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

Download

Media
831-ubc_1995-983302.pdf [ 5.75MB ]
Metadata
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
831-1.0088893-fulltext.txt
Citation
831-1.0088893.ris

Full Text

FUNDAMENTALS OF STRENGTH LOSS IN RECYCLED PAPER  by Mousa M. Nazhad  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF Doctor of Philosophy  In  THE FACULTY OF GRADUATE STUDIES DEPARTMENT OF FORESTRY  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA December, 1994 Mousa M. Nazhad, 1994  In presenting this thesis in partial fulfillment of the requirements for an advanced degree at The University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying scholarly purposes may be granted by the Head of my department or by his or her representative. It is understood that copying or publication of this thesis for fmancial gain shall not be allowed.  (signature)  Faculty  t1itof Forestry The University of British Columbia Vancouver, Canada Date: December 22, 1994  ii  ABSTRACT Considerable work has been devoted to the upgrading of recycled chemical (low yield) pulp fibers during the past decade. There is also disagreement on the effectiveness of an upgrading process regardless, whether of chemical or mechanical origin.  One  ng is serious problem which restricts sustainable progress in the field of fine-paper recycli and the lack of knowledge of the mechanism by which recycling affects the texture d arrangement of the cell wall which ultimately causes inferior properties of the recycle fibers. The deteriorative effect of recycling on fine-paper manifested itself on the loss in fibers potential bonding of recycled fibers. The loss in potential bonding of the recycled ation translated into homification (i.e., loss in fiber wet-flexibility) and/or surface deactiv than surface by recycling. The susceptibility of the fibers for hornification rather concluded deactivation during recycling is substantiated with different techniques. It is that the hornification is responsible for inferior properties of recycled fibers.  More  does not importantly, observations in the present work suggest that refining/beating on in the develop any new surface area. The effect of refining is restricted to a reducti g and rigidity of the lamellae by mechanical fatigue and subsequently, increased swellin ten or beaten) plasticization of the fiber wall. Thus, drying of never-dried fibers (unbea binds the from water pulls the lamellae toward each other by surface tension forces and e in the lamellae rich in surface by crystallization forces. These forces lead to an increas lization, is crystallization of the cell wall provided that the condition required for crystal ted again, met by the molecular orientation in the cell wall. When these fibers are re-wet ly closed. the delamination does not reverse completely, and the lamellae remain partial l surfaces This results in increased rigidity of unraveled lamellae and restricts the interna loss of fibers to access by water. The concomitant result is restricted swelling and thus, first in wet-plasticity of the fibers on recycling. Most of this change takes place in the on cycle. Repeated recycling deteriorates further the wet-plasticity of the fibers. Based  iii  these fmdings a model is proposed which explains the mechanism by which homification develops in the fiber wall during recycling.  The proposed model also provides new  information on the effects of fiber beating or refining.  I  iv  TABLE OF CONTENTS  ABSTRACT  ii  TABLE OF CONTENTS  iv viii  LIST OF FIGURES LIST OF TABLES  x  LIST OF NOMENCLATURE  xi  ACKNOWLEDGEMENT  xiv  1. INTRODUCTION  1  2. THEORETICAL BACKGROUND  4  2.2 Correlation of Cell Wall Structure and Papermaking Qualities of Fibers  4  2.1.1 Ultrastructural arrangement of the fiber wall  4  2.1.2 Crystallinity concept in polymers  7  2.1.3 Advantage of the X-ray diffraction (XRD) method  8  2.1.4 Development of XRD method for cry stallinity measurement  9  2.1.5 Crystallinity index  10  2.2 Beating (Refining) 2.2.1 Theory of the beating (refining)  11 12  2.2.1.1 Chemical theory  12  2.2.1.2 Physical theory  13  2.3 Surface area of fibers  16  2.3.1 Adsorption technique to measure the fiber surface area  17  2.3.2 External and internal surface area  19  2.4 Thermodynamics of Fiber-Water Interactions  20  2.4.1 Free energy changes  21  2.4.2 Enthalpy changes  22  V  2.4.3 Entropy chges.23 2.5 Enzymatic Hydrolysis of Recycled Fibers  24 25  3.LITERATLJRE REVIEW 3.1 Introduction  25  3.2 Postulated mechanisms for loss of strength in recycled paper  26 27  3.2.1 Fiber Flexibility 3.2.1.1 Irreversible pore closure  27  3.2.1.2 Cross-linking  28  3.2.1.3 Cellulose chain cleavage  29  3.2.1.4 Re-organization in the fiber (cell) wall  30  3.2.1.5 Bond strength  32  3.2.2 Surface Condition of Fibers  33  3.2.2.1 Hemicellulose-loss effect  33  3.2.2.2 Inactivation of the fiber surface  34  3.2.2.3 Microcompressions  36  3.2.2.4 Fiber-Water Interactions  36  3.2.2.4.1 The electric charge of pulp surfaces in water. .37 3.2.2.4.2 Changes in water structure due to fiber-water interactions  38  3.3Summary  42  4. MAThRIAL AND METHODS  44  4.1 Preparation of Pulp Samples  44  4.2 Bleaching Materials  44  4.2.1 Preparation of chlorine  44  4.2.2 preparation of chlorine dioxide  45  4.2.3 Laboratory bleaching process  46  4.3 Recycling procedure  48  vi  4.4 Fiber Length (FL) a1ysis  .  51  4.5 Molecular Weight Distribution (MWD)  52  4.6 Crystallinity and Ciystallite Size of the Fibers  54  4.6.1 D 5000 Diffractometer  54  4.6.2 Sample preparation for x-ray diffraction analysis  56  4.6.3 X-ray data analyses  57  4.7 Water Vapor Sorption  57  4.8 Measurement of the Heat of Wetting  61  4.9 Scanning Electron Microscopy (SEM)  63  4.10 Determination of Thermodynamic Properties of Fiber  64  4.11 Enzymatic Hydrolysis  65  5. RESULTS  66  6. DISCUSSION  94  6.1 Adsorption Isotherms  94  6.2 Heat of Immersion  96  6.3 Surface Analysis of Recycled Fibers  97  6.4 Thermodynamics of Fiber-Water Interactions  99  6.4.1 Free energy99 6.4.2 Differential enthalpy and hydrogen bonding  101  6.4.3 Entropy  102  6.5 Changes in Fiber Wall Ultrastructure due to Recycling  105  6.6 Enzymatic Hydrolysis  109  6.7 Theory of Mechanism of Strength Loss in Recycled Papers  112  6.8 Proposed Model for the Evolution of Fiber Wall Ultrastructure on Recycling Process 7. SUMMARY AND CONCLUSIONS  116 119  vii  119  7.1 SUMMARY  119  7.1.1 Beating  120  7.2.1 Recycling  121  7.2 Conclusions  125  8. RECOMMENDATIONS in Homification During the 8.1 Identification of the Parameters Involved  125  Bleaching Process the Cell 8.2 Role of Hemicellulose in Hornification of  Wall  125  8.3 The Role of Temperature in Recycling  126  g of Recycled Fibers 8.4 Implications of These Findings on Upgradin  127  9 BIBLIOGRAPHY  128  APPENDICES 138  A  Peak profile resolution area measurement Derivation of the formula for external surface  141  B  Integral and differential free energy changes  143  C  145  E  rential enthalpy Method for determining the integral and diffe llel surfaces Internal tension of liquid between two para  F  Calibration curve for GPC  D  GLOSSARY VITA  147 149 151  viii  LIST OF FIGURES  Fig. 1. Pictorial representation of the lamellae model for the ultrastructural arrangement 5 of lignin, cellulose and hemicelluloses in the wood cell wall Fig. 2. Schematic representation of the swelling of the fiber wall (after Scallan 1974).. .6 Fig. 3. Change in flexibility of fiber by acidity of paper. The fiber retains its flexibility 30 when it is in a neutral or slightly alkaline environment Fig. 4. Diagram showing the adherent solvent layer on surface when a surface is submerged in a liquid  38  Fig. 5. Schematic cross-section of a broken hydrogen-bonded water near cellulose surface  39  Fig. 6. Schematic representation of the adsorbed water between two cellulosic surfaces. The network was demonstrated as a distorted ice-like cluster with extension of 41 hydrogen bonds from one surface to other Fig. 7. A sketch of chlorine dioxide preparation  45  Fig. 8. A flow chart of the experimental procedure  50  Fig. 9. Block diagram of FS-200 fiber length analyzer  51  Fig. 10. Block diagram of the diffractometer in D 5000 mode  54  Fig. 11. Diffractometer beam path in 0/28 mode  55  Fig. 12. Block diagram of apparatus which was developed for conditioning the fibers at 60 different relative humidities with a constant temperature (i.e., 25±2° C) Fig. 13. A cross-section of the PARR 1451 solution calorimeter  62  Fig. 14. Changes in properties of the handsheets (made from beaten bleached kraft) due 67 to recycling Fig. 15. Comparison of fiber length (FL) distribution of unbeaten, beaten and recycled 68 fibers Fig. 16. SEM microphotographs of (a) unbeaten and (b) beaten pulp fibers  69  Fig. 16. SEM microphotographs of (c) cycle I and (d) cycle VI pulp fibers  70  Fig. 17. Comparison of the adsorption isotherms of unbeaten, beaten and recycled fibers  72  Fig. 18. Comparison of sorption isotherms of cycle I and cycle V  73  Fig. 19. Desorption isotherms of unbeaten and beaten (6000 and 12000 revs.) fibers. .75  ix Fig. 20. Comparison of critical point drying and freeze drying methods with sorption 76 potential of the fibers at 25° C Fig. 21. Comparison of the heat of wetting of unbeaten, beaten and recycled fibers  77  Fig. 22. Comparison of MWD of unbeaten, beaten and recycled fibers  78  Fig. 23. Comparison of the diffractograms of unbeaten, beaten and recycled fibers  80  Fig. 24. Comparison of the diffractograms of virgin (beaten), cycle I and cycle Ill pulp Fig. 25. Comparison of the differential free-energy of virgin and recycled fibers  83  Fig. 26. Comparison of the differential enthalpy of virgin and recycled fibers  84  Fig. 27. Comparison of the differential entropy of virgin and recycled fibers  85  Fig. 28. Comparison of scanning electron photomicrographs of hydrolyzed substrates 89 derived from (a) unbeaten and (b) puips Fig. 28. Comparison of scanning electron photomicrographs of hydrolyzed substrates 90 derived from (c) cycle I and (d) cycle VI puips Fig. 29. Comparison of MWD of hydrolyzed samples  92  Fig. 30. Comparison of the integral free-energy of virgin and recycled fiber  100  Fig. 31. Comparison of the integral entropy of virgin and recycled fiber  103  Fig. 32. Conceptual drawing of the cell wall ultrastructural arrangement in unbeaten (A), 116 beaten (B), dried (C) and recycled (D) states Fig. 33. The resolution of diffraction peaks in terms of Voigt profiles for each peak. 139 Fig. 34. Conceptual drawing of a sample of pulp in equilibrium with the vapor of a liquid. To be immersed in the liquid and thus lose the surface energy of the 146 duplex film of liquid Fig. 35. Conceptual drawing of water between two lamellae  147  K  LIST OF TABLES  Table I. Relationship of sulfuric acid concentration with relative humidity at 25° C... .58 Table II. Relative humidities over standard salt solutions at 250 C  59  Table III. Characteristics of the white spruce kraft pulp (unbleached and bleached (CEDED))  66  Table IV. Fiber length (FL) analysis of virgin (unbeaten and beaten ) and recycled fibers Table V. Surface area for virgin (unbeaten and beaten) and recycled fibers  74  Table VI. Degree of polymerization (DP) of the (unbeaten and beaten) and recycled 79 fibers Table VII. Crystallinity information on virgin (beaten) and recycled puips  81  Table Vifi. Calculated thermodynamic properties for the adsorption of water on beaten, 86 unbeaten puips dried by the CP drying method Table IX. Calculated thermodynamic properties for the adsorption of water on recycled 87 puips dried by the CP drying method Table X. Fiber length (FL) analysis of hydrolyzed (H) puips  88  Table XI. Degree of polymerization (DP) and hydrolysis yield of the hydrolyzed (H) 91 puips Table XII. Crystallinity of unhydrolyzed (U) and hydrolyzed (H) pulp samples  93  Table XIII. Development of internal tension forces between the lamellae during removal 148 of water from the cell wall  xi  LIST OF NOMENCLATURE  20:  Bragg’s angle  I:  Intensity of diffraction  CrI%:  Crystallinity Index (%)  FWHM:  Full width at half maximum Integral width (Voigt function) Integral width (Cauchy function)  g: 13  Integral width (Gaussian function) Cauchy function  I(x):  Gaussian function  1(x):  Voigt function  h:  Relative vapor pressure  M:  Moisture adsorbed at the relative vapor pressure h  Mm:  Volume of vapor adsorbed in the monolayer of adsorbate  C:  A constant depending on the heat of adsorption  ET:  Total surface area Internal surface area External surface area  a:  Effective area occupied by an adsorbed molecule (for water a  N:  Avogadro’s number (6.02 x 1023)  M:  Molecular weight of adsorbate  : 1 E  Heat of adsorption  EL:  Heat of condensation of the vapor  =  Heat of immersion of the sample equilibrated at 100% relative humidity Heat of immersion of a dry clean solid  ) 2 12.5 A  xii  R:  Universal gas constant (1.98662 calJmole.K)  T:  Absolute temperature (Kelvin temperature scale: 273.16 K + temperature )) 0 of water (T  p.:  A constant related to the amount of vapor already adsorbed  : 1 AF  Differential free energy  AF:  Integral free energy  : 1 n  Moles of water per 100 g of solid  H:  The free energy of adsorption per unit surface area Free surface-energy of the solid vocuum interface Free energy of the surface when the solid-vapor interface is constructed  : 1 zH  Differential enthalpy  z\H:  Integral enthalpy  : 1 AH  Deferential entropy Integral entropy  M:  Molecular weight of cellulose tricarbanylate  M:  Molecular weight of polystyrene  K:  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 approach  t\Hcal:  Bonding energy between water and fiber by Calorimetry approach  AHa  Heat of adsorption  DP:  Degree of polymerization  MWD:  Molecular weight distribution  FL:  Surface energy lowering produced by immersion in the liquid Surface energy of the solid  xiii  Surface energy of the solid-liquid interface F:  Surface energy lowering produced by exposure to the saturated vapor  0:  Contact angle  : 1 U  Chemical potential of the liquid at the adsorbed state  : 0 U  Chemical potential of a pure liquid  F:  Force between two lamellae under tension forces Surface tension of liquid  sg  Surface tension of solid-gas interface  11:  Intrinsic viscosity  HV:  Hydrodynamic volume  xiv  ACKNOWLEDGMENTS  As Rome is not built in one day, building the character of a man also is not built from a single act. teachers.  The cornerstones of becoming mature are parents, friends and  My parent’s love for education was the source of my inspiration and  encouragement to pursue academic study. I cannot forget on those cold days of my elementary school period how my father used his tiny body as a shield to protect me from the storm on the way to school. I am always thankful to them. The everpresent influence of my friends and teachers, past and present, during the years of education and training laid the foundation which enabled me to overcome the obstacles that I encountered and to accept future challenges along the way. Each member of my committee helped me in unique ways. I sincerely appreciate the effort they made to help this dream of mine come true. I would like to express my sincere appreciation for the assistance provided by my thesis supervisor, Professor Laszlo Paszner. He taught me patience and gave me general guidance to conduct my study. The development of my ideas presented in the final manuscript evolved through the preparation of many working papers which were always cheerfully reviewed by Dr. Paszner. His helpful criticisms were a significant factor in crystalizing my ideas. I wish to acknowledge Dr. R. S. Seth (PAPRICAN) for his contribution on the completion of this thesis. He was patient with me when I needed so much assistance; I am always grateful to him for that. At the department of Wood Science, I gratefully acknowledge Dr. S. Avramidis for his unconditional support and his generosity in providing his laboratory facilities for sorption analysis.  As well Dr. S. Ellis provided invaluable  assistance with the GPC analysis, eagerly shared his experience and ideas any time requested and carefully read this thesis. I wish also to thank the faculty and staff of the Faculty of Forestry for their help during my years of study at UBC. In particular, Ms. D.  xv  Caciman, Ms. N. Cole and Ms. C.E. Laird who, in one way or another, helped make this study a success. There are those who are due thanks for specific expertise and information.  I  thank Dr. R. J. Kerekes of PPC, at the University of British Columbia for stressing the importance of the clarity of the project objectives at the outset. He was also instrumental in 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 for helpful discussions concerning the XRD analyses. The financial assistance of the Science Council of British Columbia through the GREAT Award, and the Natural Sciences and Engineering Council of Canada are gratefully acknowledged. I would like to extend my appreciation to friends whose company and understanding made the difference at many times. I cannot name every one of them, only tell them how much I appreciate their genuine friendship.  However, I wish to  acknowledge a number of them in particular. My Special thanks to Ms. Pam Rogers and Dr. 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 they could. Their friendship is beyond words. Special thanks go to my wife Ashraf for her unfailing support throughout the many years of my graduate student life. Her charm and compassionate support have helped me maintain my sanity, even though I may not always have acknowledged it at the time. And last but not the least, to my children Arash, Iliad and llnaz for all of the missed trips, picnics, etc., when dad was absent.  xvi  DEDICATION  I dedicate this thesis to my wife Ashraf S. Moussavi. Ashraf gave up her career -  in nursing in order to minister our children during my study. Her dedicated love and encouragement have been my main source of strength. This work would not have been possible without her patience, conviction and assistance.  1  1. INTRODUCTION  Disappearance of raw material sources, rapid decrease in the number of suitable landfill sites and public awareness pose great challenges to the pulp industry to find new ways for recycling and re-using the discarded fibrous waste for sustained production of paper in the future. Paper and board represent almost 32% of discarded municipal solid waste, based on recent reports by the U.S. Environmental Protection Agency (EPA). It is estimated 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 the U.S.. If that paper had gone to a landfill it would have covered an area almost 14.3 square kilometers and 4 meters deep.  This collected waste paper could preserve 64  million trees of one tonne oven dry weight each or the old growth forest on a land area of approximately 6,000 square kilometers. The calculation is based on 58 trees of one tonne in a hectare with almost 100% yield. The latest news indicates that the recovery rate of waste paper in North America has reached to about 50%, however, most waste paper is utilized for paperboard, especially corrugated boards, newsprint and toilet rolls.  Utilization for printing and  information papers is still at a lower level. Recent rapid increase in the use of office papers for reprography and computers, as well as magazines and other printing papers, is overwhelming municipal waste management.  In the past the utilization rate of old  newspaper for malcing newsprint increased from as little as 10% in 1988 to about 50% in 1994 in only less than half a decade in North America. This might be due to the fact that the quality of mechanical pulp fiber is not affected negatively by recycling. To promote increased utilization of office waste paper, it is necessary to overcome the inferior properties of recycled fibers.  Thus, understanding the causes of such inferior fiber  quality in recycling is at the heart of a number of key operations in upgrading recycled fibers. Such findings point the way forward to probable effective methods that could  2  bring about considerable improvements in recycled paper quality.  This would thus  diminish the flow of waste streams to landfills, and thereby restrain excessive depletion of the forest resource by reducing the amount of virgin fiber required for maintaining paper supplies and quality for most grades. In fact, office waste paper could also become the fiber of choice if the inferior properties of the recycled fibers could be overcome. Inferior properties of recycled office waste on papermaking properties of fibers has been discussed thoroughly by different workers (McKee 1971, Szwarcztajn and Przybysz 1976, Howard and Bichard 1992), and an excellent review on this subject was recently published by Howard (1990). These investigations showed a dramatic loss of tensile and bursting strengths on the repeated remaking of low yield chemical pulp into paper (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 potential bonding of the fibers. On the other hand, the loss in bonding of recycled fibers could be a 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 main loss 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 original water soaked condition, the free hydroxyl groups in the cell wall are practically all satisfied by water. When the fiber dries, pairs of the fibrils/lamellae are drawn together so that the individual fibrils/lamellae adhere strongly to each other. It is obvious that the mechanism of hornification could be uncovered by pursuing the physico-chemical changes occurring during the closure of these lamellae. More explicitly, the mystery of hornification lies somewhere between these lamellae. The review (Nazhad and Paszner, 1994) also concluded that the current traditional methods of upgrading recycled fibers are  3  not adequate, and it is necessary to develop and implement new systems which will upgrade recycled papers without worsening the overall papemiaking properties. By exploring, from as many directions as possible, the changes produced in the fiber wall by recycling, it is hoped to accumulate enough data to permit formulation of a more 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 background for structural, refining and thermodynamical analysis is presented. The superstructural arrangement of the virgin fiber (i.e., the proportion of the cell wall crystalline component and crystallite thickness) was measured and compared with those of recycled fibers. The physical properties of the fibers (i.e., virgin and recycled fibers) such as sorption potential, fiber surface area, heat of wetting of the fibers were also measured.  A  comparison was made on the thermodynamics of the two systems. Finally, the recycled samples were interacted with cellulase enzymes to detect changes in the structure of the fibers by recycling.  The results of each of these investigations, in the light of the  ukrastructure of the fiber wall, are discussed separately. Cross references were made between the results obtained from the different types of investigations.  4  THEORETICAL BACKGROUND  2.1 Correlation of Cell Wall Structure and Papermaking Qualities of Fibers Evaluating the quality of fibrous materials is important in paper manufacturing. Fundamental studies of the arrangement of cellulose molecules relative to each other in a particular state of aggregation make it possible to formulate criteria for predicting the quality of the end use products. Thus, structural characteristics, such as lateral order, degree of orientation and size of crystallites influence decisively the physical and mechanical properties of the fibers. For instance, if the crystallinity of the cellulosic fiber is 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 since entrance of one molecule of water requires the breaking of many cellulose to cellulose -  -  bonds. Therefore, the elastic modulus of crystalline cellulose remains unaffected in the vicinity of a moist environment. Increased crystallinity can lead to a higher modulus elasticity in the recycled fibers.  As a consequence, flexibility, which is inversely  proportional to the modulus, will decrease affecting the conformibility and therefore the relative bonded area. In order to be able to analyze the effects of drying on the structure of the fibers, a profound knowledge of the structural features and fine structure is an essential prerequisite. Therefore, a short review of the structural concepts of cellulose fibers will be dealt with first.  2.1.1 Ultrastructural arrangement of the fiber wall The structural concept of the cell wall has naturally evolved through the collection and collation of many pieces of experimental information (Frey-Wyssling  5  1954 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 lamellar structures, 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 and Goring 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 pictorial  representation of the cell wall before and after delignification.  002 LAX  Elementary  a A AX K • -( K  fib ru  - r El a  II II  -  1 oT  — *b(I 1 —  —  Lignin Hemicelluloses Cellulose  101  L1gnin..H  matrix H emiceIlulose  Delignifjed fiber wall  Fig. 1. Pictorial representation of the lamellae model for the ultrastructural arrangement of lignin, cellulose and hemicelluloses in the wood cell wall ( adapted from Frey Wyssling 1954 and 1969, Scallan 1974, Goring 1975 and Fengel 1970).  6  The elementary fibril consists of a crystalline core that is flattened parallel to the 101 lattice plane, where these planes are oriented parallel to the lumen of the fiber wall. This shape is due to a faster growth of the 101 plane, which is more hydrophilic than the more slowly growing 101 plane (Frey-Wyssling 1954 and 1969, Stamm and Smith -1969, Kerr and Goring 1975, Krassig 1984). The crystalline core of the microfibrils is embedded in a cortex of hemicelluloses and lignin matrix. The matrix which is located on the surfaces of lamellae (i.e., sheets of microfibrils) parallel to the 101 plane is the location where water is adsorbed when cellulose fibers are delignified.  Fig. 2. Schematic representation of the swelling of the fiber wall (after Scallan 1974).  7  This concept of a multi-lamellar structure of the fiber wall has been extended by Stone and Scallan 1965, Dunning 1968, Stamm and Smith 1969, Scallan 1974, and Kerr and Goring 1975, where radial cleavage of the sheets of microfibrils may occur during swelling (Figs. 2b, 2c, 2d). These characteristics of the fiber wall suggests that there is a large area of surface on the delamination of a single fiber, which will be referred to as internal surface area.  This surface area is very great when it is compared with the  external surface area of the fibers (i.e., 200 times more in average). Upon drying the lamellae 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 when it is re-wetted. In summary, the general view is that cellulose molecules are linked together to form 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 accessible surface of those planes in any amount will reduce water uptake by the cell wall and thus the swelling potential of the fibers will be restricted. That is to say, the fibers become hornified and their wet-flexibility is lost.  2.1.2 Crystallinity concept in polymers The atoms in amorphous materials such as glasses, resins, and liquids do not have the symmetrical repeating arrangement characteristic of crystalline materials. Thus, the x-ray patterns of amorphous materials at low diffraction angles appear as one or two weak broad bands superimposed upon a continuous background.  However, a highly  ant crystalline substance has very explicit peaks at specific diffraction angles. It is import to define the terms crystalline and amorphous because in a polymer the two states can be intertwined 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.  9  urement 2.1.4 Development of the XRD method for crystallinity meas can be assessed quite The state and extent of molecular order in a fibrous polymer diffractogram. Quantitative effectively by visual examination of a wide-angle X-ray red units, is usually referred to measurement of this state of order and the size of the orde Johnson 1978). Different as crystallinity and crystallite size respectively (Hindeleh and x. The advantages and methods have been used to measure the crystallinity inde ly in the following. disadvantages of the methods employed will be described brief 8) gives an absolute The classical method of Hermans and Weiclinger (194 rated intensity under the peaks to measure of crystallinity in terms of the ratio of the integ problem with this method the integrated intensity under the complete diffractogram. The the amorphous background. is the difficulty in separating the crystalline peaks from overlap. The empirical method Furthermore this method ignores the possibility of peak date, ignores the width of the of Segal et al. (1959) which has been used extensively up to alline materials. Krassig and intensity peaks blurring the distinction between paracryst g the crystallinity. However, the Kitchen (1961) introduced the peak width for calculatin tion of accurate peak width. main drawback of the approach was the identifica rally is referred to as the “full Determination of peak width at half height which gene to the tails of a given peak. width at half maximum” (FWHM) is very sensitive relative measurements, but only Fortunately, all the work cited above is useful in terms of od in use. with respect to one particular material and the particular meth action, by employing The separation of overlapping wide-angle x-ray diffr metallurgists. These attempts Gaussian or Cauchy disthbution has long been used by which crystallite size values can were directed mostly to measure the peak widths from l 1949). Gjonnes and Norman be derived using formulae suggested by Scherrer (Hal -angle x-ray diffractograms of (1960) applied such an approach to the evaluation of wide n of overlapping reflections under cellulose substrates. This method allows the separatio ctions have the shape of Cauchy’s the assumption that the intensity curves of the refle  10  distribution.  The authors determined the variation in the state of order by a  corresponding 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 peak in order to evaluate the crystallinity and crystallite size of paracrystalline samples. In the present work, the Voigt function was employed.  Compared to  experimental peaks, the tails of Gaussian disthbution are too short, those of the Cauchy distribution are too long, but those of Voigt and Pearson distributions are closer to reality at the expense of longer computing time. The Voigt function is discussed in detail in Appendix A.  2.1.5 Crystallinity index Crystallinity index is defined as the weight fraction of the crystalline portion of polymers. The higher the degree of ordering, the more crystalline is the polymer. For real systems, virtually no amorphous substance will be completely without order. There will always be some succession of chemical bonds with fixed lengths and angles that will give rise to a particular radial density distribution function F(r) that we view as a diffractogram.  As structural features coalesce in a more ordered fashion, the x-ray  diffractogram moves away from a diffuse halo to explicit peaks whereas the total integrated area (halo plus explicit peaks) remains constant. This fact can be used to calculate the ratio of a percent crystallinity as the area of these explicit peaks and the amorphous 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.  FWHM  demonstrates the average modification of crystallite dimensions due to the influence of recycling. In the present investigation a peak resolution function was used to calculate both the crystallinity index of cellulose and the dimensions of the crystallites (Appendix A).  11  This function allows the resolution of the x-ray diffractogram into the contributions of each of the diffraction planes (i.e., 101,  loT, 002, 040, etc.) (See also Fig. 33). The  amorphous component is calculated by subtraction of the area of resolved peaks from the total area of the x-ray diffractogram. Generally, the Voigt function results in the best fit of the x-ray diffraction patterns and is routinely used for the determination of both the crystallinity index of cellulose and dimensions of the crystallite through integral breadth or full width at half maximum height (FWHM). The ciystallinity index, which is defined as the ratio of the resolved peak area to the total area under the unresolved peak profile, is calculated using the equation (Hindeleh and Johnson, 1978):  Cr1  (%)  =  192 [j 1 d9 0  /  192 ITdO]  [2.1]  x 100  where Cr1 is the crystallinity index of the sample,  ‘T  is the total intensity of diffraction  in the diffractogram, I is the total intensity of the diffraction due to the crystalline region (peaks) of the sample and 9 and °2 are the limiting values depending on the range of a diffraction angle.  2.2 Beating (Refining) Beating or refining is a mechanical treatment of the fibers. During this treatment certain changes in the structure of the fiber wall take place. Beating and refming are terms used synonymously since refiners are commonly used to beat pulp stock in current practice. The basic effects of beating are as follows: 1. Internal fibrillation variously described as delamination or swelling. -  2. External fibrillation the fiber.  -  the partial removal of the fiber wall, leaving it still attached to  12  3. 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; or removing nodes, kinks, slip planes, microcompressions from the cell wall. There is, of course, an endless list of effects which range from most significant to trivial. Although this list is not inclusive of all effects during refining, however, these are the most significant and can be found in the current literature (Emerton 1957, Atack 1977, Page 1989).  and does not indicate that, some other basic effects cannot have  occurred during refining.  2.2.1 Theory of the beating (refining)  2.2.1.1 Chemical theory The theory of beating started with the chemical theory and dominated this field for a quarter of a century. The chemical theory, or hydration theory as it came to be known, proposed that (Cross and Bevan 1920) the increase in paper strength which accompanies beating is due to the working of the surface layers of the fibers into a hydrate of cellulose. Hydrated cellulose was found to have gelatinous properties acting as an adhesive to bind the fibers together. It is important to realize that the term was originally used in a strict way.  ) is 4 For example, anhydrous copper sulfate (CuSO  hydrated with the addition of water.  Indeed, it is converted into the pentahydrate  O which is chemically distinct. ) 5H 2 (CuS . 4 0 The chemical theory of beating, in different forms, persists even to date. Henry et al. (1988) studied the mobility and interactions of the water molecules contained in the pulps. The authors concluded that the beating did not increase the interior surface of the cellulose fibers but only caused a better structuring of the water molecules along the fiber  13  surfaces.  Milichovsky (1990) suggested that beating in the aqueous medium does not  change the composition and structure of molecules forming individual pulp fibrils and microfibrils.  What is changed, however, is the composition of the phase interface  between the pulp and water medium and the molecular arrangement in single pulp formations, especially in places affected by simultaneous mechanical action. In other words, the beating process involves chemical changes in pulp beyond the level of the basic 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 nature of water molecules and their two possible orientations in hydration surfaces. If the water molecule orientation to each of the interacting surfaces is equal then, both surfaces wifi affect each other with repulsing hydration forces. On the other hand, if the orientation of water molecules is different, the surfaces will be attracted. Milichovsky’s approach to the problem of beating is limited and it does not fully explains for the differences in water molecule orientations.  2.2.1.2 Physical theory The chemical theory was inadequate and limited, therefore, the researchers on the mechanics of beating switched their attention towards the physical theory of beating. Thus, the chemical theory of beating was followed by a physical explanation first suggested by Strachan (1926).  He proposed that the strength of paper is due to the  mechanical 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 of water in the fibril pile. This pure physical theory, however, was an oversimplification in explaining the theory of beating.  14  Campbell (1930) turned his attention on the very considerable forces acting on the fibrils as a result of surface tension during drying. He suggested that during beating the fibrils are more readily teased out by beating. Furthermore, when water is removed from the paper sheet these fibrils are drawn into contact with neighboring fibers. As the paper sheet is dried hydrogen bonding occurs, to a much greater extent between fibers when fibrils 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, the constricting outer layers of the fiber (i.e., the primary and outer secondary wall) are disrupted and in part removed (external fibrillation). The disruption of the outer layers allows the fibers to swell. As a result of this swelling and the simultaneous repeated flexing by the beater and refiner bars, the fibers are internally delaminated (internal fibrillation). As a consequence of these effects, the bonds between successive coaxial lamellae of the middle secondary wall are, to some extent, severed.  However, the  mechanism of delamination caused by either flexing or swelling has not been resolved yet. 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) and supported by Scallan (1978).  Scallan proposed that the delamination, induced by  excessive swelling is achieved by chemical means. In this way, the swelling could be achieved 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 increasing stress on neighbouring bonds which would lead to the propagation and widening of the opening.  This swelling would then result in complete and uniform tangential  delamination. Once initiated in several places between radially adjacent lamellae, the delamination would grow in preference to further debonding of the microfibril sheets.  15  He supported the idea that the delamination is solely possible by swelling and that it does not necessarily require mechanical action. Conversely, Atack (1977) described the mechanism of beating by mechanical action rather than sweffing by chemical means. He suggested that the effect of beating is similar to a loaded rolling element passing over any deformable substrate. In addition to this 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 a peristaltic action.  High pressure water forced into the ‘crack tips’ of the honeycomb  structure would release fibrils from each other at points of attachment inducing the final state of coaxial delamination. The former mechanism would promote both external and internal fibrillation and the latter internal fibrillation.  The author concluded that the  operative 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 and internal fibrillation rather than cutting the fibers, is the presence of a water-film between the fibers and fiber layers. This water-film dampens the mechanical stresses which are imposed by beating.  Both Scallan (1978) and Atack (1977) tried to explain the  mechanism of delamination during refining either by chemical or mechanical means, but neither of them explained that the flexing itself could contribute to the process of refining. However, Atack (1977) identified the effect of water-film on enhancing both external and internal delamination rather than cutting the fibers during refining. Tam Doo and Kerekes (1989) extended the fatigue-failure mechanism as proposed by Atack (1977). They suggested that any given specimen of a material that is subjected to a sufficiently high level of internal cyclic stress will gradually weaken and eventually fail through fatigue. Furthermore, the authors (Tam Doo and Kerekes 1989) quantitatively measured the flexing (i.e., fatigue type) effect on the flexibility of the fibers. The increased flexibility of the pulp fibers by flexing (cyclic loading) resulted in increased sheet density and as a result the strength properties of the paper were also  16  increased. This observation strongly suggests that the flexibility is improved not only as a 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.  He  atthbuted the flexing effect to the development of cracks during refining which reduces the rigidity of the fiber walls. Furthermore, Page (1989) emphasized the combined effect of 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 the mechanism 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) as measured by Tam Doo and Kerekes (1989) suggested that flexibility could be increased before 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 agent and flexing as a fatigue phenomenon in the refining process.  2.3 Surface area of fibers A cellulosic fiber possesses a considerable volume of pores of diverse sizes and shapes.  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 from water, 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 internal surface area) scarcely exceeds the size of external surface area alone. The large internal surface of the swollen structure can be retained only by replacement of water by less polar liquids, before drying. The properties of cellulosic fibers are related to the changes  17  in internal and external surface which take place when the fibers dry from polar liquids or gases.  A knowledge of surface area, particularly the internal surface, is basic to an  understanding of the behavior of cellulosic materials. The measurement of external and  internal surface area of fibers in virgin and recycled states could reveal the potential of the fibers for water uptake. On this basis the modification on fiber flexibility or bonding potential could be assessed.  According to the literature on recycling, very little  information 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 method developed by Harkins and Jura (1944).  The internal surface area is estimated by  subtraction of external surface area from the total.  2.3.1 Adsorption technique to measure the fiber surface area  Generally, gases and vapors are adsorbed on the surfaces of solids, under appropriate conditions of temperature and pressure. It is presumed that the adsorption of water takes place, at first, in a layer one molecule thick to form a monolayer, and subsequent adsorption results in the formation of multilayers (Brunauer et al. 1938). If the amount of vapor or gas required to form a monolayer can be found and the area of each adsorbed molecule is known, the total surface area on which adsorption has taken place could be measured. The precision of the technique strongly related to the type of gas is used. For instance, the nitrogen molecules in the capilaries of 6  -  10 A could not be  detected by nitrogep gas. Probably, the proximity of opposing surfaces of capilaries will cause the nitrogen molecules to orient themselves vertically and thus either block some of the pores or not allow contineous coverage of both of the opposing surfaces. Analysis with water as the sorbed gas has been described and calaimed to be a useful tool (Weatherwax 1974, Ostberg and Salmen 1991).  18  Calculation of the amount of vapor required to form a monolayer was developed by Langmuir (1918). The theory was extended by Brunauer et al. (1938) to multilayer adsorption.  The equation derived by the authors (Brunauer et al. 1938) is based on  kinetic interpretation, which effectively describes sorption isotherms of many types. The beginning of the long linear portion of the isotherm was interpreted by the authors as completion of the monolayer and the start of formation of multilayers. The BET theory poorly explained the states of water in fiber-water interactions system (Hartley et al. 1992). However, comparison of conceptually different approaches to measuring fiber surface area suggested that the BET sorption theory could provide an estimate of the total surface area (Weatherwax 1974). This theory widely used to measure the surface area of variety of materials including cellulosic fibers. The surface areas of the fibers in this investigation is determined by application of the BET (Bruauer et al. 1938, Brunauer, 1940) equation. Generally, the BET equation is expressed as follows:  (h)/M(1h) = 1/CMm  +  (C4)h/CMm  [2.21  where h is the relative vapor pressure, M is the amount of moisture adsorbed at the relative pressure h.  Mm is the volume of vapor adsorbed when the entire adsorbent  surface is covered with a complete monomolecular layer of adsorbate, and C is a constant depending 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 at least the point where h=30%. The slope of the line is equal to (C4)/MmC and the Y intercept is equal to 1/(MmC). Therefore, the solid surface area could be calculated from the known area of one adsorbate molecule. adsorbent  ,  The total surface area per weight of  is given by:  =aNMm/Mw  [2.3]  19  where a represents the effective area occupied by an adsorbed molecule (i. e., a  =  12.5  AZ), 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 first 1 is the heat of adsorption, EL is the molecular layer, e is the base of natural logarithm, E heat 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 solvent exchange procedure, and that the total area of the swollen material may be as much as 10-fold greater than the area that results after drying from water. The calculation of the surface area using different equations derived from the adsorption isotherm has been employed 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 authors found a constant surface area over a large range of isotherms. The technique was applied to 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 work has been reported in the case of recycled fibers.  2.3.2 External and internal surface area In 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.  20  External fibrillation is generally measured by the nitrogen adsorption, silvering, permeability, heat of wetting and staining methods. The external surface area can be estimated by employing Harkins and Jura’s absolute method (1944). The advantage of the heat of wetting arises from the view that the same set of data developed for thermodynamic analysis of the fiber  -  water system can be employed to calculate the  external surface area. In this process, the film-covered surface releases a quantity of energy proportional to the area of the clean solid (See also the Appendix B). Suppose, , then the only 2 for example, that the liquid is water with a surface energy 118.5 erg/cm energy change involved is that due to the disappearance of the water surface of the adsorbed film. If the energy developed in the calorimeter, expressed in ergs, is now divided by 118.5, the quotient gives the area of the surface of the particle with its adsorbed film of water. Thus, the value of the clean solid surface is given by the relation:  =  S zHjsf,L)/ll . 8  [2.5]  where /XH(sf,L) is the heat of immersion of the sample equilibrated at 100% relative humidity expressed in ergs per amount of solid. The surfaces obtained in this manner are considered estimates of the external surface area of the solid, since the capillaries are filled with water during conditioning the samples at 100% relative humidity.  The  internal surface area of the solid is estimated from the arithmetic difference of the total and external surfaces of the sample.  21  2.4 Thermodynamics of fiber-water interactions Thermodynamics is concerned with energy relationships involving heat, mechanical energy, and other aspects of energy and energy transfer. The interaction of water and fiber (cellulose) is always accompanied by the evolution of.heat and a decrease in free energy of the system. Many aspects of the interaction can be treated by the classical methods of thermodynamics. The thermodynamics of wetting deal with the entire spectrum of solid-liquid interactions, ranging from the adsorption of a small amount of vapor by a relatively large amount of solid, to the immersion of a small quantity of solid into a large amount of liquid. An experimental investigation involving both of these processes can provide information on how much of the changes arise from the adsorbate and adsorbent, respectively. This is particularly important of the fiber-water system, in which drying reverses part of the fiber properties. In brief, the differences in the free energy to the two systems at the same relative pressure, may be due to a difference in the fiber accessible surface area (irreversible swelling), solid-liquid bonding strength (enthalpy effect), molecular ordering of adsorbate and adsorbent (entropy) or a combination of these effects.  Therefore, a  complete thermodynamic analysis of the fiberwater system can help in identifying the origin of the differences between the two systems of virgin and recycled fibers. The interpretation of adsorption isotherms of protein and cellulose has been approached using thermodynamic functions by many authors (Bull 1944), Morrison and Dzieciuch 1959, Hollenbeck et al. 1978, Argue and Maass 1935, Ostherg and Salmen 1991, Stamrn 1957). This technique was also used to explain the structure of protein and cellulose (Bull 1944, Assaff 1944, Wahba 1950). The hysteresis of the cellulose sorption isotherm was also investigated by this technique (Morisson and Dzieciuch 1959, Hollenbeck 1978).  22  Paper recycling is mostly done in water medium. Thus, studying the recycling effect by thermodynamic functions in conjunction with structural analysis by the XRD technique, fiber surface area measurement and enzymatic hydrolysis could assist further our understanding of the effect of recycling on paper properties.  2.4.1 Free energy changes The free energy of a system could be defined as the maximum energy freed in a process and made available for work (See also Appendix C). Through the adsorption isotherm the information related to both the differential and integral free energy changes could be generated. The differential free energy is given by 1 DF  =  [2.6)  RTIn(h)  where h is the relative vapor pressure, T is the experimental temperature and R is the gas constant. The integral free energy change for water vapor accompanying the sorption process is: =  Tln(h) n R 1  -  RT  h)d(h) (n / 1  [2.7]  where n 1 is the mole of water per 100 g of solid. Evaluation of the integral on the right-hand side of equation [2.7] can be /h versus h. 1 performed graphiclly from a plot of n The free surface energy of adsorption per unit of surface, H, can be obtained from the integral part of the equation [2.7] divided by the monolayer surface area (Bangham and Razouk 1937, Stamm 1957, Harkins and Jura 1944).  23  II  =  =  —  RT/E  Ih  [2.8]  h)d(h) (n / 1  Jo  where 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 interface is constructed.  2.4.2 Enthalpy changes Heat of wetting is explained in terms of surface-energy changes. When one gram of a clean solid is immersed in a liquid, it is accompanied by the evolution of a quantity of heat per gram of solid, which represents the total energy of wetting (See also Appendix 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 solid saturated at 100% relative humidity, respectively. The partial molal enthalpy change is obtained by estimating the slope of AH . 1 versus fl  2.4.3 Entropy changes The property of a system when energy transfer is not involved is generally expressed by the entropy of the system. Unlike for energy, there is no principle of conservation of entropy. Hence, when all systems taking part in a process are included,  24  the entropy either remains constant or increases. In other words, no process is possible in which 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 defining equations once the other thermodynamic variables are determined:  AS  1 AS  =  (AH z\F)/T  =  [2.10]  -  1 AF.j/T (AH  [2.11]  -  The unit of entropy, in any system, equals the ratio of the units of energy and time in that system.  2.5 Enzymatic Hydrolysis of Recycled Fibers Enzymes have been used in analysis of biological systems for more than 125 years.  Because of their specificity, it was recognized early (Whitaker, 1974) that  enzymes could be used to detect and determine the concentration of minute amounts of compounds in complex biological systems.  The availability of well-characterized  enzymes of high specifity in pure form has made enzymatic analysis more attractive during the last two decades.  Detecting the characteristics of a substrate is closely  associated with the availability of well-characterized enzymes. The field of enzymatic analysis is diverse. The analyses include determining enzyme 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 a substrate, activators, or inhibitors of enzymes. Isomeric configurations of compounds, the primary structures of complex molecules such proteins, neucleic acids, and  25  carbohydrates, the conformation of complex molecules, and the structures of cells and subcellular 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 of cellulose fibers. Pommier et al. (1989) improved the drainage quality of recycled fibers by pre-treatment with enzymes. In the present study recent findings on the mechanism of enzymatic hydrolysis of cellulose were employed to assist with the analysis of the characteristics of recycled fibers.  3. LITERATURE REVIEW ON RECYCLING  3.1 Introduction The practical effects of recycling on papermaking properties of fibers has been discussed thoroughly by different workers (Klye 1961, McKee 1971, Bovin et al. 1973, Szwarcsztajn and Przybysz 1976, Eastwood and Clark 1978, Howard 1990, Howard and Bichard 1992, Marton et al. 1993), and an excellent review on this subject was published by Howard (1990). These investigations showed a dramatic loss of tensile and bursting strengths 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 when beaten (Williams 1980).  It is thus predicted that, due to these limitations the useful  recycled fiber content in paper should be limited and a surplus of recycled paper will become available that will need to be disposed of by incineration or in cogeneration facilities, as suggested by Gotsching (1992).  26  ading of recycled paper In spite of the considerable amount of work on upgr 1982, Putz et al. 1989), only (Lundberg and de Ruvo 1978, Lindstrom and Carison . Therefore, the problem of limited improvements have been achieved in this field nce, there was a patent issued in upgrading recycled fibers is still unsolved. For insta tion of a series of soluble 1954 to Schiosser (Hunger 1978) which described the addi rittlement) after drying. The substances to puips in order to suppress hornification (emb found to work for only one best results were obtained with wetting agents, but they were cycle (Hunger 1978). method for regenerating The refining process, which could constitute an efficient same time further worsening the papermaking properties of recycled puips, causes at the tajn and Przybysz (1976) attempted of their drainage properties. For instance, Szwarcsz recycled pulp during each cycle to upgrade recycled papers by refining. They refined the content of the pulp (in spite of till the maximum breaking length was achieved. The fines during recycling by about 70% losses in the consecutive formation processes) increased t could severely limit traditional (Szwarcsztajn and Przybysz, 1976). Such a resul recycled puips by refining. The methods for regenerating the papermaking capacity of cled fibers has been confirmed negative effect of refining as a tool for upgrading of recy Naito et al. 1983). Therefore, by many others (McKee 1971, Bovin and Teder 1973, material are not adequate, and the traditional methods of upgrading of recycled paper s of recycled fibers still remains need for developing the necessary mechanical propertie ement new systems which will an enigma. It will be necessary to develop and impl rmaking properties. upgrade recycled papers without worsening their pape in recycling is at the heart Understançling the causes of such inferior fiber quality s. Such findings point the way of a number of key operations in upgrading recycled fiber considerable improvements in to probable effective methods that could bring about waste streams to landfills, and recycled paper quality thus diminishing the flow of  27  thereby restrain excessive depletion of the forest resource by reducing the amount of virgin fiber required for maintaining paper quality for most grades.  3.2 Postulated mechanisms for loss of strength in recycled paper The 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 paper strength are brought about by loss of bonding, or in other words, a reduction in the resistance of bonds to an applied force. (McKee 1971).  This is in agreement with previous reports  On the other hand, the loss of bonding, could be a function of two  parameters (Eastwood and Clarke 1978): i. Fiber flexibility*, and ii. Surface condition. It is not yet certain why loss of surface condition or flexibility of fibers occurs during recycling.  Furthermore, it has not been established which of the above factors in  recycling plays the major role in recycled paper strength loss.  The following sections  elaborate on these two parameters.  3.2.1 Fiber flexibility The effect of drying on pulp fibers has been reported by many workers (Jayme and Hunger 1958, Robertson 1963). The consensus is that the swelling capacity of fibers is lost, and irreversibility of their swelling property is increased with the level and duration 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 the moment of inertia of the body and the modulus of elasticity of the material from which the body is made.  *  28  flexibility of wet pulp fibers, or “hornification”.  The effect of drying on high-yield  mechanical pulp fibers is minor (Bovin et al. 1973, Howard and Bichard 1992). On the other hand, low-yield chemical pulp fibers show a progressively greater-loss in recovery of 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 and Bichard 1992, Marton et al. 1993). The mechanism responsible for hornification has been the subject of a long debate and is not resolved satisfactorily, as yet.  3.2.1.1 Irreversible pore closure The irreversible changes of cellulosic cell walls upon exposure to drying has historically been attributed to the closure of pores and cracks (Thode et al. 1955, Jayme and Hunger 1958, Stone and Scallan 1965 and 1968). Thode et al. (1955) reported that the hornification is caused by the irreversible closing-up of micropores and cracks during drying. Stone and Scallan (1965 and 1968) carried out comprehensive studies on the effect of drying on the structure of the cell wall.  Solute exclusion was used in  characterization of the structural changes in the cell wall. It was observed that the large and intermediate size pores were reduced upon drying. This concept was also discussed by Jayme and Hunger (1958). Hornification by both groups were defined as irreversible changes in the capillary system of the fiber cell wall. However, many questions remain still unanswered in the area of fiber science. For instance, no explanation is offered as to why 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, the solute exclusion technique suffers from some practical difficulties.  For instance, the  lower molecular weight polymer could be adsorbed easily on the fiber surface (Gray  29  1978). 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 final  conclusions proposed for changes at the structural level.  3.2.1.2 Cross-linking The auto-crosslinking hypothesis, as an origin of fiber homification during recycling, 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 natural aging or heating in an acid environment (accelerated aging) of the paper mainly promote hemiacetal type cross-linking between cellulose and hemicellulose chains. Such bonds can restrict the swelling and make the fibers brittle. Lundberg and de Ruvo (1978), by examining the equilibrium moisture content of commercial never-dried bleached kraft pulp of 44% yield, concluded that, the hindrance to swelling, after drying, in the fiber was not due to occurrence of cross-linking or the formation of sites inaccessible to water. The change in swelling was not reflected by changes 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 during recycling. They attributed this increase to the increased hydrogen bonding between the microfibrils during recycling.  3.2.1.3 Cellulose cl,iain cleavage Paper acidity, which may be present in papers because of the processes or chemicals used in their manufacture, is a major factor in hydrolysis of cellulose (McComb and Williams 1981). Acid penetrates the open amorphous regions of the fiber and cuts the carbon-oxygen glycosidic bonds that link the glucose units in the cellulose  30  chain (McComb and Williams 1981). Oxidation, a reaction between oxygen and the cellulose unit, can break carbon-hydrogen bonds, as well as carbon-oxygen bonds (McComb and Williams 1981). These reactions simultaneously liberate the portion of the fiber plastisized by humidity, lower the overall degree of polymerizatioir and make the fiber more fragile and more susceptible to breakage during subsequent beating (McComb and Williams 1981). McComb and Williams (1981) concluded that recycled fibers from alkaline paper behave more like virgin paper and can make better recycled products (See also Fig. 3).  1•1 10 0  —  o  20—  C  30—  C 0  U C  40C  500 C  o 0 0  0  60— 70 80  —  —  90100 4  I 5  6  7  8  pH of wate extract  Fig. 3. Change in flexibility of fiber by acidity of paper. The fiber retains its flexibility when it is a neutral or slightly alkaline environment (McComb and Williams 1981).  31  Stockman and Teder (1963) studied the effect of drying on the degree of polymerization (DP) of different chemical puips. They did not observe any significant change on the DP of chemical puips on heating in the range of 70-140° C. The bleached kraft 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 after  accelerated thermal aging (70° C, 50% RH, 70 days) than distilled-deionized water washed fibers.  However, it is not certain to what extent chain scission of cellulose  contributes to embrittlement of fibers at regular conditions of commercial recycling.  3.2.1.4 Re-organization in the fiber (cell) wall Reorganization and co-crystallization of cellulose chains during drying, as a source 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 fact that, upon drying, bonding forces which develop, are sufficiently large and regular to unite two or more crystalline regions as one (Ehrnrooth 1978), thus resulting in a swelling restriction in dried fibers (See also Fig. 2).  This implies that the possible  increase in crystallite size and/or order in the fiber cell wall could be responsible for the hornification effect and consequently for brittleness and loss in flexibility of the recycled fibers. It is worth mentioning that the smaller the crystallite in the fiber wall, the larger will be the surface area capable of hydrogen bonding and, therefore, the greater will be the 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 cellulose and hemicellulose, previously separated, may come together. If parts of the areas coming  32  into direct contact match sufficiently well in composition and orientation, they could well form additional crystallite zones and consequently restrict swelling on rewetting. This concept 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 during recycling. The method which they developed, was acetylation by which recovery of fiber properties was enhanced after drying.  This method was assumed to prevent co  crystallization of fiber crystallites during drying. However, the method thus adopted was indirect and did not allow formulation of accurate conclusions, because, acetylation at high degrees of substitution could convert the cellulose to a hydrophobic material with concomitant loss in bonding (Giertz 1978). Due to the possible induction of hornification by structural changes in the cell wall during drying, a systematic investigation is required. In fact, removal of lignin and hemicelluloses from the cell wall should make the fibers particularly vulnerable to such “secondary crystallization” to take place on drying.  3.2.1.5 Bond strength A strong hydrogen bond is defined as a hydrogen bond in which both members of the H-bonded complex experience significant structural changes.  Thus, when a H-  bonded complex in its minimum energy structure demonstrates a structure not too dissimilar in isolated monomers, the bond is called a “weak” hydrogen bond. According to this theory, weak hydrogen bonds are least firmly held and could be easily opened upon moisture uptake. Milichovsky (1990) suggested two or three types of qualitatively different hydrogen bonds. He proposed that these different types of hydrogen bonds are responsible for oriented (crystalline), less oriented, and unoriented (amorphous) zones in native cellulose. He suggested that the reason water causes cellulose to swell rather than dissolve it, lies in existence of strong hydrogen bonds in its oriented zones. However,  33  this does not mean that the water does not penetrate the oriented zones. Indeed, it is possible that water molecules penetrate into the oriented zones without causing any substantial changes (Milichovsky 1990).  Therefore, the ability to destroy hydrogen  bonds with water will depend not only on their strength, but also on the nature of the hydrogen bonds. Higgins (1978) interpreted the irreversibility of bonding on recycling to occurrence of stronger hydrogen bonds rather than reciystallization and/or co crystallization. He proposed that the existence of weak and strong hydrogen bonds are the basis of the whole question of recycling, and suggested that irreversible changes during recycling were brought about by the latter type of hydrogen bonds. In general, elucidating the mechanical properties of paper, by existence of different types of bonding mechanisms, has not been discused thoroughly in the literature and remains speculative. Therefore, more extensive work will be required to substantiate why certain hydrogen bonds in cellulose H-bonding are stronger than others.  3.2.2 Surface Condition of Fibers Theories in the field of surface condition of fibers are diverse. They originate from observations on fiber surface accessibility and extend to the mutual relation of fiber and water. The investigations focus on relatively few fundamental problems relating to the surface condition of recycled fibers and fiber-water interactions. For instance, the presence of accessible hemicelluloses on fiber surface, and their abundance in the cell wail, enhance both fiber-to-fiber bonding and wet-flexibility of the fibers. Disappearance of these bonding agents, either through re-distribution of fatty acids on the fiber surface or inactivation of the hydroxyl groups by any other mechanism during drying, may reverse the end use performance of the recycled fibers. Surface condition could also be affected by induction of microcompressions or a diminished surface area.  It is also  34  suggested that perhaps the formation of hydrophobic molecules are responsible for fiber surface inactivation.  It should be established what happens to the surface bonding  capacity of the fibers during drying. Does the surface bonding capacity disappear from fiber surface during recycling, or does the interaction with water create a shield around the bonding agents? Hypotheses in this field are abundant, but little is known on the real nature of the problem.  3.2.2.1 Hemicelluloses-loss effect The role of hemicelluloses in the manufacture of a strong paper sheets has been known to papermakers for a long time. The precipitation of hemicelluloses on the pulp fiber surface in the final stages of alkaline cooking improves the paper-forming properties of fibers. Precipitation of the hemicelluloses on the fiber surface also occurs during beating (Gallay 1958). Thus, some authors attributed deactivation of fiber surfaces to the loss of hemicelluloses from the fiber surface.  Simultaneous hornification on  drying of fibers high in hemicelluloses was also observed (Jappe 1958). Hillis (1984) speculated that strength loss of wood is caused by hemicellulose losses or conversion during high-temperature drying. Meiler (1947) observed on drying fibers at high temperatures that the solubility of the hemicelluloses in alkali decreased substantially. These findings were also confirmed by Davis and Thompson (1964) by heat treatment of different woods. Stamm (1964) suggested that hemicelluloses in wood are converted to a furfural polymer at high temperatures which has less affinity to swelling.  /  Eastwood and Clarke (1978) proposed that the effects of recycling on flexibility and surface condition of the fibers is possibly brought about by loss in hemicelluloses. They determined the loss of hemicelluloses quantitatively and considered them as an indicator for loss in fiber flexibility. Nevertheless, their results were not conclusive,  35  because of the complicated behavior of hemicelluloses in the strength development of paper (Cottral 1950). Indeed, it is possible to remove a certain amount of hemicelluloses from the fiber matrix without any detectable changes in the strength of paper (Cottral 1950). That is so because hemicelluloses in fibers have an optimal concentration (Cottral 1950) at which maximum paper strength is obtained. Higher than that amount will not contribute to strength 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 fiber cell wall.  Therefore, such complexities in the behavior of hemicelluloses in paper  strength development make formal conclusions, regarding their effect on the surface properties of fibers, more difficult.  3.2.2.2 Inactivation of the fiber surface It was suggested that (Hancock 1964, Chritiansen 1990) aging and hightemperature drying cause the migration of extractives to the surface. The migration of fatty acids, or some extractive component, was suggested as a main cause of surface inactivation in dried cellulosic materials. Back et al. (1967) suggested that the loss in surface accessibility of fibers upon recycling is due to re-distribution of fatty acids initially present in the pulp. Additionally, such an effect may also originate from rosin size 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 fast on the fiber surface, at paper drying temperatures. Troughton and Chow (1971) did not find any correlation between fatty acid concentration and quality of wood surface. catalytic role in the inactivation process.  They suggested that fatty acids have a  36  A critical review of the literature on over-drying of wood and fibers was published by Christiansen (1990). The review suggested that surface inactivation by heat-treatment or aging, caused basically by exudation of extractives to the surface, de activated the fiber surface and prevented the surface from hydrogen bonding. The other mechanisms, such as reorientation of wood surface molecules and closure of large and intermediate pores in cell walls during drying, which reduce surface activation and decrease the internal bonding points, were also confirmed. Reports (Howard and Bichard 1992) on mechanical pulps in contrast with chemical pulps did not show any distinguishable loss on bonding capacity  of dried  fibers. This does not coincide with reported data on wood. The disagreements could be due to the conditions of drying or many other parameters involved in defining the pulp and the process of recycling. Furthermore, the question remains as to the significance of surface inactivation during drying. If the reason is the migration of extractives to mask the fiber surface area, mechanical puips should be more susceptible than chemical puips. Therefore, larger reductions in fiber quality should be observed for mechanical puips than 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.  3.2.2.3 Microcompressions It was suggested that, recycling could cause microcompressions (Howard 1990), which could alter the surface condition of the fibers. This expectation arises because, the fibers shrink between 20-30% transversely and 1-2% longitudinally. Fiber length in paper has been 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 of longitudinally compressed fibers were found to demonstrate superior extensibility, high  37  tear and impact resistance, but reduced elastic modulus. These properties already contradict some of the findings reported on recycling. For instance, in recycling impact resistance, stretchability and elongation decreased and elastic modulus increased (McKee 197 1, Koning and Godshal 1975). Furthermore, fiber length shortening has not been detected in recycling (Jones 1990). Therefore, the possibility of changes in occurrence and intensification of microcompressions in the fibers upon recycling must diminish considerably.  There remains the question on permanency of microcompressions in  recycled fibers.  3.2.2.4 Fiber-Water Interactions For a break through in the improvement of recycling a detailed knowledge of fiber-water interactions is required. These interactions could be interpreted to dominate the 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 and Dzieciuch 1959, Goring 1978, Caulfield 1978). Studies on the properties of liquid water revealed that water is not a homogeneous solvent system, but rather, it exists in a dynamic equilibrium between the icelike* and non-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 can be 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 effect of water on the property of the agent. The subsequent section will concentrate briefly on exploring 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 did not necessarily imply that the associated water molecules have the tiidymite-like arrangement of ordinary ice. *  38  3.2.2.4.1 The electric charge of paper surfaces in water Many authors (Verwey and Overbeek 1948, Strazdins 1976) discussed the problem of fiber surface accessibility in terms of interfacial potential differences of fiber-water compositions. Generally, most substances acquire a charge when immersed in-water and migrate under an applied electric field. This is usually interpreted in terms of an ionic double layer adjacent to the adsorbing surface. The zeta potential is defmed as the potential at the surface between the freely mobile liquid and the liquid firmly adhering to the particle surface (Verwey and Overbeek 1948) (See also Fig. 4). According to the double layer theory, the thickness of the double layer should be reduced by the presence of electrolytes. Consequently, the forces of repulsion will decrease and the potential for fiber surface adsorption, will increase. Furthermore, at higher repulsion forces, a  bulk liquid  Potential at this surface Is called zetapotentlal A layer of adhered solvent molecules  \  \\  \  N\\\\  SURFACE Fig. 4. Diagram showing the adhering solvent layer on surface when a surface is submerged in a liquid.  39  significant portion of the added fillers and some of the fines will disappear from the wire section and become part of the white water.  The main drawback of this theory in  experimentation results from the vague location of the plane of shear. Generally, zetapotential is useful in the case of massive particles with a well-defined impenetrable surface, but loses a good deal of its meaning with swollen gels (Verwey and Overbeek 1978). In addition, the electrostatic interactions are not the only force governing the fiber-water interactions phenomenon (Dobbins 1970).  3.2.2.4.2 Changes in water structure due to fiber-water interactions Mutual effects of fiber and water were discussed by different authors (Goring 1978, Dobbins 1970, Morisson and Dzieciuch 1959), but because of the complexity of the interactions, the subject is not well understood. Goring (1978) considers the problem of surface condition to be due to the changes in the hydrophobic bonding system in cellulose. It was suggested that in cellulose there is a hydrophilic and hydrophobic bonding system and that perhaps changes in the hydrophobic bonding system plays an important role in formation of an irreversible fiber surface condition. Furthermore, by measuring thermal expansion of cellulose in different states (i.e., wet and dry), he concluded that a carbohydrate surface disturbs the adjacent bulk water in such a way that cellulose acts as a structure breaker (Goring 1978) (See also Fig. 5). This is actually the extension of Kauzmann (1959) and Nemethy and Scheraga’s (1962) work on proteins. The result of calculations of hydrophobic bonding in  40  /OQ  \c  o__.  Cellulose surface Fig. 5. Schematic cross-section of broken hydrogen-bonded water near the cellulose surface (Goring 1978).  proteins confirmed that a considerable part of the free energy arose from changes in the water structure, whereby the van der Waals interactions themselves contributed only part of 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 surface in condition was also emphasized by Dobbins (1970). He suggested that the increase flocculation of cellulose by rising temperature is only compatible with hydrophobic bonding, rather than hydrogen bonding which is an exothermic process.  Thus, he  concluded that the addition of simple solutes to a cellulose-water system often produces  41  effects that are difficult to explain in classical terms of coulombic association between positively charged ions, and the negatively charged cellulose interface. He suggested that there is a third class of ions that have hydrophobic character, which could explain swelling under neutral or alkaline conditions in terms of one unified mechanism. The problem with hydrophobic bonding in cellulose remains speculative and arises from unknown sources of those bonds. For instance, 35-45% of protein is made up of 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 (Kauzmann 1959, Nemethy and Scheraga 1962), whereas, the possible source of such hydrophobic bonds in cellulose is unknown. Nissan (1978) questioned the method of evaluation of the interaction of cellulose surfaces with water by Goring.  He proposed that, a higher thermal expansion for  cellulose might have resulted from its higher degree of freedom due to breaking down of its hydrogen bonding network in water. Caulfield (1978) supported the idea that water near a cellulose surface is re structured by the cellulose surface rather than destructed by it. He suggested that such a model can not explain a fractional decrease in tensile modulus, which is constant up to the fiber saturation point on adding water to a moisture content equivalent to about eight layers of absorbed water (Fig. 6). Both authors (Goring 1978, Caulfield 1978) based their hypothesis  on Frank and Wen’s (109) “flickering cluster”* theory in aqueous  solutions. Milichovsky (1990) emphasized that the water molecules rather than the solute are the actual swelling agent. He concluded that swelling which occurred on beating, is a result of the changes in the composition of the phase interface between the pulp and water 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 roomtemperature, and it is governed by local energy fluctuation.  42  dipole nature of water molecules and on their two possible orientations in hydration spheres. The enigma with Milichovsky’s approach to the problem of cellulose swelling is in understanding the possible causes of different orientations of water molecules on the surface of the fiber.  -  Cellulose surface  \ \ ,\  \  \ \  C  U C  8 0  C 0  a  0  Cellulose surface Fig. 6. Schematic representation of the adsorbed water between two cellulosic surfaces (Caulfield 1978). The network was demonstrated as a distorted ice-like cluster with extension of hydrogen bonds from one surface to other.  43  It now appears, that the bonding problems experienced in low yield chemical paper recycling are due to a number of physico-chemical factors, of which the fiber-water interaction may predominate. More explicitly, the change in surface condition could not only be brought about by loss in surface accessibility of the fibers, but also by the changes in the nature and organization of the hydrophobic and hydrophilic bonding of the fiber-water vapor system. Considerable work remains to be carried out in the area of fiber-water interactions as it may hold, at least in part, the key to resolving the problems associated with free surface inactivation in recycling and indeed papermaking itself.  3.3 Summary The importance of this study arises from the view that the lack of scientific basis for 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” or converting it into pulp and paper “sludge”. This review suggests that strength loss in recycling is brought about by the loss in flexibility and/or changes in surface condition of the recycled fibers by emphasizing hornification. It is plausible for the cell wall of chemical pulps (low yield) that drying perhaps creates new and more perfectly aligned lamellae through hydrogen bonds as in crystalline cellulose. However, little information is available on the cause of these changes.  In other words, any explanation for the cause of strength loss requires an  understanding and elucidation of structural evolution of the fibers and fiber surface inactivation during recycling.  Furthermore, a comprehensive study of the surface  inactivation of fibers requires the understanding of the phenomena related to both fiber surface and fiber-water interactions.  44  4. MATERIALS AND METHODS  4.1 Preparation of Pulp Samples A 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 the time of arrival, the moisture content was measured to ensure that it was higher than fiber saturation point (FSP). Subsequently, the log was debarked and chipped. The chips were screened, collected in plastic bags and stored in the cold room. The moisture content of the chips was 40%. The chips were pulped by the kraft process at the BCIT batch digester. About 21.18 kg (green weight) chips were cooked. The white liquor (i.e., NaOH and Na S) charge was 34.90 L and the water added was 2 27.22 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. After almost 2 h the pulp was released from the digester, washed and collected in a double plastic 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 Materials  4.2.1 Preparation of chlorine. Bleaching and preparation of the chlorine and chlorine dioxide were done in the UBC wood chemistry laboratory.  Chlorine gas was bubbled in distilled water by  connecting the tank through tubes to 4 L adsorption bottles connected in series. The apparatus was set up in the fumehood. Bubbling was continued until the color of the  45  water changed to greenish after almost 4 h. The bottles were sealed and stored in the cold room.  Before pulp chlorination, the chlorine concentration in the water was  determined in accordance with the CPPA (Canadian Pulp and Paper Association) J.22P standari  4.2.2 Preparation of chlorine dioxide Chlorine dioxide is explosive.  So, it cannot be transported.  Therefore, it is  prepared 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 as shown in Fig. 7. The bottom of the cylinder containing technical grade  [  (  C,, C,,  0 0  I  Rotameter 400 mLfmin  FumehoodJ  Fig. 7. A sketch of chlorine dioxide preparation.  46  ) was covered with glass wool. About 3 cm of the cylinder was 2 sodium chlorite (NaC1O left empty from the top. The cylinder was connected through a tube to a series of 4 L bottles (filled with distilled water) at  40  C constant temperature. The flow rate of the  nitrogen was adjusted to 400 mL/min by a rotameter and the chlorine flow rate was conducted at 23 mI/mm. It required almost 4.5 to 5 h to make 4 L of chlorine dioxide  2 concentration produced by this process was about 7.5 g/L. The solution. The C10 bottles were sealed and stored in the dark cold room at 40 C.  4.2.3 Laboratory bleaching process The bleaching procedure is normally chosen according to the pulp type based on the target brightness. Generally, for low yield chemical pulps chlorine-based procedures  are used, while for high yield chemical and mechanical pulps, lignin-preserving chemicals such as sodium peroxide and sodium dithionite are commonly employed. The successive stages of bleaching are characterized by the appropriate chemicals and conditions. Laboratory bleaching is widely used to determine the bleaching response of puips under controlled laboratory settings. The unbleached kraft brown stock pulp was bleached using a CEDED sequence. The condition for the various stages are summarized as follows:  a. Chlorination stage (C) Chlorine solution was titrated for concentration according to CPPA J.22P. Three samples 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.  The  kneading was repeated every 15 mm. After 60 mm, the pulp was filtrated and residual  47  2 and pH were determined. Finally, the puips were washed thoroughly with distilled Cl water.  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% of the applied chlorine charge for each pulp. After 2 h the puips were filtered and washed well with distilled water.  c. Dioxide stage (D) Chlorine water and chlorine dioxide solutions easily lose chlorine in the gaseous form. 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 were double sealed in plastic bags and conditioned at 6% consistency, 74° C, for 3 h. The pulp were bags were kneaded for about 90 sec every 15 miii. At the end of 3 h, the puips the filtered and washed well with distilled water. After the chlorine dioxide stage (D), on and extraction stage (E) was repeated at 12% consistency, 74° C, for 2 h. After filtrati washing thoroughly, the last D stage was performed at almost the same condition as outlined for stage 3. After the completion of the CEDED bleaching sequence, the yield was 1 determined. The freeness of the pulp was measured in accordance with CPPA C. in standard method. Subsequently, the pulp was refined (beaten) in a PFI mill mL CSF). accordance with CPPA C.7 Standard for 12, 000 revolutions (i.e., to about 250 The characteristics of the pulp (yield, Kappa number and viscosity) were determined by CPPA G. 18 and CPPA G.24P standards respectively, and are summarized in Table III.  48  4.3 Recycling procedure od for recycling was followed. Generally, Howard and Bichard’s (1992) meth According to the observations of However, the white water was not re-circulated. themselves are also hornified during Szwarcsztajn and Przybysz (1976), the fines ts in enhancing fiber surface area, recycling. The role of the recycled fines, as agen , recycling of white water during the diminishes considerably in recycling. Therefore ct on the objectives of the present process of paper-recycling does not have any impa f. work and it would rather complicate the process itsel ratory scale. The recycling was The samples were recycled five times on a labo Each time 48 g (OD) of the never-dried carried out on both beaten and unbeaten fibers. CPPA C.4 standard and after pressing pulp was made into handsheets in accordance to relative humidity (RH) and 250 C. were left in the conditioning room for 24 h at 50% , equivalent to 24 g of oven-dry sample, For example, a representative sample of the pulp water. (i.e., 1.2 percent consistency) and weighed out. It was mixed with 2 L of distilled the stock was diluted to 16 L (0.15 disintegrated at 600 rpm. After disintegration, mL of the sample was poured into the percent consistency) with water at 23 ± 2° C. 800 water level in the barrel increased to the barrel of the standard sheet machine and the the perforated plunger and the water mark. Then the stock was carefully stirred with r was laid on the sheet, blotted off with drained. Two pieces of standard blotting pape rdance with CPPA C.4 standard. The blotting paper and the pulp sheet pressed in acco the paper conditioned at RH = 50% and sheet plates were mounted on drying rings and r 24 h the handsheets were collected and temperature of 25 ± 2° C for air-drying. Afte labeled as virgin paper (cycle 0). ed in distilled water overnight. For recycling, cycle (0) handsheets were soak rding to CPPA C.6 standard and made Subsequently, the sheets were disintegrated acco ts. This procedure was repeated 5 times. into handsheets again to give cycle I handshee d in the cold room for subsequent From each cycle some of the samples were store  49  analysis of fiber crystallinity, fiber surface area and thermodynamical properties (See also Fig. 8). No beating was imposed between the cycles. This procedure was repeated 5 times (Fig. 8). In order to test and make sure that the recycling modified the strength properties of the paper, for each cycle 10 handsheets were tested according to CPPA standard methods (CPPA D.6H, CPPA D.8 and CPPA D.9) for strength properties (i.e., tensile, tear and burst respectively).  50  Fig. 8. A flow chart of the experimental procedure.  51  4.4 Fiber Length (FL) Analysis The FS-200 fiber length analyzer consists of a capillary tube through which an aqueous suspension of the fibers is passed. A light source is located on one side of the capillary, and a detector is positioned at the opposite side. As a fiber passes through the capillary tube, its image is projected onto the detector with the aid of the light source and system optics, providing information from which the lengthwise dimension of the fiber may be calculated (See also Fig. 9).  Analyzer  ‘-[ComPuter  Fig. 9. The block diagram of the FS-200 fiber length analyzer.  1  52  Fiber length distributions of both virgin and recycled pulps were determined using the Kajaani FS-200 fiber analyzer. The pulp suspension was poured into a flask and sealed with a glass stopper. The flask was shaken manually until the fibers dispersed completely. 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.  The  consistency of the suspensions was such that a minimum of 30,000 fibers were counted for each sample.  4.5 Molecular Weight Distribution (MWD) The samples were thcarbanilated according to the method suggested by Kossler et al. (1981) and Schroeder and Haigh (1979). Accordingly, 0.1 gram of the sample was placed in 500 mL of phenyl isocyanate. The flasks were then sealed with glass stoppers and placed in an oven at 80° C for 12 h. Completion of the reaction was monitored visually (i.e., clear light yellow solution) in order to prevent degradation during tricarbanilation.  This control was critical, because tricarbanilation times of different  fibers were different and depended on the pre-history of the samples. The tricarbanilated cellulose 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 dried at 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 the tricarbanilated samples prior to dissolution in TI-IF accelerated the process. Extraction of the thcarbanilates with methanol in a Soxhiet extractor completely removed the by products (N, N T diphenyl urea and methyl phenylcarbamate, and oligomers products were originally identified by Wood et al. (1986).  ). The by  These by-products can  53  produce a large peak after an elution volume of 32 mL thus disturbing the final determination 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-Physics SP8810 liquid chromatograph. Samples of cellulose tricarbanylate in THF were filtered through a Teflon membrane with a pore size of 0.45 I.tm and analyzed using a series of four TSK-GEL (type H) columns. THF was used as the eluting solvent at a flow rate of 1 mL/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 the integrator SP 4229 for peak integration and plotting the cummulative and differential distributions of molecular weight. The number and weight average DP of the cellulose was 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 tricarbanyl derivative of anhydroglucose (DP  =  MW/519). Each sample was analyzed three times by  GPC. Because of the lack of commercially available standards of cellulose tricarbanylate, the GPC carbanilation curve (i.e., the correlation of elution volume with molecular weight) was generated from the elution profile of polystyrene standards with narrow MW distributions (Appendix F). The following equation was used.  InM  =  [(1  +  cL)1nM  +  1n(K/K)]/(1 +z)  [4.11  where M and M are the MW of cellulose tricarbanylate and polystyrene, respectively. The Mark-Houwink coefficients used in the present analysis for polystyrene in THE were: K  =  1.18 x lO- and o  x 10-s and (Xc  =  =  0.74, and for cellulose tricarbanylate in THE, K  0.92, as reported by Valtasaari and Saarela (1975).  =  2.01  54  4.6 Crystallinity and Crystallite size  4.6.1 D 5000 Diffractometer An automated, Siemen’s diffractometer system equipped with a carbon monocbromator was used for the virgin and recycled fiber crystallinity analyses. The diffractometer was controlled by a computer. The diffractograms can be displayed on a printer attached to the computer. Figure 10 demonstrates block diagram of the Siemens D 5000 diffractometer system.  fl  sil  a  pLm  III  Fig. 10. Block diagram of diffractometer in D 5000 mode.  55  Generally, the emitted radiation from the line focus of the x-ray tube is diffracted at the sample and recorded by the detector. The sample rotates at a constant angular velocity such that the angle of incidence of the primary beam changes while the detector rotates at double angular velocity around the sample. The diffraction angle (2) is thus always 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 the  connected measuring electronics measure the intensity of the reflected radiation. Diffraction patterns are obtained in this way. In order that the diffracted radiation can be focused when it hits the detector, the whole effective sample surface should actually be on the focusing circle. The diffractometer beam path of the system is demonstrated in Fig. 11.  Focus X-ray tube  Detector diaphragm  Aperture diaphragm a  Scattered-radial diaphragm  Detector l  Focusing circle Sample  f  -  .easuringcircl.  —.  Fig. 11. Diffractometer beam path in /29 mode.  56  4.6.2 Sample Preparation for X-Ray Diffraction Analysis The samples, which included unbeaten and beaten (never-dried) and Cycle I  -  V  puips, were prepared as follows. The virgin (unbeaten and beaten) pulps were dried by the critical point drying (CPD) method. The advantage of the method is that the fibers retain their superstructure almost intact. Then the samples (i.e., unbeaten, virgin and Cycle I-V puips) were ground in a Willey mill to pass through a 30 mesh screen. Five samples 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, the samples were placed in a mold of 2.2 cm x 4.5 cm x 0.4 cm and pressed for 5 mm under 344 kPa pressure by the press. The thickness of the wafers was about 0.4 cm. Good distribution of the sample material was important and consequently equal thickness of the wafers was a major criterion in sample preparation. The pulp samples were analyzed almost under identical conditions. Five sheets were prepared from each sample and the samples were sealed in a plastic bag. One of the virgin paper handsheets was re-slushed, disintegrated, drained on a standard sheet machine and pressed. repeated 10 times.  This procedure was  This handsheet set has thus been recycled 10 times without  experiencing drying as used in the standard handsheet making process. The x-ray diffraction of each set of samples was recorded using a Siemens diffractometer equipped with a D 5000 rotating anode x-ray generator. The wavelength of the CuJKcx radiation source was 0.154 nm, and the spectra were obtained at 30 mA with an accelerating voltage of 40 kV.  Samples were scanned on the automated  diffractometer between 5 to 40° of 2 e (Bragg angle), with data acquisition taken at intervals of 0.04 for 1 second. The diffractograms of the virgin and recycled fibers were analyzed by a peak resolution program.  57  4.6.3 XRD data analyses A peak resolution program was used to calculate both the crystallinity index and full width at half maximum (FWHM) of the 101 and 002 peak of the traces. FWHM is inversely proportional to the crystallite size of the fibers. This program allowed the resolution 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 separation of overlapping peaks is based on the assumption that the intensity curves of the reflections have the shape of the Voigt function.  This function separates the  diffractogram pattern of native cellulose substrate to major diffraction planes (i.e., 101, 101, 021, and 002 planes). The program also incorporates an iterative procedure which ensures efficient curve fitting. Generally, the Voigt function resulted in the best fit of the x-ray diffraction patterns and is routinely used for the determination of both the crystallinity index and crystallite dimensions of polymers or metals (Krassig 1984). The program has a conversational dialog that prompts the user for necessary information about any particular calculation. The information is basically generated after the process of curve fitting is completed and a desirable peak resolution is achieved (See also Appendix A).  4.7 Water Vapor Sorption Samples of virgin and recycled fibers were collected during the recycling process as reported previously. Soda-boiled cotton was used as control. The cotton samples were subsequently dried with CPD technique and conditioned at the appropriate RH and temperature. Samples of virgin (beaten and unbeaten), recycled pulp and cotton fibers were then placed in desiccators with the humidities varying from 5 to 100%. The cotton fibers were used as control. The relative humidity inside the desiccators was controlled  58  by  varying the sulfuric acid concentration (Table I), based upon the relationship  established by Stokes and Robinson (1949).  Table I Relationship of sulfuric acid concentration with relative humidity (H) at 25° C  *  . 3 Density of sulfuric acid is 1.83 18 g/cm  The desiccators were placed in a two-level plexiglass chamber together with an analytical balance. The first level of the chamber was filled with certain salts in order to modify the relative humidity of the chamber (Table II).  59  Table II Relative humidities over saturated salt solutions at 25° 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 as much as possible. The whole apparatus was set up in the conditioning room with a temperature of 25°±2 C (Fig. 12). Water vapor sorption was followed by the gravimetric method. All weighings were performed on an analytical balance with a resolution of 0.01 mg. The weighings were made at 3, 7, 15, 20, 25, and 30 days.  60  I  j Balance  Acid  Salt  Salt  Salt  I  I  CIII room -J  Fig. 12. Block diagram of apparatus which was developed for conditioning the fibers at different relative huniidities with a constant temperature (i.e., 25±2° C).  All measurements for sorption were performed in duplicate. Two sets of samples were prepared for the investigation including three samples for each set. The first set of samples, including the virgin (unbeaten and beaten) pulp and cotton, were conditioned for water adsorption. The virgin (unbeaten and beaten) fibers were dried by different  61  methods (i.e., freeze drying and critical point drying).  The second set, including  unbeaten and beaten fibers, were conditioned for the desorption process. For the latter set, 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 relative humidities (RH). The relative humidity of the chamber was controlled by adjusting the wet and dry bulb temperatures. The sorption data was collected at relative humidities -  -  of 35, 50, 60, 80, and 90%. The data around 35% and after 80% fluctuated widely which made it difficult to keep the temperature constant for a given relative humidity. The temperature 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% relative humidity. Therefore, data collected in the conditioning chamber could not be used for the analyses in the present investigation. The sorption ratio along with the percentage of the amorphous component of the cell wall was deduced from the sorption data. The sorption ratio (SR) is defined as the ratio of the water-vapor sorption of a cellulose sample to that of cotton under the same conditions (Valentine, 1957). The fraction of amorphous cellulose (em) is then obtained from the quotient of the sorption ratio and completely accessible cellulose.  4.8 Measurement of the Heat of Wetting The 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 this investigation the heat of wetting of the fibers was measured by solution calorimetry. The heat of wetting as measured by this technique assisted in calculating the enthalpy of the fiber-water vapor system and also was used as an indicator for accessible surface area of the fibers.  62  Glass push rod  Glass cell  Stainless air can  Glass  Teflon Sample dish  Fig. 13. A cross-section of the PARR 1451 solution calorimeter.  The PARR 1451 solution calorimeter consists of a glass Dewar reaction chamber with a rotating sample cell, a thermistor probe and a specifically designed temperature measuring bridge, all assembled in a compact cabinet. Temperature changes are plotted directly in Celsius degrees using an accessory strip chart recorder. At the start of the test,  63  distilled water (100 mL) is held in the glass Dewar while the fiber (ball milled) is held in a sealed glass rotating cell which is immersed in distilled water (Fig. 13). The cell was loaded with 0.4 g of the sample each time and allowed to rotate for 10 to 15 minutes until the temperature of the sample equalized with that of the distilled water in the glass Dewar flask. The system comes to equilibrium in 10 to 15 mm with only a slight temperature drift 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 under vigorous 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 the calorimeter 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 the adsorption approach. Therefore, the integral heat of adsorption was also calculated from the data collected from solution calorimetry. The heat of wetting of a sample containing a certain amount of moisture is measured and subtracted from the heat of wetting of the oven dry sample (All  =  H Hmot). From a series of determinations, the integral heat -  of adsorption from dryness to any desired water content can be estimated.  The  differential heat of adsorption is found from the slope of the curve of integral heat of adsorption 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 method of 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 is followed by exchange of the dehydrating solvent by a transitional liquid such as carbon  64  dioxide. Subsequently, by increasing the temperature (i.e.,  400  C) and pressure ( 35  kPa), the transitional liquid is converted to the gas wherein the densities of the liquid and gas of the fluid are identical. Consequently, the liquid in the cell wall transforms from the liquid state to the gaseous state. The replacement of the liquid by gas eliminates the creation of tension forces between the fibrils/lamellae when the fibers are dried from the never-dried state.  4.10 Determination of Thermodynamic Properties of Fiber Measurement of the energies and amounts of sorption of water vapor by virgin fiber, when compared with those on recycled fibers, have led us to detail the nature of irreversible swelling of the dried fibers. The internal and external surface area of the fibers were calculated by employing the 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 and integral free energies, according to Eqs. of [2.6] and [2.7]. Equation [2.8] was used to calculate the surface-energy per unit area. The heat of wetting data, which are obtained by the calorimetric method, are used to calculate the integral enthalpy changes AH, by using equation [2.9]. The differential enthalpy 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].  65  411 Enzymatic Hydrolysis Enzymatic hydrolysis of cellulosic substrates was performed using the Cellucast Cellulase preparation (Novo Nordlisk, Denmark) supplemented with the 3-glucosidase from Novozym.  Enzymatic activities were determined in the cnzyme mixture as  described previously by Ghose (1986) and Chan et al. (1989).  The amount of enzymes  used for hydrolysis was adjusted to a final activity of 8.5 FPU (filter paper units) and 23.8 CBU (cellobiase units) per gram of cellulose. The substrates were hydrolyzed in duplicates at 2% (W/V) solids concentration in 1 of tetracyclin as a 50 mM sodium acetate buffer at pH 4.8, containing 6 mg mLpreservative, at 45° C and stirred at 150 rpm. The release of soluble sugars during hydrolysis was monitored by HPLC, using an HPX-87H column (Bio-Rad) (Schwald et al. 1988). Glucose yield was defmed as the percentage of available cellulose at time zero which 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.  The  residue was washed extensively with distilled water to remove most of the non-adsorbed enzymes and soluble sugars. The residue was then resuspended in water, collected by filtration and stored at 4° C for subsequent analysis of fiber length (FL), molecular weight distribution (MWD), scanning electron microscopy (SEM), and crystallinity.  66  5. RESULTS  Relevant characteristics of the pulp used in the present work are reported in Table ifi. Figure. 14 demonstrates changes in the physical and optical properties of the fibers as induced by recycling. According to the Fig. 14 both tensile and burst strengths  Table ifi. Characteristics of the white spruce kraft Pulp (unbleached and bleached (CEDED)).  Yield  S.  V  K  -ce1  ‘  /  Uzthlèhd 4620  2  4400  xLd  Bleached K: V: ot-cel: Hemi-cel: : nd  ‘  1140  I  17  emi*ce  (%)  8800  /  470  1200  Kappa number, Viscosity (cp), Alpha-Cellulose (%) Hemicelluloses (%) Not determined  decreased by up to 30%, whereas the tear strength increased up to 40%. The main drop in strength 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 as  measured by zero-span tensile strength was moderately affected (drop of 3.8% after the  67  0 C .t U  a£  0.  Number of cycles  Fig. 14. Changes in properties of the handsheets (made from beaten bleached kraft) due to recycling.  fifth cycle) during the course of recycling (Fig. 14). As it is demonstrated in Fig. 15 and Table IV, the fiber length profiles and average fiber lengths are almost the same for the beaten, unbeaten and recycled puips. A slight increase in the average fiber length, for the recycled pulp, may be induced by the loss of the fines. Distinct loss in tensile strength of the paper with a minor loss in fiber strength (if there is any) led to the conclusion that, the inferior strength quality of recycled papers was induced by loss in potential bonding of  68  it was reported as early as 1971 by the fibers. However, this observation is not new and McKee.  0  I 0  0.6  1.2  1.8  2.4  4.2 3.6 3 (mm) length Fiber  4.8  5.4  6  on of unbeaten, beaten and recycled Fig. 15. Comparison of fiber length (FL) distributi fibers.  69  Table IV. Fiber length (FL) analysis of virgin (unbeaten and beaten) and recycled puips  Sn: A(Ave): L(Ave): M(Ave):  Sample name Arithmetic (average) Length-Weighted (average) Mass-Weighted  Scanning electron microscopy (SEM) was employed to investigate the visible characteristics of the recycled fibers. A comparison of the SEM micrographs of unbeaten and recycled (initially beaten) fibers (Figs. 16A, 16B, 16C, 16D) demonstrates that, recycling diminishes the difference between unbeaten and beaten fibers. The external fibrillation induced on beating mostly disappeared after the first cycle (Fig. 16C). Indicating that one of the parameters (i.e., external fibrillation) which could assist the strength development of paper which is induced by mechanical treatment, is lost by recycling.  70  Fig. 16. SEM micrographs of (a) unbeaten and (b) beaten pulp fibers.  71  M 1 100L-  Fig. 16. SEM micrographs of (c) cycle I and (d) cycle VI pulp fibers.  72  To characterize further the effect of recycling, the accessibility of the fiber surface to water sorption was investigated (Figs. 17 and 18).  24  Cycle I  1  22  4,  4,  20 I  4,  18 16 4 .4  4—  Beaten pulp  14  /  o  o V 12  4..  Unbeaten pulp  10  /  ‘  .4.  1  1  /  4,  /  I  7__  4.——  Cttcrn  .4.....  8 — ——  4•  4  ••‘,.  .)I  .ie  2 0  i  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  051015202530354045505560657075808590  Relative 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 be applicable for assessment of the fiber surface area (Table V). The water sorption values  73  for both unbeaten and beaten fibers are shown in Fig. 17.  As seen, the adsorption  isotherms are identical for both unbeaten and beaten fibers, while they differ for the recycled fibers (Figs. 17 and 18). This suggests that both unbeaten and beaten samples g). On the other hand, the recycled fibers m / possess an equal fiber surface area (384 2 behaved differently during the sorption process. Their moisture uptake potential was reduced, which indicates that the total fiber surface area (i.e., internal and external surface area) was diminished by drying  (  g). The subsequent recycling m / about 238 2  further exacerbated the sorption power of the fibers (Fig. 18).  ,n.  1&  1& Cycle I (desorptlon)  14  E  Cycle V (desorptlon)  E 812  0  1o  Cycle V (adsorption) Cycle I (adsorption)  a  6  I  30  40  I  I  70 60 Relative humidity (%)  Fig. 18. Comparison of sorption isotherms of cycle I and cycle V.  80  : : : : :  :  :  : :  Vb b 1 V 1 Cf Sn C  Mm  LHBET  1 n SD  Virgin fiber (beaten) Virgin fiber (unbeaten) Cycle I Sample name Adsorption energy Total surface area Capacity of monolayer External surface area Internal surface area Bonding energy between water and fiber by BET approach (KcaL’mole) Bonding energy between water and fiber by Calorimetry approach (Kcal/mole) Mole of water per 100 g of solid Standard deviation of external and internal surface area.  Table V. Surface area for virgin (unbeaten and beaten) and recycled fibers.  75  To evaluate the behavior of beaten fibers by the sorption technique, three neverdried samples were also tested for desorption. The sets included one unbeaten and two beaten (i. e. 6000 and 12000 revolutions respectively, by PFI mill) samples. Refining up to 12000 revs, by a PH miii is a drastic mechanical treatment. of the fibers. differences were detected in the desorption isotherms of these three puips (Fig. 19).  20  15  44-  8 10 z 1;;  Unbeaten 5  Beaten (6000 & 12000)  a  0  40 io Relative humidity (%)  Fig. 19. Desorption isotherms of unbeaten and beaten (6000 and 12000 revs.).  80  No  76  This experiment further confirmed the previous observation that the refining/beating does not alter the sorption potential of the fibers. To understand further the effect of the drying process and precision of the  sorption technique, the water moisture sorption of freeze-dried an. critical point-dried puips were compared in Fig. 20. The adsorption isotherms of the fibers dried by different drying techniques showed about 20% moisture difference in the range of 20-70% relative  humidity (See also Fig. 20). This is a considerable difference between the moisture sorption potential of the fibers which were dried by two different techniques. Originally, both techniques were used by different laboratories to preserve the original structure of  24 22 20 18 16 C)  14  8 12  Beaten (freeze dried) Unbeaten (critical point  10 Beaten (critical point dried)  8 6 4  Unbeaten (freeze dried)  2 0  10  15  0  25  30  5  40 45 50 55 60 Relative humidity (%)  65707580859095  Fig. 20. Comparison of critical point drying and freeze drying methods with sorption potential of the fibers at 25° C.  77  the fiber wall. More explicitly, the structural differences between the fibers which were dried by critical point drying and freeze-drying technique should be minor. This suggests that a marginal difference in the drying condition can be reflected in distinct differences in the accessibility of fiber surfaces.  Furthermore, it is suggested that the sorption  isotherm measurement is a sensitive technique and has the potential to detect minute variations in sorptivity of the fiber. Thus, the identical responses of unbeaten and beaten fibers did not initiate from poor detectability by the technique. This observation suggests that the poor sorptivity of recycled fibers is induced by loss in accessible surface area for water.  35  ,30 ‘-4  .25 Cycle I Unbeaten  .5  [J  15 Beaten  r’  10  5.  0  I  I  I  I  2  4  6  8  I  10  I  12  I  14  I  16  I  18  20  22  Moisture content (%)  Fig. 21. Comparison of the heat of wetting of unbeaten, beaten and recycled fibers.  24  26  78  The calorimetric measurement results for the samples also followed the same trend, 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 of the experimental procedure but considerable differences were observed for the recycled samples (Fig. 21). These observations imply that the total accessible surface area of the unbeaten  and beaten fibers are the same, and both diminished when the fibers were recycled.  C’, C’,  0 >  C 0  .5  .0  0  Elutlon volume (mL)  Fig. 22. Comparison of MWD of unbeaten, beaten and recycled fibers.  79  The interdependence of structure and properties of the fibers have always been of  great interest to papermakers. Krassig (1984) proposed that the physical properties and chemical reactivity of fibrous cellulose are not only dependent on the chemical constituents, but are also affected by the spatial arrangement of the. macromolecules in the “architecture” of the fibers. Thus, the molecular and supermolecular structure of the fibers were investigated. In Fig. 22 the cbromatograms of virgin (unbeaten and beaten) and recycled pulps are superimposed on each other. Additionally, the DP of the samples remained constant during the course of recycling (Table VI). Polydispersity was also not greatly affected by  Table VI. Degree of polymerization (DP) of the virgin (unbeaten and beaten) and recycled pulps  Sn =Samplename DPw = Weight average DP DPN = number average DP P = Polydispersity  80  recycling.  The chromatograms of molecular weight distribution do not support any  changes in the population of high and low molecular weight chain fragments in the samples due to recycling. Therefore, it was concluded that, the inferior properties of recycled fibers are not induced by cellulose chain cleavage.  -  Superstructural arrangements of recycled fibers were examined by XRD technique. The diffractograms of unbeaten and beaten fibers were overlapped.  The  comparison of virgin and recycled fibers suggests distinct differences between  crystallinities of the respective samples (Fig. 23 and Table VII). Furthermore, the  0,  I 25  Bragg’s angle (2theta)  Fig. 23. Comparison of the diffractograins of unbeaten, beaten and recycled fibers.  81  data indicate that each cycle in the process of successive recycling adds a new increment of ordering in the fiber wall (Fig. 24). An increase of almost 9% in crystallinity after the first 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 first cycle. The decrease in the full width at half maximum intensity (FWHM) of the 002  peak from 2.9 to 2.57 is due either to an increase in crystallite size or crystallite perfection due to recycling. Whichever the case may be, the swelling potential of the  fibers 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 virgin  and recycled fibers. During the course of the present work, it was observed that the amorphous fraction decreased by almost 22% after the first cycle (Table VII).  Table VII. Crystallinity information on virgin (unbeaten and beaten) and recycled pulps.  Fw:11M 002 degite  Or!  (°k)  Ünbeatcn  2 018  Beaten  SD  FM1flM &grcc  SK  Fm  68 85  0+50  190  1  OA6I  2S178  6885  038  290  16  061  VycleI  1.868  7300  068  2.57  123  047  Cycle!!  1.770  xxi  nil  nil  ml  nd  Cycle!!!  1820  lIMO  083  265  nil  xxi  cycier  1774  7570  0.38  2.55  nil  aid  s’!: ,  FWHM: Full width at half maximum, Cr1: Crystallinity index SD: Standard deviation, SR: Sorption ratio, Fm: Fraction of amorphous material of cellulosic fibers  82  (I,  Ez  Ez  I 25  Bracia’s anale (2 theta  Fig. 24. Comparison of the diffractograms of virgin (beaten), cycle I and cycle Ill puips.  Interestingly, no difference between the sorption isotherms of beaten and unbeaten puips were observed, which shows that the amorphous component of the fibers has not been altered 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 observed increase in crystailinity, is induced by crystallization of amorphous phase and fuzzing up of fibrilsflamellae during recycling. This interpretation is in agreement with previous observations on the loss of accessible surface area of the fibers.  83  Thermodynamic properties of recycled fibers were also studied, with the hope that information would be collected on the structural changes of recycled fibers by interacting with water. Free energy, enthalpy and entropy of recycled fibers were  0  Moles of water per 100 got solid (ni)  Fig. 25. Comparison of the differential free-energy of virgin and recycled fibers.  84  -,  Cu .  C Iii  Moles of water per 100 g of solid (ni)  Fig. 26. Comparison of the differential enthalpy of virgin and recycled fibers.  measured and compared with virgin fibers. The overall free energy and enthalpy of the virgin 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. The higher free energy indicates that, the virgin fiber surface has more affinity to adsorb water than recycled fibers do.  85  The hydrogen bonding energy of recycled and virgin fibers were the same within the limit of experimental error. The total hydrogen bonding energy between water and cellulose 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 and virgin fibers cannot be explained by the concept of strong and weak bonds.  More  importantly, 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 moist environment.  This disorganization of virgin fibers could be explained by the high  swelling potential of the virgin fibers during water uptake.  2 w  I  Moles of water per 100 g of solid (ni)  Fig. 27. Comparison of the differential entropy of virgin and recycled fibers.  0 0.09 0.18 0.27 O.36 0.45 0.54 0.63 0.72 0.81 0.90 0.99 1.08 1.17 1.26 1.35  fli  0 0.083 0.123 0.162 0.236 0.410 0.533 0.620 0.700 0.734 0.770 0.803 0.831 0.854 0.882 0.909  P/Pa 0 1.084 1.463 1.667 1.525 1.097 1.013 1.016 1.030 1.103 1.169 1.233 1.299 1.370 1.428 1.485  n / 0 p 1 p 162 282 387 475 539 578 607 629 647 662 675 686 694 702 707  -AF 1474 1241 1078 855 528 373 283 211 183 155 130 110 93 74 56  1 -iF 154 324 478 578 684 800 900 1000 1036 1066 1118 1120 1124 1130 1140  -AH 2592 2000 1754 1522 1317 1185 1034 749 544 370 229 147 96 45 20  1 -AH —0.027 0.140 0.305 0.345 0.486 0.744 0.982 1.244 1.304 1.355 1.486 1.455 1.442 1.435 1.452  -zS 3.75 2.54 2.27 2.24 2.64 2.72 2.52 1.80 1.21 0.72 0.33 0.12 0.01 —0.09 —0.12  1 -S 0.83 1.45 1.98 2.43 2.76 2.96 3.11 3.22 3.32 3.39 3.46 3.52 3.56 3.60 3.62  IF/area  Table VI. Calculated thermodynamic properties for the adsorption of water on beaten and unbeaten cellulose dried by critical point drying method.  co  0.1485 0.2486 0.4200 0.6000 0.6657 0.7057 0.7400 0.7700 0.8050 0.8286 0.8548 0.8726 0.8936 0.9150  0.09  0.18 0.27 0.36 0.45 0.54 0.63 0.72 0.81 0.9 0.99 1.08 1.17 1.26 1.35  P/Po  0 0.0857  o  flj.  1.212 1.086 0.8571 0.7500 0.8112 0.8927 0.9600 1.0500 1.1252 1.1990 1.2600 1.3300 1.4000 1.4702  1.05  n / 0 p 1 p  278 367 428 463 487 506 522 537 550 562 572 580 586 591  161  -AF  1130 824 514 303 241 206 178 156 130 112 93 80 65 54  1455  1 —AF  324 478 584 712 812 900 960 984 1012 1031 1042 1047 1049 1054  154  -AH  1814 1515 1254 975 835 545 303 218 155 100 47 40 30 10  2450  1 -AH  0.154 0.372 0.523 0.835 1.090 1.321 1.469 1.499 1.549 1.573 1.576 1.567 1.553 1.553  —.02  -AS  2.29 2.32 2.50 2.25 1.99 1.14 0.42 0.21 0.08 —0.04 —0.15 —0.13 —0.12 —0.15  3.34  1 -AS  1.59 2.09 2.45 2.65 2.78 2.89 2.98 3.07 3.14 3.21 3.27 3.31 3.35 3.38  0.92  AF/area  Table VII. Calculated thermodynamic properties for the adsorption of water on recycled fiber. The fibers dried by critical point drying method.  88  The virgin (unbeaten and beaten) and recycled puips were also examined by enzymatic hydrolysis. The initial hydrolysis yields of virgin (unbeaten and beaten) and recycled fibers were compared and changes in the characteristics of the substrates monitored by different techniques. About one third of the beaten pulp was hydrolyzed after 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 hydrolyzed beaten pulp was, on average about twice as long as the other three puips (Table X). This observations suggested that the beaten fibers had more accessible surface area for enzyames than unbeaten and recycled fibers.  Table X. Fiber length (FL) analysis of hydrolyzed (H) puips  Sn: A(Ave): L(Ave): M(Ave):  Sample name Arithmetic (average) Length-Weighted (average) Mass-Weighted (average)  After hydrolysis, it seemed that both unbeaten and recycled puips were fragmented to a greater extent than the beaten puips (Figs. 28A-28D), correlating well with their greater reduction in fiber length during hydrolysis (Table X).  In fact, the  89  presence of intact, non-degraded fibers was detected only within partially hydrolyzed residues derived from the beaten pulp (Fig. 28B).  90  Fig. 28. Comparison of scanning electron photomicrographs of hydrolyzed substrates derived from (A) unbeaten and (B) beaten puips.  91  Fig. 28. Comparison of scanning electron photomicrographs of hydrolyzed substrates derived from (C) cycle I and (D) cycle VI puips.  92  The possible changes that could affect the degree of polymerization (DP) of each of these puips during hydrolysis (Table XI) were also examined. The beaten pulp had an initially higher weight average DP, which decreased substantially, after 4 h of enzyme hydrolysis. Both unbeaten and cycle I puips showed a similar trend, with about a twothirds reduction in DP resulting from enzymatic hydrolysis. The cycle VI pulp showed the smallest reduction in the weight average DP and virtually no change in the number  Table XI. Degree of polymerization (DP) and hydrolysis yield of the hydrolyzed (H) pulps  Sn: N(Ave): W(Ave): P: Y:  Sample name Number (average) Weight (average) Polydispersity Yield (%)  93  0 a  a  Elutlon volume (mL)  Fig. 29. Comparison of MW]) of hydrolyzed samples  average DP. After hydrolysis, the polydispersity of all the puips except the unbeaten pulp was very similar (Table XI) despite significant differences in the MWD peak distribution (Fig. 29).  It is possible that the extent of the swelling is high due to  extensive beating (i.e., 12000 revolutions by PFI mill), to allow access of the large enzyme molecules  to hydrolyze the  substrate  through  interlamellar  surfaces.  Alternatively, both unbeaten and recycled fibers, due to their more compact structure  94  (i.e., less swelling), appeared to have been hydrolyzed structurally in a more length-wise fashion rather than any direct changes in the surface of the fibers. The conclusion is supported by the observation that the fiber length of the beaten hydrolyzed pulp fibers were twice that of the fiber length of unbeaten and recycled fibers while the DPw of the beaten fibers (hydrolyzed) was half of the counterparts. The low DPw of beaten fibers with their higher average length suggested that the enzymes had more access to the cellulose surface area to attack than unbeaten and recycled fibers. When the same hydrolyzed and unhydrolyzed pulps were assayed for changes in their crystallinity (Table XII), the beaten pulp showed virtually no change in its crystallinky index (Cr1) while the other three pulps showed a substantial increase in the degree of crystallinity. The rather low crystallinity of beaten fibers clearly indicated that the enzymes had access to the crystalline component of cellulose surface. In the other words more crystalline surfaces were made available for enzymatic attack on beaten fibers when it is compared with unbeaten and recycled fibers.  Table XII. Crystallinity of hydrolyzed (H) and Unhydrolyzed (U) Pulps  Sn: FWHM: SD:  Sample name Full width at half maximum Standard deviation  95  This is consistent with the figures previously obtained for both fiber length and molecular weight distributions during hydrolysis. The lower crystallinity of the beaten fibers also support the belief that the hydrolysis is interlamellar in the case of beaten fibers. These observations bear structural information on recycled fiber wall which will be elaborated more on this in the discussion.  96  6. DISCUSSION  The aim of this dissertation has been to investigate the probable causes and mechanisms which influence the inferior properties of recycled fibers. This is evidently fundamental to the whole art of upgrading recycled fibers.  It was felt that a better  understanding of various experimental results and their interrelation could be better explained by discussing them separately. Thus, the experimental results are organized in a certain order in regard to their relevance.  The comparable work done by other  investigators are then reviewed. The discussion is terminated by connecting the common aspects 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 isotherms The adsorption isotherms of unbeaten and beaten fibers overlapped (Fig. 17). On the other hand, distinct differences were observed for the adsorption isotherms of virgin (unbeaten and beaten) and recycled fibers (Fig. 17). This suggests that beating/refining does not alter the sorption potential of the fibers, while recycling reduces the sorption potential of the fibers. This is consistent with the belief that the beating process increases the volume of the fibers but does not alter the fiber accessible surface area to water vapor, in contrast to the recycling process. The adsorption isotherms in the vicinity of monolayer completion (i.e., around RH=30% and 45% for recycled and virgin fibers respectively) suggest a distinct difference between virgin (unbeaten and beaten) and recycled fibers. This difference indicates that the virgin fibers has larger capilaries which recycled fibers are lacking. In the other words, recycled fibers contain less accessible surface area when it is compared  97  with virgin fibers. The appreciable separation in the sorption curves between virgin and recycled fibers continues up to 90% relative humidity (Fig. 17). The second section of the curve (i.e., 30 80%) corresponds to the “filling up” of -  the interstices between sheets of microfibrils in the cell wall (Fig. 17). The rigidity of lamellae would have a negative impact on the filling up of the spaces between them. The coarser* (i.e., more rigid) the lamellae, the more condensed would be the water layer and thus the volume of the pores remains intact. This is the state by which the recycling effect could be explained. Due to re-joining of the lamellae, the rigidity of the sheets of microfibrils increases, thus the swelling potential of the cell wall decreases. In the last stage (80  -  90%) relative humidity, the differences between the vapor  pressure of the water in the cellulose and the water in natural condition, as it was formed in nature, will disappear largely.  However, the difference between the virgin and  recycled fibers, has not yet been lost. Initially it was reported (Patrick, 1924) that the hysteresis observed for the al of cellulose was attributable to experimental error, being due to the incomplete remov in the last traces of adsorbed water from the sample under examination. Urquhart (1929) se a comprehensive work showed that the sorption hysteresis observed with cotton cellulo is an intrinsic property of the material which cannot be explained as being due to faulty the methods of experimentation. This observation of Urquhart (1929) seems to refute idea that the irreversible closure of recycled fibers (i.e., dried from water) can not be explained by failure in evacuating the smaller pores.  The coarseness of lamellae here refers to the thicker lamellae which are produced by ineversible closure of two or more single lamellae. *  98  6.2 Heat of immersion t produced by two The heat evolved on adding water to cellulose is the net amoun in existence and the other processes. One is the sorption of water on the cellulose surface effect is negative. is the production of new cellulose surface and swelling. The latter virgin and recycled Figure 21 represents the iinmersional adsorption isotherm of regions of the curves fibers containing various amounts of adsorbed water. The initial re content increases. describe a nearly linear decrease in the heat of wetting as the moistu to be indicative of an Zettlemoyer et al. (1955) considered this type of relationship wetting appears to depend energetically homogeneous surface. The heat evolved on distribution of energy sites simply on the amount of bare surface present. If there was a ntially adsorb vapor on the surface of the solid, the high energy sites would prefere to the amount of surface molecules and the heat of immersion would not be proportional ed. (Fig. 21) but, instead, would depend on which sites were occupi d fibers (Fig. 21). The heat of wetting of virgin pulp is greater than that of recycle a greater internal area than This is in agreement with the concept that the virgin pulp has consistent with observations the recycled pulp as indicated by sorption isotherms. This is nces between the adsorption of Argue and Maass (1935) who reported data on the differe authors proposed that the and desorption isotherms of purified cotton cellulose. These ed, had a larger surface than cellulose from which a portion of the water has been desorb dry sample. the cellulose containing the same amount of water added to a en and beaten (i.e., No detectable difference in the heat of wetting between unbeat both fibers in the ranges of 0never-dried) fibers suggests an equal fiber surface area for with the observations of 12% moisture content (See also Fig. 21). This is in agreement ible surface area by two Maass and Campbell (1939). The authors, extended the access iron surfaces; the other by ways: one, by pounding the dry cellulose between two lb/in2. A 25% increase in heat of subjecting the cellulose to a static pressure of 50000 in of wetting in either case wetting of the fiber material was obtained. The increases heat  99  were 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 unstreched and stretched nylon levelled off at 1250 and 550 cal/mole respectively, that is, about 2 times higher in the case of unstreched nylon. The authors attributed the lower water uptake of stretched nylon to its crystallization during stretching. These results and their interpretation are in excellent agreement with those obtained by determining the sorption isotherms for the pulp (virgin and recycled) fibers in Section 5. This indicates that the larger fiber surface will release a large amount of heat when immersed in liquid water. More importantly, the lower heat of wetting for recycled fibers was induced from lower accessible surface area of the recycled fibers.  6.3 Surface analysis of recycled fibers An analysis of the water vapor adsorption data indicated that the BET theory could be applied in the range of 5  -  30% relative humidity. The monolayer capacities of  0.089 and 0.055 (g of water)/(g of the fibers) obtained in this manner corresponds to a g for virgin (i.e., unbeaten and beaten) and recycled m / surface area of 384 and 238 2 fibers respectively (See also Fig. 17 and Table V). The results indicate that there is no difference 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 to water vapor. The reduction of the accessible surface area in the internal surface suggests that some of the hydrophilic surfaces (i.e., lamellae) were lost. Thus, the reduction in fiber bonding 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, the reduced (almost 36%) internal surface area of recycled fibers observed in the present work indicates that the number of hydrophilic sites is reduced by re-joining them  100  fiber irreversibly (i.e., thickening) during drying. Thus, the drying process produces ae r artifacts with fewer and consequently, coarser (thicker) lamellae. The coarse lamell e less (i.e., reduced fiber surface area) adsorb less water vapor and consequently, provid the main resistance to severe mechanical treatment. This should be recognized as difference between recycled and unbeaten (never-dried) fibers. two Generally, swelling during water uptake depends upon the relative effect of apart the opposing sets of forces. The adsorption of the swelling agent tends to push s the lamellae by overcoming interplaner attractions. The rigidity of the lamellae oppose d fibers swelling forces. Thus, the rigidity resulting from re-joining of lamellae in recycle ility of the neutralizes the forces exerted by swelling agents and consequently, the swellib fiber wall is restricted.  On exposure to the beating forces, considerable damage is  , before the undoubtedly done to such fibers as a result of their relatively high rigidity (unbeaten) beneficial results of beating can occur. On the other hand, the never-dried fibers, could fibers due to the higher water content, when compared with recycled of the respond much better during beating. During beating, a reduction in the rigidity ization of lamellae due to mechanical fatigue eases swelling and brings about the plastic deposition the fiber wall by the sorbed water. Closure and interweaving of fibers during number of on the forming wire and subsequent wet pressing can produce a greater xibility is bonding points on drying. These observations suggest that fiber wet-fle considerably lost due to re-joining of the lamellae during drying. ed by A reduction in the external fibrillation of recycled fibers was also observ g is due scanning electron microscopy (SEM). Thus, the reduced surface area of bondin Figs. 1 6A- 1 6D). to the reduction of fibrils in the outer surface of the fibers (See also /g 2 /g and 27 m 2 This is also confirmed by estimating the external surface area (i.e., 55 m te method of for virgin and recycled fibers, respectively) as suggested by the absolu Harkins and Jura (1944) (See also Table V).  101  n Emerton (1957), reported that up to 98% of the surface of moist, swolle ed during solventcellulose was obliterated by direct drying, but up to 75% was preserv in drying from exchange with ethanol. In fact, only 2% of the surface remained free in the free space water. Bull (1944) in working with egg albumin attributed the reduction The authors between protein to the crystallization of the protein when dried from water. otherwise the suggested that the protein molecules can not occur in random arrangement 6%. The free space, not occupied by the protein, would be very much greater than did not occur as authors concluded that the new arrangement of dried protein molecules copic or one crystal but that the solid mass of protein is made up of a great many micros l surface area of submicroscopic crystals. Scholz and Flath (1991) measured the interna ed that the s different cellulosic fibers by the iodine sorption method. The author observ to iodine. It was manufacturing process and pre-treatment alter the sorptivity of cellulose l surfaces of the also observed that the drying from water resulted in the loss of interna of the internal cellulose. On the other hand, the drying from methanol preserved most ,  surface area as measured by iodine adsorption.  6.4 Thermodynamics of Fiber-Water Interactions  6.4.1 Free-energy virgin Comparison of the change in integral and differential free energy of the two systems and recycled fiber demonstrates an appreciable difference between these in the range of (Figs. 25, 23 and Table VIll-IX). The free energy of virgin fibers is larger could release more 8 to 80 % relative vapor pressure. This suggests that virgin fiber energies require energy for work. However, the quantitative analysis of these It is suggested that information on fiber surface area, enthalpy and entropy of the system.  102  the free energy change is the function of the accessible sorption sites or the fiber surface area (Stamm, 1957).  Therefore, the observed differences in free energy could be  attributed to the higher surface area of virgin fibers (See also Table Vill-IX).  .4 Moles of water per 100 g of solid (ni)  Fig. 30. Comparison of the integral free-energy of virgin and recycled fibers.  The free energy change per unit area was also measured (Table VilI-IX). The result suggested that the free energy change per unit area remained equal during the entire range of adsorption for both recycled and virgin fibers. This data confirms that the  103  high sorptivity of virgin fibers is induced by its more accessible surface area when the fibers are brought into contact with liquid water. In other words, fewer sorption sites are available for recycled than virgin fibers.  6.4.2 Differential enthalpy and hydrogen bonding The differential enthalpy could be used to deduce the value of hydrogen bonding energy between the water molecules and fibers. The vertical dashed lines in Fig. 26 represents the monolayer capacities as estimated from the BET equation. At this point z\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 other words, these values represent the bond energy in excess of the normal hydrogen bonds in water. A value of H=O indicates that the bonds formed during vapor adsorption have energy equivalent to the bond strength in liquid water. The heat of condensation of water, or heat of liquefaction of water is approximately -9 kcalJmole, thus, each mole of hydrogen bonds in water has an energy of approximately -4.5 kcal/mole. Thus, the total hydrogen bond energy between water and cellulose becomes -5.72 kcallmole for the virgin and -5.79 kcal/mole for the recycled fibers (See also Table V).  These values, are  obtained calorimeirically and are reasonably comparable with the heat of adsorption energy estimated from the BET theory (namely -5.49 for virgin and -5.52 for recycled fibers). 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/mole for virgin and recycled fibers, respectively. The differences in the strength of hydrogen bonding between the virgin and recycled fibers is not significant (See the standard deviation (SD) in Table V). Thus, the differences on the enthalpy of the two systems can not be explained by the strength of the adsorption bonds. The lower enthalpies of recycled fibers when compared with virgin  104  fibers (Fig. 26) indicate a higher accessible surface area in virgin fibers for water adsorption, in line with previous observations.  6.4.3 Entropy A decrease in entropy suggests ordering of the system and an increase in the entropy of the system represents disordering. However, the entropy of a closed system either remains constant or increases but it never decreases. For clarity in this context the increase in -AS in Fig. 27 (i.e., decrease in entropy) represents ordering and the reverse indicates disordering. At early stages of adsorption the -AS in Fig. 27 is high. This is a direct result of the change of three-dimensional molecular motion to two-dimensional motion of molecules.  This is expected because according to Eqs. [2.10] and [2.11] when the  relative 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 demonstrate ordering (increase in -AS). The ordering for recycled fibers occurs at n=.3 1 and for virgin fibers at almost n:=0.50 (See also Fig. 27). The ordering in this region is atthbuted to a decrease in randomness as water is adsorbed on the fiber surface or even on some orientation of water molecules on the fiber surface. The maximum entropies for the virgin and recycled puips suggest domination of the swelling effect during water uptake (see also Fig. 27). Furthermore, Fig. 27 suggests higher swelling potential for virgin fibers. The broad range in completion of the monolayer can be attributed to the effect of higher 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 the larger ones. Then the vapor pressure gradually increases with the greater “capillary”  105  I  x  w  .4 Moles  of water per 100 g of solid (nl)  Fig. 31. Comparison of the integral entropy of virgin and recycled fibers.  cross section of the surface of water. There is then a gradual transition from water filling small capillary spaces to filling larger ones between the lamellae and eventually filling spaces between fibers.  The earlier decrease of -z\S in recycled fibers (Fig. 27) as  compared with that for virgin fibers, suggests smaller pores for the recycled fibers. A sharp  increase  in the peak also supports the belief that the pores of recycled fibers should  be smaller, where the condensed water should demonstrate higher organization than the water that is free to move. The -AS continues to drop (disordering effect) after the completion of the monolayer.  Probably, this is because of opposing effects of molecular separation,  swelling of the fibers and diminishing of the forces holding the higher layers of the  106  trend adsorbed film on the surfaces of the fibers during moisture uptake. It seems the continues up to the fiber saturation point. The comparison of the integral entropy of virgin and recycled pulp systems (Fig. entire 31) clearly demonstrates that the overall entropy of recycled fibers is higher for the ergy of the adsorption range. This further suggests higher swelling potential and free-en y of a virgin fibers when it is compared with recycled fibers. The changes in the entrop More system is caused by chnges in the structure of the fiber rather than the water. ement on expicitly, it is difficult to conceive of the water alone having a different arrang lves. In the surface of the fibers without referring to the differences of the fibers themse ate than fact, it is more usual that the adsorbent determines the arrangement of the adsorb are induced the opposite. Thus, probably the observed ordering effects on recycled fibers This view was by the changes in the structure of the fibers rather than the water. kers (1949), supported by Barkas (1942), Morisson et al. (1959), Hermans and his co-wor of hysteresis. Collins (1930) and Chakravarty (1958) on the explanation of the cause of water vapor. Barkas (1942) suggested plastic deformation of a gel on the desorption than on the Collins (1930) observed that the volume on the desorption side was greater when the crossadsorption side. The same results were reported by Chakravarty (1958) on adsorption sectional area of single jute fibers were compared on desorption with those in entropy start after at different relative humidities. More importantly, the differences n virgin and 25% relative humidity is reached. This indicates that the difference betwee than one of any recycled fibers is largely an interfibrillar microscopic phenomenon rather single microfibril on the fiber surface. zation Some of the workers explained these observations on the basis of the organi (1988) observed of water alone (Henry et al. 1988, Milichovsky 1990). Henry et al. ped. The authors during the entire process of beating fibers, no new surfaces are develo les along the interpreted their obsevation based on better structuring of the water molecu  107  fiber surfaces during beating. Thus, it is the water rather than fiber which undergone structural change during beating.  6.5 Changes in Fiber Wall Ultrastructure due to Recycling Comparison of XRD traces of unbeaten and beaten puips suggests that refining basically does not alter the crystal structure of the cellulose fibers (Fig. 23). The physical effects of refining are merely restricted to the amorphous phase of the cellulose. Not only the crystallinity, but also the full width at half maximum (FWHM) of the intensity peaks 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 virgin and recycled fibers. It was observed that the amorphous fraction decreased by almost 22% after the first cycle (Table VII). Interestingly, no differences between the sorption ratios of beaten and unbeaten puips were observed, to show that the amorphous component of the fibers had not been altered during refining (Table VII). Therefore, the decrease in the integral breadths, along with a decrease of the amorphous fraction, indicates that, probably, the observed increase in crystallinity is induced by partial crystallization of the amorphous phase of the lamellae surfaces during recycling. These observations lead to the belief that the changes in superstructure of the cellulose, at least partially, must be responsible for the origin of poor quality (hornification) of recycled fibers. These resuks are also supported by the measurements of the peak breadths. The decrease observed for the peak breadths (101 peak width) of recycled papers suggests that the crystallite size increased due to recycling. This finding is very important, since the 101 plane is known to represent the plane of lamination (Frey-Wyssling, 1954) in the fiber wall and shows conclusively the “growth’ of crystallites in recycled pulp fibers. In  108  ally along the 101 plane during drying. fact, hydrogen bonding takes place preferenti idea that irreversibility of hornification is Thus, this evidence also supports the proposed omenon brought about by irreversible joining probably related to the crystallization phen of lamellae parallel to 101 planes. wall crystallization, due to drying, In connection with the problem of cell on et al. (1993), Atalla (1992), Ingram et contradictory results have been reported. Mart iderable increase in crystallinity of the al. (1974) and Heyn (1965) reported a cons other hand, Bouchard and Deuk (1994) and cellulosic fibers due to drying. On the allinity index. Morossof (1974) observed no increase in cryst any considerable change in the Bouchard and Deuk (1994) did not observe ors concluded that the observed minor crystallinity index during recycling. The auth recycling is explainable by yield effect. The increase in crystallinity as a consequence of was uncertain due to incremental loss, if precision of the technique of yield correction of the fibers due to recycling. So, an there was any, in the non-crystalline portion the problem. alternative technique was followed to evaluate urements of crystallinity, a virgin In order to evaluate the “yield effect” on meas re-slushed up to 10 times, without drying paper was made into handsheet, pressed and s ess. Indeed, the paper is recycled for 10 time during the repeated handsheet-making proc times re-slushing and pressing, the handwithout experiencing the drying stage. After 10 according to CPPA standard. It is assumed sheet is air-dried in the conditioning room the yield effect, after 10 times recycling, the that, if the gain in crystallinity is a result of ever, the crystallinity did not increase as crystallinity index must be amplified. How that the crystallinity increase cannot be expected (Table VII). Thus, it is concluded explained by yield effect. the increase in crystallinity occurred Marton et al. (1993) observed that, most of More importantly, they observed that, re in the first cycle (5.5% after first re-wetting). also ity value to the never-dried state, but slushing not only does not restore the crystallin  109  that subsequent drying steps incrementally promote the increase in crystallinity of the fiber wall. Atalla (1992) reported that the recycling of fibers caused an increase in crystallinity, as well as a decrease in surface area. He observed that the maximum effect of recycling was observed after the first recycle, when most of the change in cellulose structure seemed to have occurred. It was also confirmed that the crystallinity was not influenced 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 widthat-half-height of the 002 reflection in the x-ray diffractogram of the pulp, did increase steadily with each progressive recycling of the pulp. The author, in light of his proposed structural model, concluded that repulping resulted in some slight enhancement of the molecular mobility of the disordered cellulose, thus allowing it to enter more readily into the domains of more coherent order at the superstructural level during the subsequent drying 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 the first-drying stage of cotton from the never-dried condition was most distinct and resulted in a greater degree of re-crystallization of never-dried cotton than obtained on subsequent drying. Heyn (1965) observed a dramatic increase in crystallinity of the freeze-dried fibers (i.e., 11% increase in crystallinity index) when re-wetted and subsequently dried overnight. 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 x ray patterns of dried and stretched keratin were identical and interpreted this by the ordering effect of drying. It could be concluded that, the re-organization of the cell wall is not confined to consolidating the fiber wall by mere zipping up the external and internal surfaces, but also 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 reorganization 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 relaminated 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 internalfibrillationand 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 recycledfibersindicate 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 thefibersthrough 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 "ligninfree" 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  115  during drying. As the water is removed, the remaining water imposes cohesive forces that draw the celluloses surfaces toward each other. These forces are due not only to the atmospheric pressures forcing the structure together as water is removed but is mainly due to a much greater extent to an internal tension in the water arising from molecular forces. With still further evaporation of water, the water layers between the cellulose particles become very thin and the reduction in vapor pressure becomes appreciable. The internal tension begins to increase and an intense force is created by which the cellulose surfaces are drawn together. This force acts between the lamellae/fibrils in the fibers for the most part and causes individual fibrils or lamellae to collapse. The force due to such internal tension is of the order of hundreds of atmospheres for water. For instance, it is 936 kgm/cm 2 for a relative vapor pressure of 50 per cent and at a relative vapor pressure of 10% it rises to about 3110 kgm/cm 2 (See also Appendix E). These forces, thus pull the hydroxyl rich planes (101 planes) close enough to make direct binding of the polysaccharide (cellulose) chains possible. The rigidity of each lamella is important in collapsing the fiber wall.  If the  capillary wall is sufficiently rigid to withstand collapsing by the internal tension forces, the water will eventually evaporate completely.  Internal tension forces are not  appreciable over large distances (See also Appendix E).  On the other hand, finer  lamellae in the cell wall will deform under tension forces. If this happens, the distance between the lamellae will decrease, on further evaporation the internal tension will be enhanced, due to the decrease in the distance between the lamellae. On further decrease in the gap between the lamellae, the tension forces will increase more and more, until eventually complete lamellar collapse will occur. In this process the finer fibrils/lamellae will be brought sufficiently close to make direct hydrogen bonding possible. Unlike the coarser lamellae, the chance for extensive hydrogen bonding for finer lamellae is higher and possibly, it may also lead to crystallization of the fiber wall occasionally. To be more specific, on drying the microfibrils (i.e., the 101 planes in adjacent coaxial  116  lamellae) will be joined together. The surface of the lamellae, if deficient in lignin, will come close enough for bonding of the polysaccharide chains to occur (Fig. 32C). If the surface of the lamellae are also hemicellulose poor or if the hemicelluloses are locally aligned parallel to the crystalline cellulose chains (See also Fig. 32C), bonding may in some cases be sufficiently extensive and regular as virtually to unite two distinct crystalline regions as one. This is particularly likely in the case of kraft fibers with originally low hemicellulose content. The presence of lignin-free zones on the lamellae surfaces is evidenced by enzymatic analysis in this investigation. Upon re-wetting, a proportion which was co-crystallized on drying is irreversible. In fact, none of the bonds so formed can be disrupted when the fiber is subsequently immersed in water. This mechanism results in a partial loss in the ability of the fiber to take up water, an effect to which the term homification has been applied.  On the other hand, the presence of  residual 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 fibers could be explained. Generally, the beating process reduces the rigidity of the fiber wall by mechanical fatigue and plasticization of the fiber wall by adsorbing water. Concomitantly the fiber flexibility increases. The striking fact is that, when the surface area of the lamellae was reduced to half, the flexibility of the fiber wall decreased 75%. This raises the point that a reduction of about 36% of accessible surface area of the fiber to water is sufficient to reduce 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 Scallan 1993), on loss in swelling potential of recycled fibers, are consistent with these observations. lt is also reported that beating and recycling processes do not change the chemical constituents of the fibers (Curran et al. 1931, Bouchard and Douek, 1993).  117  Therefore, due to the short exposure of recycled fibers in water, during the beating process, the possibility of losing the surface bonding agent, merely by recycling is hardly probable.  More importantly, the role of internal fibrillation is dominant in the  development of paper strength properties. Thus, it could be concluded that the reduction in the number of thinner wall segments by irreversibly re-joining them, results in restricted swelling. Restricted swelling, or loss in wet-flexibility of the fibers, results in poor confirmability of the fibers in the paper furnish, and paper strength properties are reduced. Emphasizing the fact that, as the internal fibrillation is a central feature of the beating effect, loss of internal fibrillation is the main cause of inferior properties of recycled fibers. This implies that the main loss in strength properties of recycled fibers induced by hornification rather than surface deactivation.  However, this kind of  reasoning does not mean that the loss in surface bonding agents is impossible. The loss of potential surface bonding during the recycling, is secondary in importance to the effect of internal fibrillation as the cause of vanishing paper strength properties on recycling.  118  6.8 Proposed Model for the Evolution of Fiber Ultrastructure on Recycling Process  These observations, with small modifications, fit with the proposed model of Scallan (1974) on the ultrastructural arrangement of the cell wall in wood.  During  chemical pulping incrusting substances are removed from between the coaxial cellulose layers. The consequence of such change is a multilayer structure such as shown in Fig. 32A.  Indeed, water replaces fully the spaces previously occupied by encrusting  substances.  Unbeaten (never—dried) ( Lignin HemicelluloSeS Cellulose  - -  —  Recycled (re-wetted after drying)  Dried  Fig. 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).  119  The beating in never-dried chemical puips, merely, increases the interstices between the lamellae (Fig. 32B). It seems that the beating does not “delaminate” the lamellae, it only expands the lamellar interstices which were filled with water after the removal of cell wall encrustants. Lack of any evidence for influence of refining at the molecular level as evidenced by XRD analysis (i.e., lack of changes in the supermolecular structure of cellulose) and overlapping of the sorption isotherms of unbeaten and beaten fibers support this point that, no laminations are created on account of refining. In other words, the pre-existing fibrils/lamellae are merely pushed apart in the course of refining (Fig. 32B). These observations suggest that the mechanical action of 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 lamellae and consequently eases swelling of the fiber by polar liquids (e.g., water). Conversely, on drying due to the development of tension forces the plasticized fibrils/lamellae are drawn together in molecular contact and fixed by extensive hydrogen bonding (Fig. 32C). If the conditions for fibrilar/lamellar orientation are such that they fulfill the requirements for co-crystalization, the cell wall cellulose will be further crystallized. 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 met the co-crystallization condition or the extensive hydrogen bonding, will be re-opened again.  However, extensive beating will not re-open the initial spacing between  fibrils/lamellae which were drawn together by extensive hydrogen bonding or co crystallization. Subsequent recycling will lead to the generation of more fines due to diminished swellibility and reduced flexibility of the fibers. This model stipulates that the refining process does not produce any new surface area by partitioning the cell wall, but merely expands the interstices between the lamellae which were pre-occupied by water, thus swelling the fibers as suggested in Fig. 21B. On  120  the other hand, recycling reduces the specific surface area due to irreversible joining of the fibrills/lamellae during drying (Fig. 32D), thus decreasing the flexibility of the fibers.  121  7. SUMMARY AND CONCLUSIONS  The 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 summary below. From these results and statements, information of immediate practical importance can be drawn. For the latter case, the correlation and interpretation of the single results must be regraded as tentative.  The theory of mechanism which has been suggested  should be regarded as simulation.  7.1. SUMMARY  7.1.1 Beating 1. The beating process does not influence the ultrastructural organization of the fiber wall. 2. The desorption isotherms of never-dried unbeaten, and beaten to 6000 and 12000 revolutions, overlapped. 3. The adsorption isotherms of unbeaten and beaten fibers overlapped in their entire range of sorption. 4. The identical shape of sorption isotherms for the unbeaten and beaten fibers, did not change with the method of drying.  122  7.1.2 Recycling 1. Drying/recycling cause re-organization of the fiber wall towards higher order crystallization. The main increase is in the cell wall ordering which occurs after the first cycle. The x-ray diffraction and sorption ratio gave consistent results with each other. Subsequent re-wetting does not restore the original organization of the cell wall. 2. Each stage of recycling after the first cycle incrementally increases the crystallinity of the fiber. 3. The integral breadth of the 101 plane of fibers decreases by recycling. The average crystallite size of the fibers is inversely proportional to the size of integral breadths of the diffraction peaks. 4. The total surface area of the fibers remained constant during the beating process, but the external surface area increased by beating. 5. The total surface area of the fibers decreased up to 36% due to drying after the first cycle. 6. Drying and recycling decreased the sorption potential of the fibers. 7. Minute differences in the condition of drying were reflected in distinct differences in adsorption isotherms of the samples. 8. The fibers dried by the method of critical point drying sorbed more water than those dried by the freeze-drying method. This is observed when the sorption isotherms of the fibers dried by those methods were compared. 9. The heat of wetting of both virgin and recycled fibers decreased linearly by the increase 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 recycled fibers. 11. The free energy of recycled fibers is less than that of virgin fibers over almost the full range of the adsorption isotherm.  123  12. The surface-energy per unit area remained constant over the whole range of the adsorption isotherm. 13. The hydrogen bond strength of recycled fibers was equal to that of virgin fibers within the limit of experimental precision. 14. The integral and differential enthalpies of virgin fibers were higher than those of the recycled 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 recycled fibers 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 recycled fibers as observed by comparing the entropy of the two systems of virgin and recycled fibers. 19. Beating increased the enzymatic hydrolysis yield by about 10% after 4 h incubation time, while recycling reduced it by almost 7% during the first cycle. 20. Successive recycling exacerbates the surface accessibility of recycled fibers, for instance 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 than that of unbeaten and recycled hydrolyzed fibers. 22. Less fragmentation of beaten fibers in comparison with beaten and recycled fibers during enzymatic hydrolysis was also observed by SEM microphotographs. 23. The DP of beaten hydrolyzed fibers was almost half of that of unbeaten and recycled hydrolyzed fibers as measured by the GPC technique. 24. The crystallinity index of recycled and beaten fibers increased, whereas that of beaten fibers remained constant during enzymatic hydrolysis.  124  7.2 Conclusions The irreversible closure of the lamellae in the fiber wall either by extensive hydrogen bonding or co-crystallization is responsible for the fiber homification during  drying. The recycling effect on low-yield chemical puips could be explained on this basis. On the other hand, refining does not produce new lamellae in the never-dried unbeaten 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 not affect the sorption potential of the fibers while an incremental difference in the condition of drying altered the sorption potential of the fibers. Indeed, all sorption and swelling of the fibers occurred at the surface of the lamellae.  Thus, reduction of the sorption  potential 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 by recycling. Approximately a 36% reduction in the internal surface area of the fibers and higher heat of wetting of the virgin fibers support these observations. Free energy and free surface energy of the fibers which are calculated by thermodynamic approaches, in the present investigation, indicate that the accessible surface area of the fibers were reduced. In addition, entropy data confirmed the ordering effect of recycling in the fiber wall. Comparison of x-ray peak intensities of virgin and recycled puips suggested that recycling reorganized the fiber wall towards a higher degree of crystallization. More importantly, the reduction in the integral width of recycled fibers suggested reuniting of 101 planes during recycling.  This observation is in close agreement with the  observations of adsorption and immersion processes and thermodynamics of fiber-water interactions. It is possible that, if the surfaces of the lamellae are lignin-free, they will come close enough to unite the two crystalline cores as one.  These united surfaces  cannot be reopened again by reslushing due to development of the high cohesion forces between them. The examination of the fibers by cellulose enzymes suggests the possible presence of lignin-free zones in the lamellae surfaces.  The observation could also  125  explain the poor performance of recycled fibers in the refming process (i.e., brittleness of the fiber in refining zone). The unbeaten and beaten (never-dried) puips behaved in a similar fashion during the adsorption. Furthermore, beating did not show any effect on the fine structure of the cell wall. These observations indicates that the fatigue phenomenon is not necessarily related to the fibrillation or sorption potential of the fibers. More importantly, the major difference between unbeaten and beaten fibers can be explained on the basis of the fiber accessible surface area to water uptake or on the number of lamellae. These observations could also explain the cause of different behaviors between mechanical and low-yield chemical puips during recycling. In mechanical pulps, the chance for irreversible closure of the lamellae is very slim due to the extensive presence of the lignin sheath on the lamellae surfaces. The drying of never-dried fibers takes place in the following manner. As water is removed from in between two adjacent finer lamellae (i.e., collapsible under water tension force) they both continuously draw toward each other until all moisture is removed and the fiber is dried. At this stage, the lamellae are very close together and form cohesive forces to activate and to reunite the lamellae with extensive hydrogen bonding. On the other hand, in the case of the coarser lamellae, if they are rigid enough to withstand the water surface tension force before the drying is completed, the condition for close packing is lost.  Eventually when the fibers are rewetted after drying, the  adsorbing water breaks the bonds and substitutes them with the same tension forces as before so that the structure gradually expands. Some of these bonds remain intact due to co-crystallization or extensive hydrogen bonding, thus the total accessible surface area of the fibers is decreased. To summarize, the tension forces created by water loss draw the fibrils/lamellae close enough to make hydrogen bonding possible. If the sheets of microfibrils are also hemicellulose poor or if the hemicelluloses are locally aligned parallel to the crystalline  126  cellulose chains, this bonding may, in some cases be sufficiently extensive and regular to cause the crystalline regions to unite as one. The rejoined planes thus formed cannot be severed when the fiber is subsequently immersed in water. This results in a partial loss of the cell wall ability to take up water, an effect to which the term homification has been applied.  This is particularly likely in the case of bleached lcraft puips with low  hemicellulose content. However, the role of hemicellulose in co-crystallization of the cell wall remains speculative and requires further work.  127  8. RECOMMENDATIONS  8.1 Identification of the Parameters Involved in Hornification During the Bleaching Process A single experiment in this laboratory showed a distinct difference in strength properties between unbleached and bleached unbeaten puips during recycling.  The  unbleached (unbeaten) puips showed poorer potential for recycling than the bleached (unbeaten) puips. Fully-bleached pulps usually contain much fewer acidic groups than unbleached pulps (Lindström 1986). Probably, acidity plays a role in the inferior quality of unbleached (unbeaten) recycled fibers. finalized yet.  However, this investigation has not been  The outcome would be interesting in regard of identification of the  parameters involved in causing the difference. The process could be followed as follows. The changes in chemical composition and strength properties of the puips in recycling could be determined for each sequence of bleaching. The results of each stage could be matched with unbleached pulp for changes in chemical composition, acidity and strength properties. This could identify the sequence in which the most deteriorative substance is removed and thus the cause of such differences could be identified. This fmding would help to track down the deteriorative substance (such as acidic groups) in the pulp and remove during papermaking. Thus, to increase the potential life of the fibers during repeated recycling.  8.2 RoLe of Hemicelluloses in Hornification of the Cell Wall The second problem which requires to be determined is the role of hemicelluloses in development of the hornification phenomenon during recycling.  The result of  recommendation A in identification of the deteriorative substance in recycling could lead to observations with respect to the differences in the hemicellulose content of the puips.  128  However, this could also be followed independently. It could be followed either on kraft pulp which has low hemicellulose content or with an organosolv pulp bearing a higher (up to 25%) hemicellulose content. The investigation could be conducted by comparing the strength properties of the paper made from alpha-cellulose pulp with those containing higher amount of hemicellulose.  The selection of the agent for removal of  hemicelluloses is important, because some agents such as caustic soda also modify the structure of the cellulose, which confounds the final conclusion. It is not yet certain that the hemicelluloses enhance homification in recycling or vice versa. Thus, the results could determine the role of hemicelluloses in recycling of fibers and consequently, the selection of appropriate agents to overcome hornification during recycling.  8.3 The Role of Temperature in Recycling It was reported that (Nazhad and Paszner, 1994) heat-drying (i.e., drying the fibers after pressing stage by heating at temperature around 900 C) affected the properties of the fibers differently, when compared with those effects noted for air-dried fibers. With heat-dried puips the strength properties increased whereas the optical properties remained the same with a slight increase. This observation posed a question on the possible development of the paper strength quality by adjusting the drying process. We asked ourselves, is there any maximum strength development which could be reached with a proper combination of temperature and moisture in the fiber during recycling? Is there any optimum combination of temperature and moisture which could minimize the deteriorative effect of recycling? The plan for the work could be set to identify these effects 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 optimum  129  condition for development of physical and optical properties and recycle potential of the fibers by suitable combination of moisture and temperature could be addressed.  8.4 Implications of These Findings on Upgrading of Recycled Fibers To reverse the recycling effects on fiber properties to the original state (i.e., virgin fiber state) requires reopening of the lamellae to increase the water uptake to the never. An dried (unbeaten) state. This could be only achieved by strong swelling agents basic optimum combination of temperature, time and concentration of swelling agent is a requirement. The second requirement for upgrading would be the condition of the fiber itself. Fibers low in hemicellulose content require swelling agents with lower affinity cases (solubilizing power) for the hemicelluloses. The loss of hemicelluloses in certain of aie harmful for strength properties. Thus, attention should be paid to prevent loss ete re hemicelluloses during the treatment. Swelling is the first condition for compl opening of the lamellae. In the second stage of reconditioning swollen fibers the rigidity of the fiber wall refining. can be reduced by imposing mechanical fatigue. This could be achieved by for The feasibility of this procedure and the economic assessment of the costs implementation, need to be addressed.  130  REFERENCES Argue, G.H. and Maass, 0., Measurement of the heats of wetting of cellulose and wood pulp. Can. J. Res., 12: 564-574, (1935). Assaf, A.G., Haas, R.H. and Purves, C.B., A new interpretation of the cellulose-water adsorption isotherm and data concerning the effect of swelling and drying on the colloidal 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 native cellulose. 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 fiber building 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 linear polymers- The effect of N-methylation on the x-ray structures and properties of linear polyarnides. 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 of steam explosion and conventional chemi-mechanical pulping processes. Pulp and Paper Can. 9 (12): 142-151, (1990). Barkas, W.W., Wood water relationships-VU. Swelling pressure and sorption hysteresis in gels. Trans. Faraday Soc., 38: 194-209, (1942). Bouchard, 3. and Deuk, M., The effect of recycling on the chemical properties of pulp s. 3. Pulp Pap. Sci. 20(5): 3131-3136, (1994). Bovin, R.H., Harder, N. and Teder, A., Changes in pul quality due to repeated papermaldng. Paper Technol. 15(10): 261-264, (1973). Brickman, W.J., Dunford, H.B., Tory, E.M., Morrison, J.L. and Brown, R.K., The reactivity of cellulose. II. Water sorption, heats of wetting, and the reactions with thallous ethylate in ether, nitration mixture, and heavy water of cotton linters alternatively wetted with 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).  131  Bull, 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 water washing on the long-term stability of cellulosic fibers. ACS, Historic Textile and Paper Materials, 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 the Interior Canada, Forest Service, Bulletin 84, pp. 36-44, (1933). Caulfield, D.F. and Moore, W.E., Effect of varying crystallinity of cellulose on enzymatic 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 relative humidity. Textile Res. 3., 28: 878-880, (1958). Chan, M., Breuil, C., Schwald, W. and Saddler, J.N., Comparison of methods for qualifying the hydrolytic potential of cellulase enzymes. Appi. Microbiol. Biotechnol., 31: 413-418, (1989). Christiansen, A. W., How overdrying wood reduces its bonding to phenol-formaldehyde adhesives: A practical review of the literature. Part I. Physical responses. Wood and Fiber Sci. 22(4): 441-459, (1990). Chung, F.H., Industrial applications of x-ray diffraction. American laboratory, 2: 144156, (1989). Clark, J., d’A, “Pulp Technology and Treatment for Paper” Second Ed., Miller Freeman Publications, 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 relative humidities. 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 chemical and physical properties of puips. J. md. and Eng. Chem., 23: 104-108, (1931).  132  Davis, W. H. and Thompson, W. S., Influence of thermal treatment of short duration on the 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 water vapor 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 in Papermaking. London, Vol. II: pp. 835-855, (1978). Ehrnrooth, E., Htun, M. and de Ruvo, A., Trans. BPBIF Symp. Fiber-Water Interactions in Paperma.king. London, Vol. II: pp. 899-915, (1978). Emerton, H.W., Fundamentals of the Beating Process. Marshal Press, London, pp. 84137, (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 ionsolvent interaction in aqueous solutions: A suggested picture of water structure. Faraday Soc. 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. Wood Sci. 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 of cellulose I and II, Acta Chem. Scand., 14(3): 683-688, (1960) Gottsching, L., Waste paper: a raw material of increasing significance. Paperi Ja Puu Paper and Timber. 4(3): 209, (1992).  133  Gray, 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 pretreatment on 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 water vapor 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 bonds with heat-treated wood., Ph. D. Thesis, University of British Columbia, Vancouver, 176, (1964). Canada, p. Harkins, W.D. and Jura, G., Surfaces of solids. XII. An absolute method for the determination 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 formation on the structure of amourphous cellulose. Appl. Polym. Symp. 28: pp. 743-752, (1976). Hatakeyama, T., Nakamura, K., Tajima, and Hatakeyama, H. International Paper Physics Conference 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 fibril structure. J.Appl.Polym. Sci. 7: 1175-1192, (1963). Henry, F., Brandt, A. Ch. and Noe, P. Untersuchung der Wechselwirkung von Wasser 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 crystallinity of cellulose fibers. A background analysis. J. Appl. phys. 19 (5): 491-506, (1948). Hess, K., Mahl, H. ‘and Gutter, E. Electronenmikroskopische Darstellung grosser Langsperioden 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 from unopened boils as studied by x-ray diffraction. 3. Poly. Sci. Part A, 3: 125 1-1265, (1965).  134  Higgins, 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 in polyamide and polyester fibers. Polymer, 19: 27-32, (1978). Hollenbeck, R.G., Peck, G.E. and Kildsig, D.O., Application of immersional calorimetry to 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. Pulp and 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, Part I: 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 Finished Product, 152-160 (1991). Jura, G. and Harkins, W., Surfaces of solids. XIV. A unitary thermodynamic theory of the 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 in Protein 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).  135  Koning, J.W., and Godshall, W.D., Repeated recycling of corrugated containers and its effect on strength properties. Tappi, 58 (9): 203-216, (1975). Kossler, I., Danhelka, M. and Netopilik, M. The carbanilate method for the determination of 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., The crystallinity 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 lines using 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 during drying 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 and properties. 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 of swelling and strength of recycled fibers. Svensk Papperstidn. 11: 355-357, (1978). Maass, 0. and Campbell, W.B., Studies in cellulose moisture phenomena. Pulp and Paper Mag. Can. 28: 108-114, (1939). Marton, R., Brown, A., Granzow, S. Keoppicus, R., and Tomlinson, S., Recycling and fiber 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. Trade J. 155: 33-40, (1971). Meller, A., Paper Trade J., The cold alkali purification of wood ceillulose II: The influence of surface properties on resistant pentosans. 125 (11): 57-60, (1947).  136  Mellon, E.F. Kom, A.H. and Hoover, S.R., Water absorption of proteins. IV. Effect of physical structure, 3. Am. Chem. Soc., 24: 2761-2764, (1949). Meyer, K.H. and Misch, L. Positions des Atomes dans le Nouveau Model Spatial de la Cellulose. 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 system cellulose-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, 197201(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 the enzymatic 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 of I: papermaking, Trans. 9th Symp. Edited by Baker, C.F. and Punton, V.W., London, Vol. 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 and the product quality in the recycled paper industry. Tappi, 72: 198-201, (1989).  137  Putz, H., Torok, I. and Gottsching, L., Making high quality board from low quality waste paper. Paper Technology, 30(6): 14-20, (1989). Pycra.ft, C.J.H. and Howarth, P. A method of increasing enzyme degradation of cellulose fibers. Paper Tech. hid. 21(9): 283-285, (1980). Ramos, L.P., Nazhad, M.M. and Saddler, J.N., Effect of enzymatic hydrolysis on the morphology 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 dry ball 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): 124127 (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 anisotropic inter-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. Gel permeation chromatographic analysis. Tappi, 62 (8): 103-105, (1979). Scolz, M. and Flath, D., Determining structure of cellulose fiber substances with the aid of iodine sorption. Textilveredlung, 26(6): 188-191, (1991). Schwald, W., Chan, M., Breuil, C. and Saddler, J.N. Comparison of HPLC and colorimetric 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 for estimating 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 change per 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 and cellulose. Wood Sci. Tech. 3: 301-323, (1969).  138  Stockman, 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 wood pulp 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. Pulp and 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 recycled pulps fractions. Cell. Chem. Technol. 10: 737-749, (1976). Tam Doo and Kerekes, The effect of beating and low-amplitude flexing on pulp fiber flexibility. JPPS. 15(1): J36-J42, (1989). Thode, E. F., Chase, A. J. and Hu, Y., Dye adsorption on wood pulp. Tappi, 38 (2): 8889, (1955). Tie-Qiang, L., Henriksson, U., Eriksson, J.C. and Odberg, L. Fundamentals of Papermaking. Ed. Baker, C.F. and Punton, V.W., Mech. Eng. Pubi. Inc. London. pp. 3945, (1989). Troughton, G. E. and Chow, S. Z., Migration of fatty acids to white spruce veneer surface during 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 by gel 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 partially saturated viscose rayon and standard cellulose in water. J. Phys. Colloid Chem. 54: 11481160, (1950).  139  Warddrop, 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 Interface Sci. 49(1): 40-47, (1974). Wood, B.F., Conner, A.H. and Hill, C.G., Jr. The effect of precipitation on the molecular weight 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 calcium carbonate fillers. Paper Trade J. 164(22): 33-34 (1980). Yamagishi, Y. and Oye, R. Influence of recycling on wood pulp fibers changes in properties 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 of crystal 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 of paper and textile of historic and artistic value. J.C.Williams (Ed.), Advances in Chemistry Series, Washington DC, 164: pp. 189-205, (1977). Zettlemoyer, A.C., Young, G.J. and Chessick, A. Studies of the surface chemistry of silicate minerals, III. Heats of immersion of bentonite in water. J. Phys. Chem. 59: 962966, (1955). Zimm, B.H., Simplified relation between thermodynamics and molecular distribution functions for a mixture. J.Chem. Phys. 21: 934-935, (1953).  140  APPENDIX A  Peak profile resolution  In order to obtain information on the atomic level such as crystal structure, crystallinity, residual stress, crystallite size, etc by XRD mathematical functions are required to represent the peak profile. It has been experimentally verified that, for the line broadening from imperfections or growth of small crystallites, neither Cauchy nor Gaussian functions are necessarily good approximations (Chung 1989). A better approximation for XRD peaks is obtained by convolution of one or more Cauchy and Gaussian functions. The synthesis of the convolution of Cauchy and Gaussian functions is called the Voigt function and was applied to the analysis of X-ray diffraction by Langford (1978) and used in this paper to calculate the crystallinity and crystallite size. The Cauchy function is given by: x) 1 ( 0  =  {A.1J  nw  [A.2j  where =  and  0 is full width at half maximum (FWHM) and x is f3 is integral breadth, 2w  peak position. The Gaussian function is given by: Iq(X)  =  Texp(7tx2/(g)2)  [A.3]  where  g 13  and  g 3  =  Wg(lt/1fl2)h1’2  is the integral breadth and  [A.4]  is FWHM for the Gaussian function.  141  From the two functions of I (x) and g t Cx) of Cauchy and Gaussian one forms their convolution Io*Ig = I as the function I (x) such that: (00  1(x)  Ic(U)Ig(XU)dU  [A.5J  O0  where  I,  and  g t  are the Cauchy and Gaussian components of the Voigt function. The  equations [A.l], [A.3J and [A.5] have a Fourier transform. An approximate value for the  0  40  Fig. 33. The resolution of diffraction peaks in terms of Voigt profiles for each peak.  integral breadth,  of equation [A.5j is deduced as  142  the integral breadth of the Voigt function() is  f32  I13  +  [A.6]  (3g)2  and the full width at half maximum (FWHM), 2w, of the Voigt function is approximated in the form of  2w  {[()2  +  2]_l/2}/p,• (f3 ) 0  [A.7]  The crystallinity is measured as the ratio of the total area under the resolved peaks to the total scatter under the diffractogram. The apparent crystallite size also could be obtained by using either the integral breadth or FWHIvI. To obtain the actual crystallite size, 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 the Voigt function into a single peaks. It is worth mentioning that a computer program is available commercially to search for better fit by using the Voigt function.  143  APPENDIX B  Derivation of the formula for external surface area measurement  The heat of wetting or diminution of heat content, which accompanies immersion in the liquid, is given by the Gibbs-Helmholtz equation: (I) FL  /E  =  (FL  —  [B. 1 ‘  TdFL/dT)  represents the surface-energy lowering produced by immersion in the liquid. FL =  where  —  [B.2]  sL  ? is the free surface energy of the solid,  L  the free interfacial energy of the  solid-liquid interface, and T is the absolute temperature. According to the first law of the thermodynamics, the energy released will be the same whichever way we proceed. Thus, zH  (s/L)  /E could also be obtained by the  sum 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, defined  by: [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: [B.4]  +?Vcos 8 FL=FV  dFL/dT = dF/dT + dRLV cose)/dT  or  [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 have zMI  (s/L)  /E = (F  —  TdF/dT) +  (?,Lv  cosO  -  Td[? cosO] /dT) [B. 6)  144  where (F,  —  TdF/dT) is the heat of adsorption Ha, and the term in the second  bracket is the heat generated by immersion of an already saturated solid. The heat of immersion of this wet solid is designated by  (sf/L)  /Z where f indicates that the  solid possesses a “film” of adsorbed liquid. Thus, tH (sf/L) /  =  (, cosO Td[) cosO]/dT) —  In the special case where q  =  [B.7J  0, the equation [B.7] would be the total surface energy of  the normal liquid. So AH  (sf/L)  /  =  (2,  —  Td[?] /cIT)  [B.81  The right hand side of equation [B.8] is the surface-energy of the normal liquid. If the liquid is water, thus the surface-energy of water is 118.5 ergs/cm , thus 2 =  AHj(sf,L)/llB.5  [B.9]  The surface area obtained in this manner is an estimate of the external surface  area of the solid material since the capillaries and some intra-aggregate spaces are filled with water.  145  APPENDIX C  Integral and differential free energy changes.  At a given temperature, the free energy changes of a system, LP, associated with the adsorption of a adsorbate on an adsorbent is given as:  [C. 11  AF n 1 + 2 Ar =n4F  2 are the partial molal free energy changes of the adsorbate and where AF 1 and AF  2 are the numbers of moles of each, respectively. 1 and n adsorbent, respectively, and n , is given as the difference 1 The differential free energy change for the vapor, AF  between the chemical potential of the liquid in the adsorbed state and pure liquid.  1 AF  =  1 U  -  =  RT1n(h)  [C.2]  where u 10 are the chemical potentials of the adsorbate and pure liquid 1 and u respectively, h is the relative vapor pressure, T is absolute temperature and R is the gas constant.  The  ‘2  2 AF  =  is obtained by the Gibbs-Duhem relation: [C.31  n d(AF / 1 -(n ) 2 )  An expression for the integral free energy change can be obtained by substitution  of Gibbs-Duhem equation and Eq. [C.2j into Eq. [C. 1]. Thus, the integral free energy change for water vapor accompanying the sorption process is: Ar  =  Tln(h) n R 1  -  RT  [ln(h)]} {n d 1  [C.4]  146  Evaluation of the integral on the right-hand side of Eq. [C.3] can be performed h versus h by replacing d[In(h)] with its derivative n / graphically from a plot of 1 (1/h)d(h)  =  in Eq. [C.41.  Tin(h) n R 1  -  RT  [C.5]  h)d(h) (n / 1  The error associated with low pressure extrapolation to P  =  0 is smaller with this  1 versus in (h). technique than with a plot of n The differential free energy change has units of energy per mole of adsorbate, and the integral free energy change simply has units of energy, because the entire process is specified for a certain number of moles of each component.  I  147  APPENDIX D  Method for determining the integral and differential enthalpy  Figure 23 is an schematic representation of the enthalpy changes associated with the wetting of a solid by a liquid. Step 1  Represents the immersion of a dry clean solid into a liquid. The heat of  =  immersion of this solid is designated by zH Step 4  =  (s/L)  Demonstrates the immersion of a solid covered by a film, and it is represented  by IHj(sf,L)  According to the first law of thermodynamics, the energy released will be independent of the path. Therefore, Step 1 is equal to the sum of the Step 2 to 4. A + /Ha + AHj 1 (six.) = fl  (sf/L)  [D. 1]  where 2. is the molar heat of vaporization of the liquid and AHa is the heat of adsorption. The net integral enthalpy change associated with the adsorption process is the heat change in excess of the normal condensation of the liquid.  =  —  ) 1 (—n  [D.21  Thus, from Eq. [D.1] ( 51 =zH H )  —  /) AH( f 5  [D.3]  It is to be noted that in Eq. [D.31 both of the terms are measurable by the calorimetric method as reported. The integral enthalpy An, thus would be generated by measurement of the heats of immersion of solids with varying moisture contents. This procedure renders the entire  148  range of  iS.H  values. The differential enthalpy change is obtained by estimating the  slope of  IH  . 1 versus n  clean solid in vacuum  +  1  moles water of vapor flj  -4  solid with 1 mole of n adsorbed water  2 1  I I  I L_  I I  I __..J  Fig. 34. Conceptual drawing of a sample of pulp in equilibrium with the vapor of a liquid. To be immersed in the liquid and thus lose the surface energy of the duplex film of liquid.  149  APPENDiX E  Internal tension of liquid between two parallel surfaces  Assume two parallel lamellae with a distance x apart. As water evaporates, the water between the lamellae assume the shape droplets. Thus, the liquid appears as a flattened drop having a radius of r in the plane of lamellae. Let the volume of the drop be V and its surface tension 1 dynes per cm. Further, it is assumed that the liquid wets the lamellae, so that the edge of the liquid meets the lamellae at zero angle.  j b r 2  Fig. 35. Conceptual drawing of water between two lamellae.  If the lamellae displaced an increment of dx, the change in potential energy of the system would be: Fdx  =  1 + )dS  where F is the force, 2L, solid-gas,  (sL  and  1 dS  =  -  -  2 ?4S  [E.1]  is the surface tension of liquid, solid-liquid and  2 is the change in the cfS is the change in the area of the liquid surface and c1S  area of the dry solid. 1 The area of the liquid-gas interface equals: S  =  rx 2 it  2 The area of wetted lamellae surface equals: S /dd.x 1 dS  =  rx), 2 d/dd,.c (iu  where V  =  =  2 2itr , or rx itr x 2  =  150  dx dS / 1  =  (2/2)’j/x  dx dS / 2  =  2 _2V/x  =  [E.2j  /2 t r 2  and [E.3j  _2itr2/x  Thus the force drawing two lamellae under tension forces would be F  =  dx (dS / 1  -  dx) dS / 2  =  )(it2r/2  -  x) 2itr / 2  [E.4J  If the value of x is to small in comparison with r F  =  [E.5]  2itr2/x  thus Internai tension  =  2 Fhr  =  22Jx dynes/cm 2  =  2 2Jxg grams/cm  or Equation [E.5] shows that a drop of water with sorption tension of 75 dynes/cm is held between two lamellae with a diameter of 1 cm increase with decrease in the distance between the lamellae. The internal tension force is enough at certain distances to pull the adjacent surfaces toward each other enough to make the operation of molecular forces possible. Campbell (1933) was the first to identify the role of these forces on bond development of a paper sheet.  Table 13. Development of internal tension forces between the lamellae during removal of water from the cell wall.  Internal tensb of  0 P/P  water (N/cm)  I6  0:99  Th1CCSS f watev between the laycrs  (em)  Lix IO 49x IO  +  (>  /  9360 31 JOG  jØ4  151  APPENDIX F  Calibration curve for GPC  It has already been stated that the GPC retention volume gives a measure of the size of the dissolved molecule. The exact correlation exists through use of standard narrow MWD samples having a known molecular weight at their peak retention volumes. However, for the higher molecular weight polymers only polystyrene has been produced on a sufficiently large scale having a narrow enough MWD for use as calibration standards. For other polymers a method of indirect calibration is required which can convert the polystyrene calibration to one applicable to the measured polymer. The GPC  separation 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 have different conformations in a given solvent. Also a polymer will assume different sizes in different solvent media at different temperatures. The hydrodynamic volume (HV) of a polymer is measured as  liv  [Tfl  =  [F.1}  zc M  where [1] is intrinsic viscosity, and M is molecular weight. This parameter forms the basis of the “universal” calibration. The Mark-Houwink relationship explains the interrelation of M and [ii] empirically as follow: [ii]  =  K  [F.2]  x Ma  The hydrodynamic’volume (HV) of polystyrene and cellulose tricarbanylate (CTC) is given by:  liv  of polystrene  =  , x 1 [ri]  =  (1 + CLI,)  152  HVOfCTC  []cxMcccMc  1W of polystrene KpMp (1 +  a)  =  ) 0 (1 + a  M 0 K  1flK + (1 + a)lflM  =  , 0 ) inN 0 0 + (1 + cz inK  ) + (1. + cz,)inM 0 ifl(K/K =  thus  WI of CTC, =  (l+cz)  =  0 (1 + a ) inN 0  ) 0 ) + (1 + CXp)iflNp)/(1 + CL 0 [ifl(K/K  or or or [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 valid for other polymers, too.  I  153  GLOSSARY Beating: Mechanical treatment of wet fibers which brings about desirable modifications to the sheet properties. Cellulose: The chief substance in the cell wall of plants. It is the fibrous substance that remains 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 wood chips in a digester. Kraft (sulphate) or suiphite process are examples of chemical pulping process. Co-crystallization: Uniting of two crystalline cores of the two adjacent elementary fibrils is 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 change in enthalpy. Therefore, the heat of transformation in any change of phase is equal to the difference between the enthalpies of the system in the two phases. If a system performs a cyclic process, the initial and final enthalpies are equal and the net enthalpy change in the process is zero. Entropy: Entropy in the present context is interpreted in terms of disorder or randomness of a system. External Fibrillation: Disruption of the outer layers of fibers during beating is referred as external fibrillation. As a result of this increase in external surface, the area available for contact and bonding with adjacent fibers increased. Fiber Wall: Fiber wall consists of concentric lamina made up of small repeating units 100 A by 100 A, consisting of a microcrystalline core surrounded by an amorphous sheath. Flexibility: It is defmed as the inverse of the product of the moment of inertia of the body and 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 cluster formation and dissolution in a liquid water at room temperature, and it is governed by local energy fluctuation. Free Energy: The free energy of a system could be defined as the maximum energy freed in a process and made available for work. Hornification: frreversible loss in swelling potential of the fibers is referred as hornification. Internal Fibrillation: Increase in volume during beating generally termed swelling or internal fibrillation.  154  Kraft Pulp: Wood pulp produced by the sulphate chemical process using cooking liquor which is made up primarily of sodium hydroxide (NaOH) and sodium, suiphide ( S). Also known as sulphate pulp. 2 Na Laminar Structure of Fibers: Wood fibers are made up of a series of concentric layers with variations in thickness, fibril orientation and chemical composition. Lignin: A brown-colored organic substance which is separated chemically during the cooking process which releases cellulose fibers. It is removed along with other organic materials in the spent liquor during subsequent washing and bleaching stages. Never-Dried Pulp Fiber: A fiber which is collected from a fresh tree and never experienced drying. Recrystallization: The concept of recrystallization which is frequently evoked to explain the increase in crystallinity index after recycling. Recycled Fiber: A pulp fiber which has been rewetted from the dry state and made into paper at least once. Secondary Fiber: Pulp recovered from a paper product which has already served a commercial purpose. Secondary fiber may be obtained from an intermediate processor, such as a printer or converter, from a final user, such as a homeowner or from its end destination; the municipal dump. Surface Deactivation: The loss in surface bonding potential of the fibers in the present context is referred as surface deactivation or surface inactivation. Thermodynamics: Thermodynamics is concerned with energy relationships involving heat, mechanical energy, and other aspects of energy and energy transfer.  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0088893/manifest

Comment

Related Items