wet-plasticity of the fibers on recycling. Most of this change takes place in the firstcycle. Repeated recycling deteriorates further the wet-plasticity of the fibers. Based oniiithese fmdings a model is proposed which explains the mechanism by which homificationdevelops in the fiber wall during recycling. The proposed model also provides newinformation on the effects of fiber beating or refining.IivTABLE OF CONTENTSABSTRACT iiTABLE OF CONTENTS ivLIST OF FIGURES viiiLIST OF TABLES xLIST OF NOMENCLATURE xiACKNOWLEDGEMENT xiv1. INTRODUCTION 12. THEORETICAL BACKGROUND 42.2 Correlation of Cell Wall Structure and Papermaking Qualities of Fibers 42.1.1 Ultrastructural arrangement of the fiber wall 42.1.2 Crystallinity concept in polymers 72.1.3 Advantage of the X-ray diffraction (XRD) method 82.1.4 Development of XRD method for crystallinity measurement 92.1.5 Crystallinity index 102.2 Beating (Refining) 112.2.1 Theory of the beating (refining) Chemical theory Physical theory 132.3 Surface area of fibers 162.3.1 Adsorption technique to measure the fiber surface area 172.3.2 External and internal surface area 192.4 Thermodynamics of Fiber-Water Interactions 202.4.1 Free energy changes 212.4.2 Enthalpy changes 22V2.4.3 Entropychges.232.5 Enzymatic Hydrolysis of Recycled Fibers 243.LITERATLJRE REVIEW 253.1 Introduction 253.2 Postulated mechanisms for loss of strength in recycled paper 263.2.1 Fiber Flexibility Irreversible pore closure Cross-linking Cellulose chain cleavage Re-organization in the fiber (cell) wall 303.2.1.5 Bond strength 323.2.2 Surface Condition of Fibers 333.2.2.1 Hemicellulose-loss effect 333.2.2.2 Inactivation of the fiber surface 343.2.2.3 Microcompressions 363.2.2.4 Fiber-Water Interactions 363. The electric charge of pulp surfaces in water. .373. Changes in water structure due to fiber-waterinteractions 383.3Summary 424. MAThRIAL AND METHODS 444.1 Preparation of Pulp Samples 444.2 Bleaching Materials 444.2.1 Preparation of chlorine 444.2.2 preparation of chlorine dioxide 454.2.3 Laboratory bleaching process 464.3 Recycling procedure 48vi4.4 Fiber Length (FL) a1ysis . 514.5 Molecular Weight Distribution (MWD) 524.6 Crystallinity and Ciystallite Size of the Fibers 544.6.1 D 5000 Diffractometer 544.6.2 Sample preparation for x-ray diffraction analysis 564.6.3 X-ray data analyses 574.7 Water Vapor Sorption 574.8 Measurement of the Heat of Wetting 614.9 Scanning Electron Microscopy (SEM) 634.10 Determination of Thermodynamic Properties of Fiber 644.11 Enzymatic Hydrolysis 655. RESULTS 666. DISCUSSION 946.1 Adsorption Isotherms 946.2 Heat of Immersion 966.3 Surface Analysis of Recycled Fibers 976.4 Thermodynamics of Fiber-Water Interactions 996.4.1 Free energy996.4.2 Differential enthalpy and hydrogen bonding 1016.4.3 Entropy 1026.5 Changes in Fiber Wall Ultrastructure due to Recycling 1056.6 Enzymatic Hydrolysis 1096.7 Theory of Mechanism of Strength Loss in Recycled Papers 1126.8 Proposed Model for the Evolution of Fiber Wall Ultrastructure on RecyclingProcess 1167. SUMMARY AND CONCLUSIONS 119vii7.1 SUMMARY1197.1.1 Beating1197.2.1 Recycling1207.2 Conclusions1218. RECOMMENDATIONS1258.1 Identification of the Parameters Involved in HomificationDuring theBleaching Process1258.2 Role of Hemicellulose in Hornification of the Cell Wall1258.3 The Role of Temperature in Recycling1268.4 Implications of These Findings on Upgrading of RecycledFibers 1279 BIBLIOGRAPHY128APPENDICESA Peak profile resolution138B Derivation of the formula for external surface area measurement 141C Integral and differential free energy changes143D Method for determining the integral and differentialenthalpy 145E Internal tension of liquid between two parallel surfaces 147F Calibration curve for GPC149GLOSSARY151VITAviiiLIST OF FIGURESFig. 1. Pictorial representation of the lamellae model for the ultrastructural arrangementof lignin, cellulose and hemicelluloses in the wood cell wall 5Fig. 2. Schematic representation of the swelling of the fiber wall (after Scallan 1974).. .6Fig. 3. Change in flexibility of fiber by acidity of paper. The fiber retains its flexibilitywhen it is in a neutral or slightly alkaline environment 30Fig. 4. Diagram showing the adherent solvent layer on surface when a surface issubmerged in a liquid 38Fig. 5. Schematic cross-section of a broken hydrogen-bonded water near cellulosesurface 39Fig. 6. Schematic representation of the adsorbed water between two cellulosic surfaces.The network was demonstrated as a distorted ice-like cluster with extension ofhydrogen bonds from one surface to other 41Fig. 7. A sketch of chlorine dioxide preparation 45Fig. 8. A flow chart of the experimental procedure 50Fig. 9. Block diagram of FS-200 fiber length analyzer 51Fig. 10. Block diagram of the diffractometer in D 5000 mode 54Fig. 11. Diffractometer beam path in 0/28 mode 55Fig. 12. Block diagram of apparatus which was developed for conditioning the fibers atdifferent relative humidities with a constant temperature (i.e., 25±2° C) 60Fig. 13. A cross-section of the PARR 1451 solution calorimeter 62Fig. 14. Changes in properties of the handsheets (made from beaten bleached kraft) dueto recycling 67Fig. 15. Comparison of fiber length (FL) distribution of unbeaten, beaten and recycledfibers 68Fig. 16. SEM microphotographs of (a) unbeaten and (b) beaten pulp fibers 69Fig. 16. SEM microphotographs of (c) cycle I and (d) cycle VI pulp fibers 70Fig. 17. Comparison of the adsorption isotherms of unbeaten, beaten and recycledfibers 72Fig. 18. Comparison of sorption isotherms of cycle I and cycle V 73Fig. 19. Desorption isotherms of unbeaten and beaten (6000 and 12000 revs.) fibers. .75ixFig. 20. Comparison of critical point drying and freeze drying methods with sorptionpotential of the fibers at 25° C 76Fig. 21. Comparison of the heat of wetting of unbeaten, beaten and recycled fibers 77Fig. 22. Comparison of MWD of unbeaten, beaten and recycled fibers 78Fig. 23. Comparison of the diffractograms of unbeaten, beaten and recycled fibers 80Fig. 24. Comparison of the diffractograms of virgin (beaten), cycle I and cycle Ill pulpFig. 25. Comparison of the differential free-energy of virgin and recycled fibers 83Fig. 26. Comparison of the differential enthalpy of virgin and recycled fibers 84Fig. 27. Comparison of the differential entropy of virgin and recycled fibers 85Fig. 28. Comparison of scanning electron photomicrographs of hydrolyzed substratesderived from (a) unbeaten and (b) puips 89Fig. 28. Comparison of scanning electron photomicrographs of hydrolyzed substratesderived from (c) cycle I and (d) cycle VI puips 90Fig. 29. Comparison of MWD of hydrolyzed samples 92Fig. 30. Comparison of the integral free-energy of virgin and recycled fiber 100Fig. 31. Comparison of the integral entropy of virgin and recycled fiber 103Fig. 32. Conceptual drawing of the cell wall ultrastructural arrangement in unbeaten (A),beaten (B), dried (C) and recycled (D) states 116Fig. 33. The resolution of diffraction peaks in terms of Voigt profiles for each peak. 139Fig. 34. Conceptual drawing of a sample of pulp in equilibrium with the vapor of aliquid. To be immersed in the liquid and thus lose the surface energy of theduplex film of liquid 146Fig. 35. Conceptual drawing of water between two lamellae 147KLIST OF TABLESTable I. Relationship of sulfuric acid concentration with relative humidity at 25° C... .58Table II. Relative humidities over standard salt solutions at 250 C 59Table III. Characteristics of the white spruce kraft pulp (unbleached and bleached(CEDED)) 66Table IV. Fiber length (FL) analysis of virgin (unbeaten and beaten ) and recycled fibersTable V. Surface area for virgin (unbeaten and beaten) and recycled fibers 74Table VI. Degree of polymerization (DP) of the (unbeaten and beaten) and recycledfibers 79Table VII. Crystallinity information on virgin (beaten) and recycled puips 81Table Vifi. Calculated thermodynamic properties for the adsorption of water on beaten,unbeaten puips dried by the CP drying method 86Table IX. Calculated thermodynamic properties for the adsorption of water on recycledpuips dried by the CP drying method 87Table X. Fiber length (FL) analysis of hydrolyzed (H) puips 88Table XI. Degree of polymerization (DP) and hydrolysis yield of the hydrolyzed (H)puips 91Table XII. Crystallinity of unhydrolyzed (U) and hydrolyzed (H) pulp samples 93Table XIII. Development of internal tension forces between the lamellae during removalof water from the cell wall 148xiLIST OF NOMENCLATURE20: Bragg’s angleI: Intensity of diffractionCrI%: Crystallinity Index (%)FWHM: Full width at half maximumIntegral width (Voigt function)Integral width (Cauchy function)13g: Integral width (Gaussian function)Cauchy functionI(x): Gaussian function1(x): Voigt functionh: Relative vapor pressureM: Moisture adsorbed at the relative vapor pressure hMm: Volume of vapor adsorbed in the monolayer of adsorbateC: A constant depending on the heat of adsorptionET: Total surface areaInternal surface areaExternal surface areaa: Effective area occupied by an adsorbed molecule (for water a = 12.5 A2)N: Avogadro’s number (6.02 x 1023)M: Molecular weight of adsorbateE1: Heat of adsorptionEL: Heat of condensation of the vaporHeat of immersion of the sample equilibrated at 100% relativehumidityHeat of immersion of a dry clean solidxiiR: Universal gas constant (1.98662 calJmole.K)T: Absolute temperature (Kelvin temperature scale: 273.16 K + temperatureof water (T0))p.: A constant related to the amount of vapor already adsorbedAF1: Differential free energyAF: Integral free energyn1: Moles of water per 100 g of solidH: The free energy of adsorption per unit surface areaFree surface-energy of the solid vocuum interfaceFree energy of the surface when the solid-vapor interface is constructedzH1: Differential enthalpyz\H: Integral enthalpyAH1: Deferential entropyIntegral entropyM: Molecular weight of cellulose tricarbanylateM: Molecular weight of polystyreneK: Mark-Houwink Coefficient of cellulose tricarbanylate (2.01 x l0)K: Mark-Houwink Coefficient of polystyrene (1.18 x 10)(Xe: Mark-Houwink Coefficient of cellulose tricarbanylate (0.92)Mark-Houwink Coefficient of polystyrene (0.74)JIBET: Bonding energy between water and fiber by BET approacht\Hcal: Bonding energy between water and fiber by Calorimetry approachAHa Heat of adsorptionDP: Degree of polymerizationMWD: Molecular weight distributionFL: Surface energy lowering produced by immersion in the liquidSurface energy of the solidxiiiSurface energy of the solid-liquid interfaceF: Surface energy lowering produced by exposure to the saturated vapor0: Contact angleU1: Chemical potential of the liquid at the adsorbed stateU0: Chemical potential of a pure liquidF: Force between two lamellae under tension forcesSurface tension of liquidsg Surface tension of solid-gas interface11: Intrinsic viscosityHV: Hydrodynamic volumexivACKNOWLEDGMENTSAs Rome is not built in one day, building the character of a man also is not builtfrom a single act. The cornerstones of becoming mature are parents, friends andteachers. My parent’s love for education was the source of my inspiration andencouragement to pursue academic study. I cannot forget on those cold days of myelementary school period how my father used his tiny body as a shield to protect me fromthe storm on the way to school. I am always thankful to them. The everpresent influenceof my friends and teachers, past and present, during the years of education and traininglaid the foundation which enabled me to overcome the obstacles that I encountered and toaccept future challenges along the way.Each member of my committee helped me in unique ways. I sincerely appreciatethe effort they made to help this dream of mine come true. I would like to express mysincere appreciation for the assistance provided by my thesis supervisor, Professor LaszloPaszner. He taught me patience and gave me general guidance to conduct my study. Thedevelopment of my ideas presented in the final manuscript evolved through thepreparation of many working papers which were always cheerfully reviewed by Dr.Paszner. His helpful criticisms were a significant factor in crystalizing my ideas. I wishto acknowledge Dr. R. S. Seth (PAPRICAN) for his contribution on the completion ofthis thesis. He was patient with me when I needed so much assistance; I am alwaysgrateful to him for that. At the department of Wood Science, I gratefully acknowledgeDr. S. Avramidis for his unconditional support and his generosity in providing hislaboratory facilities for sorption analysis. As well Dr. S. Ellis provided invaluableassistance with the GPC analysis, eagerly shared his experience and ideas any timerequested and carefully read this thesis. I wish also to thank the faculty and staff of theFaculty of Forestry for their help during my years of study at UBC. In particular, Ms. D.xvCaciman, Ms. N. Cole and Ms. C.E. Laird who, in one way or another, helped make thisstudy a success.There are those who are due thanks for specific expertise and information. Ithank Dr. R. J. Kerekes of PPC, at the University of British Columbia for stressing theimportance of the clarity of the project objectives at the outset. He was also instrumentalin providing facilities for part of this work. Dr. L. Groat of the Department of Geology(UBC) deserve recognition for the permission to use his laboratory facilities and forhelpful discussions concerning the XRD analyses.The financial assistance of the Science Council of British Columbia through theGREAT Award, and the Natural Sciences and Engineering Council of Canada aregratefully acknowledged.I would like to extend my appreciation to friends whose company andunderstanding made the difference at many times. I cannot name every one of them, onlytell them how much I appreciate their genuine friendship. However, I wish toacknowledge a number of them in particular. My Special thanks to Ms. Pam Rogers andDr. G. Voss, Ms. Touran Vahedi, Mr. J. Gofraniha, Dr. I. Nelson, Mr. D. Aquino, Dr. G.Nercessian and Mr. A.A. Karim for helping me enormously in the many ways that theycould. Their friendship is beyond words.Special thanks go to my wife Ashraf for her unfailing support throughout themany years of my graduate student life. Her charm and compassionate support havehelped me maintain my sanity, even though I may not always have acknowledged it atthe time. And last but not the least, to my children Arash, Iliad and llnaz for all of themissed trips, picnics, etc., when dad was absent.xviDEDICATIONI dedicate this thesis to my wife - Ashraf S. Moussavi. Ashraf gave up her careerin nursing in order to minister our children during my study. Her dedicated love andencouragement have been my main source of strength. This work would not have beenpossible without her patience, conviction and assistance.11. INTRODUCTIONDisappearance of raw material sources, rapid decrease in the number of suitablelandfill sites and public awareness pose great challenges to the pulp industry to find newways for recycling and re-using the discarded fibrous waste for sustained production ofpaper in the future. Paper and board represent almost 32% of discarded municipal solidwaste, based on recent reports by the U.S. Environmental Protection Agency (EPA). It isestimated that each tonne of recovered paper saves 2.4 cubic meters of landfill space(O’Brein, 1993). Around 35 million tones of waste paper were recycled in 1992 in theU.S.. If that paper had gone to a landfill it would have covered an area almost 14.3square kilometers and 4 meters deep. This collected waste paper could preserve 64million trees of one tonne oven dry weight each or the old growth forest on a land area ofapproximately 6,000 square kilometers. The calculation is based on 58 trees of one tonnein a hectare with almost 100% yield.The latest news indicates that the recovery rate of waste paper in North Americahas reached to about 50%, however, most waste paper is utilized for paperboard,especially corrugated boards, newsprint and toilet rolls. Utilization for printing andinformation papers is still at a lower level. Recent rapid increase in the use of officepapers for reprography and computers, as well as magazines and other printing papers, isoverwhelming municipal waste management. In the past the utilization rate of oldnewspaper for malcing newsprint increased from as little as 10% in 1988 to about 50% in1994 in only less than half a decade in North America. This might be due to the fact thatthe quality of mechanical pulp fiber is not affected negatively by recycling. To promoteincreased utilization of office waste paper, it is necessary to overcome the inferiorproperties of recycled fibers. Thus, understanding the causes of such inferior fiberquality in recycling is at the heart of a number of key operations in upgrading recycledfibers. Such findings point the way forward to probable effective methods that could2bring about considerable improvements in recycled paper quality. This would thusdiminish the flow of waste streams to landfills, and thereby restrain excessive depletionof the forest resource by reducing the amount of virgin fiber required for maintainingpaper supplies and quality for most grades. In fact, office waste paper could also becomethe fiber of choice if the inferior properties of the recycled fibers could be overcome.Inferior properties of recycled office waste on papermaking properties of fibershas been discussed thoroughly by different workers (McKee 1971, Szwarcztajn andPrzybysz 1976, Howard and Bichard 1992), and an excellent review on this subject wasrecently published by Howard (1990). These investigations showed a dramatic loss oftensile and bursting strengths on the repeated remaking of low yield chemical pulp intopaper (See also Fig. 14).A review of recycling of chemical puips (Ehrnrooth et al. 1978, Howard 1990),clearly reveals that the losses in paper strength are brought about by loss in potentialbonding of the fibers. On the other hand, the loss in bonding of recycled fibers could bea function of two parameters (Eastwood and Clarke, 1978), (a) fiber flexibility and (b)surface condition.In a recent review by Nazhad and Paszner (1994) it was proposed that the mainloss in potential of fiber bonding is brought about by homification or Verhornung (i.e.,loss in wet-flexibility) of the fibers during recycling. Generally, for fibers in the originalwater soaked condition, the free hydroxyl groups in the cell wall are practically allsatisfied by water. When the fiber dries, pairs of the fibrils/lamellae are drawn togetherso that the individual fibrils/lamellae adhere strongly to each other. It is obvious that themechanism of hornification could be uncovered by pursuing the physico-chemicalchanges occurring during the closure of these lamellae. More explicitly, the mystery ofhornification lies somewhere between these lamellae. The review (Nazhad and Paszner,1994) also concluded that the current traditional methods of upgrading recycled fibers are3not adequate, and it is necessary to develop and implement new systems which willupgrade recycled papers without worsening the overall papemiaking properties.By exploring, from as many directions as possible, the changes produced in thefiber wall by recycling, it is hoped to accumulate enough data to permit formulation of amore definite picture of the causes of inferior properties of recycled fibers. On this basis,the ultrastructure of the fiber wall must be fully reviewed. The theoretical backgroundfor structural, refining and thermodynamical analysis is presented. The superstructuralarrangement of the virgin fiber (i.e., the proportion of the cell wall crystalline componentand crystallite thickness) was measured and compared with those of recycled fibers. Thephysical properties of the fibers (i.e., virgin and recycled fibers) such as sorptionpotential, fiber surface area, heat of wetting of the fibers were also measured. Acomparison was made on the thermodynamics of the two systems. Finally, the recycledsamples were interacted with cellulase enzymes to detect changes in the structure of thefibers by recycling. The results of each of these investigations, in the light of theukrastructure of the fiber wall, are discussed separately. Cross references were madebetween the results obtained from the different types of investigations.4THEORETICAL BACKGROUND2.1 Correlation of Cell Wall Structure and Papermaking Qualities of FibersEvaluating the quality of fibrous materials is important in paper manufacturing.Fundamental studies of the arrangement of cellulose molecules relative to each other in aparticular state of aggregation make it possible to formulate criteria for predicting thequality of the end use products. Thus, structural characteristics, such as lateral order,degree of orientation and size of crystallites influence decisively the physical andmechanical properties of the fibers. For instance, if the crystallinity of the cellulosic fiberis increased, the flexibility of the fibers could be affected inversely. It is well known(Salmen, 1988) that water does not penetrate into crystalline domains of cellulose sinceentrance of one molecule of water requires the breaking of many cellulose - to - cellulosebonds. Therefore, the elastic modulus of crystalline cellulose remains unaffected in thevicinity of a moist environment. Increased crystallinity can lead to a higher moduluselasticity in the recycled fibers. As a consequence, flexibility, which is inverselyproportional to the modulus, will decrease affecting the conformibility and therefore therelative bonded area.In order to be able to analyze the effects of drying on the structure of the fibers, aprofound knowledge of the structural features and fine structure is an essentialprerequisite. Therefore, a short review of the structural concepts of cellulose fibers willbe dealt with first.2.1.1 Ultrastructural arrangement of the fiber wallThe structural concept of the cell wall has naturally evolved through thecollection and collation of many pieces of experimental information (Frey-Wyssling51954 and 1969, Hess et al. 1957, Hearle 1963, Stone and Scallan 1965, Mclntash 1967,Stamm and Smith 1969, Fengel 1970, Scallan 1974, Kerr and Goring 1975).The wall of delignified fibers could be visualized as an array of lamellarstructures, each composed of a sheet of cellulose micmfibrils (Kerr and Goring 1975,Stone and Scallan 1965, Dunning 1969, Stamm and Smith 1969, Scallan 1974, Kerr andGoring 1975). The sheets of microfibrils are parallel to the planes of 101 in the fiber wall(Frey-Wyssling 1954 and 1969, Wardrop 1964). Fig. (1) represents a pictorialrepresentation of the cell wall before and after delignification.Lignin- Hemicelluloses— CelluloseFig. 1. Pictorial representation of the lamellae model for the ultrastructural arrangementof lignin, cellulose and hemicelluloses in the wood cell wall ( adapted from FreyWyssling 1954 and 1969, Scallan 1974, Goring 1975 and Fengel 1970).LAXII II002*b(Ia A AX K - r 1 —— 1 oTElementary-( K • El a101fib ruDelignifjedfiber wallL1gnin..HmatrixH emiceIlulose6The elementary fibril consists of a crystalline core that is flattened parallel to the 101lattice plane, where these planes are oriented parallel to the lumen of the fiber wall. Thisshape is due to a faster growth of the 101 plane, which is more hydrophilic than the moreslowly growing 101 plane (Frey-Wyssling 1954 and 1969, Stamm and Smith -1969, Kerrand Goring 1975, Krassig 1984). The crystalline core of the microfibrils is embedded ina cortex of hemicelluloses and lignin matrix. The matrix which is located on the surfacesof lamellae (i.e., sheets of microfibrils) parallel to the 101 plane is the location wherewater is adsorbed when cellulose fibers are delignified.Fig. 2. Schematic representation of the swelling of the fiber wall (after Scallan 1974).7This concept of a multi-lamellar structure of the fiber wall has been extended byStone and Scallan 1965, Dunning 1968, Stamm and Smith 1969, Scallan 1974, and Kerrand Goring 1975, where radial cleavage of the sheets of microfibrils may occur duringswelling (Figs. 2b, 2c, 2d). These characteristics of the fiber wall suggests that there is alarge area of surface on the delamination of a single fiber, which will be referred to asinternal surface area. This surface area is very great when it is compared with theexternal surface area of the fibers (i.e., 200 times more in average). Upon drying thelamellae join together to form a single continuous layer with a negligible void volume(Fig. 2a). The lamellar structure of the dried fiber wall is not completely reversible whenit is re-wetted.In summary, the general view is that cellulose molecules are linked together toform concentric 101 planes parallel to fiber axis whose exposed surfaces are hydrophilic(Frey-Wyssling 1954 and 1968, Emerton 1957 and Wardrop 1964, Dunning 1968).Water is adsorbed between these planes (Stamm and Smith 1969). Loss in accessiblesurface of those planes in any amount will reduce water uptake by the cell wall and thusthe swelling potential of the fibers will be restricted. That is to say, the fibers becomehornified and their wet-flexibility is lost.2.1.2 Crystallinity concept in polymersThe atoms in amorphous materials such as glasses, resins, and liquids do not havethe symmetrical repeating arrangement characteristic of crystalline materials. Thus, thex-ray patterns of amorphous materials at low diffraction angles appear as one or twoweak broad bands superimposed upon a continuous background. However, a highlycrystalline substance has very explicit peaks at specific diffraction angles. It is importantto define the terms crystalline and amorphous because in a polymer the two states can beintertwined in many ways yielding states described as quasi-crystalline, para-crystalline,8 semi-ordered, or even randomly ordered (Wims and Mayers 1986). An important factor determining the characteristics of polycrystalline polymers is the amount, orientation and size of the crystallites in the material, and the changes in these quantities when the polymer is subjected to various external influences such as drying, heating and stressing. Direct quantitative information on the changes in polycrystalline behavior, brought about by various external influences, is important in the study of physical and mechanical properties of polymers and their end use products. Different authors looked at the crystallinity of recycled fibers (Bouchard and Douek 1994, Guest and Voss 1981, Yamagishi and Oye 1981). However, little attention was paid to the method of sample preparation and reliability of the methods employed. This is a first extensive work on the cellulose crystallization effect of paper recycling. The sample preparation adopted was reproducible. The correlation of crystallization with the behavior of recycled fibers in paper furnishes is discussed. 2.1.3 Advantage of the X-ray diffraction (XRD) method The XRD method has several advantages over other methods in that it provides a rather rapid, convenient, and accurate method of crystallinity measurement provided the sample preparation is precise. Modern XRD systems have a major advantage in that the instrumentation is computerized which eliminates using the time-consuming film recording techniques. The data can be collected automatically, and analyzed with advanced software. Furthermore, the modern XRD technique does not require any other corrections previously required for the film techniques. Although there are six different techniques, x-ray diffraction (XRD), nuclear magnetic resonance (Proton and deuterium NMR), differential scanning calorimetry (DSC), infrared (IR), sorption ratio (SR), and density (D) to determine the crystallinity of polymers, only the XRD technique analyses directly the polycrystalline material in terms of the fundamental definition of its crystals. 92.1.4 Development of the XRD method for crystallinity measurementThe state and extent of molecular order in a fibrous polymer can be assessed quiteeffectively by visual examination of a wide-angle X-ray diffractogram. Quantitativemeasurement of this state of order and the size of the ordered units, is usually referred toas crystallinity and crystallite size respectively (Hindeleh and Johnson 1978). Differentmethods have been used to measure the crystallinity index. The advantages anddisadvantages of the methods employed will be described briefly in the following.The classical method of Hermans and Weiclinger (1948) gives an absolutemeasure of crystallinity in terms of the ratio of the integrated intensity under the peaks tothe integrated intensity under the complete diffractogram. The problem with this methodis the difficulty in separating the crystalline peaks from the amorphous background.Furthermore this method ignores the possibility of peak overlap. The empirical methodof Segal et al. (1959) which has been used extensively up to date, ignores the width of theintensity peaks blurring the distinction between paracrystalline materials. Krassig andKitchen (1961) introduced the peak width for calculating the crystallinity. However, themain drawback of the approach was the identification of accurate peak width.Determination of peak width at half height which generally is referred toas the “fullwidth at half maximum” (FWHM) is very sensitive to the tails of a given peak.Fortunately, all the work cited above is useful in terms of relative measurements, but onlywith respect to one particular material and the particular method in use.The separation of overlapping wide-angle x-ray diffraction, by employingGaussian or Cauchy disthbution has long been used by metallurgists. These attemptswere directed mostly to measure the peak widths from which crystallite size values canbe derived using formulae suggested by Scherrer (Hall 1949). Gjonnes and Norman(1960) applied such an approach to the evaluation of wide-angle x-ray diffractograms ofcellulose substrates. This method allows the separation of overlapping reflections underthe assumption that the intensity curves of the reflections have the shape of Cauchy’s10distribution. The authors determined the variation in the state of order by acorresponding change in width of the peaks in x-ray diffractograms of the sample.Hindeleh and Johnson (1978) used a combined Gaussian-Cauchy function for each peakin order to evaluate the crystallinity and crystallite size of paracrystalline samples.In the present work, the Voigt function was employed. Compared toexperimental peaks, the tails of Gaussian disthbution are too short, those of the Cauchydistribution are too long, but those of Voigt and Pearson distributions are closer to realityat the expense of longer computing time. The Voigt function is discussed in detail inAppendix A.2.1.5 Crystallinity indexCrystallinity index is defined as the weight fraction of the crystalline portion ofpolymers. The higher the degree of ordering, the more crystalline is the polymer. Forreal systems, virtually no amorphous substance will be completely without order. Therewill always be some succession of chemical bonds with fixed lengths and angles that willgive rise to a particular radial density distribution function F(r) that we view as adiffractogram. As structural features coalesce in a more ordered fashion, the x-raydiffractogram moves away from a diffuse halo to explicit peaks whereas the totalintegrated area (halo plus explicit peaks) remains constant. This fact can be used tocalculate the ratio of a percent crystallinity as the area of these explicit peaks and theamorphous content as the area of the diffuse halo. Full width at half maximum (FWHM)or integral breadth could also be determined by peak resolution programs. FWHMdemonstrates the average modification of crystallite dimensions due to the influence ofrecycling.In the present investigation a peak resolution function was used to calculate boththe crystallinity index of cellulose and the dimensions of the crystallites (Appendix A).11This function allows the resolution of the x-ray diffractogram into the contributions ofeach of the diffraction planes (i.e., 101, loT, 002, 040, etc.) (See also Fig. 33). Theamorphous component is calculated by subtraction of the area of resolved peaks from thetotal area of the x-ray diffractogram. Generally, the Voigt function results in the best fitof the x-ray diffraction patterns and is routinely used for the determination of both thecrystallinity index of cellulose and dimensions of the crystallite through integral breadthor full width at half maximum height (FWHM). The ciystallinity index, which is definedas the ratio of the resolved peak area to the total area under the unresolved peak profile, iscalculated using the equation (Hindeleh and Johnson, 1978):192 192Cr1 (%)=[j 10d9 / ITdO] x 100 [2.1]where Cr1 is the crystallinity index of the sample, ‘T is the total intensity of diffractionin the diffractogram, I is the total intensity of the diffraction due to the crystallineregion (peaks) of the sample and 9 and °2 are the limiting values depending on the rangeof a diffraction angle.2.2 Beating (Refining)Beating or refining is a mechanical treatment of the fibers. During this treatmentcertain changes in the structure of the fiber wall take place. Beating and refming areterms used synonymously since refiners are commonly used to beat pulp stock in currentpractice.The basic effects of beating are as follows:1. Internal fibrillation - variously described as delamination or swelling.2. External fibrillation - the partial removal of the fiber wall, leaving it still attached tothe fiber.123. Cutting or shortening of the fibers and fines production.4. Redistribution of hemicellulose from the interior of the fiber to the exterior.5. Introduction of nodes, kinks, slip planes, microcompressions in the cell wall; orremoving nodes, kinks, slip planes, microcompressions from the cell wall.There is, of course, an endless list of effects which range from most significant totrivial. Although this list is not inclusive of all effects during refining, however, these arethe most significant and can be found in the current literature (Emerton 1957, Atack1977, Page 1989). and does not indicate that, some other basic effects cannot haveoccurred during refining.2.2.1 Theory of the beating (refining) Chemical theoryThe theory of beating started with the chemical theory and dominated this fieldfor a quarter of a century. The chemical theory, or hydration theory as it came to beknown, proposed that (Cross and Bevan 1920) the increase in paper strength whichaccompanies beating is due to the working of the surface layers of the fibers into ahydrate of cellulose. Hydrated cellulose was found to have gelatinous properties actingas an adhesive to bind the fibers together. It is important to realize that the term wasoriginally used in a strict way. For example, anhydrous copper sulfate (CuSO4) ishydrated with the addition of water. Indeed, it is converted into the pentahydrate(CuSO4.5H20)which is chemically distinct.The chemical theory of beating, in different forms, persists even to date. Henry etal. (1988) studied the mobility and interactions of the water molecules contained in thepulps. The authors concluded that the beating did not increase the interior surface of thecellulose fibers but only caused a better structuring of the water molecules along the fiber13surfaces. Milichovsky (1990) suggested that beating in the aqueous medium does notchange the composition and structure of molecules forming individual pulp fibrils andmicrofibrils. What is changed, however, is the composition of the phase interfacebetween the pulp and water medium and the molecular arrangement in single pulpformations, especially in places affected by simultaneous mechanical action. In otherwords, the beating process involves chemical changes in pulp beyond the level of thebasic molecular structure.The author based this concept of beating on a theory which is called SCHL(structural changes in hydration layers) theory. The theory is based on the dipolar natureof water molecules and their two possible orientations in hydration surfaces. If the watermolecule orientation to each of the interacting surfaces is equal then, both surfaces wifiaffect each other with repulsing hydration forces. On the other hand, if the orientation ofwater molecules is different, the surfaces will be attracted. Milichovsky’s approach to theproblem of beating is limited and it does not fully explains for the differences in watermolecule orientations. Physical theoryThe chemical theory was inadequate and limited, therefore, the researchers on themechanics of beating switched their attention towards the physical theory of beating.Thus, the chemical theory of beating was followed by a physical explanation firstsuggested by Strachan (1926). He proposed that the strength of paper is due to themechanical entanglement of fibrils raised as a velvet-like pile on the surface of the fibers.According to this theory the increase of wemess is due to retention of greater amounts ofwater in the fibril pile. This pure physical theory, however, was an oversimplification inexplaining the theory of beating.14Campbell (1930) turned his attention on the very considerable forces acting on thefibrils as a result of surface tension during drying. He suggested that during beating thefibrils are more readily teased out by beating. Furthermore, when water is removed fromthe paper sheet these fibrils are drawn into contact with neighboring fibers. As the papersheet is dried hydrogen bonding occurs, to a much greater extent between fibers whenfibrils are exposed on the surface.The modem theory of beating is based upon a hypothesis put forward by Emerton(1957). Emerton (1957) suggested that in the early stage of beating or refining, theconstricting outer layers of the fiber (i.e., the primary and outer secondary wall) aredisrupted and in part removed (external fibrillation). The disruption of the outer layersallows the fibers to swell. As a result of this swelling and the simultaneous repeatedflexing by the beater and refiner bars, the fibers are internally delaminated (internalfibrillation). As a consequence of these effects, the bonds between successive coaxiallamellae of the middle secondary wall are, to some extent, severed. However, themechanism of delamination caused by either flexing or swelling has not been resolvedyet. The proposed theory of Emerton (1957) is probably substantially correct, however,the process of swelling and flexing of fibers is not clearly understood.The mechanism of delamination was originally suggested by Emerton (1957) andsupported by Scallan (1978). Scallan proposed that the delamination, induced byexcessive swelling is achieved by chemical means. In this way, the swelling could beachieved either by powerful swelling agents or after beating. The author (Scallan 1978)suggested that the breakage of one bond within the structure would result in an increasingstress on neighbouring bonds which would lead to the propagation and widening of theopening. This swelling would then result in complete and uniform tangentialdelamination. Once initiated in several places between radially adjacent lamellae, thedelamination would grow in preference to further debonding of the microfibril sheets.15He supported the idea that the delamination is solely possible by swelling and that it doesnot necessarily require mechanical action.Conversely, Atack (1977) described the mechanism of beating by mechanicalaction rather than sweffing by chemical means. He suggested that the effect of beating issimilar to a loaded rolling element passing over any deformable substrate. In addition tothis type of straining, he claimed that as the pressure pulse moves along the draped fibers,water is pumped through the honeycomb structure of the middle secondaiy wall by aperistaltic action. High pressure water forced into the ‘crack tips’ of the honeycombstructure would release fibrils from each other at points of attachment inducing the finalstate of coaxial delamination. The former mechanism would promote both external andinternal fibrillation and the latter internal fibrillation. The author concluded that theoperative stress level of both processes has led to a fatigue type failure mechanism.Furthermore, he emphasized that the reason for the promotion of both external andinternal fibrillation rather than cutting the fibers, is the presence of a water-film betweenthe fibers and fiber layers. This water-film dampens the mechanical stresses which areimposed by beating. Both Scallan (1978) and Atack (1977) tried to explain themechanism of delamination during refining either by chemical or mechanical means, butneither of them explained that the flexing itself could contribute to the process ofrefining. However, Atack (1977) identified the effect of water-film on enhancing bothexternal and internal delamination rather than cutting the fibers during refining.Tam Doo and Kerekes (1989) extended the fatigue-failure mechanism asproposed by Atack (1977). They suggested that any given specimen of a material that issubjected to a sufficiently high level of internal cyclic stress will gradually weaken andeventually fail through fatigue. Furthermore, the authors (Tam Doo and Kerekes 1989)quantitatively measured the flexing (i.e., fatigue type) effect on the flexibility of thefibers. The increased flexibility of the pulp fibers by flexing (cyclic loading) resulted inincreased sheet density and as a result the strength properties of the paper were also16increased. This observation strongly suggests that the flexibility is improved not only asa result of delamination but also as a result of mechanical action (i.e., flexing). Page(1989) also confirmed the effect of flexing on development of fiber flexibility. Heatthbuted the flexing effect to the development of cracks during refining which reducesthe rigidity of the fiber walls. Furthermore, Page (1989) emphasized the combined effectof swelling by chemical and mechanical means during refining.In summary, the original proposal by Emerton (1957) was improved by Scallan(1974) and Mack (1977), but the problem of fully explaining or understanding themechanism of swelling and flexing remained unresolved. Scallan (1978) and Atack(1977) looked at the problem of refining from a single theoretical point of view:delamination by either swelling via chemical means or flexing via mechanical means.The crack hypothesis of Page (1989) and correlation of flexing (cyclic loading) asmeasured by Tam Doo and Kerekes (1989) suggested that flexibility could be increasedbefore the final bond failure occurs, be that single or collective secondary bond breakage.None of these theories account for the mutual interactions of water as a swelling agentand flexing as a fatigue phenomenon in the refining process.2.3 Surface area of fibersA cellulosic fiber possesses a considerable volume of pores of diverse sizes andshapes. The surface area of the fibers is related to the previous treatment of them.Cellulosic fibers in the vicinity of polar molecules swell. When the fibers are dried fromwater, the large and medium size pores disappear (Stone and Scallan 1968, Stamm 1964).It was observed that the total surface area of dried fiber (i.e., both external and internalsurface area) scarcely exceeds the size of external surface area alone. The large internalsurface of the swollen structure can be retained only by replacement of water by lesspolar liquids, before drying. The properties of cellulosic fibers are related to the changes17in internal and external surface which take place when the fibers dry from polar liquids orgases. A knowledge of surface area, particularly the internal surface, is basic to anunderstanding of the behavior of cellulosic materials. The measurement of external andinternal surface area of fibers in virgin and recycled states could reveal the potential ofthe fibers for water uptake. On this basis the modification on fiber flexibility or bondingpotential could be assessed. According to the literature on recycling, very littleinformation is available on this parameter (Usuda, 1982).In this context, total surface area of the fibers was measured by BET’s equation(Brunauer, Emmet and Teller 1938) and the external surface area by the methoddeveloped by Harkins and Jura (1944). The internal surface area is estimated bysubtraction of external surface area from the total.2.3.1 Adsorption technique to measure the fiber surface areaGenerally, gases and vapors are adsorbed on the surfaces of solids, underappropriate conditions of temperature and pressure. It is presumed that the adsorption ofwater takes place, at first, in a layer one molecule thick to form a monolayer, andsubsequent adsorption results in the formation of multilayers (Brunauer et al. 1938). Ifthe amount of vapor or gas required to form a monolayer can be found and the area ofeach adsorbed molecule is known, the total surface area on which adsorption has takenplace could be measured. The precision of the technique strongly related to the type ofgas is used. For instance, the nitrogen molecules in the capilaries of 6 - 10 A could not bedetected by nitrogep gas. Probably, the proximity of opposing surfaces of capilaries willcause the nitrogen molecules to orient themselves vertically and thus either block someof the pores or not allow contineous coverage of both of the opposing surfaces. Analysiswith water as the sorbed gas has been described and calaimed to be a useful tool(Weatherwax 1974, Ostberg and Salmen 1991).18Calculation of the amount of vapor required to form a monolayer was developedby Langmuir (1918). The theory was extended by Brunauer et al. (1938) to multilayeradsorption. The equation derived by the authors (Brunauer et al. 1938) is based onkinetic interpretation, which effectively describes sorption isotherms of many types. Thebeginning of the long linear portion of the isotherm was interpreted by the authors ascompletion of the monolayer and the start of formation of multilayers. The BET theorypoorly explained the states of water in fiber-water interactions system (Hartley et al.1992). However, comparison of conceptually different approaches to measuring fibersurface area suggested that the BET sorption theory could provide an estimate of the totalsurface area (Weatherwax 1974). This theory widely used to measure the surface area ofvariety of materials including cellulosic fibers.The surface areas of the fibers in this investigation is determined by application ofthe BET (Bruauer et al. 1938, Brunauer, 1940) equation. Generally, the BET equation isexpressed as follows:(h)/M(1h) = 1/CMm + (C4)h/CMm [2.21where h is the relative vapor pressure, M is the amount of moisture adsorbed at therelative pressure h. Mm is the volume of vapor adsorbed when the entire adsorbentsurface is covered with a complete monomolecular layer of adsorbate, and C is a constantdepending on the heat of adsorption in the first adsorbed layer. Materials such as fiber,glass and charcoal give good straight line plots between the h/M(1-h) against Ii up to atleast the point where h=30%. The slope of the line is equal to (C4)/MmC and the Yintercept is equal to 1/(MmC). Therefore, the solid surface area could be calculated fromthe known area of one adsorbate molecule. The total surface area per weight ofadsorbent , is given by:=aNMm/Mw [2.3]19where a represents the effective area occupied by an adsorbed molecule (i. e., a = 12.5AZ), M is the molecular weight of adsorbate and N is Avogadro’s number (6.O2x1O).The heat of adsorption in the first adsorbed layer is given by:.c= - EL)/RT [2.4]where .t is a constant related to the amount of vapor already adsorbed in the firstmolecular layer, e is the base of natural logarithm, E1 is the heat of adsorption, EL is theheat of condensation of the vapor, R is universal constant and T is absolute temperature.It is evident that the expanded structure is largely retained by the solventexchange procedure, and that the total area of the swollen material may be as much as10-fold greater than the area that results after drying from water. The calculation of thesurface area using different equations derived from the adsorption isotherm has beenemployed by various workers. Weatherwax (1974), used the BET equation (1938), the“t’ procedure of Hagamassy et al. (1969) and the cluster theory of Zimrn and Lundberg(1953) to calculate the surface area from the moisture adsorption isotherm. The authorsfound a constant surface area over a large range of isotherms. The technique was appliedto study the beating effect on fiber surface area by many workers (Tie-qiang et al. 1989,Seborg and Stamm 1931, Campbell and Pidgeon 1930, Henry et al. 1988), but no workhas been reported in the case of recycled fibers.2.3.2 External and internal surface areaIn connection with the beating effect, the fiber surface is classified as:a) external surface area, the area available for contact and bonding with adjacent fibers;and b) the internal surface area, i.e., the increase in volume, generally termed swelling.20External fibrillation is generally measured by the nitrogen adsorption, silvering,permeability, heat of wetting and staining methods. The external surface area can beestimated by employing Harkins and Jura’s absolute method (1944). The advantage ofthe heat of wetting arises from the view that the same set of data developed forthermodynamic analysis of the fiber - water system can be employed to calculate theexternal surface area. In this process, the film-covered surface releases a quantity ofenergy proportional to the area of the clean solid (See also the Appendix B). Suppose,for example, that the liquid is water with a surface energy 118.5 erg/cm2, then the onlyenergy change involved is that due to the disappearance of the water surface of theadsorbed film. If the energy developed in the calorimeter, expressed in ergs, is nowdivided by 118.5, the quotient gives the area of the surface of the particle with itsadsorbed film of water. Thus, the value of the clean solid surface is given by the relation:= zHjsf,L)/ll8.S [2.5]where /XH(sf,L) is the heat of immersion of the sample equilibrated at 100% relativehumidity expressed in ergs per amount of solid. The surfaces obtained in this manner areconsidered estimates of the external surface area of the solid, since the capillaries arefilled with water during conditioning the samples at 100% relative humidity. Theinternal surface area of the solid is estimated from the arithmetic difference of the totaland external surfaces of the sample.212.4 Thermodynamics of fiber-water interactionsThermodynamics is concerned with energy relationships involving heat,mechanical energy, and other aspects of energy and energy transfer. The interaction ofwater and fiber (cellulose) is always accompanied by the evolution of.heat and a decreasein free energy of the system. Many aspects of the interaction can be treated by theclassical methods of thermodynamics.The thermodynamics of wetting deal with the entire spectrum of solid-liquidinteractions, ranging from the adsorption of a small amount of vapor by a relatively largeamount of solid, to the immersion of a small quantity of solid into a large amount ofliquid. An experimental investigation involving both of these processes can provideinformation on how much of the changes arise from the adsorbate and adsorbent,respectively. This is particularly important of the fiber-water system, in which dryingreverses part of the fiber properties.In brief, the differences in the free energy to the two systems at the same relativepressure, may be due to a difference in the fiber accessible surface area (irreversibleswelling), solid-liquid bonding strength (enthalpy effect), molecular ordering ofadsorbate and adsorbent (entropy) or a combination of these effects. Therefore, acomplete thermodynamic analysis of the fiberwater system can help in identifying theorigin of the differences between the two systems of virgin and recycled fibers.The interpretation of adsorption isotherms of protein and cellulose has beenapproached using thermodynamic functions by many authors (Bull 1944), Morrison andDzieciuch 1959, Hollenbeck et al. 1978, Argue and Maass 1935, Ostherg and Salmen1991, Stamrn 1957). This technique was also used to explain the structure of protein andcellulose (Bull 1944, Assaff 1944, Wahba 1950). The hysteresis of the cellulose sorptionisotherm was also investigated by this technique (Morisson and Dzieciuch 1959,Hollenbeck 1978).22Paper recycling is mostly done in water medium. Thus, studying the recyclingeffect by thermodynamic functions in conjunction with structural analysis by the XRDtechnique, fiber surface area measurement and enzymatic hydrolysis could assist furtherour understanding of the effect of recycling on paper properties.2.4.1 Free energy changesThe free energy of a system could be defined as the maximum energy freed in aprocess and made available for work (See also Appendix C).Through the adsorption isotherm the information related to both the differentialand integral free energy changes could be generated. The differential free energy isgiven byDF1 = RTIn(h) [2.6)where h is the relative vapor pressure, T is the experimental temperature and R is the gasconstant.The integral free energy change for water vapor accompanying the sorptionprocess is:=n1RTln(h) - RT (n1/h)d(h) [2.7]where n1 is the mole of water per 100 g of solid.Evaluation of the integral on the right-hand side of equation [2.7] can beperformed graphiclly from a plot ofn1/h versus h.The free surface energy of adsorption per unit of surface, H, can be obtained fromthe integral part of the equation [2.7] divided by the monolayer surface area (Banghamand Razouk 1937, Stamm 1957, Harkins and Jura 1944).23IhII= —= RT/E (n1/h)d(h) [2.8]Jowhere E is the surface area of the monolayer, y8 is the free surface-energy of the solid-vacuum interface and y is the free energy of the surface when the solid-vapor interfaceis constructed.2.4.2 Enthalpy changesHeat of wetting is explained in terms of surface-energy changes. When one gramof a clean solid is immersed in a liquid, it is accompanied by the evolution of a quantityof heat per gram of solid, which represents the total energy of wetting (See alsoAppendix D).Measurement of the heats of immersion of solids with varying moisture contents,renders the entire range of AH values.tH = - z&HI(Sf,L) [2.9]where AHI(s,L) and zHj(st,J are the heats of immersion of a dry clean solid and a solidsaturated at 100% relative humidity, respectively.The partial molal enthalpy change is obtained by estimating the slope of AHversus fl1.2.4.3 Entropy changesThe property of a system when energy transfer is not involved is generallyexpressed by the entropy of the system. Unlike for energy, there is no principle ofconservation of entropy. Hence, when all systems taking part in a process are included,24the entropy either remains constant or increases. In other words, no process is possible inwhich the entropy decreases, when all systems taking part in the process are included.Both the integral and differential net entropies can be obtained from the definingequations once the other thermodynamic variables are determined:AS = (AH - z\F)/T [2.10]AS1 = (AH1 - AF.j/T [2.11]The unit of entropy, in any system, equals the ratio of the units of energy and timein that system.2.5 Enzymatic Hydrolysis of Recycled FibersEnzymes have been used in analysis of biological systems for more than 125years. Because of their specificity, it was recognized early (Whitaker, 1974) thatenzymes could be used to detect and determine the concentration of minute amounts ofcompounds in complex biological systems. The availability of well-characterizedenzymes of high specifity in pure form has made enzymatic analysis more attractiveduring the last two decades. Detecting the characteristics of a substrate is closelyassociated with the availability of well-characterized enzymes.The field of enzymatic analysis is diverse. The analyses include determiningenzyme amounts for inborn errors of metabolism, nongenetically associated diseases,adequate heat treatment of foods, and in measuring soil quality (Whitaker, 1985).Enzymes were also used to determine quantitatively the specific compounds of asubstrate, activators, or inhibitors of enzymes. Isomeric configurations of compounds,the primary structures of complex molecules such proteins, neucleic acids, and25carbohydrates, the conformation of complex molecules, and the structures of cells andsubcellular organelles can be addressed by enzymatic analysis (Whitaker, 1985).This technique was also used to determine the characteristics of recycled fibers.Oltus et a!. (1987) treated different grades of virgin and recycled papers with cellulase.Pycraft and Howarth (1980) examined the effect of recycling on the hydrolysis rate.Howarth et al. (1983) investigated the effect of enzymes on the papermaking process ofcellulose fibers. Pommier et al. (1989) improved the drainage quality of recycled fibersby pre-treatment with enzymes. In the present study recent findings on the mechanism ofenzymatic hydrolysis of cellulose were employed to assist with the analysis of thecharacteristics of recycled fibers.3. LITERATURE REVIEW ON RECYCLING3.1 IntroductionThe practical effects of recycling on papermaking properties of fibers has beendiscussed thoroughly by different workers (Klye 1961, McKee 1971, Bovin et al. 1973,Szwarcsztajn and Przybysz 1976, Eastwood and Clark 1978, Howard 1990, Howard andBichard 1992, Marton et al. 1993), and an excellent review on this subject was publishedby Howard (1990). These investigations showed a dramatic loss of tensile and burstingstrengths on repeated remaking of low-yield chemical pulp (See also Fig. 14) into paper.Furthermore, recycled fibers become more brittle and break like match sticks whenbeaten (Williams 1980). It is thus predicted that, due to these limitations the usefulrecycled fiber content in paper should be limited and a surplus of recycled paper willbecome available that will need to be disposed of by incineration or in cogenerationfacilities, as suggested by Gotsching (1992).26In spite of the considerable amount of work on upgrading of recycled paper(Lundberg and de Ruvo 1978, Lindstrom and Carison 1982, Putz et al. 1989), onlylimited improvements have been achieved in this field. Therefore, theproblem ofupgrading recycled fibers is still unsolved. For instance, there was a patent issued in1954 to Schiosser (Hunger 1978) which described the addition of a series of solublesubstances to puips in order to suppress hornification (embrittlement) after drying. Thebest results were obtained with wetting agents, but they were found to workfor only onecycle (Hunger 1978).The refining process, which could constitute an efficient method for regeneratingthe papermaking properties of recycled puips, causes at the same time further worseningof their drainage properties. For instance, Szwarcsztajn and Przybysz (1976) attemptedto upgrade recycled papers by refining. They refined the recycled pulp during each cycletill the maximum breaking length was achieved. The fines content of the pulp (in spite oflosses in the consecutive formation processes) increased during recycling by about 70%(Szwarcsztajn and Przybysz, 1976). Such a result could severely limit traditionalmethods for regenerating the papermaking capacity of recycled puips byrefining. Thenegative effect of refining as a tool for upgrading of recycled fibers has been confirmedby many others (McKee 1971, Bovin and Teder 1973, Naito et al. 1983). Therefore,traditional methods of upgrading of recycled paper material are not adequate, and theneed for developing the necessary mechanical properties of recycled fibers still remainsan enigma. It will be necessary to develop and implement new systems which willupgrade recycled papers without worsening their papermaking properties.Understançling the causes of such inferior fiber quality in recycling is atthe heartof a number of key operations in upgrading recycled fibers. Such findingspoint the wayto probable effective methods that could bring about considerable improvements inrecycled paper quality thus diminishing the flow of waste streams tolandfills, and27thereby restrain excessive depletion of the forest resource by reducing the amount ofvirgin fiber required for maintaining paper quality for most grades.3.2 Postulated mechanisms for loss of strength in recycled paperThe review of recycling of chemical puips (Klye 1961, McKee 1971, Bovin et a!.1973, Szwarcsztajn and Przybysz 1976, Eastwood and Clark 1978, Howard 1990,Howard and Bichard 1992, Marton et al. 1993), clearly reveals that the losses in paperstrength are brought about by loss of bonding, or in other words, a reduction in theresistance of bonds to an applied force. This is in agreement with previous reports(McKee 1971). On the other hand, the loss of bonding, could be a function of twoparameters (Eastwood and Clarke 1978):i. Fiber flexibility*, andii. Surface condition.It is not yet certain why loss of surface condition or flexibility of fibers occurs duringrecycling. Furthermore, it has not been established which of the above factors inrecycling plays the major role in recycled paper strength loss. The following sectionselaborate on these two parameters.3.2.1 Fiber flexibilityThe effect of drying on pulp fibers has been reported by many workers (Jaymeand Hunger 1958, Robertson 1963). The consensus is that the swelling capacity of fibersis lost, and irreversibility of their swelling property is increased with the level andduration of drying. This irreversible fiber swelling property is called the loss in* Flexibility in this context refers to the wet-plasticity of pulp fibers. It is defined as the inverse of the product of themoment of inertia of the body and the modulus of elasticity of the material from which the body is made.28flexibility of wet pulp fibers, or “hornification”. The effect of drying on high-yieldmechanical pulp fibers is minor (Bovin et al. 1973, Howard and Bichard 1992). On theother hand, low-yield chemical pulp fibers show a progressively greater-loss in recoveryof associated water with the extent of drying (Klye 1961, McKee 1971, Bovin et al. 1973,Szwaresztajn and Przybysz 1976, Eastwood and Clark 1978, Howard 1990, Howard andBichard 1992, Marton et al. 1993). The mechanism responsible for hornification hasbeen the subject of a long debate and is not resolved satisfactorily, as yet. Irreversible pore closureThe irreversible changes of cellulosic cell walls upon exposure to drying hashistorically been attributed to the closure of pores and cracks (Thode et al. 1955, Jaymeand Hunger 1958, Stone and Scallan 1965 and 1968). Thode et al. (1955) reported thatthe hornification is caused by the irreversible closing-up of micropores and cracks duringdrying.Stone and Scallan (1965 and 1968) carried out comprehensive studies on theeffect of drying on the structure of the cell wall. Solute exclusion was used incharacterization of the structural changes in the cell wall. It was observed that the largeand intermediate size pores were reduced upon drying. This concept was also discussedby Jayme and Hunger (1958). Hornification by both groups were defined as irreversiblechanges in the capillary system of the fiber cell wall. However, many questions remainstill unanswered in the area of fiber science. For instance, no explanation is offered as towhy the cell wall pores are lost and do not open again when the fibers are rewetted(Lundberg and de Ruvo 1978). The effect of wood components, such as hemicelluloses,in the closing up of the pores, were not detailed comprehensively. Furthermore, thesolute exclusion technique suffers from some practical difficulties. For instance, thelower molecular weight polymer could be adsorbed easily on the fiber surface (Gray291978). Aside from this problem, molecules or particles in the size range 1000-10000 A°are not easy to handle (Page 1989). These shortcomings could also cloud the finalconclusions proposed for changes at the structural level. Cross-linkingThe auto-crosslinking hypothesis, as an origin of fiber homification duringrecycling, was supported by several workers (Back and Klinga 1963, Back et al. 1967).Back and his co-workers (Back and Klinga 1963, Back et a!. 1967) believe that naturalaging or heating in an acid environment (accelerated aging) of the paper mainly promotehemiacetal type cross-linking between cellulose and hemicellulose chains. Such bondscan restrict the swelling and make the fibers brittle.Lundberg and de Ruvo (1978), by examining the equilibrium moisture content ofcommercial never-dried bleached kraft pulp of 44% yield, concluded that, the hindranceto swelling, after drying, in the fiber was not due to occurrence of cross-linking or theformation of sites inaccessible to water. The change in swelling was not reflected bychanges in the equilibrium moisture content. Scallan and Tigerson (1991) observed that,for low-yield bleached kraft pulp the elastic modulus of the wet fiber is doubled duringrecycling. They attributed this increase to the increased hydrogen bonding between themicrofibrils during recycling. Cellulose cl,iain cleavagePaper acidity, which may be present in papers because of the processes orchemicals used in their manufacture, is a major factor in hydrolysis of cellulose(McComb and Williams 1981). Acid penetrates the open amorphous regions of the fiberand cuts the carbon-oxygen glycosidic bonds that link the glucose units in the cellulose30chain (McComb and Williams 1981). Oxidation, a reaction between oxygen and thecellulose unit, can break carbon-hydrogen bonds, as well as carbon-oxygen bonds(McComb and Williams 1981). These reactions simultaneously liberate the portion ofthe fiber plastisized by humidity, lower the overall degree of polymerizatioir and makethe fiber more fragile and more susceptible to breakage during subsequent beating(McComb and Williams 1981). McComb and Williams (1981) concluded that recycledfibers from alkaline paper behave more like virgin paper and can make better recycledproducts (See also Fig. 3).1•110 —0o 20—C0C 30—UC40-C50-0Co 60—00070 —80 —90-100 I4 5 6 7 8pH of wate extractFig. 3. Change in flexibility of fiber by acidity of paper. The fiber retains its flexibilitywhen it is a neutral or slightly alkaline environment (McComb and Williams 1981).31Stockman and Teder (1963) studied the effect of drying on the degree ofpolymerization (DP) of different chemical puips. They did not observe any significantchange on the DP of chemical puips on heating in the range of 70-140° C. The bleachedkraft pulp, however, was very sensitive to temperatures exceeding 1400 C.Other factors have been observed to influence chain cleavage. Burgess (1986)reported that tap water or calcium-sulfate washed fibers exhibited less degradation afteraccelerated thermal aging (70° C, 50% RH, 70 days) than distilled-deionized waterwashed fibers. However, it is not certain to what extent chain scission of cellulosecontributes to embrittlement of fibers at regular conditions of commercial recycling. Re-organization in the fiber (cell) wallReorganization and co-crystallization of cellulose chains during drying, as asource of hornification, was discussed by some authors (Ehmrooth 1978, Ingram et al.1974, Kulshreshtha et al. 1973, Morosoff 1974). Such a possibility initiates from the factthat, upon drying, bonding forces which develop, are sufficiently large and regular tounite two or more crystalline regions as one (Ehrnrooth 1978), thus resulting in aswelling restriction in dried fibers (See also Fig. 2). This implies that the possibleincrease in crystallite size and/or order in the fiber cell wall could be responsible for thehornification effect and consequently for brittleness and loss in flexibility of the recycledfibers. It is worth mentioning that the smaller the crystallite in the fiber wall, the largerwill be the surface area capable of hydrogen bonding and, therefore, the greater will bethe moisture sorption for a given weight fraction of polycrystalline material (Ingram et al.1974).Clark (1985) suggested that, when the fiber is dried, adjacent surfaces of celluloseand hemicellulose, previously separated, may come together. If parts of the areas coming32into direct contact match sufficiently well in composition and orientation, they could wellform additional crystallite zones and consequently restrict swelling on rewetting. Thisconcept was also suggested by Peterlin (1974) and examined by Ebrnrooth et al. (1978)on paper recycling, in order to evaluate the possible co-crystallization.of the fibers duringrecycling. The method which they developed, was acetylation by which recovery of fiberproperties was enhanced after drying. This method was assumed to prevent cocrystallization of fiber crystallites during drying. However, the method thus adopted wasindirect and did not allow formulation of accurate conclusions, because, acetylation athigh degrees of substitution could convert the cellulose to a hydrophobic material withconcomitant loss in bonding (Giertz 1978). Due to the possible induction of hornificationby structural changes in the cell wall during drying, a systematic investigation isrequired. In fact, removal of lignin and hemicelluloses from the cell wall should makethe fibers particularly vulnerable to such “secondary crystallization” to take place ondrying. Bond strengthA strong hydrogen bond is defined as a hydrogen bond in which both members ofthe H-bonded complex experience significant structural changes. Thus, when a H-bonded complex in its minimum energy structure demonstrates a structure not toodissimilar in isolated monomers, the bond is called a “weak” hydrogen bond. Accordingto this theory, weak hydrogen bonds are least firmly held and could be easily openedupon moisture uptake. Milichovsky (1990) suggested two or three types of qualitativelydifferent hydrogen bonds. He proposed that these different types of hydrogen bonds areresponsible for oriented (crystalline), less oriented, and unoriented (amorphous) zones innative cellulose. He suggested that the reason water causes cellulose to swell rather thandissolve it, lies in existence of strong hydrogen bonds in its oriented zones. However,33this does not mean that the water does not penetrate the oriented zones. Indeed, it ispossible that water molecules penetrate into the oriented zones without causing anysubstantial changes (Milichovsky 1990). Therefore, the ability to destroy hydrogenbonds with water will depend not only on their strength, but also on the nature of thehydrogen bonds. Higgins (1978) interpreted the irreversibility of bonding on recycling tooccurrence of stronger hydrogen bonds rather than reciystallization and/or cocrystallization. He proposed that the existence of weak and strong hydrogen bonds arethe basis of the whole question of recycling, and suggested that irreversible changesduring recycling were brought about by the latter type of hydrogen bonds.In general, elucidating the mechanical properties of paper, by existence ofdifferent types of bonding mechanisms, has not been discused thoroughly in the literatureand remains speculative. Therefore, more extensive work will be required to substantiatewhy certain hydrogen bonds in cellulose H-bonding are stronger than others.3.2.2 Surface Condition of FibersTheories in the field of surface condition of fibers are diverse. They originatefrom observations on fiber surface accessibility and extend to the mutual relation of fiberand water. The investigations focus on relatively few fundamental problems relating tothe surface condition of recycled fibers and fiber-water interactions. For instance, thepresence of accessible hemicelluloses on fiber surface, and their abundance in the cellwail, enhance both fiber-to-fiber bonding and wet-flexibility of the fibers. Disappearanceof these bonding agents, either through re-distribution of fatty acids on the fiber surfaceor inactivation of the hydroxyl groups by any other mechanism during drying, mayreverse the end use performance of the recycled fibers. Surface condition could also beaffected by induction of microcompressions or a diminished surface area. It is also34suggested that perhaps the formation of hydrophobic molecules are responsible for fibersurface inactivation. It should be established what happens to the surface bondingcapacity of the fibers during drying. Does the surface bonding capacity disappear fromfiber surface during recycling, or does the interaction with water create a shield aroundthe bonding agents? Hypotheses in this field are abundant, but little is known on the realnature of the problem. Hemicelluloses-loss effectThe role of hemicelluloses in the manufacture of a strong paper sheets has beenknown to papermakers for a long time. The precipitation of hemicelluloses on the pulpfiber surface in the final stages of alkaline cooking improves the paper-formingproperties of fibers. Precipitation of the hemicelluloses on the fiber surface also occursduring beating (Gallay 1958). Thus, some authors attributed deactivation of fiber surfacesto the loss of hemicelluloses from the fiber surface. Simultaneous hornification ondrying of fibers high in hemicelluloses was also observed (Jappe 1958).Hillis (1984) speculated that strength loss of wood is caused by hemicellulose lossesor conversion during high-temperature drying. Meiler (1947) observed on drying fibersat high temperatures that the solubility of the hemicelluloses in alkali decreasedsubstantially. These findings were also confirmed by Davis and Thompson (1964) byheat treatment of different woods. Stamm (1964) suggested that hemicelluloses in woodare converted to a furfural polymer at high temperatures which has less affinity toswelling. /Eastwood and Clarke (1978) proposed that the effects of recycling on flexibilityand surface condition of the fibers is possibly brought about by loss in hemicelluloses.They determined the loss of hemicelluloses quantitatively and considered them as anindicator for loss in fiber flexibility. Nevertheless, their results were not conclusive,35because of the complicated behavior of hemicelluloses in the strength development ofpaper (Cottral 1950).Indeed, it is possible to remove a certain amount of hemicelluloses from the fibermatrix without any detectable changes in the strength of paper (Cottral 1950). That is sobecause hemicelluloses in fibers have an optimal concentration (Cottral 1950) at whichmaximum paper strength is obtained. Higher than that amount will not contribute tostrength development, but it can inversely affect the strength properties of the paper(Cottral 1950) due to lowering the proportion of large molecule populations in the fibercell wall. Therefore, such complexities in the behavior of hemicelluloses in paperstrength development make formal conclusions, regarding their effect on the surfaceproperties of fibers, more difficult. Inactivation of the fiber surfaceIt was suggested that (Hancock 1964, Chritiansen 1990) aging and high-temperature drying cause the migration of extractives to the surface. The migration offatty acids, or some extractive component, was suggested as a main cause of surfaceinactivation in dried cellulosic materials. Back et al. (1967) suggested that the loss insurface accessibility of fibers upon recycling is due to re-distribution of fatty acidsinitially present in the pulp. Additionally, such an effect may also originate from rosinsize and resin acids which may have been added to the paper during the sizing operation.Nissan (1978) confirmed this hypothesis, and added that these reactions occur very faston the fiber surface, at paper drying temperatures.Troughton and Chow (1971) did not find any correlation between fatty acidconcentration and quality of wood surface. They suggested that fatty acids have acatalytic role in the inactivation process.36A critical review of the literature on over-drying of wood and fibers waspublished by Christiansen (1990). The review suggested that surface inactivation byheat-treatment or aging, caused basically by exudation of extractives to the surface, deactivated the fiber surface and prevented the surface from hydrogen bonding. The othermechanisms, such as reorientation of wood surface molecules and closure of large andintermediate pores in cell walls during drying, which reduce surface activation anddecrease the internal bonding points, were also confirmed.Reports (Howard and Bichard 1992) on mechanical pulps in contrast withchemical pulps did not show any distinguishable loss on bonding capacity of driedfibers. This does not coincide with reported data on wood. The disagreements could bedue to the conditions of drying or many other parameters involved in defining the pulpand the process of recycling. Furthermore, the question remains as to the significance ofsurface inactivation during drying. If the reason is the migration of extractives to maskthe fiber surface area, mechanical puips should be more susceptible than chemical puips.Therefore, larger reductions in fiber quality should be observed for mechanical puipsthan chemical pulps. This is not the case in recycling. It seems more plausible,therefore, to state that, recycling mostly suffers from changes in the fiber structure(stiffening) rather than changes in surface condition. MicrocompressionsIt was suggested that, recycling could cause microcompressions (Howard 1990), whichcould alter the surface condition of the fibers. This expectation arises because, the fibersshrink between 20-30% transversely and 1-2% longitudinally. Fiber length in paper hasbeen observed to shorten by as much as 20% during drying (Page and Tydeman 1961).According to literature citations (Page et al. 1985) sheets containing large proportions oflongitudinally compressed fibers were found to demonstrate superior extensibility, high37tear and impact resistance, but reduced elastic modulus. These properties alreadycontradict some of the findings reported on recycling. For instance, in recycling impactresistance, stretchability and elongation decreased and elastic modulus increased (McKee197 1, Koning and Godshal 1975). Furthermore, fiber length shortening has not beendetected in recycling (Jones 1990). Therefore, the possibility of changes in occurrenceand intensification of microcompressions in the fibers upon recycling must diminishconsiderably. There remains the question on permanency of microcompressions inrecycled fibers. Fiber-Water InteractionsFor a break through in the improvement of recycling a detailed knowledge offiber-water interactions is required. These interactions could be interpreted to dominatethe role played by the fiber surface in defining the structure and properties of the water,as well as influence of water in imposing a structure on the fiber (Morisson andDzieciuch 1959, Goring 1978, Caulfield 1978).Studies on the properties of liquid water revealed that water is not a homogeneoussolvent system, but rather, it exists in a dynamic equilibrium between the icelike* andnon-ice-like forms which are present in room-temperature water (Frank and Wen 1957).Indeed the cellulose surface is very sensitive to this equilibrium (Dobbins 1970) and canbe influenced profoundly by the addition of any third component such as additives.Thus, the reactivity of any surface agents on fiber interfaces depends on the mutual effectof water on the property of the agent. The subsequent section will concentrate briefly onexploring the significance of proposed theories on fiber-water interactions.* The term Ice-like” was used in describing the extent of hydrogen bonding in a liquid water at room temperature. It didnot necessarily imply that the associated water molecules have the tiidymite-like arrangement of ordinary ice.383. The electric charge of paper surfaces in waterMany authors (Verwey and Overbeek 1948, Strazdins 1976) discussed the problem offiber surface accessibility in terms of interfacial potential differences of fiber-watercompositions. Generally, most substances acquire a charge when immersed in-water andmigrate under an applied electric field. This is usually interpreted in terms of an ionicdouble layer adjacent to the adsorbing surface. The zeta potential is defmed as thepotential at the surface between the freely mobile liquid and the liquid firmly adhering tothe particle surface (Verwey and Overbeek 1948) (See also Fig. 4). According to thedouble layer theory, the thickness of the double layer should be reduced by the presenceof electrolytes. Consequently, the forces of repulsion will decrease and the potential forfiber surface adsorption, will increase. Furthermore, at higher repulsion forces, abulk liquidPotential at this surfaceIs called zetapotentlalA layer of adhered solvent molecules\ \\ \ N\\\\SURFACEFig. 4. Diagram showing the adhering solvent layer on surface when a surface issubmerged in a liquid.39significant portion of the added fillers and some of the fines will disappear from the wiresection and become part of the white water. The main drawback of this theory inexperimentation results from the vague location of the plane of shear. Generally, zeta-potential is useful in the case of massive particles with a well-defined impenetrablesurface, but loses a good deal of its meaning with swollen gels (Verwey and Overbeek1978). In addition, the electrostatic interactions are not the only force governing thefiber-water interactions phenomenon (Dobbins 1970). Changes in water structure due to fiber-water interactionsMutual effects of fiber and water were discussed by different authors (Goring1978, Dobbins 1970, Morisson and Dzieciuch 1959), but because of the complexity ofthe interactions, the subject is not well understood.Goring (1978) considers the problem of surface condition to be due to the changesin the hydrophobic bonding system in cellulose. It was suggested that in cellulose thereis a hydrophilic and hydrophobic bonding system and that perhaps changes in thehydrophobic bonding system plays an important role in formation of an irreversible fibersurface condition. Furthermore, by measuring thermal expansion of cellulose in differentstates (i.e., wet and dry), he concluded that a carbohydrate surface disturbs the adjacentbulk water in such a way that cellulose acts as a structure breaker (Goring 1978) (Seealso Fig. 5). This is actually the extension of Kauzmann (1959) and Nemethy andScheraga’s (1962) work on proteins. The result of calculations of hydrophobic bonding in40/OQ \c o__.Cellulose surfaceFig. 5. Schematic cross-section of broken hydrogen-bonded water near the cellulosesurface (Goring 1978).proteins confirmed that a considerable part of the free energy arose from changes in thewater structure, whereby the van der Waals interactions themselves contributed only partof the total energy of formation of the hydrophobic bond (Nemethy and Scheraga 1962).The role of hydrophobic bonds in fiber-water interactions on the fiber surfacecondition was also emphasized by Dobbins (1970). He suggested that the increase inflocculation of cellulose by rising temperature is only compatible with hydrophobicbonding, rather than hydrogen bonding which is an exothermic process. Thus, heconcluded that the addition of simple solutes to a cellulose-water system often produces41effects that are difficult to explain in classical terms of coulombic association betweenpositively charged ions, and the negatively charged cellulose interface. He suggested thatthere is a third class of ions that have hydrophobic character, which could explainswelling under neutral or alkaline conditions in terms of one unified mechanism.The problem with hydrophobic bonding in cellulose remains speculative andarises from unknown sources of those bonds. For instance, 35-45% of protein is made upof non-polar side-chains which have low affinity for water. In an aqueous environment,the non-polar groups adhere to one another forming hydrophobic bonds (Kauzmann1959, Nemethy and Scheraga 1962), whereas, the possible source of such hydrophobicbonds in cellulose is unknown.Nissan (1978) questioned the method of evaluation of the interaction of cellulosesurfaces with water by Goring. He proposed that, a higher thermal expansion forcellulose might have resulted from its higher degree of freedom due to breaking down ofits hydrogen bonding network in water.Caulfield (1978) supported the idea that water near a cellulose surface is restructured by the cellulose surface rather than destructed by it. He suggested that such amodel can not explain a fractional decrease in tensile modulus, which is constant up tothe fiber saturation point on adding water to a moisture content equivalent to about eightlayers of absorbed water (Fig. 6). Both authors (Goring 1978, Caulfield 1978) basedtheir hypothesis on Frank and Wen’s (109) “flickering cluster”* theory in aqueoussolutions.Milichovsky (1990) emphasized that the water molecules rather than the soluteare the actual swelling agent. He concluded that swelling which occurred on beating, is aresult of the changes in the composition of the phase interface between the pulp andwater medium rather than the structural changes in the fiber. This theory is based on the* The term is used to explain the co-operative nature of duster formation and dissolution in a liquid water at room-temperature, and it is governed by local energy fluctuation.42dipole nature of water molecules and on their two possible orientations in hydrationspheres. The enigma with Milichovsky’s approach to the problem of cellulose swelling isin understanding the possible causes of different orientations of water molecules on thesurface of the fiber. -\ \Cellulose surface,\ \ \ \Cellulose surfaceFig. 6. Schematic representation of the adsorbed water between two cellulosic surfaces(Caulfield 1978). The network was demonstrated as a distorted ice-like cluster withextension of hydrogen bonds from one surface to other.CUC80C0a043It now appears, that the bonding problems experienced in low yield chemicalpaper recycling are due to a number of physico-chemical factors, of which the fiber-waterinteraction may predominate. More explicitly, the change in surface condition could notonly be brought about by loss in surface accessibility of the fibers, but also by thechanges in the nature and organization of the hydrophobic and hydrophilic bonding of thefiber-water vapor system.Considerable work remains to be carried out in the area of fiber-water interactionsas it may hold, at least in part, the key to resolving the problems associated with freesurface inactivation in recycling and indeed papermaking itself.3.3 SummaryThe importance of this study arises from the view that the lack of scientific basisfor explaining the causes of strength loss (bonding capacity) on recycling of fibers,prevents papermakers from developing of an effective method for upgrading and, indeed,wide spread use of recycled fibers to avoid burning waste paper as “non-recyclable” orconverting it into pulp and paper “sludge”.This review suggests that strength loss in recycling is brought about by the loss inflexibility and/or changes in surface condition of the recycled fibers by emphasizinghornification. It is plausible for the cell wall of chemical pulps (low yield) that dryingperhaps creates new and more perfectly aligned lamellae through hydrogen bonds as incrystalline cellulose. However, little information is available on the cause of thesechanges. In other words, any explanation for the cause of strength loss requires anunderstanding and elucidation of structural evolution of the fibers and fiber surfaceinactivation during recycling. Furthermore, a comprehensive study of the surfaceinactivation of fibers requires the understanding of the phenomena related to both fibersurface and fiber-water interactions.444. MATERIALS AND METHODS4.1 Preparation of Pulp SamplesA fresh white spruce (Picea glauca) tree, 75 years of age from Williams Lake,BC was cut and transported to UBC. Care was taken to keep the wood in green state.Therefore, the transportation was arranged in a way to receive the log in two days. At thetime of arrival, the moisture content was measured to ensure that it was higher than fibersaturation point (FSP). Subsequently, the log was debarked and chipped. The chips werescreened, collected in plastic bags and stored in the cold room.The moisture content of the chips was 40%. The chips were pulped by the kraftprocess at the BCIT batch digester. About 21.18 kg (green weight) chips were cooked.The white liquor (i.e., NaOH and Na2S) charge was 34.90 L and the water added was27.22 kg. The ratio of liquor to oven dry wood was 4.5. The temperature was set at 165°C and pressure attained was 1068 kPa. The digester was completely computerized. Afteralmost 2 h the pulp was released from the digester, washed and collected in a doubleplastic bag. The pulp was then screened and stored in the cold room for future use.After screening, the pulp was ready for the bleaching stage.4.2 Bleaching Materials4.2.1 Preparation of chlorine.Bleaching and preparation of the chlorine and chlorine dioxide were done in theUBC wood chemistry laboratory. Chlorine gas was bubbled in distilled water byconnecting the tank through tubes to 4 L adsorption bottles connected in series. Theapparatus was set up in the fumehood. Bubbling was continued until the color of the45water changed to greenish after almost 4 h. The bottles were sealed and stored in thecold room. Before pulp chlorination, the chlorine concentration in the water wasdetermined in accordance with the CPPA (Canadian Pulp and Paper Association) J.22Pstandari4.2.2 Preparation of chlorine dioxideChlorine dioxide is explosive. So, it cannot be transported. Therefore, it isprepared in the paper mills. An apparatus according to Fig. 7 was built in the fumehood.The chlorine and Nitrogen tanks were connected through rotameters to a 1 L cylinder asshown in Fig. 7. The bottom of the cylinder containing technical grade_.—.--C,,C,,00[I(FumehoodJRotameter400 mLfminFig. 7. A sketch of chlorine dioxide preparation.46sodium chlorite (NaC1O2)was covered with glass wool. About 3 cm of the cylinder wasleft empty from the top. The cylinder was connected through a tube to a series of 4 Lbottles (filled with distilled water) at 40 C constant temperature. The flow rate of thenitrogen was adjusted to 400 mL/min by a rotameter and the chlorine flow rate wasconducted at 23 mI/mm. It required almost 4.5 to 5 h to make 4 L of chlorine dioxidesolution. The C102 concentration produced by this process was about 7.5 g/L. Thebottles were sealed and stored in the dark cold room at 40 C.4.2.3 Laboratory bleaching processThe bleaching procedure is normally chosen according to the pulp type based onthe target brightness. Generally, for low yield chemical pulps chlorine-based proceduresare used, while for high yield chemical and mechanical pulps, lignin-preservingchemicals such as sodium peroxide and sodium dithionite are commonly employed. Thesuccessive stages of bleaching are characterized by the appropriate chemicals andconditions. Laboratory bleaching is widely used to determine the bleaching response ofpuips under controlled laboratory settings. The unbleached kraft brown stock pulp wasbleached using a CEDED sequence. The condition for the various stages are summarizedas follows:a. Chlorination stage (C)Chlorine solution was titrated for concentration according to CPPA J.22P. Threesamples of 30 g oven dry pulp were sealed in plastic bags. The chlorine addition was 7%of oven dry weight of the pulp. The bleaching conditions were: 3% consistency, 20° C,and 60 mm. To ensure well mixing, the pulp was kneaded for about 2 mm. Thekneading was repeated every 15 mm. After 60 mm, the pulp was filtrated and residual47Cl2 and pH were determined. Finally, the puips were washed thoroughly with distilledwater.b. Extraction stage (E)The pulps were sealed in the plastic bags and conditioned for extraction.Subsequently the puips were extracted at 12% consistency, 74° C, for 2 h using 60% ofthe applied chlorine charge for each pulp. After 2 h the puips were filtered and washedwell with distilled water.c. Dioxide stage (D)Chlorine water and chlorine dioxide solutions easily lose chlorine in the gaseousform. Therefore, the concentration of the chlorine dioxide solution was measured again.The chlorine dioxide addition was almost 1% per 30 g oven dry pulp. The samples weredouble sealed in plastic bags and conditioned at 6% consistency, 74° C, for 3 h. The pulpbags were kneaded for about 90 sec every 15 miii. At the end of 3 h, the puips werefiltered and washed well with distilled water. After the chlorine dioxide stage (D), theextraction stage (E) was repeated at 12% consistency, 74° C, for 2 h. After filtration andwashing thoroughly, the last D stage was performed at almost the same condition asoutlined for stage 3.After the completion of the CEDED bleaching sequence, the yield wasdetermined. The freeness of the pulp was measured in accordance with CPPA C. 1standard method. Subsequently, the pulp was refined (beaten) in a PFI mill inaccordance with CPPA C.7 Standard for 12, 000 revolutions (i.e., to about 250 mL CSF).The characteristics of the pulp (yield, Kappa number and viscosity) were determined byCPPA G. 18 and CPPA G.24P standards respectively, and are summarized in Table III.484.3 Recycling procedureGenerally, Howard and Bichard’s (1992) method for recycling was followed.However, the white water was not re-circulated. According tothe observations ofSzwarcsztajn and Przybysz (1976), the fines themselves are also hornified duringrecycling. The role of the recycled fines, as agents in enhancing fiber surface area,diminishes considerably in recycling. Therefore, recycling of white water during theprocess of paper-recycling does not have any impact on the objectives of the presentwork and it would rather complicate the process itself.The samples were recycled five times on a laboratory scale. Therecycling wascarried out on both beaten and unbeaten fibers. Each time 48 g (OD) of the never-driedpulp was made into handsheets in accordance to CPPA C.4 standard and after pressingwere left in the conditioning room for 24 h at 50% relative humidity (RH) and 250 C.For example, a representative sample of the pulp, equivalent to 24 gof oven-dry sample,weighed out. It was mixed with 2 L of distilled water. (i.e., 1.2 percent consistency) anddisintegrated at 600 rpm. After disintegration, the stock was diluted to 16 L (0.15percent consistency) with water at 23 ± 2° C. 800 mL of the sample was pouredinto thebarrel of the standard sheet machine and the water level in thebarrel increased to themark. Then the stock was carefully stirred with the perforatedplunger and the waterdrained. Two pieces of standard blotting paper was laid on thesheet, blotted off withblotting paper and the pulp sheet pressed in accordance with CPPA C.4 standard. Thesheet plates were mounted on drying rings and the paper conditioned at RH = 50% andtemperature of 25 ± 2° C for air-drying. After 24 h the handsheets were collected andlabeled as virgin paper (cycle 0).For recycling, cycle (0) handsheets were soaked in distilled water overnight.Subsequently, the sheets were disintegrated according to CPPA C.6 standard and madeinto handsheets again to give cycle I handsheets. This procedurewas repeated 5 times.From each cycle some of the samples were stored in the cold room for subsequent49analysis of fiber crystallinity, fiber surface area and thermodynamical properties (Seealso Fig. 8). No beating was imposed between the cycles. This procedure was repeated 5times (Fig. 8). In order to test and make sure that the recycling modified the strengthproperties of the paper, for each cycle 10 handsheets were tested according to CPPAstandard methods (CPPA D.6H, CPPA D.8 and CPPA D.9) for strength properties (i.e.,tensile, tear and burst respectively).50Fig. 8. A flow chart of the experimental procedure.514.4 Fiber Length (FL) AnalysisThe FS-200 fiber length analyzer consists of a capillary tube through which anaqueous suspension of the fibers is passed. A light source is located on one side of thecapillary, and a detector is positioned at the opposite side. As a fiber passes through thecapillary tube, its image is projected onto the detector with the aid of the light source andsystem optics, providing information from which the lengthwise dimension of the fibermay be calculated (See also Fig. 9).Analyzer‘-[ComPuter 1Fig. 9. The block diagram of the FS-200 fiber length analyzer.52Fiber length distributions of both virgin and recycled pulps were determinedusing the Kajaani FS-200 fiber analyzer. The pulp suspension was poured into a flaskand sealed with a glass stopper. The flask was shaken manually until the fibers dispersedcompletely. Care was taken, by gentle stirring, to ensure that no fiber cutting occurred.The recycled samples were soaked in water for not less than 4 h prior to disintegration.Two measurements of the same sample were made by the FS-200 analyzer. Theconsistency of the suspensions was such that a minimum of 30,000 fibers were countedfor each sample.4.5 Molecular Weight Distribution (MWD)The samples were thcarbanilated according to the method suggested by Kossler etal. (1981) and Schroeder and Haigh (1979). Accordingly, 0.1 gram of the sample wasplaced in 500 mL of phenyl isocyanate. The flasks were then sealed with glass stoppersand placed in an oven at 80° C for 12 h. Completion of the reaction was monitoredvisually (i.e., clear light yellow solution) in order to prevent degradation duringtricarbanilation. This control was critical, because tricarbanilation times of differentfibers were different and depended on the pre-history of the samples. The tricarbanilatedcellulose was recovered according to the procedure suggested by Kossler et al. (1981)rather than by evaporation of the reaction solvents as suggested by Wood et al. (1986).All the samples were tricarbanilated in duplicate. The tricarbanilated samples were driedat 60° C overnight and then solubilized in tetrahydrofuran (THE) (2.5 mg/mi). The dry(60° C) tricarbanilated samples dissolved very slowly in THE. Freeze drying of thetricarbanilated samples prior to dissolution in TI-IF accelerated the process. Extraction ofthe thcarbanilates with methanol in a Soxhiet extractor completely removed the byproducts (N, NT diphenyl urea and methyl phenylcarbamate, and oligomers ). The byproducts were originally identified by Wood et al. (1986). These by-products can53produce a large peak after an elution volume of 32 mL thus disturbing the finaldetermination of both molecular weight distribution and polydispersity (Nazhad et al.1994). The samples were analyzed by gel permeation chromatography (GPC).The GPC of the tricarbanyl derivatives was carried out on a Spectra-PhysicsSP8810 liquid chromatograph. Samples of cellulose tricarbanylate in THF were filteredthrough a Teflon membrane with a pore size of 0.45 I.tm and analyzed using a series offour TSK-GEL (type H) columns. THF was used as the eluting solvent at a flow rate of 1mL/min. The solute concentration in the eluent was detected by a UV spectrophotometer(Spectroflow 757) at the wavelength 235 nm. The signal from the detector was fed to theintegrator SP 4229 for peak integration and plotting the cummulative and differentialdistributions of molecular weight. The number and weight average DP of the cellulosewas obtained by dividing the molecular weight (MW) of the tricarbanilated polymer (i.e.,both number and weight average DP) by the corresponding MW of the tricarbanylderivative of anhydroglucose (DP = MW/519). Each sample was analyzed three times byGPC.Because of the lack of commercially available standards of cellulosetricarbanylate, the GPC carbanilation curve (i.e., the correlation of elution volume withmolecular weight) was generated from the elution profile of polystyrene standards withnarrow MW distributions (Appendix F). The following equation was used.InM = [(1 + cL)1nM + 1n(K/K)]/(1 +z) [4.11where M and M are the MW of cellulose tricarbanylate and polystyrene, respectively.The Mark-Houwink coefficients used in the present analysis for polystyrene in THEwere: K = 1.18 x lO- and o = 0.74, and for cellulose tricarbanylate in THE, K = 2.01x 10-s and (Xc = 0.92, as reported by Valtasaari and Saarela (1975).544.6 Crystallinity and Crystallite size4.6.1 D 5000 DiffractometerAn automated, Siemen’s diffractometer system equipped with a carbonmonocbromator was used for the virgin and recycled fiber crystallinity analyses. Thediffractometer was controlled by a computer. The diffractograms can be displayed on aprinter attached to the computer. Figure 10 demonstrates block diagram of the SiemensD 5000 diffractometer system.fl sil apLmIIIFig. 10. Block diagram of diffractometer in D 5000 mode.55Generally, the emitted radiation from the line focus of the x-ray tube is diffractedat the sample and recorded by the detector. The sample rotates at a constant angularvelocity such that the angle of incidence of the primary beam changes while the detectorrotates at double angular velocity around the sample. The diffraction angle (2) is thusalways equal to twice the glancing angle (8). Each time the Bragg condition is satisfied,the primary beam is reflected from the sample to the detector. The detector and theconnected measuring electronics measure the intensity of the reflected radiation.Diffraction patterns are obtained in this way. In order that the diffracted radiation can befocused when it hits the detector, the whole effective sample surface should actually beon the focusing circle. The diffractometer beam path of the system is demonstrated inFig. 11.FocusX-ray tubeAperturediaphragm Scattered-radialdiaphragmaDetectordiaphragml Focusing circleSampleDetector-.easuringcircl. —.fFig. 11. Diffractometer beam path in /29 mode.564.6.2 Sample Preparation for X-Ray Diffraction AnalysisThe samples, which included unbeaten and beaten (never-dried) and Cycle I - Vpuips, were prepared as follows. The virgin (unbeaten and beaten) pulps were dried bythe critical point drying (CPD) method. The advantage of the method is that the fibersretain their superstructure almost intact. Then the samples (i.e., unbeaten, virgin andCycle I-V puips) were ground in a Willey mill to pass through a 30 mesh screen. Fivesamples of 0.15 g oven-dry weight were prepared for XRD analysis from each pulp.Great care was taken to keep all the samples at equal weight and thickness. Then, thesamples were placed in a mold of 2.2 cm x 4.5 cm x 0.4 cm and pressed for 5 mm under344 kPa pressure by the press. The thickness of the wafers was about 0.4 cm. Gooddistribution of the sample material was important and consequently equal thickness of thewafers was a major criterion in sample preparation. The pulp samples were analyzedalmost under identical conditions. Five sheets were prepared from each sample and thesamples were sealed in a plastic bag. One of the virgin paper handsheets was re-slushed,disintegrated, drained on a standard sheet machine and pressed. This procedure wasrepeated 10 times. This handsheet set has thus been recycled 10 times withoutexperiencing drying as used in the standard handsheet making process.The x-ray diffraction of each set of samples was recorded using a Siemensdiffractometer equipped with a D 5000 rotating anode x-ray generator. The wavelengthof the CuJKcx radiation source was 0.154 nm, and the spectra were obtained at 30 mAwith an accelerating voltage of 40 kV. Samples were scanned on the automateddiffractometer between 5 to 40° of 2 e (Bragg angle), with data acquisition taken atintervals of 0.04 for 1 second. The diffractograms of the virgin and recycled fibers wereanalyzed by a peak resolution program.574.6.3 XRD data analysesA peak resolution program was used to calculate both the crystallinity index andfull width at half maximum (FWHM) of the 101 and 002 peak of the traces. FWHM isinversely proportional to the crystallite size of the fibers. This program allowed theresolution of the diffractograms into the contribution of each of the diffraction planes.The Voigt function provided a better fit of the x-ray diffraction patterns. The separationof overlapping peaks is based on the assumption that the intensity curves of thereflections have the shape of the Voigt function. This function separates thediffractogram pattern of native cellulose substrate to major diffraction planes (i.e., 101,101, 021, and 002 planes). The program also incorporates an iterative procedure whichensures efficient curve fitting. Generally, the Voigt function resulted in the best fit of thex-ray diffraction patterns and is routinely used for the determination of both thecrystallinity index and crystallite dimensions of polymers or metals (Krassig 1984). Theprogram has a conversational dialog that prompts the user for necessary informationabout any particular calculation. The information is basically generated after the processof curve fitting is completed and a desirable peak resolution is achieved (See alsoAppendix A).4.7 Water Vapor SorptionSamples of virgin and recycled fibers were collected during the recycling processas reported previously. Soda-boiled cotton was used as control. The cotton sampleswere subsequently dried with CPD technique and conditioned at the appropriate RH andtemperature.Samples of virgin (beaten and unbeaten), recycled pulp and cotton fibers werethen placed in desiccators with the humidities varying from 5 to 100%. The cottonfibers were used as control. The relative humidity inside the desiccators was controlled58by varying the sulfuric acid concentration (Table I), based upon the relationshipestablished by Stokes and Robinson (1949).Table I Relationship of sulfuric acid concentration withrelative humidity (H) at 25° C* Density of sulfuric acid is 1.83 18 g/cm3.The desiccators were placed in a two-level plexiglass chamber together with ananalytical balance. The first level of the chamber was filled with certain salts in order tomodify the relative humidity of the chamber (Table II).59Table II Relative humidities over saturated salt solutions at25° C (Browning 1967).This salt solution was prepared to match a given relative humidity inside the desiccators.The main purpose of this instrumentation was to increase the accuracy of the process asmuch as possible. The whole apparatus was set up in the conditioning room with atemperature of 25°±2 C (Fig. 12). Water vapor sorption was followed by the gravimetricmethod. All weighings were performed on an analytical balance with a resolution of 0.01mg. The weighings were made at 3, 7, 15, 20, 25, and 30 days.60j II ICIII room-JAll measurements for sorption were performed in duplicate. Two sets of sampleswere prepared for the investigation including three samples for each set. The first set ofsamples, including the virgin (unbeaten and beaten) pulp and cotton, were conditionedfor water adsorption. The virgin (unbeaten and beaten) fibers were dried by differentBalanceAcidSalt Salt SaltFig. 12. Block diagram of apparatus which was developed for conditioning the fibers atdifferent relative huniidities with a constant temperature (i.e., 25±2° C).61methods (i.e., freeze drying and critical point drying). The second set, includingunbeaten and beaten fibers, were conditioned for the desorption process. For the latterset, the fibers were beaten to two different freeness degrees (i.e., 6000 and 12000 rev.).Samples of cycle I-V were also conditioned in the chamber at different relativehumidities (RH). The relative humidity of the chamber was controlled by adjusting thewet - and - dry bulb temperatures. The sorption data was collected at relative humiditiesof 35, 50, 60, 80, and 90%. The data around 35% and after 80% fluctuated widely whichmade it difficult to keep the temperature constant for a given relative humidity. Thetemperature was to be kept constant at 25±2° C during collection of the sorption data.The conditioning chamber did not allow collection of data at less than 35% relativehumidity. Therefore, data collected in the conditioning chamber could not be used forthe analyses in the present investigation.The sorption ratio along with the percentage of the amorphous component of thecell wall was deduced from the sorption data. The sorption ratio (SR) is defined as theratio of the water-vapor sorption of a cellulose sample to that of cotton under the sameconditions (Valentine, 1957). The fraction of amorphous cellulose (em) is then obtainedfrom the quotient of the sorption ratio and completely accessible cellulose.4.8 Measurement of the Heat of WettingThe heat of wetting was measured by a PARR 1451 solution calorimeter (Fig.13). The calorimeter can measure heats of reaction of many different systems. In thisinvestigation the heat of wetting of the fibers was measured by solution calorimetry. Theheat of wetting as measured by this technique assisted in calculating the enthalpy of thefiber-water vapor system and also was used as an indicator for accessible surface area ofthe fibers.62GlasscellFig. 13. A cross-section of the PARR 1451 solution calorimeter.The PARR 1451 solution calorimeter consists of a glass Dewar reaction chamberwith a rotating sample cell, a thermistor probe and a specifically designed temperaturemeasuring bridge, all assembled in a compact cabinet. Temperature changes are plotteddirectly in Celsius degrees using an accessory strip chart recorder. At the start of the test,Glasspush rodGlassStainlessair canTeflonSampledish63distilled water (100 mL) is held in the glass Dewar while the fiber (ball milled) is held ina sealed glass rotating cell which is immersed in distilled water (Fig. 13). The cell wasloaded with 0.4 g of the sample each time and allowed to rotate for 10 to 15 minutes untilthe temperature of the sample equalized with that of the distilled water in the glass Dewarflask. The system comes to equilibrium in 10 to 15 mm with only a slight temperaturedrift from the heat of stirring and from any heat leaking into or out of the calorimeter.After the equilibrium is reached the push rod is depressed to drop the contents of the cell(pulp fibers) into the surrounding water. The reaction then proceeds to completion undervigorous stirring action of the rotating cell. The temperature rise was measured (ATe).The heat generated by the reaction is calculated from the temperature rise of thecalorimeter and the heat capacities of the various components in the calorimeter (e =116.57 calJT) and normalized to the amount of the sample used.The integral heat of adsorption (integral enthalpy) is difficult to measure by theadsorption approach. Therefore, the integral heat of adsorption was also calculated fromthe data collected from solution calorimetry. The heat of wetting of a sample containinga certain amount of moisture is measured and subtracted from the heat of wetting of theoven dry sample (All = H- Hmot). From a series of determinations, the integral heatof adsorption from dryness to any desired water content can be estimated. Thedifferential heat of adsorption is found from the slope of the curve of integral heat ofadsorption versus moles of water per 100 g of fiber.4.9 Scanning Electron Microscopy (SEM)The virgin (unbeaten and beaten) and recycled samples were dried by the methodof critical point drying (CPD). The procedure reported by Sachs (1986) was followed.The method is based on dehydrating the fibers by solvent exchange with ethanol which isfollowed by exchange of the dehydrating solvent by a transitional liquid such as carbon64dioxide. Subsequently, by increasing the temperature (i.e., 400 C) and pressure ( 35kPa), the transitional liquid is converted to the gas wherein the densities of the liquid andgas of the fluid are identical. Consequently, the liquid in the cell wall transforms fromthe liquid state to the gaseous state. The replacement of the liquid by gas eliminates thecreation of tension forces between the fibrils/lamellae when the fibers are dried from thenever-dried state.4.10 Determination of Thermodynamic Properties of FiberMeasurement of the energies and amounts of sorption of water vapor by virginfiber, when compared with those on recycled fibers, have led us to detail the nature ofirreversible swelling of the dried fibers.The internal and external surface area of the fibers were calculated by employingthe BET (1938) equation and absolute method of Harkins and Jura (1944) Eqs. [2.3] and[2.5].The adsorption isotherm (Fig. 17) was used to calculate the differential andintegral free energies, according to Eqs. of [2.6] and [2.7]. Equation [2.8] was used tocalculate the surface-energy per unit area.The heat of wetting data, which are obtained by the calorimetric method, are usedto calculate the integral enthalpy changes AH, by using equation [2.9]. The differentialenthalpy change is obtained by estimating the slope of All versus nj.Finally, the integral and differential entropies are calculated from Eqs. [2.10] and[2.11].65411 Enzymatic HydrolysisEnzymatic hydrolysis of cellulosic substrates was performed using the CellucastCellulase preparation (Novo Nordlisk, Denmark) supplemented with the 3-glucosidasefrom Novozym. Enzymatic activities were determined in the cnzyme mixture asdescribed previously by Ghose (1986) and Chan et al. (1989). The amount of enzymesused for hydrolysis was adjusted to a final activity of 8.5 FPU (filter paper units) and23.8 CBU (cellobiase units) per gram of cellulose.The substrates were hydrolyzed in duplicates at 2% (W/V) solids concentration in50 mM sodium acetate buffer at pH 4.8, containing 6 mg mL-1 of tetracyclin as apreservative, at 45° C and stirred at 150 rpm. The release of soluble sugars duringhydrolysis was monitored by HPLC, using an HPX-87H column (Bio-Rad) (Schwald etal. 1988). Glucose yield was defmed as the percentage of available cellulose at time zerowhich was released as glucose after the indicated incubation time.Partially hydrolyzed residues were obtained by stopping the hydrolysis after 4 h.After the mixture was cooled in ice water, the residue was isolated by filtration. Theresidue was washed extensively with distilled water to remove most of the non-adsorbedenzymes and soluble sugars. The residue was then resuspended in water, collected byfiltration and stored at 4° C for subsequent analysis of fiber length (FL), molecularweight distribution (MWD), scanning electron microscopy (SEM), and crystallinity.665. RESULTSRelevant characteristics of the pulp used in the present work are reported in Tableifi. Figure. 14 demonstrates changes in the physical and optical properties of the fibersas induced by recycling. According to the Fig. 14 both tensile and burst strengthsTable ifi. Characteristics of the white spruce kraft Pulp (unbleached and bleached(CEDED)).Yield K V -ce1 emi*ceS.‘(%) //Uzthlèhd 4620 2 ‘ I 1140 470Bleached 4400 xLd 17 8800 1200K: Kappa number,V: Viscosity (cp),ot-cel: Alpha-Cellulose (%)Hemi-cel: Hemicelluloses (%)nd : Not determineddecreased by up to 30%, whereas the tear strength increased up to 40%. The main drop instrength properties of the paper (i.e., 12-17%) is observed in the first Cycle (Fig. 14).This is in agreement with the data reported by McKee (1971) and Howard and Bichard(1992). The scattering coefficient, which is an indicator of the non-bonded area,increases with the drop in tensile and burst. (Fig. 14). Moreover, fiber strength asmeasured by zero-span tensile strength was moderately affected (drop of 3.8% after the670C.tUa£0.Fig. 14. Changes in properties of the handsheets (made from beaten bleached kraft) dueto recycling.fifth cycle) during the course of recycling (Fig. 14). As it is demonstrated in Fig. 15 andTable IV, the fiber length profiles and average fiber lengths are almost the same for thebeaten, unbeaten and recycled puips. A slight increase in the average fiber length, for therecycled pulp, may be induced by the loss of the fines. Distinct loss in tensile strength ofthe paper with a minor loss in fiber strength (if there is any) led to the conclusion that, theinferior strength quality of recycled papers was induced by loss in potential bonding ofNumber of cycles68the fibers. However, this observation is not new and it was reported asearly as 1971 byMcKee.0IFig. 15. Comparison of fiber length (FL) distribution of unbeaten, beaten and recycledfibers.0 0.6 1.2 1.8 2.4 3 3.6 4.2 4.85.4 6Fiber length (mm)69Scanning electron microscopy (SEM) was employed to investigate the visiblecharacteristics of the recycled fibers. A comparison of the SEM micrographs of unbeatenand recycled (initially beaten) fibers (Figs. 16A, 16B, 16C, 16D) demonstrates that,recycling diminishes the difference between unbeaten and beaten fibers. The externalfibrillation induced on beating mostly disappeared after the first cycle (Fig. 16C).Indicating that one of the parameters (i.e., external fibrillation) which could assist thestrength development of paper which is induced by mechanical treatment, is lost byrecycling.Table IV. Fiber length (FL) analysis of virgin (unbeaten and beaten) and recycled puipsSn:A(Ave):L(Ave):M(Ave):Sample nameArithmetic (average)Length-Weighted (average)Mass-Weighted70Fig. 16. SEM micrographs of (a) unbeaten and (b) beaten pulp fibers.71100L-MFig. 16. SEM micrographs of (c) cycle I and (d) cycle VI pulp fibers.72To characterize further the effect of recycling, the accessibility of the fibersurface to water sorption was investigated (Figs. 17 and 18).24Cycle I22 14,4,20I18 4,16 /4— 4 4,.4 I14 Beaten pulp / /o 1o 4.. 1V 12 Unbeaten pulp‘ /10.4. 4.——7__ Cttcrn.4.....8———4• .)I4••‘,..ie20 i I I I I I I I I I I I I I I I051015202530354045505560657075808590Relative humidity (%)Fig. 17. Comparison of the adsorption isotherms of unbeaten, beaten and recycled fibers.Water sorption assisted by BET analysis is a useful technique and has been proven to beapplicable for assessment of the fiber surface area (Table V). The water sorption values73for both unbeaten and beaten fibers are shown in Fig. 17. As seen, the adsorptionisotherms are identical for both unbeaten and beaten fibers, while they differ for therecycled fibers (Figs. 17 and 18). This suggests that both unbeaten and beaten samplespossess an equal fiber surface area (384 m2/g). On the other hand, the recycled fibersbehaved differently during the sorption process. Their moisture uptake potential wasreduced, which indicates that the total fiber surface area (i.e., internal and externalsurface area) was diminished by drying ( about 238 m2/g). The subsequent recyclingfurther exacerbated the sorption power of the fibers (Fig. 18).,n.1 &1 &14EE81201oa6Cycle I (desorptlon)Cycle V (desorptlon)Cycle V (adsorption)Cycle I (adsorption)I I I30 40 60 70 80Relative humidity (%)Fig. 18. Comparison of sorption isotherms of cycle I and cycle V.TableV.Surfaceareaforvirgin(unbeaten andbeaten) andrecycledfibers.Vb:Virginfiber(beaten)V11b:Virginfiber(unbeaten)Cf:CycleISn:SamplenameC:AdsorptionenergyTotal surfaceareaMm:CapacityofmonolayerExternalsurfaceareaInternalsurfaceareaLHBET:BondingenergybetweenwaterandfiberbyBETapproach(KcaL’mole)BondingenergybetweenwaterandfiberbyCalorimetryapproach(Kcal/mole)n1:Moleofwaterper100gofsolidSD:Standarddeviationofexternalandinternalsurfacearea.75To evaluate the behavior of beaten fibers by the sorption technique, three never-dried samples were also tested for desorption. The sets included one unbeaten and twobeaten (i. e. 6000 and 12000 revolutions respectively, by PFI mill) samples. Refining upto 12000 revs, by a PH miii is a drastic mechanical treatment. of the fibers. Nodifferences were detected in the desorption isotherms of these three puips (Fig. 19).20154-4-8 10z1;;5aUnbeatenBeaten (6000 & 12000)0 40 ioRelative humidity (%)80Fig. 19. Desorption isotherms of unbeaten and beaten (6000 and 12000 revs.).76This experiment further confirmed the previous observation that the refining/beating doesnot alter the sorption potential of the fibers.To understand further the effect of the drying process and precision of thesorption technique, the water moisture sorption of freeze-dried an. critical point-driedpuips were compared in Fig. 20. The adsorption isotherms of the fibers dried by differentdrying techniques showed about 20% moisture difference in the range of 20-70% relativehumidity (See also Fig. 20). This is a considerable difference between the moisturesorption potential of the fibers which were dried by two different techniques. Originally,both techniques were used by different laboratories to preserve the original structure of2422201816C) 148121086420 10 15 0 25 30 5 40 45 50 55 60Relative humidity (%)Fig. 20. Comparison of critical point drying and freeze drying methods with sorptionpotential of the fibers at 25° C.Unbeaten (critical pointBeaten (critical point dried)Beaten (freeze dried)Unbeaten (freeze dried)6570758085909577the fiber wall. More explicitly, the structural differences between the fibers which weredried by critical point drying and freeze-drying technique should be minor. This suggeststhat a marginal difference in the drying condition can be reflected in distinct differencesin the accessibility of fiber surfaces. Furthermore, it is suggested that the sorptionisotherm measurement is a sensitive technique and has the potential to detect minutevariations in sorptivity of the fiber. Thus, the identical responses of unbeaten and beatenfibers did not initiate from poor detectability by the technique. This observation suggeststhat the poor sorptivity of recycled fibers is induced by loss in accessible surface area forwater.35,30‘-4.25.51510Cycle IUnbeaten[J5.Beatenr’0I I I I I I I I I2 4 6 8 10 12 14 16 18 20 22 24Moisture content (%) 26Fig. 21. Comparison of the heat of wetting of unbeaten, beaten and recycledfibers.78The calorimetric measurement results for the samples also followed the sametrend, as shown for surface measurement and water vapor sorption processes (Fig. 21).The behavior of both beaten and unbeaten fibers was almost the same at the precision ofthe experimental procedure but considerable differences were observed for the recycledsamples (Fig. 21).These observations imply that the total accessible surface area of the unbeatenand beaten fibers are the same, and both diminished when the fibers were recycled.C’,C’,0>C0.5.00Elutlon volume (mL)Fig. 22. Comparison of MWD of unbeaten, beaten and recycled fibers.79The interdependence of structure and properties of the fibers have always been ofgreat interest to papermakers. Krassig (1984) proposed that the physical properties andchemical reactivity of fibrous cellulose are not only dependent on the chemicalconstituents, but are also affected by the spatial arrangement of the. macromolecules inthe “architecture” of the fibers. Thus, the molecular and supermolecular structure of thefibers were investigated.In Fig. 22 the cbromatograms of virgin (unbeaten and beaten) and recycled pulpsare superimposed on each other. Additionally, the DP of the samples remained constantduring the course of recycling (Table VI). Polydispersity was also not greatly affected byTable VI. Degree of polymerization (DP) of the virgin (unbeaten and beaten) andrecycled pulpsSn =SamplenameDPw = Weight average DPDPN = number average DPP = Polydispersity80recycling. The chromatograms of molecular weight distribution do not support anychanges in the population of high and low molecular weight chain fragments in thesamples due to recycling. Therefore, it was concluded that, the inferior properties ofrecycled fibers are not induced by cellulose chain cleavage. -Superstructural arrangements of recycled fibers were examined by XRDtechnique. The diffractograms of unbeaten and beaten fibers were overlapped. Thecomparison of virgin and recycled fibers suggests distinct differences betweencrystallinities of the respective samples (Fig. 23 and Table VII). Furthermore, theFig. 23. Comparison of the diffractograins of unbeaten, beaten and recycledfibers.0,I25Bragg’s angle (2theta)81data indicate that each cycle in the process of successive recycling adds a new incrementof ordering in the fiber wall (Fig. 24). An increase of almost 9% in crystallinity after thefirst cycle is followed by about 1.2% increase in crystallinity for the successive cycles.This suggests that a major cellulose reorganization (crystallization) occurs in the firstcycle. The decrease in the full width at half maximum intensity (FWHM) of the 002peak from 2.9 to 2.57 is due either to an increase in crystallite size or crystalliteperfection due to recycling. Whichever the case may be, the swelling potential of thefibers will be reduced due to the higher re-organization of the fiber wall.Moisture regain was used as an indicator for the noncrystafline fraction of virginand recycled fibers. During the course of the present work, it was observed that theamorphous fraction decreased by almost 22% after the first cycle (Table VII).Table VII. Crystallinity information on virgin (unbeaten and beaten) and recycled pulps.Fw:11M002 Or! FM1flMs’!: degite (°k) SD &grcc SK Fm, Ünbeatcn 2 018 68 85 0+50 190 1 OA6IBeaten 2S178 6885 038 290 16 061VycleI 1.868 7300 068 2.57 123 047Cycle!! 1.770 xxi nil nil ml ndCycle!!! 1820 lIMO 083 265 nil xxicycier 1774 7570 0.38 2.55 nil aidFWHM: Full width at half maximum, Cr1: Crystallinity index SD: Standard deviation,SR: Sorption ratio, Fm: Fraction of amorphous material of cellulosic fibers82(I,EzEzIFig. 24. Comparison of the diffractograms of virgin (beaten), cycle I and cycle Illpuips.Interestingly, no difference between the sorption isotherms of beaten and unbeaten puipswere observed, which shows that the amorphous component of the fibers has not beenaltered during refining (Figs. 17, 19, 20). Therefore, the decrease in the integral breadths,along with a decrease of the amorphous fraction, indicate that, probably, the observedincrease in crystailinity, is induced by crystallization of amorphous phase and fuzzing upof fibrilsflamellae during recycling. This interpretation is in agreement with previousobservations on the loss of accessible surface area of the fibers.25Bracia’s anale (2 theta83Thermodynamic properties of recycled fibers were also studied, with the hopethat information would be collected on the structural changes of recycled fibers byinteracting with water. Free energy, enthalpy and entropy of recycled fibers were0Moles of water per 100 got solid (ni)Fig. 25. Comparison of the differential free-energy of virgin and recycled fibers.84-,Cu.CIiiFig. 26. Comparison of the differential enthalpy of virgin and recycled fibers.measured and compared with virgin fibers. The overall free energy and enthalpy of thevirgin fibers were higher than those of recycled fibers (Fig. 25, 26 and Tables VIll-IX).Fig. 27 shows that differential entropy differences depended on the moisture ratio. Thehigher free energy indicates that, the virgin fiber surface has more affinity to adsorbwater than recycled fibers do.Moles of water per 100 g of solid (ni)85IThe hydrogen bonding energy of recycled and virgin fibers were the same withinthe limit of experimental error. The total hydrogen bonding energy between water andcellulose is -5.72 kcal/mole for the virgin and -5.79 kcal/mole for the recycled fibers(Table V). This indicates that, the difference in the strong properties of recycled andvirgin fibers cannot be explained by the concept of strong and weak bonds. Moreimportantly, the higher entropy of recycled fibers in comparison with virgin fibers (Fig.27) suggests a degree of disorganization of virgin fibers in the vicinity of the moistenvironment. This disorganization of virgin fibers could be explained by the highswelling potential of the virgin fibers during water uptake.2wMoles of water per 100 g of solid (ni)Fig. 27. Comparison of the differential entropy of virgin and recycled fibers.000TableVI.Calculatedthermodynamicpropertiesfortheadsorptionof wateronbeatenandunbeatencellulosedriedbycritical pointdryingmethod.fliP/Pan1p0/p-AF-iF1-AH-AH1-zS-S1IF/area0.090.0831.08416214741542592—0.0273.750.830.180.1231.463282124132420000.1402.541.450.270.1621.667387107847817540.3052.271.98O.360.2361.52547585557815220.3452.242.430.450.4101.09753952868413170.4862.642.760.540.5331.01357837380011850.7442.722.960.630.6201.01660728390010340.9822.523.110.720.7001.03062921110007491.2441.803.220.810.7341.10364718310365441.3041.213.320.900.7701.16966215510663701.3550.723.390.990.8031.23367513011182291.4860.333.461.080.8311.29968611011201471.4550.123.521.170.8541.370694931124961.4420.013.561.260.8821.428702741130451.435—0.093.601.350.9091.485707561140201.452—0.123.62coTableVII.Calculatedthermodynamicpropertiesfortheadsorptionofwateronrecycledfiber.Thefibersdriedbycritical pointdryingmethod.flj.P/Pon1p0/p-AF—AF1-AH-AH1-AS-AS1AF/areao00.090.08571.0516114551542450—.023.340.920.180.14851.212278113032418140.1542.291.590.270.24861.08636782447815150.3722.322.090.360.42000.857142851458412540.5232.502.450.450.60000.75004633037129750.8352.252.650.540.66570.81124872418128351.0901.992.780.630.70570.89275062069005451.3211.142.890.720.74000.96005221789603031.4690.422.980.810.77001.05005371569842181.4990.———0.133.311.260.89361.4000586651049301.553—0.123.351.350.91501.4702591541054101.553—0.153.3888The virgin (unbeaten and beaten) and recycled puips were also examined byenzymatic hydrolysis. The initial hydrolysis yields of virgin (unbeaten and beaten) andrecycled fibers were compared and changes in the characteristics of the substratesmonitored by different techniques. About one third of the beaten pulp was hydrolyzedafter 4 h, compared to only 20% of the cycle VI pulp (Table XI). It was observed that,regardless of similarities in their initial fiber length distribution, the partially hydrolyzedbeaten pulp was, on average about twice as long as the other three puips (Table X). Thisobservations suggested that the beaten fibers had more accessible surface area forenzyames than unbeaten and recycled fibers.Table X. Fiber length (FL) analysis of hydrolyzed (H) puipsSn:A(Ave):L(Ave):M(Ave):Sample nameArithmetic (average)Length-Weighted (average)Mass-Weighted (average)After hydrolysis, it seemed that both unbeaten and recycled puips werefragmented to a greater extent than the beaten puips (Figs. 28A-28D), correlating wellwith their greater reduction in fiber length during hydrolysis (Table X). In fact, the89presence of intact, non-degraded fibers was detected only within partially hydrolyzedresidues derived from the beaten pulp (Fig. 28B).90Fig. 28. Comparison of scanning electron photomicrographs of hydrolyzed substratesderived from (A) unbeaten and (B) beaten puips.91Fig. 28. Comparison of scanning electron photomicrographs of hydrolyzed substratesderived from (C) cycle I and (D) cycle VI puips.92The possible changes that could affect the degree of polymerization (DP) of eachof these puips during hydrolysis (Table XI) were also examined. The beaten pulp had aninitially higher weight average DP, which decreased substantially, after 4 h of enzymehydrolysis. Both unbeaten and cycle I puips showed a similar trend, with about a two-thirds reduction in DP resulting from enzymatic hydrolysis. The cycle VI pulp showedthe smallest reduction in the weight average DP and virtually no change in the numberTable XI. Degree of polymerization (DP) and hydrolysis yield of the hydrolyzed (H)pulpsSample nameNumber (average)Weight (average)PolydispersityYield (%)Sn:N(Ave):W(Ave):P:Y:930aaFig. 29. Comparison of MW]) of hydrolyzed samplesaverage DP. After hydrolysis, the polydispersity of all the puips except the unbeatenpulp was very similar (Table XI) despite significant differences in the MWD peakdistribution (Fig. 29). It is possible that the extent of the swelling is high due toextensive beating (i.e., 12000 revolutions by PFI mill), to allow access of the largeenzyme molecules to hydrolyze the substrate through interlamellar surfaces.Alternatively, both unbeaten and recycled fibers, due to their more compact structureElutlon volume (mL)94(i.e., less swelling), appeared to have been hydrolyzed structurally in a more length-wisefashion rather than any direct changes in the surface of the fibers. The conclusion issupported by the observation that the fiber length of the beaten hydrolyzed pulp fiberswere twice that of the fiber length of unbeaten and recycled fibers while the DPw of thebeaten fibers (hydrolyzed) was half of the counterparts. The low DPw of beaten fiberswith their higher average length suggested that the enzymes had more access to thecellulose surface area to attack than unbeaten and recycled fibers.When the same hydrolyzed and unhydrolyzed pulps were assayed for changes intheir crystallinity (Table XII), the beaten pulp showed virtually no change in itscrystallinky index (Cr1) while the other three pulps showed a substantial increase in thedegree of crystallinity. The rather low crystallinity of beaten fibers clearly indicated thatthe enzymes had access to the crystalline component of cellulose surface. In the otherwords more crystalline surfaces were made available for enzymatic attack on beatenfibers when it is compared with unbeaten and recycled fibers.Table XII. Crystallinity of hydrolyzed (H) and Unhydrolyzed (U) PulpsSn: Sample nameFWHM: Full width at half maximumSD: Standard deviation95This is consistent with the figures previously obtained for both fiber length andmolecular weight distributions during hydrolysis. The lower crystallinity of the beatenfibers also support the belief that the hydrolysis is interlamellar in the case of beatenfibers. These observations bear structural information on recycled fiber wall which willbe elaborated more on this in the discussion.966. DISCUSSIONThe aim of this dissertation has been to investigate the probable causes andmechanisms which influence the inferior properties of recycled fibers. This is evidentlyfundamental to the whole art of upgrading recycled fibers. It was felt that a betterunderstanding of various experimental results and their interrelation could be betterexplained by discussing them separately. Thus, the experimental results are organized ina certain order in regard to their relevance. The comparable work done by otherinvestigators are then reviewed. The discussion is terminated by connecting the commonaspects of the findings as the “ Theory of mechanism of strength loss in recycled fibers”and a model proposed based on these results.6.1 Adsorption isothermsThe adsorption isotherms of unbeaten and beaten fibers overlapped (Fig. 17). Onthe other hand, distinct differences were observed for the adsorption isotherms of virgin(unbeaten and beaten) and recycled fibers (Fig. 17). This suggests that beating/refiningdoes not alter the sorption potential of the fibers, while recycling reduces the sorptionpotential of the fibers. This is consistent with the belief that the beating process increasesthe volume of the fibers but does not alter the fiber accessible surface area to watervapor, in contrast to the recycling process.The adsorption isotherms in the vicinity of monolayer completion (i.e., aroundRH=30% and 45% for recycled and virgin fibers respectively) suggest a distinctdifference between virgin (unbeaten and beaten) and recycled fibers. This differenceindicates that the virgin fibers has larger capilaries which recycled fibers are lacking. Inthe other words, recycled fibers contain less accessible surface area when it is compared97with virgin fibers. The appreciable separation in the sorption curves between virgin andrecycled fibers continues up to 90% relative humidity (Fig. 17).The second section of the curve (i.e., 30 - 80%) corresponds to the “filling up” ofthe interstices between sheets of microfibrils in the cell wall (Fig. 17). The rigidity oflamellae would have a negative impact on the filling up of the spaces between them. Thecoarser* (i.e., more rigid) the lamellae, the more condensed would be the water layer andthus the volume of the pores remains intact. This is the state by which the recyclingeffect could be explained. Due to re-joining of the lamellae, the rigidity of the sheets ofmicrofibrils increases, thus the swelling potential of the cell wall decreases.In the last stage (80 - 90%) relative humidity, the differences between the vaporpressure of the water in the cellulose and the water in natural condition, as it was formedin nature, will disappear largely. However, the difference between the virgin andrecycled fibers, has not yet been lost.Initially it was reported (Patrick, 1924) that the hysteresis observed for thecellulose was attributable to experimental error, being due to the incomplete removal ofthe last traces of adsorbed water from the sample under examination. Urquhart (1929) ina comprehensive work showed that the sorption hysteresis observed with cotton celluloseis an intrinsic property of the material which cannot be explained as being due to faultymethods of experimentation. This observation of Urquhart (1929) seems to refute theidea that the irreversible closure of recycled fibers (i.e., dried from water) can not beexplained by failure in evacuating the smaller pores.* The coarseness of lamellae here refers to the thicker lamellae which are produced by ineversible closureof two or more single lamellae.986.2 Heat of immersionThe heat evolved on adding water to cellulose is the net amount produced by twoprocesses. One is the sorption of water on the cellulose surface in existence and the otheris the production of new cellulose surface and swelling. The latter effect is negative.Figure 21 represents the iinmersional adsorption isotherm of virgin and recycledfibers containing various amounts of adsorbed water. The initial regions of the curvesdescribe a nearly linear decrease in the heat of wetting as the moisture content increases.Zettlemoyer et al. (1955) considered this type of relationship to be indicative of anenergetically homogeneous surface. The heat evolved on wetting appears to dependsimply on the amount of bare surface present. If there was a distribution of energy siteson the surface of the solid, the high energy sites would preferentially adsorb vapormolecules and the heat of immersion would not be proportional to the amount of surface(Fig. 21) but, instead, would depend on which sites were occupied.The heat of wetting of virgin pulp is greater than that of recycled fibers (Fig. 21).This is in agreement with the concept that the virgin pulp has a greater internal areathanthe recycled pulp as indicated by sorption isotherms. This is consistent with observationsof Argue and Maass (1935) who reported data on the differences between the adsorptionand desorption isotherms of purified cotton cellulose. These authors proposed that thecellulose from which a portion of the water has been desorbed, had a larger surfacethanthe cellulose containing the same amount of water added to a dry sample.No detectable difference in the heat of wetting between unbeaten and beaten (i.e.,never-dried) fibers suggests an equal fiber surface area for both fibers in the ranges of 0-12% moisture content (See also Fig. 21). This is in agreement with the observations ofMaass and Campbell (1939). The authors, extended the accessible surface area by twoways: one, by pounding the dry cellulose between two iron surfaces; the otherbysubjecting the cellulose to a static pressure of 50000 lb/in2. A 25% increase in heat ofwetting of the fiber material was obtained. The increases in heat of wetting in eithercase99were interpreted to be due to the increased surface area. Dole and McLaren (1947)measured the heat of wetting of stretched and unstreched nylon. MI values of unstrechedand stretched nylon levelled off at 1250 and 550 cal/mole respectively, that is, about 2times higher in the case of unstreched nylon. The authors attributed the lower wateruptake of stretched nylon to its crystallization during stretching.These results and their interpretation are in excellent agreement with thoseobtained by determining the sorption isotherms for the pulp (virgin and recycled) fibersin Section 5. This indicates that the larger fiber surface will release a large amount ofheat when immersed in liquid water. More importantly, the lower heat of wetting forrecycled fibers was induced from lower accessible surface area of the recycled fibers.6.3 Surface analysis of recycled fibersAn analysis of the water vapor adsorption data indicated that the BET theorycould be applied in the range of 5 - 30% relative humidity. The monolayer capacities of0.089 and 0.055 (g of water)/(g of the fibers) obtained in this manner corresponds to asurface area of 384 and 238 m2/g for virgin (i.e., unbeaten and beaten) and recycledfibers respectively (See also Fig. 17 and Table V). The results indicate that there is nodifference in the surface area of unbeaten (never-dried) and beaten (never-dried) fibers.Unlike the effects of beating, recycling of the fibers reduces the accessible surface towater vapor.The reduction of the accessible surface area in the internal surface suggests thatsome of the hydrophilic surfaces (i.e., lamellae) were lost. Thus, the reduction in fiberbonding potential is induced by a reduction in the fiber internal hydrophiic surface area.The internal surface area has a substantial effect on flexibility of the fibers. In fact, thereduced (almost 36%) internal surface area of recycled fibers observed in the presentwork indicates that the number of hydrophilic sites is reduced by re-joining them100irreversibly (i.e., thickening) during drying. Thus, the drying process produces fiberartifacts with fewer and consequently, coarser (thicker) lamellae. The coarser lamellae(i.e., reduced fiber surface area) adsorb less water vapor and consequently, provide lessresistance to severe mechanical treatment. This should be recognized as the maindifference between recycled and unbeaten (never-dried) fibers.Generally, swelling during water uptake depends upon the relative effect of twoopposing sets of forces. The adsorption of the swelling agent tends to push apart thelamellae by overcoming interplaner attractions. The rigidity of the lamellae opposes theswelling forces. Thus, the rigidity resulting from re-joining of lamellae in recycled fibersneutralizes the forces exerted by swelling agents and consequently, the swellibility of thefiber wall is restricted. On exposure to the beating forces, considerable damage isundoubtedly done to such fibers as a result of their relatively high rigidity, before thebeneficial results of beating can occur. On the other hand, the never-dried (unbeaten)fibers due to the higher water content, when compared with recycled fibers, couldrespond much better during beating. During beating, a reduction in the rigidity of thelamellae due to mechanical fatigue eases swelling and brings about the plasticization ofthe fiber wall by the sorbed water. Closure and interweaving of fibers during depositionon the forming wire and subsequent wet pressing can produce a greater number ofbonding points on drying. These observations suggest that fiber wet-flexibility isconsiderably lost due to re-joining of the lamellae during drying.A reduction in the external fibrillation of recycled fibers was also observed byscanning electron microscopy (SEM). Thus, the reduced surface area of bonding is dueto the reduction of fibrils in the outer surface of the fibers (See also Figs. 1 6A- 1 6D).This is also confirmed by estimating the external surface area (i.e., 55 m2/g and 27 m2/gfor virgin and recycled fibers, respectively) as suggested by the absolute method ofHarkins and Jura (1944) (See also Table V).101Emerton (1957), reported that up to 98% of the surface of moist, swollencellulose was obliterated by direct drying, but up to 75% was preserved during solvent-exchange with ethanol. In fact, only 2% of the surface remained free in drying fromwater. Bull (1944) in working with egg albumin attributed the reduction in the free spacebetween protein to the crystallization of the protein when dried from water. The authorssuggested that the protein molecules can not occur in random arrangement , otherwise thefree space, not occupied by the protein, would be very much greater than 6%. Theauthors concluded that the new arrangement of dried protein molecules did not occur asone crystal but that the solid mass of protein is made up of a great many microscopic orsubmicroscopic crystals. Scholz and Flath (1991) measured the internal surface area ofdifferent cellulosic fibers by the iodine sorption method. The authors observed that themanufacturing process and pre-treatment alter the sorptivity of cellulose to iodine. It wasalso observed that the drying from water resulted in the loss of internal surfaces of thecellulose. On the other hand, the drying from methanol preserved most of the internalsurface area as measured by iodine adsorption.6.4 Thermodynamics of Fiber-Water Interactions6.4.1 Free-energyComparison of the change in integral and differential free energy of the virginand recycled fiber demonstrates an appreciable difference between these two systems(Figs. 25, 23 and Table VIll-IX). The free energy of virgin fibers is larger in the range of8 to 80 % relative vapor pressure. This suggests that virgin fiber could release moreenergy for work. However, the quantitative analysis of these energies requireinformation on fiber surface area, enthalpy and entropy of the system. It is suggested that102the free energy change is the function of the accessible sorption sites or the fiber surfacearea (Stamm, 1957). Therefore, the observed differences in free energy could beattributed to the higher surface area of virgin fibers (See also Table Vill-IX).Fig. 30. Comparison of the integral free-energy of virgin and recycled fibers..4The free energy change per unit area was also measured (Table VilI-IX). Theresult suggested that the free energy change per unit area remained equal during theentire range of adsorption for both recycled and virgin fibers. This data confirms that theMoles of water per 100 g of solid (ni)103high sorptivity of virgin fibers is induced by its more accessible surface area when thefibers are brought into contact with liquid water. In other words, fewer sorption sites areavailable for recycled than virgin fibers.6.4.2 Differential enthalpy and hydrogen bondingThe differential enthalpy could be used to deduce the value of hydrogen bondingenergy between the water molecules and fibers. The vertical dashed lines in Fig. 26represents the monolayer capacities as estimated from the BET equation. At this pointz\H has the value of -1.22 kcal/mole for virgin and -1.29 kcal/mole for recycled fibers.These are net values in excess of the heat of condensation to liquid water. In the otherwords, these values represent the bond energy in excess of the normal hydrogen bonds inwater. A value of H=O indicates that the bonds formed during vapor adsorption haveenergy equivalent to the bond strength in liquid water. The heat of condensation ofwater, or heat of liquefaction of water is approximately -9 kcalJmole, thus, each mole ofhydrogen bonds in water has an energy of approximately -4.5 kcal/mole. Thus, the totalhydrogen bond energy between water and cellulose becomes -5.72 kcallmole for thevirgin and -5.79 kcal/mole for the recycled fibers (See also Table V). These values, areobtained calorimeirically and are reasonably comparable with the heat of adsorptionenergy estimated from the BET theory (namely -5.49 for virgin and -5.52 for recycledfibers). The parameter C had a value of 4.38 for virgin and 5.61 for recycled fiber.Using these values in Eq. (4), leads to the values of -0.99 kcal/mole and -1.02 kcal/molefor virgin and recycled fibers, respectively.The differences in the strength of hydrogen bonding between the virgin andrecycled fibers is not significant (See the standard deviation (SD) in Table V). Thus, thedifferences on the enthalpy of the two systems can not be explained by the strength of theadsorption bonds. The lower enthalpies of recycled fibers when compared with virgin104fibers (Fig. 26) indicate a higher accessible surface area in virgin fibers for wateradsorption, in line with previous observations.6.4.3 EntropyA decrease in entropy suggests ordering of the system and an increase in theentropy of the system represents disordering. However, the entropy of a closed systemeither remains constant or increases but it never decreases. For clarity in this context theincrease in -AS in Fig. 27 (i.e., decrease in entropy) represents ordering and the reverseindicates disordering.At early stages of adsorption the -AS in Fig. 27 is high. This is a direct result ofthe change of three-dimensional molecular motion to two-dimensional motion ofmolecules. This is expected because according to Eqs. [2.10] and [2.11] when therelative vapor pressure approaches zero, the differential free energy goes to infinity,which suggests minimum entropy, thus a high -AS (ordering).On completion of the monolayer both virgin and recycled fibers demonstrateordering (increase in -AS). The ordering for recycled fibers occurs at n=.3 1 and forvirgin fibers at almost n:=0.50 (See also Fig. 27). The ordering in this region is atthbutedto a decrease in randomness as water is adsorbed on the fiber surface or even on someorientation of water molecules on the fiber surface. The maximum entropies for thevirgin and recycled puips suggest domination of the swelling effect during water uptake(see also Fig. 27). Furthermore, Fig. 27 suggests higher swelling potential for virginfibers. The broad range in completion of the monolayer can be attributed to the effect ofhigher potential interfibrillar adsorption of virgin fibers which recycled fibers are lacking.Due to the nature of capillary forces, the smaller surfaces are filled up first, and then thelarger ones. Then the vapor pressure gradually increases with the greater “capillary”105IxwFig. 31. Comparison of the integral entropy of virgin and recycled fibers..4cross section of the surface of water. There is then a gradual transition from water fillingsmall capillary spaces to filling larger ones between the lamellae and eventually fillingspaces between fibers. The earlier decrease of -z\S in recycled fibers (Fig. 27) ascompared with that for virgin fibers, suggests smaller pores for the recycled fibers. Asharp increase in the peak also supports the belief that the pores of recycled fibers shouldbe smaller, where the condensed water should demonstrate higher organization than thewater that is free to move.The -AS continues to drop (disordering effect) after the completion of themonolayer. Probably, this is because of opposing effects of molecular separation,swelling of the fibers and diminishing of the forces holding the higher layers of theMoles of water per 100 g of solid (nl)106adsorbed film on the surfaces of the fibers during moisture uptake. It seems the trendcontinues up to the fiber saturation point.The comparison of the integral entropy of virgin and recycled pulp systems (Fig.31) clearly demonstrates that the overall entropy of recycled fibers is higher for the entireadsorption range. This further suggests higher swelling potential and free-energy of thevirgin fibers when it is compared with recycled fibers. The changes in the entropy of asystem is caused by chnges in the structure of the fiber rather than the water. Moreexpicitly, it is difficult to conceive of the water alone having a different arrangement onthe surface of the fibers without referring to the differences of the fibers themselves. Infact, it is more usual that the adsorbent determines the arrangement of the adsorbate thanthe opposite. Thus, probably the observed ordering effects on recycled fibers are inducedby the changes in the structure of the fibers rather than the water. This view wassupported by Barkas (1942), Morisson et al. (1959), Hermans and his co-workers (1949),Collins (1930) and Chakravarty (1958) on the explanation of the cause of hysteresis.Barkas (1942) suggested plastic deformation of a gel on the desorption of water vapor.Collins (1930) observed that the volume on the desorption side was greater than on theadsorption side. The same results were reported by Chakravarty (1958) when the cross-sectional area of single jute fibers were compared on desorption with those on adsorptionat different relative humidities. More importantly, the differences in entropy start after25% relative humidity is reached. This indicates that the difference between virgin andrecycled fibers is largely an interfibrillar microscopic phenomenon rather than one of anysingle microfibril on the fiber surface.Some of the workers explained these observations on the basis of the organizationof water alone (Henry et al. 1988, Milichovsky 1990). Henry et al. (1988) observedduring the entire process of beating fibers, no new surfaces are developed. The authorsinterpreted their obsevation based on better structuring of the water molecules along the107fiber surfaces during beating. Thus, it is the water rather than fiber which undergonestructural change during beating.6.5 Changes in Fiber Wall Ultrastructure due to RecyclingComparison of XRD traces of unbeaten and beaten puips suggests that refiningbasically does not alter the crystal structure of the cellulose fibers (Fig. 23). The physicaleffects of refining are merely restricted to the amorphous phase of the cellulose. Notonly the crystallinity, but also the full width at half maximum (FWHM) of the intensitypeaks of x-ray diffractograms remain constant values (Table VII), confirming that,refining does not affect the crystalline order of the fibers.Moisture regain was used as an indicator for the noncrystalline fraction of virginand recycled fibers. It was observed that the amorphous fraction decreased by almost22% after the first cycle (Table VII). Interestingly, no differences between the sorptionratios of beaten and unbeaten puips were observed, to show that the amorphouscomponent of the fibers had not been altered during refining (Table VII). Therefore, thedecrease in the integral breadths, along with a decrease of the amorphous fraction,indicates that, probably, the observed increase in crystallinity is induced by partialcrystallization of the amorphous phase of the lamellae surfaces during recycling. Theseobservations lead to the belief that the changes in superstructure of the cellulose, at leastpartially, must be responsible for the origin of poor quality (hornification) of recycledfibers.These resuks are also supported by the measurements of the peak breadths. Thedecrease observed for the peak breadths (101 peak width) of recycled papers suggeststhat the crystallite size increased due to recycling. This finding is very important, sincethe 101 plane is known to represent the plane of lamination (Frey-Wyssling, 1954) in thefiber wall and shows conclusively the “growth’ of crystallites in recycled pulp fibers. In108fact, hydrogen bonding takes place preferentially along the 101 plane during drying.Thus, this evidence also supports the proposed idea that irreversibility of hornification isprobably related to the crystallization phenomenon brought about by irreversible joiningof lamellae parallel to 101 planes.In connection with the problem of cell wall crystallization, due to drying,contradictory results have been reported. Marton et al. (1993), Atalla (1992), Ingram etal. (1974) and Heyn (1965) reported a considerable increase in crystallinity of thecellulosic fibers due to drying. On the other hand, Bouchard and Deuk (1994) andMorossof (1974) observed no increase in crystallinity index.Bouchard and Deuk (1994) did not observe any considerable changein thecrystallinity index during recycling. The authors concluded that the observed minorincrease in crystallinity as a consequence of recycling is explainable by yield effect. Theprecision of the technique of yield correction was uncertain due to incremental loss, ifthere was any, in the non-crystalline portion of the fibers due to recycling. So, analternative technique was followed to evaluate the problem.In order to evaluate the “yield effect” on measurements of crystallinity, a virginpaper was made into handsheet, pressed and re-slushed up to10 times, without dryingduring the repeated handsheet-making process. Indeed, thepaper is recycled for 10 timeswithout experiencing the drying stage. After 10 times re-slushing and pressing, the hand-sheet is air-dried in the conditioning room according to CPPA standard. It is assumedthat, if the gain in crystallinity is a result of the yield effect, after 10 times recycling, thecrystallinity index must be amplified. However, the crystallinity did not increase asexpected (Table VII). Thus, it is concluded that the crystallinity increase cannot beexplained by yield effect.Marton et al. (1993) observed that, most of the increase in crystallinity occurredin the first cycle (5.5% after first re-wetting). More importantly, they observed that, reslushing not only does not restore the crystallinity value to thenever-dried state, but also109that subsequent drying steps incrementally promote the increase in crystallinity of thefiber wall.Atalla (1992) reported that the recycling of fibers caused an increase incrystallinity, as well as a decrease in surface area. He observed that the maximum effectof recycling was observed after the first recycle, when most of the change in cellulosestructure seemed to have occurred. It was also confirmed that the crystallinity was notinfluenced by beating or the amount of hemicellulose in the pulp. More importantly,Atalla (1992) observed that the crystallinity, as measured by the reciprocal of the width-at-half-height of the 002 reflection in the x-ray diffractogram of the pulp, did increasesteadily with each progressive recycling of the pulp. The author, in light of his proposedstructural model, concluded that repulping resulted in some slight enhancement of themolecular mobility of the disordered cellulose, thus allowing it to enter more readily intothe domains of more coherent order at the superstructural level during the subsequentdrying cycle. These observations are in agreement with those reported by Hatakeyama(1988), Okayama et al. (1982), Atalla (1978), Guest and Voss (1983).Heyn (1965) in examining the drying effect on cotton fibers also reported that thefirst-drying stage of cotton from the never-dried condition was most distinct and resultedin a greater degree of re-crystallization of never-dried cotton than obtained on subsequentdrying. Heyn (1965) observed a dramatic increase in crystallinity of the freeze-driedfibers (i.e., 11% increase in crystallinity index) when re-wetted and subsequently driedovernight. Similar observations on cotton drying were also reported by Zeronian (1977),Ingram et al. (1974), Kulshreshtha et al. (1973). Mellon et al. (1949) observed that the xray patterns of dried and stretched keratin were identical and interpreted this by theordering effect of drying.It could be concluded that, the re-organization of the cell wall is not confined toconsolidating the fiber wall by mere zipping up the external and internal surfaces, butalso extends to re-crystallization of cellulose in the cell wall (Figs. 23-24 and Table VII)in which it makes the process partially irreversible. Thus hornification, at least partially, should be recognized as a higher order (supermolecular) change which occurs in uniting the sheets of microfibrils or crystallization in the amorphous phase due to drying. This finding is in agreement with previous observations at present dissertation. Scallan and Tigerstrom (1992) reported that the elastic modulus of wet-fibers was doubled by recycling. An increase in the elastic modulus of wet-fibers was attributed to the enhanced hydrogen bonding between the microfibrils. This observation could also be explained by the increase in crystallinity of the fibers. It is well known that water does not penetrate into crystalline domains of cellulose since entrance of one water molecule will require the breaking of many cellulose-to-cellulose bonds (Salmen 1988). Therefore, its elastic modulus remains intact in the vicinity of a moist environment. These points demonstrate plausibly that the increased modulus is induced by increased crystallinity due to recycling. Since wet-flexibility is inversely proportional to the elasticity, the observed hornification (i.e., loss in wet-flexibility), due to recycling, is tied closely to the crystallization phenomenon of the fiber wall during drying. The comparison of the diffractograms of unbeaten and beaten (virgin) pulps (i.e., crystallinity and peak broadening) reveal that, refining/beating has only a minor impact, if it has any, on the fiber ultrastructural organization. Its effect is restricted to extending the distances between the fibrils/lamellae through the combined effects of swelling and mechanical fatigue. This finding is also in agreement with observations of Hatakeyama et al. (1987). The authors concluded that, the molecular order of cellulose remains intact even after considerable beating. Inversely, drying plays a significant role in the re-organization of,the supermolecular structure of the delignified fibers by increased crystallization and crystallite size. In agreement with this hypothesis, Baker and Fuller (1943) reported that the introduction of methyl groups on nylon amide increased the vapor phase sorption by opening up the structure. Assaf et al. (1944) explained the obliteration of the cellulose surface during drying by the action of "slide fastener" as it unites two serrated edges. The authors concluded that hydrogen bonds formed between two glucose residues on opposite sides of the fissure, and that the corresponding macromolecules, held together in this way at one point, then crystalize imperfectly but extensively against each other. This view was also supported by Bull (1944) in explaining the lower hydrophilicity of denaturated proteins. Brickman et al. (1953) attributed the reduced water sorptivity to the physical arrangement of the fibrils and/or macromolecular chains which restricted the penetration by the water vapor during the time allotted for the adsorption. As it is discussed, the delaminated fiber wall will be re-joined during drying by crystallizing forces (i.e., extensive hydrogen bonding). The severing of such coherent forces requires an appreciable amount of energy which can not be rendered by water per se. For instance, each square centimeter of crystal cellulose is held together by 1.43 x 1015 hydrogen bonds, which corresponds to approximately, 3.29 x 1016 kJ/mole or 7.86 x 1015 kcal/mole. Thus, re-slushing them would result in a partial delamination of the fiber wall. The lamellae which were fused by crystallizing forces will not be re-laminated on re-wetting. Thus, the average pore size will be reduced significantly. Reversal of this effect requires swelling agents with higher swelling power than water to re-open the cell wall. 6.6 Enzymatic Hydrolysis Previously, when the modes of enzymatic hydrolysis of steam exploded and kraft pulped wood chips were compared (Ramos et al. 1993), it was apparent that the two cellulosic residues were hydrolyzed in distinct ways and the mechanism appeared to be influenced by the fine structure of the substrate. In the work reported here I had hoped to obtain further information on the structural differences between virgin and recycled fibers by subjecting them to 4 h enzymatic hydrolysis using cellulase enzymes (Novo Nordisk, Denmark). It was shown by scanning electron microscopy (SEM) that the hydrolysis of the unbeaten and recycled pulps resulted in small particles with each of the fragments of almost equal size (Figs. 28A-28D). Therefore, the initial action of the enzyme treatment was to attack the less ordered parts of the fibers resulting in a rapid reduction in particle size. Despite the lower degree of fragmentation of the beaten pulp, as evidenced by both SEM and fiber length analysis, a greater hydrolysis yield was obtained (30% yield). Similar results have been reported by other workers when using steam explosion or ball milling (Grous et al. 1986, Caulfield and Moore 1974, Rivers and Emert 1987). Each of these pretreatment processes has a mechanical component which acts in a similar fashion to pulp refining. For example, steam explosion has been shown to result in partial delamination of the cell wall, considerable defibrilation and higher flexibility (Yamashiki et al. 1990, Barbe et al. 1990). It is recognized that beating or refining results in both external and internal fibrillation and consequently enhances swelling and fiber flexibility. Therefore, fibrillation in the fibers rather than direct reduction in particle size probably results in better accessibility of the cell wall to enzyme hydrolysis. The higher average fiber length of the beaten fibers along with their lowered DP, when they are compared with unbeaten and recycled fibers indicate that interlamellar hydrolysis is occurring in the beaten fibers (Tables X-XI). This indicates that beating assists swelling of the fibers, enhances the lamellar distances, thus, increases the available surface area to enzymes. The swelling facilitates the entrance of the enzyme molecules to the lamellae interstices and allows digestion of the fibers through the surface. The foregoing interpretation could also help to explain why the lesser digestibility of the recycled pulp is reflected in a reduction in the potential bonding of these fibers. In the present work, the crystalhnity index of unbeaten and recycled pulps demonstrated a residue with a substantially increased crystallinity. In contrast, the crystallinity index of the more readily hydrolyzed beaten pulp did not change appreciably and, in fact, decreased slightly (Table XII). It is possible that changes in the structural organization of the crystallites resulted in an increase in the surface area accessible to the enzymes. The results of the present work have shown that the beating/refining of the pulp increased the accessible surface area without any significant impact on the fiber length, DP or crystallinity characteristics (Tables IV, VI, VII). In contrast, drying, as would occur in papermaking, results in a drastic reduction in surface area which is exacerbated further by recycling (Table V). Thus, the chance for the enzymes to penetrate inter-lamellae and attack the surfaces of lamellae in recycled fibers, was reduced. That is why the DP of the beaten fibers was decreased dramatically in comparison with unbeaten and recycled fibers. The fiber length is affected differently. Lack of considerable changes in crystallinity index of beaten fibers is also explainable by fiber swelling. Probably, beating/refining exposed more crystalline surfaces to the enzymes than the unbeaten and recycled fibers. This also clearly shows that the internal fibrillation induced by recycling is a critical requirement for enzymatic attack, which was exacerbated by recycling. More importantly this indicates that lamellae surfaces are partially free from lignin which allows the degradation of cellulose fibers and consequently reduction of the average DP (Table XI). Fig. 32 could also be used to explain the mechanism by which the structure of the substrate enhances the effectiveness of enzymatic hydrolysis. Beating enhances the swelling of the fibers and makes "lignin-free" surfaces accessible to enzymes. The hydrolysis of such lignin-free sites on the surface of the lamellae breaks down the cellulose chains and thus results in DP reduction. This observation suggests that the exposed surfaces (lignin-free sites) are mostiy crystalline. Thus the slight reduction in crystallinity index of the beaten fibers is caused by hydrolysis of those surfaces by enzymes. It seems probable that, the lignin-free sites are responsible for irreversible closure of these lamellae during drying. This interpretation could also help explain how the lesser digestibility of the recycled pulp is related in a reduction in fiber surface area. Plausibly, the decrease in the total surface area of the fibers results in a decrease in the contact area between the fibers, eventually bringing about lesser bonding and diminished strength for the recycled paper. 6.7 Theory of Mechanism of Strength Loss in Recycled Papers The nature of the inter-lamellar adhesion within fibers has perhaps not yet been fully understood. This initiates from the considerable amount of ambiguity about the nature of the fiber wall aggregation during drying. However, accumulated evidence suggests (Nissan, 1990) that hydrogen bonds are mostly responsible for the cohesion of cellulose molecules in every single plane. Individually, hydrogen bonds are weak, only about one-sixth of the strength of glycosidic bonds between successive glucose units in the cellulose chain molecules (Nissan, 1990). It is through their frequent occurrence along the fibrils and fibers that the reasonably strong adhesion of fibers in paper can be attributed to. Molecular forces of all recognized types, such as hydrogen bonding, do not operate at greater distances than a few Angstroms (A). If the binding forces become appreciable within less than 4A, then the required bonding area between the fibers would be extremely small. There should be a mechanism by which, the fibers could be drawn close enough to make hydrogen bonding possible. This mechanism was comprehensively explained by Campbell (1933). During the wetting process, water penetrates between the sheets of microfibrils, and breaks the hydrogen bonds whilist locates itself on surfaces of lamellae, whereby the lamellae are pushed apart. Depending on the rigidity of the wall and polarity of the agent, the effect would be either distinctive or difficult to detect. The reverse effect acts 115during drying. As the water is removed, the remaining water imposes cohesive forcesthat draw the celluloses surfaces toward each other. These forces are due not only to theatmospheric pressures forcing the structure together as water is removed but is mainlydue to a much greater extent to an internal tension in the water arising from molecularforces. With still further evaporation of water, the water layers between the celluloseparticles become very thin and the reduction in vapor pressure becomes appreciable. Theinternal tension begins to increase and an intense force is created by which the cellulosesurfaces are drawn together. This force acts between the lamellae/fibrils in the fibers forthe most part and causes individual fibrils or lamellae to collapse. The force due to suchinternal tension is of the order of hundreds of atmospheres for water. For instance, it is936 kgm/cm2for a relative vapor pressure of 50 per cent and at a relative vapor pressureof 10% it rises to about 3110 kgm/cm2 (See also Appendix E). These forces, thus pullthe hydroxyl rich planes (101 planes) close enough to make direct binding of thepolysaccharide (cellulose) chains possible.The rigidity of each lamella is important in collapsing the fiber wall. If thecapillary wall is sufficiently rigid to withstand collapsing by the internal tension forces,the water will eventually evaporate completely. Internal tension forces are notappreciable over large distances (See also Appendix E). On the other hand, finerlamellae in the cell wall will deform under tension forces. If this happens, the distancebetween the lamellae will decrease, on further evaporation the internal tension will beenhanced, due to the decrease in the distance between the lamellae. On further decreasein the gap between the lamellae, the tension forces will increase more and more, untileventually complete lamellar collapse will occur. In this process the finer fibrils/lamellaewill be brought sufficiently close to make direct hydrogen bonding possible. Unlike thecoarser lamellae, the chance for extensive hydrogen bonding for finer lamellae is higherand possibly, it may also lead to crystallization of the fiber wall occasionally. To bemore specific, on drying the microfibrils (i.e., the 101 planes in adjacent coaxial116lamellae) will be joined together. The surface of the lamellae, if deficient in lignin, willcome close enough for bonding of the polysaccharide chains to occur (Fig. 32C). If thesurface of the lamellae are also hemicellulose poor or if the hemicelluloses are locallyaligned parallel to the crystalline cellulose chains (See also Fig. 32C), bonding may insome cases be sufficiently extensive and regular as virtually to unite two distinctcrystalline regions as one. This is particularly likely in the case of kraft fibers withoriginally low hemicellulose content. The presence of lignin-free zones on the lamellaesurfaces is evidenced by enzymatic analysis in this investigation. Upon re-wetting, aproportion which was co-crystallized on drying is irreversible. In fact, none of the bondsso formed can be disrupted when the fiber is subsequently immersed in water. Thismechanism results in a partial loss in the ability of the fiber to take up water, an effect towhich the term homification has been applied. On the other hand, the presence ofresidual lignin, and perhaps to some extent the branched hemicellulose chain molecules,present between the microfibrils will tend to reduce the extent of irreversible bonding(Fig. 32C). This is also the case by which reversible swelling of mechanical pulp fiberscould be explained.Generally, the beating process reduces the rigidity of the fiber wall by mechanicalfatigue and plasticization of the fiber wall by adsorbing water. Concomitantly the fiberflexibility increases. The striking fact is that, when the surface area of the lamellae wasreduced to half, the flexibility of the fiber wall decreased 75%. This raises the point thata reduction of about 36% of accessible surface area of the fiber to water is sufficient toreduce the flexibility of the fiber wall to an appreciable amount (about 50%) and,therefore, reduce the strength quality of the paper. Reported data in the literature(McKee 1971, Szwarcsztajn and Przybysz 1976, Howard 1990, and Laivins and Scallan1993), on loss in swelling potential of recycled fibers, are consistent with theseobservations. lt is also reported that beating and recycling processes do not change thechemical constituents of the fibers (Curran et al. 1931, Bouchard and Douek, 1993).117Therefore, due to the short exposure of recycled fibers in water, during the beatingprocess, the possibility of losing the surface bonding agent, merely by recycling is hardlyprobable. More importantly, the role of internal fibrillation is dominant in thedevelopment of paper strength properties. Thus, it could be concluded that the reductionin the number of thinner wall segments by irreversibly re-joining them, results inrestricted swelling. Restricted swelling, or loss in wet-flexibility of the fibers, results inpoor confirmability of the fibers in the paper furnish, and paper strength properties arereduced.Emphasizing the fact that, as the internal fibrillation is a central feature of thebeating effect, loss of internal fibrillation is the main cause of inferior properties ofrecycled fibers. This implies that the main loss in strength properties of recycled fibersinduced by hornification rather than surface deactivation. However, this kind ofreasoning does not mean that the loss in surface bonding agents is impossible. The lossof potential surface bonding during the recycling, is secondary in importance to the effectof internal fibrillation as the cause of vanishing paper strength properties on recycling.1186.8 Proposed Model for the Evolution of Fiber Ultrastructure on Recycling ProcessThese observations, with small modifications, fit with the proposed model ofScallan (1974) on the ultrastructural arrangement of the cell wall in wood. Duringchemical pulping incrusting substances are removed from between the coaxial celluloselayers. The consequence of such change is a multilayer structure such as shown in Fig.32A. Indeed, water replaces fully the spaces previously occupied by encrustingsubstances.( Lignin-- HemicelluloSeS— CelluloseFig. 32. Conceptual drawing of the cell wall ultrastructural arrangement in unbeaten (A),beaten (B), dried (C) and recycled (D) states (Adopted in part from Scallan,1974).Unbeaten (never—dried)Recycled (re-wetted after drying) Dried119The beating in never-dried chemical puips, merely, increases the intersticesbetween the lamellae (Fig. 32B). It seems that the beating does not “delaminate” thelamellae, it only expands the lamellar interstices which were filled with water after theremoval of cell wall encrustants. Lack of any evidence for influence of refining at themolecular level as evidenced by XRD analysis (i.e., lack of changes in thesupermolecular structure of cellulose) and overlapping of the sorption isotherms ofunbeaten and beaten fibers support this point that, no laminations are created on accountof refining. In other words, the pre-existing fibrils/lamellae are merely pushed apart inthe course of refining (Fig. 32B). These observations suggest that the mechanical actionof refining should be sought in weakening of the fiber wall via flexing (cyclic loading)rather than final fatigue failure (fibrillation). Flexing reduces the rigidity of the lamellaeand consequently eases swelling of the fiber by polar liquids (e.g., water).Conversely, on drying due to the development of tension forces the plasticizedfibrils/lamellae are drawn together in molecular contact and fixed by extensive hydrogenbonding (Fig. 32C). If the conditions for fibrilar/lamellar orientation are such that theyfulfill the requirements for co-crystalization, the cell wall cellulose will be furthercrystallized. Re-wetting of the dried fibers results in a structure demonstrated in Fig.32D, allowing little moisture sorption. Indeed, some fibrills/lamellae which have not metthe co-crystallization condition or the extensive hydrogen bonding, will be re-openedagain. However, extensive beating will not re-open the initial spacing betweenfibrils/lamellae which were drawn together by extensive hydrogen bonding or cocrystallization. Subsequent recycling will lead to the generation of more fines due todiminished swellibility and reduced flexibility of the fibers.This model stipulates that the refining process does not produce any new surfacearea by partitioning the cell wall, but merely expands the interstices between the lamellaewhich were pre-occupied by water, thus swelling the fibers as suggested in Fig. 21B. On120the other hand, recycling reduces the specific surface area due to irreversible joining ofthe fibrills/lamellae during drying (Fig. 32D), thus decreasing the flexibility of the fibers.1217. SUMMARY AND CONCLUSIONSThe conclusions to be drawn from the research described herein are of two kinds,(a) experimental results or statement of facts, and (b) theories of supposed mechanisms.The former will be given first and in as a concise a manner as possible in the summarybelow. From these results and statements, information of immediate practical importancecan be drawn. For the latter case, the correlation and interpretation of the single resultsmust be regraded as tentative. The theory of mechanism which has been suggestedshould be regarded as simulation.7.1. SUMMARY7.1.1 Beating1. The beating process does not influence the ultrastructural organization of the fiberwall.2. The desorption isotherms of never-dried unbeaten, and beaten to 6000 and 12000revolutions, overlapped.3. The adsorption isotherms of unbeaten and beaten fibers overlapped in their entirerange of sorption.4. The identical shape of sorption isotherms for the unbeaten and beaten fibers, did notchange with the method of drying.1227.1.2 Recycling1. Drying/recycling cause re-organization of the fiber wall towards higher ordercrystallization. The main increase is in the cell wall ordering which occurs afterthe first cycle. The x-ray diffraction and sorption ratio gave consistent resultswith each other. Subsequent re-wetting does not restore the original organizationof the cell wall.2. Each stage of recycling after the first cycle incrementally increases the crystallinity ofthe fiber.3. The integral breadth of the 101 plane of fibers decreases by recycling. The averagecrystallite size of the fibers is inversely proportional to the size of integralbreadths of the diffraction peaks.4. The total surface area of the fibers remained constant during the beating process, butthe external surface area increased by beating.5. The total surface area of the fibers decreased up to 36% due to drying after the firstcycle.6. Drying and recycling decreased the sorption potential of the fibers.7. Minute differences in the condition of drying were reflected in distinct differences inadsorption isotherms of the samples.8. The fibers dried by the method of critical point drying sorbed more water than thosedried by the freeze-drying method. This is observed when the sorption isothermsof the fibers dried by those methods were compared.9. The heat of wetting of both virgin and recycled fibers decreased linearly by theincrease in oismre content of the fibers up to 12% moisture content.10. The heat of wetting of virgin fibers was higher than determined for the recycledfibers.11. The free energy of recycled fibers is less than that of virgin fibers over almost the fullrange of the adsorption isotherm.12312. The surface-energy per unit area remained constant over the whole range of theadsorption isotherm.13. The hydrogen bond strength of recycled fibers was equal to that of virgin fiberswithin the limit of experimental precision.14. The integral and differential enthalpies of virgin fibers were higher than those of therecycled fibers during the adsorption of water vapor.15. The entropy of the virgin fibers was higher than that of the recycled fibers.16. The monolayer completion on water adsorption occurred earlier for the recycledfibers than for the virgin fibers.17. The maximum ordering effect was about 0.31 for recycled and 0.50 for virgin fibers.18. The ordering of virgin fibers disappeared much slower than that for the recycledfibers as observed by comparing the entropy of the two systems of virgin andrecycled fibers.19. Beating increased the enzymatic hydrolysis yield by about 10% after 4 h incubationtime, while recycling reduced it by almost 7% during the first cycle.20. Successive recycling exacerbates the surface accessibility of recycled fibers, forinstance the enzymatic hydrolysis yield decreased up to 23% by cycle VI.21. The average fiber length of beaten hydrolyzed fibers was almost twice as long thanthat of unbeaten and recycled hydrolyzed fibers.22. Less fragmentation of beaten fibers in comparison with beaten and recycled fibersduring enzymatic hydrolysis was also observed by SEM microphotographs.23. The DP of beaten hydrolyzed fibers was almost half of that of unbeaten and recycledhydrolyzed fibers as measured by the GPC technique.24. The crystallinity index of recycled and beaten fibers increased, whereas that of beatenfibers remained constant during enzymatic hydrolysis.1247.2 ConclusionsThe irreversible closure of the lamellae in the fiber wall either by extensivehydrogen bonding or co-crystallization is responsible for the fiber homification duringdrying. The recycling effect on low-yield chemical puips could be explained on thisbasis. On the other hand, refining does not produce new lamellae in the never-driedunbeaten pulp, but only facilitates the condition of cell wall swelling.Beating the fibers up to 12000 revolutions by PH mill (i.e., 250 mL CSF) did notaffect the sorption potential of the fibers while an incremental difference in the conditionof drying altered the sorption potential of the fibers. Indeed, all sorption and swelling ofthe fibers occurred at the surface of the lamellae. Thus, reduction of the sorptionpotential of the fibers induced by recycling could be caused by closure of the lamellae.In other words, the accessible surfaces of the fibers to take up water are reduced byrecycling. Approximately a 36% reduction in the internal surface area of the fibers andhigher heat of wetting of the virgin fibers support these observations. Free energy andfree surface energy of the fibers which are calculated by thermodynamic approaches, inthe present investigation, indicate that the accessible surface area of the fibers werereduced. In addition, entropy data confirmed the ordering effect of recycling in the fiberwall. Comparison of x-ray peak intensities of virgin and recycled puips suggested thatrecycling reorganized the fiber wall towards a higher degree of crystallization. Moreimportantly, the reduction in the integral width of recycled fibers suggested reuniting of101 planes during recycling. This observation is in close agreement with theobservations of adsorption and immersion processes and thermodynamics of fiber-waterinteractions. It is possible that, if the surfaces of the lamellae are lignin-free, they willcome close enough to unite the two crystalline cores as one. These united surfacescannot be reopened again by reslushing due to development of the high cohesion forcesbetween them. The examination of the fibers by cellulose enzymes suggests the possiblepresence of lignin-free zones in the lamellae surfaces. The observation could also125explain the poor performance of recycled fibers in the refming process (i.e., brittleness ofthe fiber in refining zone).The unbeaten and beaten (never-dried) puips behaved in a similar fashion duringthe adsorption. Furthermore, beating did not show any effect on the fine structure of thecell wall. These observations indicates that the fatigue phenomenon is not necessarilyrelated to the fibrillation or sorption potential of the fibers. More importantly, the majordifference between unbeaten and beaten fibers can be explained on the basis of the fiberaccessible surface area to water uptake or on the number of lamellae. These observationscould also explain the cause of different behaviors between mechanical and low-yieldchemical puips during recycling. In mechanical pulps, the chance for irreversible closureof the lamellae is very slim due to the extensive presence of the lignin sheath on thelamellae surfaces.The drying of never-dried fibers takes place in the following manner. As water isremoved from in between two adjacent finer lamellae (i.e., collapsible under watertension force) they both continuously draw toward each other until all moisture isremoved and the fiber is dried. At this stage, the lamellae are very close together andform cohesive forces to activate and to reunite the lamellae with extensive hydrogenbonding. On the other hand, in the case of the coarser lamellae, if they are rigid enoughto withstand the water surface tension force before the drying is completed, the conditionfor close packing is lost. Eventually when the fibers are rewetted after drying, theadsorbing water breaks the bonds and substitutes them with the same tension forces asbefore so that the structure gradually expands. Some of these bonds remain intact due toco-crystallization or extensive hydrogen bonding, thus the total accessible surface area ofthe fibers is decreased.To summarize, the tension forces created by water loss draw the fibrils/lamellaeclose enough to make hydrogen bonding possible. If the sheets of microfibrils are alsohemicellulose poor or if the hemicelluloses are locally aligned parallel to the crystalline126cellulose chains, this bonding may, in some cases be sufficiently extensive and regular tocause the crystalline regions to unite as one. The rejoined planes thus formed cannot besevered when the fiber is subsequently immersed in water. This results in a partial lossof the cell wall ability to take up water, an effect to which the term homification has beenapplied. This is particularly likely in the case of bleached lcraft puips with lowhemicellulose content. However, the role of hemicellulose in co-crystallization of thecell wall remains speculative and requires further work.1278. RECOMMENDATIONS8.1 Identification of the Parameters Involved in Hornification During the BleachingProcessA single experiment in this laboratory showed a distinct difference in strengthproperties between unbleached and bleached unbeaten puips during recycling. Theunbleached (unbeaten) puips showed poorer potential for recycling than the bleached(unbeaten) puips. Fully-bleached pulps usually contain much fewer acidic groups thanunbleached pulps (Lindström 1986). Probably, acidity plays a role in the inferior qualityof unbleached (unbeaten) recycled fibers. However, this investigation has not beenfinalized yet. The outcome would be interesting in regard of identification of theparameters involved in causing the difference. The process could be followed as follows.The changes in chemical composition and strength properties of the puips in recyclingcould be determined for each sequence of bleaching. The results of each stage could bematched with unbleached pulp for changes in chemical composition, acidity and strengthproperties. This could identify the sequence in which the most deteriorative substance isremoved and thus the cause of such differences could be identified. This fmding wouldhelp to track down the deteriorative substance (such as acidic groups) in the pulp andremove during papermaking. Thus, to increase the potential life of the fibers duringrepeated recycling.8.2 RoLe of Hemicelluloses in Hornification of the Cell WallThe second problem which requires to be determined is the role of hemicellulosesin development of the hornification phenomenon during recycling. The result ofrecommendation A in identification of the deteriorative substance in recycling could leadto observations with respect to the differences in the hemicellulose content of the puips.128However, this could also be followed independently. It could be followed either on kraftpulp which has low hemicellulose content or with an organosolv pulp bearing a higher(up to 25%) hemicellulose content. The investigation could be conducted by comparingthe strength properties of the paper made from alpha-cellulose pulp with those containinghigher amount of hemicellulose. The selection of the agent for removal ofhemicelluloses is important, because some agents such as caustic soda also modify thestructure of the cellulose, which confounds the final conclusion. It is not yet certain thatthe hemicelluloses enhance homification in recycling or vice versa. Thus, the resultscould determine the role of hemicelluloses in recycling of fibers and consequently, theselection of appropriate agents to overcome hornification during recycling.8.3 The Role of Temperature in RecyclingIt was reported that (Nazhad and Paszner, 1994) heat-drying (i.e., drying thefibers after pressing stage by heating at temperature around 900 C) affected the propertiesof the fibers differently, when compared with those effects noted for air-dried fibers.With heat-dried puips the strength properties increased whereas the optical propertiesremained the same with a slight increase. This observation posed a question on thepossible development of the paper strength quality by adjusting the drying process. Weasked ourselves, is there any maximum strength development which could be reachedwith a proper combination of temperature and moisture in the fiber during recycling? Isthere any optimum combination of temperature and moisture which could minimize thedeteriorative effect of recycling? The plan for the work could be set to identify theseeffects and concomitantly to optimize the drying effect and recyclibility of the fibers.Handsheets could be made with different water content (namely different moisture).These handsheets could be conditioned in different drying temperatures. An optimum129condition for development of physical and optical properties and recycle potential of thefibers by suitable combination of moisture and temperature could be addressed.8.4 Implications of These Findings on Upgrading of Recycled FibersTo reverse the recycling effects on fiber properties to the original state (i.e., virginfiber state) requires reopening of the lamellae to increase the water uptake to the never-dried (unbeaten) state. This could be only achieved by strong swelling agents. Anoptimum combination of temperature, time and concentration of swelling agent is a basicrequirement.The second requirement for upgrading would be the condition of the fiber itself.Fibers low in hemicellulose content require swelling agents with lower affinity(solubilizing power) for the hemicelluloses. The loss of hemicelluloses in certain casesaie harmful for strength properties. Thus, attention should be paid to prevent loss ofhemicelluloses during the treatment. Swelling is the first condition for complete reopening of the lamellae.In the second stage of reconditioning swollen fibers the rigidity of the fiber wallcan be reduced by imposing mechanical fatigue. This could be achieved by refining.The feasibility of this procedure and the economic assessment of the costs forimplementation, need to be addressed.130REFERENCESArgue, G.H. and Maass, 0., Measurement of the heats of wetting of cellulose and woodpulp. Can. J. Res., 12: 564-574, (1935).Assaf, A.G., Haas, R.H. and Purves, C.B., A new interpretation of the cellulose-wateradsorption isotherm and data concerning the effect of swelling and drying on thecolloidal surface of cellulose. 3. Am. Chem. Soc., 66: 66-73, (1944).Atalla, R.H., The influence of elevated temperatures on structure in the isolation of nativecellulose. 3. Poly. Sci.: Poly. ed., 16: 601-605, (1978).Atalla, R.H., Structural changes in cellulose during papermaking and recycling. Mat. Res.Soc. Symp. Proc. 266: 10-14, (1992).Back, E.L., Discussion, , Trans. BPBJF Symp. Fiber-Water Interactions in Papermalcing.London, Vol. II: p. 873, (1978).Back, E.L. and Klinga, 0. Reactions in dimensional stabilization of paper and fiberbuilding board by heat treatment. Svensk Papperstidning, 66(15): 745-753, (1963).Back, E.L., Htun, T., Jackson, M. and Johnson, F. Tappi, 50(11) 542-546 (1967).Baker, C. and Fuller, G. Intermolecular forces and chain configuration in linearpolymers- The effect of N-methylation on the x-ray structures and properties of linearpolyarnides. J.Am. Chem. Soc., 65: 1120-1 128, (1943).Bangham, D.H. and Razouk, R.I., Adsorption and the wettability of solid surfaces. Trans.Faraday Soc., 75: 1459-147 1, (1937).Barbe, M.C., Kokta, B.V., Lavallee, H-C. and Taylor, J. Aspen pulping: a comparison ofsteam explosion and conventional chemi-mechanical pulping processes. Pulp and PaperCan. 9 (12): 142-151, (1990).Barkas, W.W., Wood water relationships-VU. Swelling pressure and sorption hysteresisin gels. Trans. Faraday Soc., 38: 194-209, (1942).Bouchard, 3. and Deuk, M., The effect of recycling on the chemical properties of pulps.3. Pulp Pap. Sci. 20(5): 3131-3136, (1994).Bovin, R.H., Harder, N. and Teder, A., Changes in pul quality due to repeatedpapermaldng. Paper Technol. 15(10): 261-264, (1973).Brickman, W.J., Dunford, H.B., Tory, E.M., Morrison, J.L. and Brown, R.K., Thereactivity of cellulose. II. Water sorption, heats of wetting, and the reactions with thallousethylate in ether, nitration mixture, and heavy water of cotton linters alternatively wettedwith water and dried. Can. 3. Chem., 31: 550-563, (1953).Browning, B.L., Methods of wood chemistry. V: I, Interscience Publisher, New York, p.202, (1967).Brunauer, S., Emmett, P.H., and Teller, E., Adsorption of gases in multimolecular layers.3. Am. Chem. Soc. 60: 309-319, (1938).131Bull, H.B., Adsorption of water vapor by proteins. J. Am. Chem. Soc., 66: 1499-1507,(1944).Burgess, H. D. Gel permeation chromatography: Use in estimating the effect of waterwashing on the long-term stability of cellulosic fibers. ACS, Historic Textile and PaperMaterials, 20: 363-376, (1986).Campbell, W.B., The mechanism of bonding. Tappi, 42(12): 999-1001, (1959).Campbell, W.B. and Pidgeon, L.M., Hydration of cellulose by beating. Pulp Pap. Mag.,6: 185-190, (1930).Campbell, W.B., The cellulose-water relationship in paper-making. Department of theInterior Canada, Forest Service, Bulletin 84, pp. 36-44, (1933).Caulfield, D.F. and Moore, W.E., Effect of varying crystallinity of cellulose onenzymatic hydrolysis, Wood Sci. 6 (4): 375-379, (1974).Caulfield, D.F., Trans. BPBIF Symp. Fiber-Water Interactions in Papermaking. London,Vol. I: pp. 63-67, (1978). (1977).Chakravarty, A.C., Change of cross-sectional area of single jute filaments with relativehumidity. Textile Res. 3., 28: 878-880, (1958).Chan, M., Breuil, C., Schwald, W. and Saddler, J.N., Comparison of methods forqualifying the hydrolytic potential of cellulase enzymes. Appi. Microbiol. Biotechnol.,31: 413-418, (1989).Christiansen, A. W., How overdrying wood reduces its bonding to phenol-formaldehydeadhesives: A practical review of the literature. Part I. Physical responses. Wood and FiberSci. 22(4): 441-459, (1990).Chung, F.H., Industrial applications of x-ray diffraction. American laboratory, 2: 144-156, (1989).Clark, J., d’A, “Pulp Technology and Treatment for Paper” , Second Ed., Miller FreemanPublications, pp. 615 (1985).Clark, I., d’A, New thoughts on cellulose bonding. Tappi, 67(12): 82-83, (1984).Collins, G.E., The swelling of cotton hairs in water and in air at various relativehumidities. J.Textile Inst. 21: T31 1-T315, (1930).Cottral, L.G. Hemicelluloses effect on papermaking. Tappi, 33 (9): 47 1-480, (1950).Cross, C. and Bevan, E.J. A textbook of papermaking. 5th edn. E.and F.N. Spon, London,p. 206, (1920).Curran, C.E., Simmonds, F.A., and Chang, H.M., Effect of beating upon certain chemicaland physical properties of puips. J. md. and Eng. Chem., 23: 104-108, (1931).132Davis, W. H. and Thompson, W. S., Influence of thermal treatment of short duration onthe toughness and chemical composition of wood. Forest Prod. J. 14 (8): 350-356,(1964).Dobbins, R., The role of water in cellulose-solute interactions. Tappi, .53(12): 2284-2290,(1970).Dole, M. and McLaren, A.D., The free energy, heat and entropy of sorption of watervapor by proteins and high polymers. 3. Am. Chem. Soc., 69: 65 1-657, (1947).Dunford, H.B. and Morrison, J.L., The heat of wetting of silk fibroin by water. Can. J.Chem., 33: 904-912, (1955).Dunning, C.E., Cell-wall morphology of longleaf pine latewood. Wood Sci. 1(2): 65-76,(1968).Eastwood, F.G. and Clarke, B., Trans. BPBIF Symp. Fiber-Water Interactions inPapermaking. London, Vol. II: pp. 835-855, (1978).Ehrnrooth, E., Htun, M. and de Ruvo, A., Trans. BPBIF Symp. Fiber-Water Interactionsin Paperma.king. London, Vol. II: pp. 899-915, (1978).Emerton, H.W., Fundamentals of the Beating Process. Marshal Press, London, pp. 84-137, (1957).Fengel, D., Ultrastructural behavior of cell wall polysaccharides. 53(3): 497-503, (1970).Frank, H. S. and Wen, W. Y., III. Ion-solvent interaction - Structural aspects of ion-solvent interaction in aqueous solutions: A suggested picture of water structure. FaradaySoc. 24: 133-140, (1957).Frey-Wyssling, A., The fine structure of cellulose microfibrils. Science, 119: 80-82,(1954).Frey-Wyssling, A., Laminar sorption and sweffing theory for wood and cellulose. WoodSci. Thech. 3: 301-323, (1969).Gallay, W., Fundamentals of Paper Making Fibers, Tech. Sect. B.P. and B.M.A., 5 8-65,(1958).Ghose, T.K., Measurement of ceilulase activities. Pure and Appi. Chem. 59, 257-268,(1986).Giertz, H., Trans. BPBIF Symp. Fiber-Water Interactions in Papermaking. London, Vol.II: p. 912, (1978).Gjonnes, J. and Norman, N., X-ray investigations on cellulose II and mixtures ofcellulose I and II, Acta Chem. Scand., 14(3): 683-688, (1960)Gottsching, L., Waste paper: a raw material of increasing significance. Paperi Ja PuuPaper and Timber. 4(3): 209, (1992).133Gray, D.G., Trans. BPBIF Symp. Fiber-Water Interactions in Papermaking. London, Vol.I: pp. 31-36, (1978).Grous, W.E., Converse, A.O. and Grethlein, H.E., Effect of steam explosion pretreatmenton pore size and enzymatic hydrolysis of poplar. Enzym. Microb. Technol. 8: 274-280,(1986).Guest, D.A. and Voss, G.P., Improving the quality of recycled fiber. Paper Tech. md. 24(7): 256-261, (Nov. 1983).Hagamassy, J., Jr., Brunaauer, S. and Milhail, R.S., Pore structure analysis by watervapor adsorption I. t-curves for water vapor. J. Colloid Interface Sci. 29: 485-49 1, (1969).Hall. W.H., X-ray diffraction. Proc. Phy. Soc., A62, p. 741, (1949).Hancock, W. V., The influence of native fatty acids on the formation of glue bondswith heat-treated wood., Ph. D. Thesis, University of British Columbia, Vancouver,Canada, p. 176, (1964).Harkins, W.D. and Jura, G., Surfaces of solids. XII. An absolute method for thedetermination of the area of a finely divided crystalline solid. J. Am. Chem. Soc., 66:1362-1366, (1944).Hartly, I.D., Kamke, F.A. Peemoeller, H., Cluster theory for water sorption in wood.Wood Sci. Technol. 26: 83-99, (1992).Hatakeyaina, H., Hatakeyama, T. and Nakano, 3. The effect of hydrogen-bond formationon the structure of amourphous cellulose. Appl. Polym. Symp. 28: pp. 743-752, (1976).Hatakeyama, T., Nakamura, K., Tajima, and Hatakeyama, H. International Paper PhysicsConference held in Mont-Rolland, Quebec, CPPA, Tappi, PITA, pp. 87-9 1, (1987).Hearle, J.W.S., The fine structure of fibers and crystalline polymers, I. Fringed fibrilstructure. J.Appl.Polym. Sci. 7: 1175-1192, (1963).Henry, F., Brandt, A. Ch. and Noe, P. Untersuchung der Wechselwirkung vonWasser und Cellulose in Gemahlenen Zelistoffen Durch mikrowellenspektroskopie. 42(9): 499-5 10, (1988).Hermans, P.H., Physics and Chemistry of Cellulose Fibers, Elsevier Pub. Co., Inc., N.Y.pp. 176-220, (1949).Hermans, P. H. and Weidinger, A. Quantitative x-ray investigations on the crystallinityof cellulose fibers. A background analysis. J. Appl. phys. 19 (5): 491-506, (1948).Hess, K., Mahl, H. ‘and Gutter, E. Electronenmikroskopische Darstellung grosserLangsperioden in Zelluiosefasem und ihr Vergleich mit den Perioden anderer Faserarten.Kolloid-Z. 155: pp. 1-19, (1957).Heyn, A.N.J., Crystalline state of cellulose in fresh and dried mature cotton fiber fromunopened boils as studied by x-ray diffraction. 3. Poly. Sci. Part A, 3: 125 1-1265, (1965).134Higgins, H.G., Trans. BPBIF Symp. Fiber-Water Interactions in Papermaking. London,Vol. II: p. 913, (1978).(64). Hillis, W. E., High temperature and chemical effects on wood stability. Part I,Wood Sci. Technol. 18: 281-293, (1984).Hindeleh, A.M. and Johnson, D.J. Crystallinity and crystallite size measurement inpolyamide and polyester fibers. Polymer, 19: 27-32, (1978).Hollenbeck, R.G., Peck, G.E. and Kildsig, D.O., Application of immersional calorimetryto investigation of solid-liquid interactions: microcrystalline cellulose-water system. J.Pharma. Sci., Vol. 67 (11): 1599-1606, (1978).Howard, R.C., The effects of recycling on paper quality. JPPS, 16(5): 143-149, (1990).Howard, R.C. and Bichard, W., The basic effects of recycling on pulp properties. J. Pulpand Paper Sci. 18 (4), J151-J159, (1992).Howarth, P., Skerry, A.M. and Mann, S. The effect of paper making on cellulose fibres.Paper Tech. md. 24(7): 126-130, (1983).Howsmon, J.A., in Cellulose and Cellulose Derivatives, 2nd Ed., Edited by Ott, E.,Spurling, H.M. and Grafflin, M.W. Interscience, New York. p. 45, (1954).Hunger, G., Trans. BPBIF Symp. Fiber-Water Interactions in Papermaking. London, Vol.II: p. 913, (1978).Ingram, P., Woods, D.K., Peterlin, A. and Williams, J.L., Never-dried cotton fibers, PartI: Morphology and Transport Properties. Text. Res. J. 44(2): 96-106, (1974).Jappe, N. A., Tappi, 41(5): 224-231, (1958).Jayme, G. and Hunger, G. in “Fundamentals of Papermaking Fibers”, ed. by Bolman, F.BPBMA, Kenley, Eng., (1958).Jones, G.L., Simulating end-use preformance. Recycling Paper: From Fiber to FinishedProduct, 152-160 (1991).Jura, G. and Harkins, W., Surfaces of solids. XIV. A unitary thermodynamic theory ofthe adsorption of vapors on solids and of insoluble films on liquid subphases. J. Am.Chem. Soc. 68: 1941-1952, (1946).Kauzman, W., Some factors in the interpretation of protein denaturation. Advances inProtein Chemistry, Vol. 14, pp. 1-63, (1959).Nemethy, G., and S,cheraga, H., Structure of water and hydrophobic bonding propteins.I. a model for the thermodynamic properties of liquid water. I. Phys. Chem., Vol. 66, pp.1773-1789, (1962).Kerr, A.J. and Goring, D.A.I., The ultrastructural arrangement of the wood cell wall.Cellul. Chem. Technol. 9(6): 563-573, (1975).Klye, R.C., The effect of drying in pulp strengths. Appita, 14 (6): xxi-xxxv, (1961).135Koning, J.W., and Godshall, W.D., Repeated recycling of corrugated containers and itseffect on strength properties. Tappi, 58 (9): 203-216, (1975).Kossler, I., Danhelka, M. and Netopilik, M. The carbanilate method for the determinationof the degree of polymerization of cellulose. Svensk paperstidin. 18: R137-r139, (1981).Krassig, H. and Kitchen, W., Factors influencing tensile properties of cellulose fibers, J.Polym. Sci., 123- 172, (1961).Krassig, H. Structure of cellulose and its relation to properties of cellulose fibers. 38 (12):57 1-582, (1984).Kulshreshtha, A.K., Patel, K.F., Patel, A.R., Patel, M.M. and Baddi, N.T., Thecrystallinity of never-dried cotton. Cell. Chem. Tech. 7: 343-346, (1973).Laivins, G.V. and Scallan, A.M., The mechanism of homification of wood puips. JPPS,(1994).Langford, J.I., A rapid method for analysing the breadths of diffraction and spectral linesusing the Voigt function. J. Appl. Cryst. 13: 10-14, (1978).Langmuir, I, The adsorption of gases on plane surfaces of glass, mica and platinum. J.Am. Chem. Soc. 40: 1361-1403, (1918).Lindstrom, T. and Carlson, G., The effect of carboxyl groups and their ionic form duringdrying on the hornification of cellulose fibers. Svensk Papperstidn., 85 (3): R146-R151,(1982).Lindstrom, T., The concept and measurement of fiber swelling, in ‘Paper structur andproperties. edited by Bristow, J.A. and Kolseth, P. Marcel Dekker, Inc. New York, pp.75-97, (1986).Lundberg, R. and de Ruvo, A., The influence of drying condition on the recovery ofswelling and strength of recycled fibers. Svensk Papperstidn. 11: 355-357, (1978).Maass, 0. and Campbell, W.B., Studies in cellulose moisture phenomena. Pulp and PaperMag. Can. 28: 108-114, (1939).Marton, R., Brown, A., Granzow, S. Keoppicus, R., and Tomlinson, S., Recycling andfiber structure. Progress in Paper Recycling, 2: 58-70 (1993).McComb, R.E. and Williams, J.C., The value of alkaline papers. Tappi, 64 (4): 93-96,(1981)McIntosh, D.C., The effect of refining on the structure of the fiber wall. Tappi, 50(10):482-488, (1967). /McKee, R.C., Effect of repulping on sheet properties and fiber characteristics. Pap. TradeJ. 155: 33-40, (1971).Meller, A., Paper Trade J., The cold alkali purification of wood ceillulose II: Theinfluence of surface properties on resistant pentosans. 125 (11): 57-60, (1947).136Mellon, E.F. Kom, A.H. and Hoover, S.R., Water absorption of proteins. IV. Effect ofphysical structure, 3. Am. Chem. Soc., 24: 2761-2764, (1949).Meyer, K.H. and Misch, L. Positions des Atomes dans le Nouveau Model Spatial de laCellulose. Helv. Chim. Acta, 20: 232-244, (1937).Milichovsky, M., A new concept of chemistry refining processes. Tappi, 73 (8): 22 1-232,(1990).Morrison, J.L. and Dzieciuch, M. A., The thermodynamic properties of the systemcellulose-water vapor. Can. J. Chem. 37: 1379-1390, (1959).Morosoff, N., Never-dried cotton fibers. III. Crystallinity and crystallite size. 3. Appi.Polym. Sci. 18: 1837-1854, (1974).Naito, T, Usuda, M. and Kadoya, T., International Paper Physics Conference, 197-201(1983).Nazhad, M.M. and Paszner, L. Fundamentals of strength loss in recycled paper. Tappi,77(9): 171-179, (1994).Nazhad, M. M., Ramos, R.P., Paszner, L. and Saddler, J.N. Structural constraints in theenzymatic hydrolysis of cellulosic substrates. Enzyme Microb. Technol., in press, (1994).Nissan, A.H., Trans. BPBIF Symp. Fiber-Water Interactions in Papermaking. London,Vol. I: p. 60, (1978).Nissan, A. and Batten, G.L.Jr. On the primacy of the hydrogen bond in paper mechanics,Tappi, 68(2): 159-164, (1990).O’Brien, J., U.S. paper industry sets 50 percent paper recovery goal for 2000. Paperage,2: 30-31, (1994).Okayama, T., Okada, Y. and Oye, R., Infuence of recycling on wood pulp fibers IV.Effects of conditions for water removal during recycling. Jappan Tappi, 36(3): 42-53,(1982).Oltus, E., Mato, 3. Bouer, S. and Farkas, V. Enzymatic hydrolsis of waste paper.Cellulose Chem. Tech. 21: 663-675, (1987).Ostberg and Salmen, Effects of fibrilation of wood fibers on their interaction with water,Nordic Pulp and Paper Research Institute, 12: 94-96, (1991).Page, D.H., The beating of chemical puips- the action and effects, Fundamentals ofpapermaking, Trans. 9th Symp. Edited by Baker, C.F. and Punton, V.W., London, Vol. I:pp. 121-152, (1989).Patrick, A. M., Physical Chemistry, Edited by Taylor, F., London, Vol. 11: p. 1278,(1924).Pommier, 3-Cl., Fuentes, J-L. and Goma, G. Using enzymes to improve the process andthe product quality in the recycled paper industry. Tappi, 72: 198-201, (1989).137Putz, H., Torok, I. and Gottsching, L., Making high quality board from low quality wastepaper. Paper Technology, 30(6): 14-20, (1989).Pycra.ft, C.J.H. and Howarth, P. A method of increasing enzyme degradation of cellulosefibers. Paper Tech. hid. 21(9): 283-285, (1980).Ramos, L.P., Nazhad, M.M. and Saddler, J.N., Effect of enzymatic hydrolysis on themorphology and fine structure of pretreared cellulosic residues. Enzyme Microb.Technol. 15: 821-831, (1993).Rivers, D.B. and Emert, G.H., Lignocellulose pretreatment: a comparison of wet and dryball attrition. Biotechnol. Lett. 9: 365-368, (1987).Robertson, A. A., Interaction of liquids with cellulose. Tappi, 53(7): 133 1-1339, (1970).Sachs, I.B., Retaining raised fibrils and microfibrils on fiber surfaces. Tappi 69 (11): 124-127 (1986).Salmen, L., Crstallinity effects on mechanical properties of H-bond-dominated solids.Comments on the recent article by Batten and Nissan. Tappi, 70(12): 190-193, (1988).Scallan, A.M., The structure of the cell wall of wood - A consequence of anisotropicinter-microfibrillar bonding. Wood Science, 6(3): 266-27 1, (1974).Scallan, A.M., Trans. BPBIF Symp. Fiber-Water Interactions in Papennaking. London,Vol. I: pp. 1-23, (1978).Scallan, A.M. and Tigerson, A.C., Swelling and elasticity of the cell walls of pulp fibers,J.Pulp Paper Sci. 18: 188-192, (1992).Schroeder, L.R. and Haigh, F.C. Cellulose and wood pulp polysaccharides. Gelpermeation chromatographic analysis. Tappi, 62 (8): 103-105, (1979).Scolz, M. and Flath, D., Determining structure of cellulose fiber substances with the aidof iodine sorption. Textilveredlung, 26(6): 188-191, (1991).Schwald, W., Chan, M., Breuil, C. and Saddler, J.N. Comparison of HPLC andcolorimetric methods for measuring cellulolytic activity. Appi. Microbiol. 28: 398-403,(1988).Seborg, C.O. and Stanim, A.J., Sorption of water vapor by paper-making materials, I.Effect of beating. md. Eng. Chem. 23(11): 1271-1275, (1931).Segal, L., Creely, J.J., Martine, Jr, A.E. and Conrad, C.M., An empirical method forestimating the degree of crystallinity of native cellulose using the x-ray diffractometer,Text. Res. J., 786-794 (Oct. 1959)Stamm, A.J., Adsorption in swelling versus nonswelling systems, II. Free energy changeper unit artea of effective molecular contact. Tappi 40 (9): 765-770, (1957).Stamm, A. J. Wood and Cellulose Science, Ronald Press Co., New York, 1964.Stamm, A.J. and Smith, W.E., Laminar sorption and swelling theory of wood andcellulose. Wood Sci. Tech. 3: 301-323, (1969).138Stockman, and Teder, A., The effect of drying on the properties of papermakin puips.Part 2. The effect of heat-treatment on the mechanical properties. Svensk Papperstidn.,66 (20): 822-832, (1963).Stokes, R.H. and Robinson, R.A., Procedure for control of relative humidity. md. Eng.Chem. 41: 2013-2016, (1949).Stone, J.E. and Scallan, A.M., A structural model for the cell wall of water-swollen woodpulp fibers based on their accessibility to macromolecules. Cell. Chem. Technol. 2(3):343-358, (1968).Stone, J.E. and Scallan, A.M., A study of cell wall structure by nitrogen adsorption. Pulpand Paper Mag. Can., 65 (7): T407-T414, (1965).Strachan, J., Proc. Tech. Sec., BPBMA, 6(2): 139-142, (1926). (Cited in Emerton 1957)Strazdins, E., Factors affecting retention of wet-end additives, Tappi, 53 (1): 80-83,(1970).Szwarcsztajn, E. and Przybysz, K., Investigation on changes in the properties of recycledpulps fractions. Cell. Chem. Technol. 10: 737-749, (1976).Tam Doo and Kerekes, The effect of beating and low-amplitude flexing on pulp fiberflexibility. JPPS. 15(1): J36-J42, (1989).Thode, E. F., Chase, A. J. and Hu, Y., Dye adsorption on wood pulp. Tappi, 38 (2): 88-89, (1955).Tie-Qiang, L., Henriksson, U., Eriksson, J.C. and Odberg, L. Fundamentals ofPapermaking. Ed. Baker, C.F. and Punton, V.W., Mech. Eng. Pubi. Inc. London. pp. 39-45, (1989).Troughton, G. E. and Chow, S. Z., Migration of fatty acids to white spruce veneer surfaceduring drying: Relevance to theories of inactivation. Wood Sci., 3 (3): 129-133, (1971).Urquhart, A.R., Adsorption Hysteresis, II. Shirley Institute, Memoires, 8: 19-26, (1929).Usuda, M. Surface of cellulose fiber. Japan Tappi, 57(2): 423-433, (1982).Valentine, L. Studies on the sorption of moisture by polymers, I. Effect of crystallinity.Text. Res. J. 24: 313-333, (1957).Vakasaari, L. and Saarela, K. Determination of chain length distribution of cellulose bygel permeation chromatography using the tricarbanilate derivative. Cell. Chem. Tech.,21, 663-672, (1975).Verwey, E.J.W., and Overbeek, J.Th.G., “Theory of the Stability of Lyophobic Coiloids”,Elsevier Publishing Co., Amesterdam, (1948).Verwey, E.J.W., Trans. BPBIF Symp. Fiber-Water Interactions in Papermaking. London,Vol. I: pp. 55-63, (1978).Wahba, M., Moisture relationships of cellulose. II The heats of wetting of partiallysaturated viscose rayon and standard cellulose in water. J. Phys. Colloid Chem. 54: 1148-1160, (1950).139Warddrop, A.B. The phase of lignification in differentiation of wood fibers. Tappi, 60(4):225-243, (1957)Weatherwax, R.C., Transient pore structure of cellulosic materials. J. Colloid InterfaceSci. 49(1): 40-47, (1974).Wood, B.F., Conner, A.H. and Hill, C.G., Jr. The effect of precipitation on the molecularweight distribution of cellulose tricarbanylate. 3. Appi. Polym. Sd. 32: 3703-3712,(1986).Whitaker, J.R. Analytical application of enzymes. Biotech. Bioeng. 26: 31-38, (1985).Wims, A.M. and Mayers, M.E. Principles of crystallinity measurements. Ad. X-ray Anal.29: 281-290, (1986).Williams, J.C., Retaining the strength of secondary fibers with alkaline calciumcarbonate fillers. Paper Trade J. 164(22): 33-34 (1980).Yamagishi, Y. and Oye, R. Influence of recycling on wood pulp fibers - changes inproperties of wood pulp fibers with recycling. Jappan Tappi, 56(9): 33-43, (1981).Yamashild, T., Matsui, T., Saitoh, M., Matsuda, Y., Okajima, K. and Kamide, K.Characterization of cellulose treated by the steam explosion method. Part Ill: Effect ofcrystal forms (cellulose I, II, and III) of original cellulose on changes in morphology,degree of polymerization, solubility and supermolecular structure by steam explosion.British Polym. J. 22: 201-212, (1990).Zeronian, S.H., Heat-induced changes in the properties of cotton fibers, Preservation ofpaper and textile of historic and artistic value. J.C.Williams (Ed.), Advances inChemistry Series, Washington DC, 164: pp. 189-205, (1977).Zettlemoyer, A.C., Young, G.J. and Chessick, A. Studies of the surface chemistry ofsilicate minerals, III. Heats of immersion of bentonite in water. J. Phys. Chem. 59: 962-966, (1955).Zimm, B.H., Simplified relation between thermodynamics and molecular distributionfunctions for a mixture. J.Chem. Phys. 21: 934-935, (1953).140APPENDIX APeak profile resolutionIn order to obtain information on the atomic level such as crystal structure,crystallinity, residual stress, crystallite size, etc by XRD mathematical functions arerequired to represent the peak profile. It has been experimentally verified that, for theline broadening from imperfections or growth of small crystallites, neither Cauchy norGaussian functions are necessarily good approximations (Chung 1989). A betterapproximation for XRD peaks is obtained by convolution of one or more Cauchy andGaussian functions. The synthesis of the convolution of Cauchy and Gaussian functionsis called the Voigt function and was applied to the analysis of X-ray diffraction byLangford (1978) and used in this paper to calculate the crystallinity and crystallite size.The Cauchy function is given by:10(x) = {A.1Jwhere= nw [A.2jand f3 is integral breadth, 2w0 is full width at half maximum (FWHM) and x ispeak position.The Gaussian function is given by:Iq(X) = Texp(7tx2/(g)2) [A.3]where13g = Wg(lt/1fl2)h1’2 [A.4]and 3g is the integral breadth and is FWHM for the Gaussian function.141From the two functions of I (x) and tg Cx) of Cauchy and Gaussian oneforms their convolution Io*Ig = I as the function I (x) such that:(001(x) Ic(U)Ig(XU)dUO0[A.5Jwhere I, and tg are the Cauchy and Gaussian components of the Voigt function. Theequations [A.l], [A.3J and [A.5] have a Fourier transform. An approximate value for theFig. 33. The resolution of diffraction peaks in terms of Voigt profiles for each peak.integral breadth, of equation [A.5j is deduced as0 40142the integral breadth of the Voigt function() isf32 I13 + (3g)2 [A.6]and the full width at half maximum (FWHM), 2w, of the Voigt function isapproximated in the form of2w {[()2 + (f30)2]_l/2}/p,• [A.7]The crystallinity is measured as the ratio of the total area under the resolved peaksto the total scatter under the diffractogram. The apparent crystallite size also could beobtained by using either the integral breadth or FWHIvI. To obtain the actual crystallitesize, the apparent size values must be multiplied by the appropriate Scherrer constant(1949). Fig. Al shows how overlapped peaks of diffractogram could be resolved by theVoigt function into a single peaks. It is worth mentioning that a computer program isavailable commercially to search for better fit by using the Voigt function.143APPENDIX BDerivation of the formula for external surface area measurementThe heat of wetting or diminution of heat content, which accompanies immersionin the liquid, is given by the Gibbs-Helmholtz equation:(I) /E = (FL — TdFL/dT) [B. 1 ‘FL represents the surface-energy lowering produced by immersion in the liquid.FL = — sL[B.2]where ? is the free surface energy of the solid, L the free interfacial energy of thesolid-liquid interface, and T is the absolute temperature.According to the first law of the thermodynamics, the energy released will be thesame whichever way we proceed. Thus, zH (s/L) /E could also be obtained by thesum of heat of adsorption (LHa) and heat generated when the solid, already saturated,is immersed into the liquid.F is the surface-energy lowering produced by exposure to the saturated vapor, definedby:—[B.3]where stands for the surface free energy of unit area of the solid-vapor.FL and F are related by the following equations:FL=FV+?Vcos8 [B.4] ordFL/dT = dF/dT + dRLV cose)/dT [B.5]where 0 is the contact angle of the liquid on the solid.Combining equations [B.4j and [B.5j with equation [B.l] we havezMI (s/L) /E = (F — TdF/dT) + (?,Lv cosO - Td[? cosO] /dT)[B. 6)144where (F,— TdF/dT) is the heat of adsorption Ha, and the term in the secondbracket is the heat generated by immersion of an already saturated solid. The heat ofimmersion of this wet solid is designated by (sf/L) /Z where f indicates that thesolid possesses a “film” of adsorbed liquid. Thus,tH (sf/L) / = (, cosO — Td[) cosO]/dT) [B.7JIn the special case where q = 0, the equation [B.7] would be the total surface energy ofthe normal liquid. SoAH (sf/L) / = (2, — Td[?] /cIT) [B.81The right hand side of equation [B.8] is the surface-energy of the normal liquid. If theliquid is water, thus the surface-energy of water is 118.5 ergs/cm2,thus= AHj(sf,L)/llB.5 [B.9]The surface area obtained in this manner is an estimate of the external surfacearea of the solid material since the capillaries and some intra-aggregate spaces are filledwith water.145APPENDIX CIntegral and differential free energy changes.At a given temperature, the free energy changes of a system, LP, associated withthe adsorption of a adsorbate on an adsorbent is given as:Ar =n4F1 + n2AF [C. 11where AF1 and AF2 are the partial molal free energy changes of the adsorbate andadsorbent, respectively, and n1 and n2 are the numbers of moles of each, respectively.The differential free energy change for the vapor, AF1, is given as the differencebetween the chemical potential of the liquid in the adsorbed state and pure liquid.AF1 = U1 - = RT1n(h) [C.2]where u1 and u10 are the chemical potentials of the adsorbate and pure liquidrespectively, h is the relative vapor pressure, T is absolute temperature and R is thegas constant.The‘2 is obtained by the Gibbs-Duhem relation:AF2 = -(n1/n2)d(AF [C.31An expression for the integral free energy change can be obtained by substitutionof Gibbs-Duhem equation and Eq. [C.2j into Eq. [C. 1]. Thus, the integral free energychange for water vapor accompanying the sorption process is:Ar =n1RTln(h) - RT {n1d[ln(h)]} [C.4]146Evaluation of the integral on the right-hand side of Eq. [C.3] can be performedgraphically from a plot of n1/h versus h by replacing d[In(h)] with its derivative(1/h)d(h) in Eq. [C.41.=n1RTin(h) - RT (n1/h)d(h) [C.5]The error associated with low pressure extrapolation to P = 0 is smaller with thistechnique than with a plot of n1 versus in (h).The differential free energy change has units of energy per mole of adsorbate, andthe integral free energy change simply has units of energy, because the entire process isspecified for a certain number of moles of each component.I147APPENDIX DMethod for determining the integral and differential enthalpyFigure 23 is an schematic representation of the enthalpy changes associated withthe wetting of a solid by a liquid.Step 1 = Represents the immersion of a dry clean solid into a liquid. The heat ofimmersion of this solid is designated by zH (s/L)Step 4 = Demonstrates the immersion of a solid covered by a film, and it is representedby IHj(sf,L)According to the first law of thermodynamics, the energy released will beindependent of the path. Therefore, Step 1 is equal to the sum of the Step 2 to 4.(six.) = fl1A + /Ha + AHj (sf/L) [D. 1]where 2. is the molar heat of vaporization of the liquid and AHa is the heat ofadsorption.The net integral enthalpy change associated with the adsorption process is theheat change in excess of the normal condensation of the liquid.=—(—n1) [D.21Thus, from Eq. [D.1]H =zH(51)—AH(5f/) [D.3]It is to be noted that in Eq. [D.31 both of the terms are measurable by thecalorimetric method as reported.The integral enthalpy An, thus would be generated by measurement of the heatsof immersion of solids with varying moisture contents. This procedure renders the entirerange of iS.H values. The differential enthalpy change is obtained by estimating theslope of IH versus n1.148clean solidin vacuum1+ flj molesof watervapor-4 solid withn1 mole ofadsorbed waterFig. 34. Conceptual drawing of a sample of pulp in equilibrium with the vapor of aliquid. To be immersed in the liquid and thus lose the surface energy of the duplex filmof liquid.21I II II IL_ __..J149APPENDiX EInternal tension of liquid between two parallel surfacesAssume two parallel lamellae with a distance x apart. As water evaporates, thewater between the lamellae assume the shape droplets. Thus, the liquid appears as aflattened drop having a radius of r in the plane of lamellae. Let the volume of the dropbe V and its surface tension 1 dynes per cm. Further, it is assumed that the liquid wets thelamellae, so that the edge of the liquid meets the lamellae at zero angle.b2rjFig. 35. Conceptual drawing of water between two lamellae.If the lamellae displaced an increment of dx, the change in potential energy of thesystem would be:Fdx = )dS1 + (sL - = dS1 - ?4S2 [E.1]where F is the force, 2L, and is the surface tension of liquid, solid-liquid andsolid-gas, cfS is the change in the area of the liquid surface and c1S2 is the change in thearea of the dry solid.The area of the liquid-gas interface equals: S1 = it2rxThe area of wetted lamellae surface equals: S2 = 2itrdS1/dd.x = d/dd,.c (iu2rx), where V = itr2x, or rx =150dS1/dx = (2/2)’j/x = t2r/2 [E.2janddS2/dx = _2V/x _2itr2/x [E.3jThus the force drawing two lamellae under tension forces would beF = (dS1/ x - dS2/dx) = )(it2r/2 - 2itr/x) [E.4JIf the value of x is to small in comparison with rF = 2itr2/x [E.5]thusInternai tension = Fhr2 = 22Jx dynes/cm2= 2Jxg grams/cm2orEquation [E.5] shows that a drop of water with sorption tension of 75 dynes/cm is heldbetween two lamellae with a diameter of 1 cm increase with decrease in the distancebetween the lamellae. The internal tension force is enough at certain distances to pull theadjacent surfaces toward each other enough to make the operation of molecular forcespossible. Campbell (1933) was the first to identify the role of these forces on bonddevelopment of a paper sheet.Table 13. Development of internal tension forces between the lamellae during removal ofwater from the cell wall.P/P0 Internal tensb of Th1CCSS fwatevwater (N/cm) between the laycrs(em)0:99 I6 Lix IO49x IO+ (> /936031 JOG jØ4151APPENDIX FCalibration curve for GPCIt has already been stated that the GPC retention volume gives a measure of thesize of the dissolved molecule. The exact correlation exists through use of standardnarrow MWD samples having a known molecular weight at their peak retention volumes.However, for the higher molecular weight polymers only polystyrene has been producedon a sufficiently large scale having a narrow enough MWD for use as calibrationstandards. For other polymers a method of indirect calibration is required which canconvert the polystyrene calibration to one applicable to the measured polymer. The GPCseparation is determined by the size and shape of the polymer molecule in solution i.e.,its hydro-dynaniic volume. Different polymers of the same molecular weight can havedifferent conformations in a given solvent. Also a polymer will assume different sizes indifferent solvent media at different temperatures. The hydrodynamic volume (HV) of apolymer is measured asliv = [Tfl zc M [F.1}where [1] is intrinsic viscosity, and M is molecular weight. This parameter forms thebasis of the “universal” calibration. The Mark-Houwink relationship explains theinterrelation of M and [ii] empirically as follow:[ii] = K x Ma [F.2]The hydrodynamic’volume (HV) of polystyrene and cellulose tricarbanylate (CTC) isgiven by:liv of polystrene = [ri]1, x = (1 + CLI,)152HVOfCTC []cxMcccMc (l+cz)1W of polystrene = WI of CTC, thusKpMp (1 + a) = K0M (1 + a0) or1flK + (1 + a)lflM = inK0 + (1 + cz0) inN0, orifl(K/K0) + (1. + cz,)inM = (1 + a0) inN0 or= [ifl(K/K0) + (1 + CXp)iflNp)/(1 + CL0) [F.3]Mark-Houwink coefficients are obtained from the literature and manipulated in Eq. [F.3]and calibration curve will be. built accordingly for CTC samples. The Eq. [F.3] is validfor other polymers, too.I153GLOSSARYBeating: Mechanical treatment of wet fibers which brings about desirable modificationsto the sheet properties.Cellulose: The chief substance in the cell wall of plants. It is the fibrous substance thatremains after the non-fibrous portions, such as lignin and some carbohydrates,have been removed during the cooking and bleaching operation of a pulp.Chemical pulp: Wood pulp produced by removing lignin during the cooking of woodchips in a digester. Kraft (sulphate) or suiphite process are examples of chemicalpulping process.Co-crystallization: Uniting of two crystalline cores of the two adjacent elementary fibrilsis called co-crystallization.Crystallinity Index: The percent crystalline material in the total fiber is expressed as“crystallinity index’.Dried Pulp Fiber: A fiber which has been dried at least once.Enthalpy Change: The heat flow in any reversible isobaric process is equal to the changein enthalpy. Therefore, the heat of transformation in any change of phase is equalto the difference between the enthalpies of the system in the two phases. If asystem performs a cyclic process, the initial and final enthalpies are equal and thenet enthalpy change in the process is zero.Entropy: Entropy in the present context is interpreted in terms of disorder or randomnessof a system.External Fibrillation: Disruption of the outer layers of fibers during beating is referred asexternal fibrillation. As a result of this increase in external surface, the areaavailable for contact and bonding with adjacent fibers increased.Fiber Wall: Fiber wall consists of concentric lamina made up of small repeating units 100A by 100 A, consisting of a microcrystalline core surrounded by an amorphoussheath.Flexibility: It is defmed as the inverse of the product of the moment of inertia of the bodyand the modulus of elasticity of the material from which the body is made.Flickering Cluster: The term is used to explain the co-operative nature of clusterformation and dissolution in a liquid water at room temperature, and it isgoverned by local energy fluctuation.Free Energy: The free energy of a system could be defined as the maximum energy freedin a process and made available for work.Hornification: frreversible loss in swelling potential of the fibers is referred ashornification.Internal Fibrillation: Increase in volume during beating generally termed swelling orinternal fibrillation.154Kraft Pulp: Wood pulp produced by the sulphate chemical process using cooking liquorwhich is made up primarily of sodium hydroxide (NaOH) and sodium, suiphide (Na2S). Also known as sulphate pulp.Laminar Structure of Fibers: Wood fibers are made up of a series of concentric layerswith variations in thickness, fibril orientation and chemical composition.Lignin: A brown-colored organic substance which is separated chemically during thecooking process which releases cellulose fibers. It is removed along with otherorganic materials in the spent liquor during subsequent washing and bleachingstages.Never-Dried Pulp Fiber: A fiber which is collected from a fresh tree and neverexperienced drying.Recrystallization: The concept of recrystallization which is frequently evoked to explainthe increase in crystallinity index after recycling.Recycled Fiber: A pulp fiber which has been rewetted from the dry state and made intopaper at least once.Secondary Fiber: Pulp recovered from a paper product which has already served acommercial purpose. Secondary fiber may be obtained from an intermediateprocessor, such as a printer or converter, from a final user, such as a homeowneror from its end destination; the municipal dump.Surface Deactivation: The loss in surface bonding potential of the fibers in the presentcontext is referred as surface deactivation or surface inactivation.Thermodynamics: Thermodynamics is concerned with energy relationships involvingheat, mechanical energy, and other aspects of energy and energy transfer.


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items