"Applied Science, Faculty of"@en . "Chemical and Biological Engineering, Department of"@en . "DSpace"@en . "UBCV"@en . "Chen, Margaret Ying Ting"@en . "2009-05-04T20:52:04Z"@en . "1998"@en . "Master of Applied Science - MASc"@en . "University of British Columbia"@en . "A redox cycle involving anthraquinone 2-sulfonic acid - - denoted as AQS, and oxygen in an\r\naqueous alkaline media was investigated for its ability to promote the in-situ electroreduction of\r\noxygen to hydrogen peroxide with the coincident promotion of the brightening of thermo-mechanical\r\npulp. Current applied at a carbon cloth cathode was used to reduce the AQS to\r\nhydroanthraquinone - - denoted as H\u00E2\u0082\u0082AQS. The reaction of H\u00E2\u0082\u0082AQS with oxygen was shown to\r\nproduce hydrogen peroxide. Evidence presented in this study indicates the operation of a redox\r\ncycle involving AQS/ H\u00E2\u0082\u0082AQS, oxygen and current to produce hydrogen peroxide. It is thought\r\nhydrogen peroxide produced in this redox cycle was responsible for the larger brightness gains and\r\nyellowness reduction of the thermomechanical pulp observed in experimental runs employing the\r\nAQS/oxygen redox cycle, than in corresponding runs in the absence of the AQS/oxygen redox\r\ncycle.\r\nThe batch electrochemical reactor employed in this study consisted of a platinized titanium anode\r\nand carbon cloth cathode. Major modifications of the reactor design, which included increasing\r\nthe area of the working electrode (the cathode), decreasing the area of the counter electrode\r\n(anode), employing membrane/fritted glass separators and providing adequate mixing, were found\r\nto suppress undesirable electrochemical reactions and promote even current distribution in the\r\nelectrochemical cell, resulting in an increased current efficiency and production rate of peroxide.\r\nAll runs were conducted at 60 \u00C2\u00B0C pH 10.5-11 and at atmospheric pressure.\r\nSome other important quantitative findings were:\r\n\u00E2\u0080\u00A2 In a solution containing 0.1 M Na\u00E2\u0082\u0082SO\u00E2\u0082\u0084, 4.7mM Na\u00E2\u0082\u0082SiO\u00E2\u0082\u0083, 0.32mM MgSO\u00E2\u0082\u0084, the hydrogen\r\nperoxide produced in 3 hours with the use of the 5mM AQS and oxygen redox cycle was\r\napproximately 49.5(+/-) 2.5 mM, which is 28(+/-) 2.5mM more than in a corresponding run in\r\nthe absence of the AQS. \u00E2\u0080\u00A2 In a solution containing 0.1 M Na\u00E2\u0082\u0082SO\u00E2\u0082\u0084, 4.7mM Na\u00E2\u0082\u0082SiO\u00E2\u0082\u0083, 0.32mM MgSO\u00E2\u0082\u0084, with oxygen purged\r\nfor 3 hours (at 1.5A without pulp), subsequently followed by oxygen purged for 5 hours (at 0 A\r\nwith 1 wt. % thermomechanical pulp), the presence of 5mM AQS brought about 21 (%ISO)\r\ngain and 15 (%) yellowness reduction more than for a corresponding run in the absence of the\r\nAQS.\r\n\u00E2\u0080\u00A2 A 2\u00C2\u00B3 factorial experiment was performed to illuminate the three (3) key process variables:\r\ncurrent (0.3A to 1.5A), gas (Air-Oxygen), the AQS concentration (OmM to 5mM). From this\r\nfactorial experiment, it is found that the presence of two key factors, namely oxygen and\r\ncurrent, are essential and the presence of catalyst factor, such as AQS, is beneficial in the in-situ\r\nelectrochemical generation of hydrogen peroxide in the AQS/oxygen system.\r\nAlthough the AQS/oxygen redox cycle, used in this project, generated hydrogen peroxide at a high\r\ncurrent efficiency and produced a brightened pulp that is comparable to that obtained using\r\nmerchant peroxide, this system cannot be realized in the industry due to the severe fouling\r\nproblems associated with the use of its working electrode. It is thought that the AQS crystallizes\r\nand blocks electro-active areas of the carbon cloth electrode, which makes it unsuitable for any\r\nsubsequent runs. Due to the fouling problems posed by the carbon cloth cathode with the AQS\r\nredox couple, the electroreduction of oxygen mediated by AQS is impractical for industrial use\r\nwithout further study."@en . "https://circle.library.ubc.ca/rest/handle/2429/7834?expand=metadata"@en . "4983648 bytes"@en . "application/pdf"@en . "The In-Situ Brightening of Pulp by the Electro-reduction of Oxygen, Mediated with Anthraquinone-2-Sulfonic Acid By Margaret Ying Ting Chen B.A. Sc. (Chemical Engineering) University of British Columbia A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF Master of Applied Science in THE FACULTY OF GRADUATE STUDIES Department of Chemical Engineering We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA APRIL, 1998 \u00C2\u00A9 Margaret Ying Ting Chen, 1998 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 the permission for extensive copying of this thesis for scholarly purpose may be granted by the Head of my Department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of Chemical Engineering The University of British Columbia 2216 Main M a l l Vancouver, B . C . V 6 T 1Z4 Canada Date: A p r i l 25 t h , 1998 Date: Apr i l , 1998 Abstract A redox cycle involving anthraquinone 2-sulfonic acid - - denoted as AQS, and oxygen in an aqueous alkaline media was investigated for its ability to promote the in-situ electroreduction of oxygen to hydrogen peroxide with the coincident promotion of the brightening of thermo-mechanical pulp. Current applied at a carbon cloth cathode was used to reduce the AQS to hydroanthraquinone - - denoted as H 2 A Q S . The reaction of H 2 A Q S with oxygen was shown to produce hydrogen peroxide. Evidence presented in this study indicates the operation of a redox cycle involving A Q S / H 2 A Q S , oxygen and current to produce hydrogen peroxide. It is thought hydrogen peroxide produced in this redox cycle was responsible for the larger brightness gains and yellowness reduction of the thermomechanical pulp observed in experimental runs employing the AQS/oxygen redox cycle, than in corresponding runs in the absence of the AQS/oxygen redox cycle. The batch electrochemical reactor employed in this study consisted of a platinized titanium anode and carbon cloth cathode. Major modifications of the reactor design, which included increasing the area of the working electrode (the cathode), decreasing the area of the counter electrode (anode), employing membrane/fritted glass separators and providing adequate mixing, were found to suppress undesirable electrochemical reactions and promote even current distribution in the electrochemical cell, resulting in an increased current efficiency and production rate of peroxide. A l l runs were conducted at 60 \u00C2\u00B0C pH 10.5-11 and at atmospheric pressure. Some other important quantitative findings were: \u00E2\u0080\u00A2 In a solution containing 0.1 M Na2SC>4, 4.7mM Na2SiC>3, 0.32mM MgSC\u00C2\u00BB4, the hydrogen peroxide produced in 3 hours with the use of the 5mM AQS and oxygen redox cycle was approximately 49.5(+/-) 2.5 mM, which is 28(+/-) 2.5mM more than in a corresponding run in the absence of the AQS. ii \u00E2\u0080\u00A2 In a solution containing 0.1 M Na 2S04, 4.7mM Na 2Si03, 0.32mM M g S 0 4 , with oxygen purged for 3 hours (at 1.5A without pulp), subsequently followed by oxygen purged for 5 hours (at 0 A with 1 wt. % thermomechanical pulp), the presence of 5mM AQS brought about 21 (%ISO) gain and 15 (%) yellowness reduction more than for a corresponding run in the absence of the AQS. \u00E2\u0080\u00A2 A 2 factorial experiment was performed to illuminate the three (3) key process variables: current (0.3A to 1.5A), gas (Air-Oxygen), the AQS concentration (OmM to 5mM). From this factorial experiment, it is found that the presence of two key factors, namely oxygen and current, are essential and the presence of catalyst factor, such as AQS, is beneficial in the in-situ electrochemical generation of hydrogen peroxide in the AQS/oxygen system. Although the AQS/oxygen redox cycle, used in this project, generated hydrogen peroxide at a high current efficiency and produced a brightened pulp that is comparable to that obtained using merchant peroxide, this system cannot be realized in the industry due to the severe fouling problems associated with the use of its working electrode. It is thought that the AQS crystallizes and blocks electro-active areas of the carbon cloth electrode, which makes it unsuitable for any subsequent runs. Due to the fouling problems posed by the carbon cloth cathode with the AQS redox couple, the electroreduction of oxygen mediated by AQS is impractical for industrial use without further study. iii Table of Contents Page No. ABSTRACT ii TABLE OF CONTENTS v LIST OF FIGURES vii LIST OF TABLES xi NOMENCLATURE x CHAPTER 1.0. INTRODUCTION 1 CHAPTER 2.0. BACKGROUND AND LITERATURE REVIEW 5 2.1. Structure and Chemical Composition of Wood 2.1.1.Chemical Composition of Wood 5 2.1.2.Structure and Chemical Composition of Wood Fiber 8 2.2. Mechanical Pulping and Brightening 9 2.3. Mechanical Pulp Brightening by Hydrogen Peroxide 11 2.4. Brightness 13 2.5. Brightening Chemicals 14 2.6. Hydrogen Peroxide Brightening 15 2.7. Factors Affecting Hydrogen Peroxide Responses 2.7.1. Temperature/Retention Time 16 2.7.2. Consistency of Pulp 17 2.7.3. Mixing 17 2.7.4. Alkalinity of Hydrogen Peroxide Brightening 18 2.7.5. Peroxide Charge 19 iv Table of Contents Page No. 2.7.6. Peroxide Brightening Additives/Stabilizers 20 2.7.7. Fiber Properties 22 2.8. Electrochemical Production of Hydrogen Peroxide 23 2.8.1. General Electrochemistry 23 2.8.2. In-situ Electrochemical Production of Hydrogen Peroxide 24 2.8.3. Electrode Materials 24 CHAPTER 3.0. PROPOSED RESEARCH OBJECTIVES 28 CHAPTER 4.0 EXPERIMENTAL APPARATUS AND PROCEDURES 29 4.1. Cyclic Voltammetry Apparatus 29 4.2. Experimental Procedures 31 4.3. Electrochemical Reactors 32 4.3.1. H-Cell Divided Glass Reactor 32 4.3.2. Pulp Brightening Reactors 34 4.3.2.1. Undivided Reactor -Configuration 1 4.3.2.2. Undivided Reactor - Configuration 2 4.3.3.3. Undivided Reactor- Configuration 3 4.4. Preparation of Chemicals and Pulp 3 8 4.5. H 2 0 2 Generation and TMP Brightening 39 CHAPTER 5.0. EXPERIMENTAL RESULTS AND DISCUSSIONS 40 5.1. Electrochemistry of Anthraquinone 2-Sulfonic Acid (AQS) 41 5.1.1. Cyclic Voltammogram (CV) 41 5.1.2. Current, Electrode Potential vs. SCE Curve Study 45 5.1.3. Bulk Electrolysis 47 v Table of Contents Page No. 5.2. Electrochemical Generation of H2O2 48 5.2.1.. Stage I: Preliminary Experiments 50 5.2.1.1. Determine the Effect of Anthraquinone 2-Sulfonic Acid 5.2.1.2. Effect of Current Density 5.2.1.3. The Effect of Anthraquinone 2, 6 Disulfonic Acid 5.2.2. Stagell: 57 Preliminary Investigation 5.2.3. Stage III: 59 5.2.3.1. Effect of Cation Membrane 5.2.3.2. Effect of Current Density 5.2.4. Stage IV: 61 5.2.4.1. Effect of Fritted Glass Separation 5.2.4.2. Carbon Felt as an Alternative Cathode Material 5.2.5. Stage V: Number of Carbon Cloths vs. Rate of H2O2 Production 67 5.2.6. Stage VI: Final Modification in Reactor Design 70 5.3. Electrochemical Brightening of Pulp 72 5.3.1. Hydrogen Peroxide Charges vs. Brightening Effects 73 5.3.2. Ex-situ Brightening vs. In-situ Brightening 74 5.3.3. Effect of Residual Carbon Fibres 80 5.3.5. Brightening Effects of In-situ Generated Peroxide vs. Merchant Peroxide \u00E2\u0080\u0094 81 5.4. Factorial Design Experiments 83 5.5. Aging of the Electrode 87 CHAPTER 6.0. GENERAL DISCUSSIONS 88 CHAPTER 7.0. CONCLUSIONS AND RECOMMENDATIONS 91 vi Table Of Contents Page No. BIBLIOGRAPHY 93 APPENDICES 96 Appendix I. Analysis of Factorial Design Appendix II. Current Efficiency of Hydrogen Peroxide Generation Appendix III. Cyclic Voltammetry (CV) Appendix IV. Hydrogen Peroxide Titration Procedures Appendix V. Experimental Results from Section 5.2. Appendix VI. Experimental Results from Section 5.3. Appendix VII. Experimental Results from Section 5.4. Appendix VIII. Glossary vn L i s t o f F igures PAGE NO. Figure 2.1.1. Chemical Components of Wood 5 Figure 2.1.2. Structure of Cellulose 6 Figure 2.1.3. Structure of Lignin 7 Figure 2.2.1. The Basic Chromophoric structures 10 Figure 4.1. Cyclic Voltammetry Set-up 29 Figure 4.3.1. H- Cell Divided Reactor 32 Figure 4.3.2.1. Pulp Brightening Reactor (Configuration #1) 34 Figure 4.3.2.2. Pulp Brightening Reactor (Configuration #2) 35 Figure 4.3.3.3. Pulp Brightening Reactor (Configuration #3) 36 Figure 5.1.1. Cyclic Voltammogram for the Anthraquinone 2 - Sulfonic 42 Redox Couple in the Presence of Nitrogen Figure 5.1.2. Cyclic Voltammogram for the Anthraquinone 2-Sulfonic 43 Redox Couple Formed in the Presence of Oxygen Figure 5.1.3. (I, E vs. SCE); Current vs. the Potential of the Working 45 Electrod Figure 5.2.1. Effect of 5mMAQS vs. OmM AQS on the Rate of H 2 0 2 51 Generation Figure 5.2.2. Effect of Current Density on the Rate of H 2 0 2 Generation 53 Figure 5.2.3. Effect of AQ2S vs. AQS On the Rate of H 2 0 2 Generation 55 Figure 5.2.4. Effect of Scaled-up Reactor/Cation Membrane on the Rate Of 58 H 2 0 2 Generation Figure 5.2.5. Effect of Current Density on the Rate Of H2O2 Generation.... 60 ix PAGE NO. Figure 5.2.6. Effect of Medium Sized Fritted Glass on the Rate of H2O2 62 Generation Figure 5.2.7. Effect of Carbon Felt (Cathode) on the Rate Of Ff 20 2 65 Generation Figure 5.2.8. H 2 0 2 in Three (3) Hours vs. Number of Carbon Cloth 68 Figure 5.3.1. Ex-Situ Brightening Effects vs. Time 75 Figure 5.3.2. In-Situ Brightening Effects vs. time 77 Figure 5.3.3. Brightening Effects of Merchant Peroxide vs. 81 (Merchant Peroxide+5mMAQS) vs. Ex-situ Generated H 2 0 2 Figure 5.3.4. Yellowness-Reducing Effect of Merchant Peroxide vs. 82 (Merchant H202+5mM AQS) vs. Ex-situ Generated H 2 0 2 Figure 5.4.1. Cube Plot of Hydrogen Peroxide Concentration as a Function 84 of O2, Catalyst and Current Density levels x List of Tables PAGE NO. Table 2.1. Chemical Composition of Wood 8 Table 2.5. Typical Brightening Processes and Their Results 14 Table 4.2. Experimental Procedures 31 Table 5.2.0. Outline of the Six Progressive Stages 49 Table 5.2.1. Flowsheet of Stage I 50 Table 5.3.0. Flowsheet of the Overall Experimental Plan, Section 5.3 72 Table 5.2.2. The Flowsheet of Experimental Plant, Stage IV 61 Table 5.3.1. Initial Hydrogen Peroxide Concentration vs. In-situ Brightening Effects 73 In Three Hours Table 5.3.2. Procedures for Both In-situ and Ex-situ Brightening 74 Table 5.3.3. Effect of Carbon Fibre Residuals on Ex-situ Brightening Processes... 80 Table 5.4.1. Variables and Their Level in the 23 Factorial Experiments 83 Table 5.4.2. Main and Interaction Effects in 2 Factorial Experiments 85 Table 5.5. Aging of the Electrode 87 xi Chapter 1.0. - Introduction Mechanical pulp brightening technology has evolved, due in part to more stringent environmental regulation and the rising market demands for environmentally friendly paper products. Through this evolution, hydrogen peroxide has become one of the most important chemicals for mechanical pulp brightening in Canada. Hydrogen peroxide consumption in Canadian pulp mills increased from 108 kiloton in 1995 to 125 kiloton in 1996 and was forecast to increase further in 1997. As a result, new hydrogen peroxide manufacturing facilities have been and will be started in the 1996-98 period, e.g. Du Pont's North American peroxide capacity will increase by about 30% to 190 kt/y, Chemyprox is doubling its capacity at the Beconcour facility to 66kt/y and in eastern Canada Degussa will start building a 40kt/y plant. Hydrogen peroxide is employed as an oxidizing agent in various organic and inorganic processes. It is favored for brightening mechanical pulp to values up to 80% ISO for use in high volume products such as newsprint and tissue. Brightening of mechanical pulp is a lignin retention process, maintaining the desired high yield. The low capital cost associated with a peroxide stage [28], ease of application and the relatively non-toxic and innocuous nature of its reaction products in the brightening process make hydrogen peroxide a desirable environmentally friendly, brightening agent. Below are the two major industrial processes for producing hydrogen peroxide, which include the hydrolysis of peroxydisulfuric acid, formed by the electro-oxidation of a concentrated sulfate solution (now obsolete), and the auto-oxidation of anthraquinone. 1 1. Catalytic Anthraquinone Process 2. Electrochemical Persulfate (or Persulfuric acid) Process (developed in the early 1900s) 2S0 4 2 \" ==> S 2 0 8 2\" + 2e\" S 2 0 8 2\" + 2 H 2 0 ==>2S04 2\" + H 2 0 2 + 2 H + . The electrochemical persulfate process was the major commercial route to hydrogen peroxide [2] until about the late 1950's. Due to high capital and high energy cost, the persulfate process became obsolete for merchant production of concentrated hydrogen peroxide. Instead, the conventional process of auto oxidation of anthraquinols, developed during the Second World War, has become the most important source of peroxide. The auto oxidation of anthraquinols is a cyclic process wherein a working solution containing an alkyl anthraquinone dissolved in an organic solvent (e.g. 50:50 mixture of benzene and C7 and Cn secondary alcohol) is sequentially hydrogenated, oxidized, subjected to a liquid -liquid extraction and recycled back to the hydrogenation unit. [15]. The peroxide solution obtained by water extraction of hydrogen peroxide from the organic solvent is further purified concentrated and stabilized [29]. The several disadvantages associated with conventional anthraquinone include high capital cost, i.e. a large number of stages, costly unit operations, catalyst poisoning problems, and hazardous, costly organic solvent treatment. 2 Because of the problems inherent in the conventional hydrogen peroxide generating processes, alternative methods have been researched. Direct combination of H2 and O2 , for example, has been the investigated in by DuPont Co (USA), Albchem Industries Ltd (Canada) and companies in Japan, but this process is not yet commercialized [25]. Electrosysnthesis of alkaline peroxide from O2 and NaOH,, has also been closely examined as an alternative peroxide generating method, particularly for the pulp and paper mills, where dilute solutions of alkaline peroxide are required for brightening and bleaching purposes. The electro-reduction of oxygen in alkaline media is summarized as below: 3. Electrochemical Reduction of Oxygen in Alkaline media 2 Faraday NaOH + ^ O j => N a H 0 2 The product is a mixture of H.202 in excess NaOH. In this alkaline electro-reduction of oxygen system, the reduction of HO2\" to OH\" is thermodynamically favored over the reduction of 0 2 to HO2\". Thus, a cathode, which has good intrinsic properties for promoting the desired reactions, is pre-requisite to achieve a good current efficiency for H 0 2 \" . During the earlier investigation of this process, the gas diffusion 0 2 cathodes were used. However, serious fouling problems associated with the gas diffusion cathodes arise as a result of prolonged use, due to crystallisation of N a H 0 2 . n H 2 0 in the cathode pores [2]. Subsequent research, at the University of British Columbia (UBC), introduced the trickle-bed cathode in which O2 gas and NaOH solution pass through a fixed-bed electrode in co-current flow [21,24] A similar research was carried out by Dow Chemical Co and later by H-D Tech Inc., introducing the composite chip fixed-bed cathode composed of composite chip fixed-bed cathode composed of graphite chips coated with a mixture of carbon black and Teflon [4,19]. Concurrently, the E-TEK reactor [8,9], uses a gas diffusion cathode composed of layers of carbon black/Teflon mixture pasted onto a substrate of graphite cloth. For this particular E-TEK process, the 3 electrolyte and 0 2 gas are introduced to opposite faces of the cloth and reaction occurs along the gas/liquid/solid menisci in the pores. All the electro-synthesis of alkaline peroxide processes, mentioned above, nevertheless, cannot economically compete with the existing anthraquinone oxidation process [14,18], because of high capital and power costs. Furthermore, those methods also produce a solution with pH>12 and NaOH/H2C\u00C2\u00BB2ratio>1.5 and thus are not suitable for direct application in pulp brightening. In this thesis project, the electro-synthesis of the relatively expensive oxidant H2O2 (0.5$/kg) from the cheaper oxidant 0 2 (0.1$/kg), will be examined with three major process modifications. These three modifications include the addition of a redox couple promoter (an anthraquinone compound), the attempt of reusing and recycling the chemicals, and the possibility of electro-synthesizing H 2 0 2 at brightening pH (pH=10.5-ll). Finally, this project will investigate the electro-brightening of TMP by H2O2 generated in-situ via the anthraquionone 2-sulfonic acid mediated electro-reduction of oxygen. 4 Chapter 2.0. - Background and Literature Review 2.1. Structure and Chemical Composition of Wood Fiber and Unbrightened pulp 2.1.1 Chemical Composition of Wood Wood is a plentiful source of fiber, and is essentially the only source of fiber utilized in North America for the paper making industry. The three polymeric chemical components of wood are cellulose, hemicellulose and lignin. The chemical composition of wood is illustrated in Figure 2.1.1.[34]. WOOD L I G N I N EXTRACTIVES CARBOHYDRATES Terpencf Resin Acid* (Softwood*) Fatty Acid* Phenofa U r r W r p o n i f w a W c i C E L L U L O S E H E M I C E L L U L O S E Glucose Glucose Mannose Galactose Xylose Arabirwse Figure 2.1.1 -- Chemical Components of Wood 5 Cellulose Cellulose is the major component of wood fiber walls, which determines the characteristics of the fiber, thereby influencing its functional use in paper making. It is a polysaccharide and a high molecular weight stereo regular linear polymer. The fiber bundles it forms impart a high tensile strength to the wood. The structure of cellulose is shown below in Figure 2.1.2 [34]. Figure 2.1.2 Structure of Cellulose Hemicellulose Hemicellulose is a short chain, branched polymer of five sugar monomers: glucose, mannose, galactose,xylose and arabinose. In comparison to cellulose, hemicellulose has shorter polymer chains that are more easily hydrolysed by acids to basic monomeric components. It serves as the supportive matrix for the cellulose microfibrils. 6 Lignin Lignin is an important mechanical reinforcement agent for the entire tree. It makes up approximately 28 percent of wood. Lignin is a highly branched, three-dimensional polymer network consisting of aromatic rather than saccharide building units. It forms the middle lamella, the intercellular material which cements the fibers together. The chemical components of lignin include methoxy, aliphatic and phenolic groups in a three dimensional polymer linked by C-O-C and C-C bonds.. The structure of lignin is illustrated in Figure 2.1.3.[16]. OH o HC CH\u00E2\u0080\u0094O HC- \u00E2\u0080\u00A2CH HC-OH 0 -Figure 2.1.3. Structure of Lignin [151 7 2.1.2. Structure and Chemical Composition of Wood Fiber Wood and fiber characteristics vary depending upon species, genetics, tree age, climate, natural fluctuations in the environment and soil fertility. Woods are classified into two major groups, softwoods (conifers) or hardwoods (broad-leafed) tress. Generally, as indicated in Table 2.1, hardwoods contain a larger fraction of cellulose, hemicellulose and extractives and less lignin than soft woods. Table 2.1. Chemical Composition of Wood [34] Component Hardwood % Softwood % Cellulose 45(\u00C2\u00B1)2 42(\u00C2\u00B1)2 Hemicellulose 30(\u00C2\u00B1)5 27(\u00C2\u00B1)2 Lignin 20(\u00C2\u00B1)4 28(\u00C2\u00B1)3 Extractives 5(\u00C2\u00B1)3 3(\u00C2\u00B1)2 Fibers from softwoods and hardwoods also differ. On the average, softwood fibres are hollow, about 3.5 mm in length, 35 um in diameter with a wall thickness of 2-7 um. The hardwood fibres, however, tend to be smaller, the fibres are about 1.5 mm in length and 20 microns in diameter. A typical fibre wall contains four main layers. The primary wall is a thin outer layer surrounding the fibre. The secondary wall is composed of three separate layers, the outer (SI), middle (S2) and inner (S3). These three layers, i.e. SI, S2, S3 are built up by lamellae formed by almost parallel microfibrils (cellulose bundles) between which the lignin and hemicellulose are located. Approximately 80 percent of this secondary-wall region is composed of holocellulose, and 20 percent is lignin. The diameter of the lumen, i.e. the hollow air space inside the fiber varies with the species of the tree and ranges from 3 to 10 microns. 8 2.2 Chemical Principle of Brightening The objective of brightening is to increase the brightness of mechanical or chemi-mechanical pulp, which is often brownish yellow. In contrast to pulp bleaching, which is to substantially dissolve the residual color-producing lignin and leave a relatively white cellulosic fibre, the brightening process selectively changes color-producing chromophores to relatively colorless substances without dissolving the major components of wood. The main advantage of mechanical pulping and chemi-mechanical pulping is therefore, high yield. However, because of the limit of brightening - 85% ISO brightness value, and poor brightness stability, the use of mechanical pulp and chemi-mechanical pulp is still limited. Although the detailed mechanism of brightening is not well understood, the general principle of brightening is to sever the conjugation between individual chromophores, change the chemical structure of chromophores, eliminate auxochromes, or break the chemical bonds between auxochromes and chromophores. Chromophore groups in lignin are composed of mainly carbonyl groups and double bonds. When they are combined with benzene rings and auxochromes in a certain way, they form chromophoric structures which shift the wavelength of light absorption from the ultraviolet region into the visible region, and generate the yellow-brown color of wood [27]. Three major types of basic chromophoric structure [7] in wood are: 1) Ortho- and para- quinone (la, lb) 2) Ortho-hydroxl- and para-hydroxyl-phenyl ketones (Ha, lib) 3) para-quinone methides and carbonium ions (Ilia, 111b) 9 These structures are illustrated in Figure 2.2.1. Figure 2.2.1. Three Basic Chromophoric Structures [7] 10 2.3 Mechanical Pulping and Brightening The two basic types of mechanical pulping are stone groundwood. in which fiber is removed from wood by pressing a log against a grindstone, and chip refining, in which wood chips are broken into fibers on passing between rotating disks having bars on their surface. The objectives of mechanical pulp brightening are summarized as: 1) To decolorize lignin and other extraneous components without solubilizing them, 2) To preserve yield commensurate with the above, 3) To do the least mechanical damage to the fibers, 4) To inhibit the formation of carbonyl groups, and 5) To be cost effective Mechanical pulp brightening employs chemicals that selectively attack and decolorize some of the chromophoric groups but do not dissolve lignin. This is a lignin retention process in which yield is preserved, i.e. usually around 90 - 95%.[34] Reductive agents such as sodium dithionite and oxidative agents such as hydrogen peroxide are commonly used for brightening mechanical pulp. Softwoods and hardwoods respond differently to brightening chemicals. The brightening process is a function of wood species. All species give different unbrightened brightnesses and respond differently to brightening depending on their natural color, their quality and their extractive content and type. Generally, only those species yielding pulp with an initial brightness value over 55% ISO are utilized for mechanical pulping. 11 Although the brightness gains can be substantial, no known method of selective brightening produces a permanent effect; exposure to light and atmospheric oxygen causes the lignin to rapidly discolor, as can be readily observed with old newspapers (34) 12 2.4 Brightness Brightness and yellowness are two important optical properties of pulp. Brightness is defined as the reflectance of blue light with a specified spectral distribution with peak at 457 nm from an opaque surface of pulp sheets compared to a specified reflecting, diffusion standard surface [4]. Visual brightness is subjective and there is no general correlation between papermaker's brightness and visual brightness. Two widely accepted methods of measuring brightness are that of General Electric (Tappi Standard T452, U.S.A.) and the Zeiss Elrepho (official standard worldwide, except the U.S.A.). In this thesis project, the CPPA method (Zeiss Elrepho) is used. This method specifies that the sample should be diffusely illuminated with a highly reflecting, integrating sphere. Reflected light is measure 90\u00C2\u00B0 to the sample, and reflectance is compared to absolute reflectance from an imaginary perfectly reflecting, diffusing surface, and the ratio of the reflectance to absolute reflectance is taken as brightness. The standard is opal glass. The brightness obtained is expressed as a %ISO. For general commercial uses, MgO powder is used as a brightness standard, which is about 98 - 99% of absolute reflectance. In this case, the brightness is Elrepho brightness or MgO brightness. There is no relationship between the GE and the ISO brightness scale, but generally GE brightness is about 1% lower than ISO brightness. 13 2.5 Brightening Chemicals Commercial brightening chemicals include sodium bisulfite (or other sulfur dioxide derivatives), sodium dithionite and hydrogen peroxide. Sodium bisulfite and sodium dithionite are reducing brightening agents, while hydrogen peroxide is an oxidizing agent. Sodium borohydride is not a commercial brightening chemical, but only used for on-site generation of sodium dithionite from SO2, as in the Borol Process. Some typical brightening processes and their brightening results are summarized as below: Brightness Goal Effective Brightening Process Brightness - 65 to 75% ISO or Brightness Gain of 12 to 14% ISO Refiner dithionite brightening for refiner based mills Brightness - 70 to 80% ISO or Brightness gain of 15 to 18% ISO Two Stages Brightening: Peroxide-dithionite brightening Brightness of more than 80% ISO or Brightness gain of more than 20 % ISO Two Stages Brightening: Peroxide - peroxide brightening Table 2.5. Brightening Chemicals 14 2.6 Hydrogen Peroxide Brightening Traditionally, hydrogen peroxide brightening has been a single-stage process carried out over a range of consistencies up to 15%. In recent years, demand for higher-quality, brighter mechanical pulps has promoted the development of more sophisticated processes. The trend is thus moving toward two-stage, high consistency systems with recycle of chemical. Single-stage systems are usually adequate for brightness gains up to 15 points. If only moderate gain is required (up to 10 points), a medium consistency system is ample. For brightness gains above 10 points, a high-consistency system is generally required. Although a shorter retention time and lower chemical dosage are needed for a high-consistency system, the equipment costs are higher in comparison to a low consistency system. Presses are required to reach 25 - 40% consistency before the peroxide is added. 2.7. Factors Affecting Hydrogen Peroxide Brightening Response Hydrogen peroxide is an effective oxidative brightening agent. It improves the brightness of mechanical pulp while preserving yield. The results of hydrogen peroxide brightening of mechanical/chemimechanical pulp are dependent on various factors. For a given mechanical pulp brightening operation in industry, the temperature, retention time, consistency of pulp, mixing, alkalinity/pH, peroxide charge and additives/stabilzers are empirically chosen and optimized in accordance to the desired brightness and other specific operational constraints [32]. 15 2.7.1. Temperature/Retention time Both temperature and retention time are two important factors, which may improve brightness results. Temperature Industrial brightening of mechanical pulp usually proceeds in the 40 - 70 \u00C2\u00B0C range. The brightening reaction is accelerated at higher temperatures. However, as the temperature approaches 80 \u00C2\u00B0C, brightening efficiency decreases drastically. At high temperatures, the amount of alkali present becomes critical. High concentrations of alkali will promote the decomposition of hydrogen peroxide. In other word, when the total alkali is too high, rapid decomposition of peroxide occurs; when the total alkali is too low, the brightening reaction is significantly retarded. It has also been shown that the amount of alkali required for optimum brightening response decreases as the temperature rises. Retention Time Retention time is also an important factor in the brightening process. Excessive retention periods will result in alkali darkening and therefore, brightening reversion due to the presence of insufficient peroxide. The retention time employed in industrial brightening reactions is 1-5 hours. The most rapid brightening effect is observed in the early stages of the brightening process where the concentration of chemicals is high and the brightness of the pulp is low [33]. 16 2.7.2. Consistency of pulp Brightening efficiency increases as the pulp consistency increases. This is due to an increase in concentration of brightening chemicals resulting from the reduced volume of the aqueous phase [3] in a high consistency pulp brightening system. Although hydrogen peroxide can be used to brighten pulp of all consistencies, most commercial installations brighten pulps with consistencies in the medium consistency (10% to 15%) or low consistency (3% to 6%) range. Medium consistency pulp brightening is adequate when a brightness gain of up to 12 points is required. For brightness gains above 12 points, high consistency pulp brightening is generally required. Even though a shorter retention time and lower chemical dosage is required in the high consistency brightening, the equipment costs are higher. In summary, the consistency used will depend upon the balance between capital cost and the cost of chemicals. 2.7.3 Mixing Good mixing is critical for effective brightening. It has been shown that improved mixing reduces chemical consumption and leads to a more uniform product [26]. However, good mixing is difficult to achieve, because even measuring or characterizing mixing can be difficult [13]. When non-uniform mixing occurs, pulp may receive a mal-distribution of chemicals, which results in less desirable brightening reactions taking place. Some regions of pulp may receive an excessive amount of chemicals while other regions receive insufficient chemicals and thus, are not adequately brightened. The net effect can be overcome by adding an excess of chemicals. As a result of adding excessive chemicals, a desired degree of brightening pulp with a lesser strength is produced [37]. 17 2.7.4. Alkalinity of Hydrogen Peroxide Brightening If insufficient alkali is present, the pH may fall to a degree where brightening ceases. On the contrary, if an excessive amount of alkali is present, peroxide may rapidly decompose to water and oxygen via Equation 1. H 2 0 2 = \u00E2\u0080\u0094 > H 2 0 + 1/2 0 2 (1) The oxygen generated from peroxide decomposition, in conjunction with alkali, may further react with lignin to form new chromophores. Reichert and Pete [30] recommended that the total alkali should lie in the range of 1.2 to 1.9 % by weight based on OD pulp. Higher alkalinity was determined to be more advantageous for low consistency-low temperature brightening while the lower alkalinity was found to be preferable high consistency- high temperature brightening. The lignin preserving brightening effect of hydrogen peroxide in mechanical pulp brightening is attributed to the oxidative action of perhydroxyl ion (HO2\"), which is dissociated from hydrogen peroxide in alkaline conditions via Equation 2 [4]. H 2 0 2 \u00E2\u0080\u0094 > H02\" + H + pKa=11.67at25\u00C2\u00B0C (2) The rate of hydrogen peroxide generation not only increases with increasing alkalinity, the dissociation reaction of hydrogen peroxide to perhydroxyl ions (Equation 1) is also promoted at high pH levels. At a pH of 10.5, less than 10% of the peroxide exists as the perhydroxyl ion [15]. However, hydrogen peroxide is also susceptible to decomposition with increasing alkalinity. High alkalinity will lead to the formation of new chromophores through alkali darkening and promote the decomposition reaction of hydrogen peroxide to form water and oxygen via Equation 2. Due to these two limiting factors, the brightening liquor must be sufficiently alkaline to maintain an adequate concentration of perhydroxyl ions, but not so alkaline as to cause excessive peroxide decomposition and pulp darkening. 18 Thus, pH must be carefully chosen to remain within these two constraints. The alkalinity of brightening liquor is provided by caustic soda and sodium silicate. In general, the brightening of mechanical pulps is usually carried out at a pH between 10-12 and at a temperature between 40-70 \u00C2\u00B0C [32]. 2.7.5 Peroxide Charge A suitable balance between peroxide and free alkali should be maintained in brightening systems. The brightness attained in pulp increases as the amount of peroxide applied is increased, but the incremental gains diminish to a point where further increases in peroxide charge can not be economically justified. The leveling off effect in brightness becomes apparent especially at charges greater than 1 % peroxide by weight on OD pulp. Economic considerations limit peroxide charges in most commercial operations to the range of 1 - 2% peroxide by weight on OD pulp, with brightness gains amounting to 10 - 15 points. 19 2.7.6. Peroxide Brightening Additives/Stabilizers Equation 1 (section 2.7.4). is catalyzed by the presence of certain metal ions such as iron, manganese and copper. The sources of these heavy metals include process water, equipment, and the pulp and wood fibers themselves [5]. For example, wood can often contain 100 PPM of manganese which is ample to cause serious decomposition of hydrogen peroxide in a brightening process [16]. Stabilization of Brightness After the completion of the brightening process, the stock should be stabilized by reducing the pH to the 5 to 6 range. This prevents alkali darkening and loss of brightness when the residual hydrogen peroxide disappears. The pH can be reduced with either sulfuric acid or sulfur dioxide. When sulfur dioxide is used, the residual hydrogen peroxide is also destroyed according to Equation 3. H 2 0 2 +S02 =====> H 2 S0 4 (3) When sulfuric acid is used to adjust the pH, the residual peroxide remains in the stock. These small amounts of peroxide do not interfere with subsequent paper-making operations. A .Chelating Agents The various heavy metals, mentioned above, deplete hydrogen peroxide, reduce brightness gains and increase brightness reversion. Their detrimental effects can be subdued by the addition of chelating agents, such as STPP (sodium tripolyphosphate), TSSP (tetrasodium pyrophosphate) or the more costly sodium salts of EDTA (ethylenediamine tetraacetic acid) and DTPA (diethylene triamine pentaacetic acid). The addition of 0.1 - 0.5 wt % DTPA on OD (oven dry) pulp, for example, produces higher brightness by 0.5 - 1% Iso [11]. Although the amount of required chelant depends on the metal concentration, as a rule of thumb, approximately 0.25% of DTPA 20 by weight (based on OD pulp) is added to the solution prior to the addition of hydrogen peroxide. To maximize the effectiveness of chelating agents, it is best to allow retention time for metal chelation to occur [12]. B. Sodium Silicate Sodium silicate is added to the hydrogen peroxide brightening liquors as a source of alkali and as a buffering agent. It also stabilizes hydrogen peroxide in the presence of heavy metals as it forms insoluble heavy metal silicates or adsorbs these metals in calcium and magnesium silicate floes [35]. The addition of silicate to peroxide solutions in concentration ranging from 1 to 5% by weight on OD pulp also appears to increase the level of brightness. The mechanism of this function is not well understood. C. Magnesium Sulfate Small amounts of magnesium sulfate may also be added to brightening liquors when the mill/process water is not sufficiently hard. The magnesium acts as another stabilizer. It may form insoluble floes by adsorbing or co-precipitating heavy metal ions. Usually, about 0.05% weight percent magnesium sulfate (based on OD pulp) is added to the brightening liquor. 21 2.7.7. Fiber Properties Resin removal is important in brightened TMP used for tissue and absorbency products. Effective peroxide treatment of pulp will reduce the resin and fatty acid content from greater than 1% to less than 0.3%. Peroxide brightening increases fiber flexibility, improving the tensile index by about 25%. Fiber to fiber bonding is increased, reducing the amount of dust produced during the tissue manufacturing and converting operations. Peroxide brightened pulp can be stored several months without causing disintegration problems. 22 2.8 Electrochemical Production of Hydrogen Peroxide 2.8.1 General Electrochemistry: A simple electrochemical system comprises two electrodes and an electrolyte. The electrode at which oxidation occurs (eg. R => 0 + + n e\") is the anode and the electrode at which a reduction occurs (eg. 0 + + n e\" =>R) is the cathode. Electrochemical reactions involve mass transfer, adsorption/desorption, and electron transfer processes at the electrode/electrolyte interface. The rate of an electrochemical reaction is dependent on the electrode material, reactant concentration, electrode potential and temperature. Current efficiency of an electrochemical process can be calculated by Equation 4, given as below: CE=nFR/I (4) Where n = Electro-stoichiometry for desired reaction F=Faraday number (96480 kcoulombs/kmole) R=Production rate of desire product (kmole/s) I=Total current (kAmp) 23 2.8.2. In-situ Electrochemical Production of Hydrogen Peroxide In most systems, multiple reactions occur at a given electrode. For the interest of the present study, the electrochemical reactions of the cathode - - Reaction 5 & 6, the chemical reaction of the bulk solution - - Reaction 7, and the electrochemical reactions of the anode - - Reaction 9,10 are summarized below: Cathode Desirable Reaction: 02+H20+2e\" = > H02\" + OH\" Eo= -0.076 VSHE (5) AQS+ 2e\" + 2H+ ==>H2AQS E\u00C2\u00B0 = -0.66 VSHE (6) H 2AQS+0 2 ======> H 20 2+AQS (7) Undesired: H02\"+H20+2e\" ======> 3OH\" Eo=-0.878 VSHE (8) Anode: Major reaction: 02+2H20+4e\"<==== 40H\" Eo= +0.401 SHE (9) Undesired: 02+H20+2e\" <====H02\" +OH\" Eo=+0.15VSHE (10) The electrochemical reduction of oxygen in alkaline solutions was studied by Oloman et. al. in electrolysers with fixed and fluidized beds of graphite particles [22,23]. Although oxygen can be electrochemically reduced to hydrogen peroxide on the carbon cloth cathode via Reaction 5 [10], the electrochemical method of producing hydrogen peroxide is not a common commercial practice due to oxygen's limited solubility in aqueous media at STP within 1.1 mM. Several operating factors that may overcome a limited oxygen concentration in the solution are to be investigated in this research. These factors include choosing proper electrode materials, improving the mixing/reactor system and using redox mediating catalysts. Oxygen transfer in the electrolyte is facilitated by sparging it through the bulk electrolyte. Rapid mass transfer is achieved by the flow of bubbles purged through the bulk electrolyte generating turbulent behavior. The mass transfer of bulk solution to the cathode surface is promoted by 24 mixing. Finally, a porous carbon cloth/graphite felt were chosen to increase the reactive area and thereby suppress the mass transport limitation in the system. The electrochemical reduction of oxygen on carbon cloth occurs in two major steps. The transfer of the reactive gas at the gas (bubble)-liquid interface proceeds followed by the liquid (solubilized reactive gas)-solid mass transfer. For pure oxygen, the gas side of (inside the bubble) mass transfer resistance is usually negligible when compared to the other two resistance, i.e. gas-liquid and liquid-solid mass transfer resistances, mentioned above [36,6]. The undesired reaction on the anode, (reaction 10), which wastes the peroxide generated is suppressed by utilizing a separator such as a cation membrane or fritted glass or by reducing the size of the anode used. The separators suppress the diffusion and the migration of the desirable peroxide ions onto the anode surfaces, where undesirable oxidation of peroxide ions occurs. Similarly, a relatively small sized anode will also suppress the undesirable oxidation reactions. Furthermore, the maximum current that is suitable for an electrochemical cell operating only via the electroreduction of oxygen (reaction 5) is limited. Once the current exceeds an optimal level, (which is dictated by its,electrode potential, oxygen mass transfer and process conditions), the polarity of the electrode will become excessively negative so that peroxide will be further reduced to hydroxide via reaction 8. Because of this, a cathode with a large effective surface area is employed to lower the real current density. In the system proposed for the present study, hydrogen peroxide generation may be enhanced by combining the electrochemical reduction of oxygen on the cathode represented by reaction 5 and electro-reduction of AQS to H2AQS on the cathode (Reaction 6), followed by the H2AQS reaction with oxygen given by reaction 7. By adding AQS, a higher current density can be used and hydrogen peroxide is produced simultaneously by reaction 5 and by reaction 6 to 7. It is also important to note that the solubility of AQS in distilled water was experimentally measured to be in the range of 5-10mM at 60\u00C2\u00B0C, and thus, the maximum applicable concentration of the catalyst, AQS, is lOmM. 25 In summary, the effectiveness of this system is determined by the following two major rate determining factors: 1) The concentration of AQS in the electrolyte - - this fixes the AQS mass transfer rate and the limiting current density in reaction 6, which should be substantially higher than that used in Reaction 5 for useful results. 2) The rate of Reaction 7, which must produce AQS at a rate to match its use in Reaction 6. This rate is dependent on the intrinsic kinetics of Reaction 7 in addition to the rate of mass transfer of oxygen from the gas to the bulk electrolyte. 2.8.3. Electrode Materials The electrocatalytic properties of the electrode material are important in large, energy intensive electrolytic industrial process, in which the cost of power is a critical factor in determining process economics. The rate of electrocatalytic reaction can be increased by increasing the effective surface area of the electrode by using carbon cloth. Three-dimensional electrodes are used to increase the space-time yield of electrochemical reactors. Void fraction for electrical current flow in the electrolyte is increased by increasing the porosity of the electrode, which leads to a decrease in the effective conductivity of the electrode matrix. Good matrix conductivity, high internal surface area and high porosity are some of the basic pre-requisite for designing an effective three-dimensional electrode [39]. Supporting electrolytes which carry current but do not engage in electrode reactions are employed, if necessary, to increase the conductivity of the electrolyte. The conductivity of electrolyte solutions increases with increasing temperature. 26 The overall rate of industrial electrochemical processes is often under mass transfer control because the electrode material is chosen for fast electrode kinetics. The transport of the reactants and products between the electrode surface and the bulk electrolyte thus is a key factor which determines the process efficiency. Other important factors which affect the process efficiency include reactant concentration, mixing and the electrochemical reactor design. The mass transfer factor becomes even more critical in the industrial processes where high mass transfer rates are required to minimize capital investments. In those cases, large electrode surfaces are often employed to achieve the commercial production targets. 27 Chapter 3.0. - Proposed Research Objectives The proposed work involves the generation of hydrogen peroxide by the reactions shown in Reaction 5 and Reaction 6 to Reaction 7 (Section 2.8) and the brightening of mechanical pulp in-situ with a carbon fibre cloth cathode with pH in the range 10.5 to 11. No previous work has been done on in-situ hydrogen peroxide brightening of mechanical pulp by electroreduction of oxygen with an AQS redox mediators employing a graphite cloth cathode. Two different anthraquinone compounds, types of cathodes at various operating conditions and reactor configurations will be evaluated for peroxide generation. The potential benefit of such a system is the in-situ generation of relatively expensive hydrogen peroxide from oxygen in an alkali environment suitable for mechanical pulp brightening. The continuous supply of hydrogen peroxide prevents peroxide depletion ~ hence may generate higher brightness gains in mechanical pulp brightening. The proposed system also integrates two different processes, i.e. generating of hydrogen peroxide and mechanical pulp brightening, in a single operation. The work described in the following chapters was carried out to investigate the possibility of generating hydrogen peroxide in-situ to brighten mechanical pulp. The experimental plan encompassed examining the basic electrochemical behavior, the rate of hydrogen peroxide generation, and the brightening effect of TMP with the proposed AQS/Oxygen redox cycle. The experimental conditions: pH (10.5-11), temperature (60 \u00C2\u00B0C) and additives were chosen to approximate the conditions found in the conventional peroxide brightening procedures. The choice of the electrode materials was based on their availability and the experience of the project supervisor (Prof. C. Oloman). Other variables such as the current, mixing, and four (4) reactor designs were explored by trial and error. Finally, the pulp consistency was chosen at 1% to enable thorough mixing in the proposed reactors, while yielding enough pulp for subsequent brightness and yellowness analyses. 28 Chapter 4- Experimental Apparatus and Procedure Three major types of equipment employed in this study were cyclic voltammetry equipment, electrochemical reactors, and standard testing equipment for both the hydrogen peroxide analysis and mechanical pulp analysis (i.e. brightness and yellowness tests). 4.1 Cyclic Voltammetry (CV) The apparatus employed for the cyclic voltammetry study, also denoted as CV, consists of a Pine Potentiostat RDE 3 in conjunction with a Watanabe WX446 X-Y recorder. This system has three electrodes ~ working, counter and reference electrodes. A platinised titanium electrode served as a counter electrode with a saturated calomel electrode (SCE) as the reference electrode. The working electrode was a polyethylene tube wrapped around with 2 layers of 184 cm2 GC-12 carbon cloth. The equipment set up is illustrated below: Pine Potentiostat RDE 3 WE CE SCE Watanabe WX446 X-Y Recorder WE-Working Electrode CE-Counter Electrode SCE-Standard Calomel Electrode N 2 or 0 2 Figure 4.1. Cyclic Voltammetry Set-up 29 In order to reduce the ohmic potential drop between the working and reference electrode, the calomel reference electrode was placed in a Luggin capillary filled with saturated KC1 solution and having a porous plug at its tip. The Luggin capillary was placed a few millimeter away from the working electrode. To ensure the cleanliness of the electrodes, several pretreatments were carried out before the cyclic voltammetry tests. The working electrode, i.e. the carbon cloth electrode, was first placed in a weak HNO3, ( i.e. approximately 5 %), at room temperature for approximately 15 minutes followed by a thorough rinse with distilled water. Subsequently, the counter electrode, i.e. the platinized anode was pretreated by slow anodic scanned in 0.5M H 2 S O 4 solution, in order to oxidize the traces of impurities still present [17]. Finally, it was again rinsed with distilled water. The cyclic voltammetry measurement always commenced with the recording of the background current from the supporting electrolyte. This confirms the cleanliness of the electrode. The 1000 ml supporting electrolyte solution containing 0.1M Na2SC\u00C2\u00AB4, 4.7mM Na2SiC\u00C2\u00BB3 and 0.32mM MgSC>4 was then added. For all tests, the pH and the temperature of the solution were maintained at 10.5, and 60 \u00C2\u00B0C respectively. For the nitrogen runs, the oxygen was removed from the cell by purging nitrogen for approximately 5 to 10 minutes. The three electrodes were placed in the cell which was then closed and the potentiostat turned on to record the cyclic votammograms at a scan rate of lOOmV/s. The CV of anthraquinone 2 sulfonic acid on a GC-12 carbon cloth electrode was performed for Oxygen/5mM AQS, Nitrogen/5mM AQS, and Oxygen/OmM AQS at 60 \u00C2\u00B0C. 30 4.2. Experimental Procedures For the electrochemical generation of peroxide in absence of pulp, the stepwise procedure for the preparation of the bulk electrolyte solution are illustrated in step one and two. For pulp brightening experiments, please refer to step one through four. (Table 4.2): Step One 1. OraM AQS or 5 mM AQS, 0.1M of Na2S04, 4.7mM Na2Si03, 0.22mM DTP A and 0.32mM MgS0 4 were added to 1 liter of distilled water. 2. The solution was allowed to attain the temperature required in the brightening run. Step Two 1. The solution was placed in a 1000 ml beaker with oxygen or nitrogen purged at atmospheric pressure. 2. The solution is subjected to the desired current and thoroughly mixed for the reaction duration. Step Three a. in-situ brightening b: ex-situ brightening The unbrightened TMP pulp was placed in the beaker at 1% consistency. Current and oxygen flows were continued during the TMP brightening reaction. The current and gas were shut off when a desirable level of peroxide concentration had been reached. This brightening liquor was filtered before adding the pulp. Current and oxygen were not used during the TMP brightening reaction. Step Four: The pH of the brightened pulp was lowered to approximately 5-6 with a 15 minutes weak sulfuric acid rinse at room temperature before filtration. Table 4.2. Experimental Procedures 31 4.3 Electrochemical Reactors Hydrogen peroxide generation and in-situ pulp brightening were investigated using four different reactor designs, which included a small H-shaped reactor and an undivided reactor with three different configurations. 4.3.1. Small scale H-shaped divided glass reactor The small scale H-shaped, 250 ml, divided glass reactor, employed for the preliminary study of AQS, is shown in Figure 4.3.1.. Legend 1: Cathode 2:Anode 3: separator (e.g. membrane) 4. saturated calomel reference electrode (SCE) 5. magnetic stirrer Figure 4.3.1. H-Cell divided reactor 32 Due to small volume of this reactor, it was only used for the study of hydrogen peroxide generation. The cathodic chamber of the reactor has a 500 ml capacity and has an outer diameter and height of 6cm and 15cm respectively. The O.D. and height for the anodic chamber are 4 cm and 15 cm respectively. The cathode and anode are separated by a medium sized fritted glass with an O.D. of 2 cm. The purpose of the fritted glass is to suppress the transport of HO2\" and H2O2 to the anode and its destruction via Reaction 9 (Section 2.9). The cathode material used in this reactor was carbon cloth wrapped around a chemically inert polyethylene tube. The area of the platinized titanium anode was 0.0008m2 (with height of 5.7 cm and width of 1.4 cm). A DC power supply (Anatek 6030) was used for the experiments involving currents of up to and including 4.OA. Gas was purged in through a glass tube, which was connected to the source cylinder and a flow meter with Teflon tubing. The temperature of the system was regulated by placing the reactor in a constant temperature bath. 4.3.2. Pulp Brightening Reactors The electrochemical brightening was investigated in a 1000 ml glass beaker. The design of this reactor was modified three times in the attempt to increase the rate of hydrogen peroxide generation and thereby improve the brightening process. The three different set ups of the brightening reactor used for the generation of hydrogen peroxide and brightening runs are illustrated in Figure 4.3.2.1., Figure 4.3.2.2. and Figure 4.3.2.3.. 33 FLOW METER Figure 4.3.2.1. Pulp brightening reactor (configuration #1) 34 Figure 4.3.2.2. Pulp brightening reactor (configuration #2) 35 36 The reactor used for all three configurations was a 1000ml glass beaker (with a diameter of 12cm and height of 14cm). The anode material was platinized titanium, whereas the cathode material was either carbon cloth or carbon felt. A DC power supply (Anatek 6030), involving currents of up to and including 1.5 A, was connected to the carbon cloth/felt cathode by a single clamp. A variable speed motor with a three blade plastic stirrer provided the mixing during the experiments. Gas was purged into a glass tube connecting to the source cylinder and a flow meter with Teflon tubing. The temperature of the system was regulated by immersion of the reactor in a constant temperature bath. Reactor Configuration #1 In this reactor design, the cathode and the anode used have the same dimensions as those employed in the H-cell reactor. The cathode consisted of two layers of GC-12 carbon cloth with dimensions - - 23 cm by 8 cm wrapped around the diameter of the polyethylene tube; the anode was 5.7cm by 1.4 cm platinized titanium. At times, separators, such as cation membrane and fritted glass were placed between the anode and the cathode of reactor configuration #1. Reactor Configuration #2 & #3 For configurations #2 and #3, the carbon cloth was positioned along the wall of the glass beaker and a PVC screen, with a mesh size of 1.2 cm by 2 cm was bent to hold the carbon cloth on the wall of the beaker. The area of the anode was also reduced from 8 cm2 (5.7cm by 1.4cm) to 3.5 cm (2.5cm by 1.4cm). The modification in the cathode design shown in configurations #2 and #3 was to increase the actual cathode area to reactor volume ratio. As can be seen in Figure 4.3.2.2. and Figure 4.3.2.3., configuration #3 differs from #2 in the positioning of the mixer and anode. The mixer was inserted at a 30\u00C2\u00B0 angle so that it is directly beneath the anode. The purpose for this modification is that both the mixer and the anode be placed in the center of reactor to allow an even distribution of current as well as better mass transfer of reactants to the cathode. 37 4.4 Preparation of Chemicals and Pulp Anthraquinone 2 -sulfonic acid was obtained from the Aldrich Chemical company. The other chemicals used (sodium sulfate, sulfuric acid, magnesium sulfate, DTPA and sodium silicate) were obtained through BDH Chemicals. All chemicals were reagent grade and were used without purification. All solutions were prepared using 1 litre of distilled water. For all runs, 1 molar sodium hydroxide solution was added dropwise to bring the bulk 0.1M sodium sulfate solution to pH 10.5\u00E2\u0080\u009411. Subsequently, 0.32mM magnesium sulphate, 0.22mM DTPA, and 4.7mM sodium silicate were weighed and added to the bulk solution. The solution was then allowed to attain the temperature required in the brightening run. Due to its low solubility, all chemicals were mixed at an elevated temperature to ensure that the anthraquinone compound was fully dissolved in the bulk solution. The solution was then mixed in the glass beaker with 10 grams of oven dry thermo-mechanical pulp. The consistency of the pulp suspension was about 1 wt %. The pulp slurry was purged with gas (O2 or N2) at atmospheric pressure, mixed homogeneously and subjected to the desired current for the reaction duration. The pulp used in the experiments was obtained from previous researcher, Mr. Raj Ragapogal of the electrochemical brightening research group. The original properties were ISO brightness 44.7 (+/) 1 and R457 yellowness 33.3 (+/-) .6. The pulp was a mixture of 35% hemlock, 30% fir and larch, 15% cedar, 10% spruce and balsam fir and 10%) lodgepole pine. The pulp samples were prepared by weighing 48.5 (+/-)5 g of pulp from the refrigerator (=10 g of oven dry pulp). After the brightening run, the pulp was treated with weak sulfuric acid (0.01M) at room temperature for approximately 15 minutes. The treated brightened pulp was then thoroughly dispersed, rinsed and prepared as a handsheet in a handsheet former. 38 4.5 Hydrogen Peroxide Generation and TMP Brightening Hydrogen peroxide generation trials were conducted in all four reactor configurations with GC-12 carbon cloth or carbon felt cathodes and platinized titanium anode. The bulk solution consisted of a small amount of 1 molar NaOH solution (for pH adjustment), 0.1M sodium sulfate, 0.32mM magnesium sulfate, 0.22mM DTPA and 4.7mM sodium silicate. Due to solubility limitation, the maximum anthraquinone 2 sulfonic acid concentration that could be completely dissolved in distilled water at 60 \u00C2\u00B0C was approximately 5 ~ lOmM. The reactor was placed in the constant temperature bath and runs were conducted at 60 \u00C2\u00B0C and pH 10.5\u00E2\u0080\u009411. The bulk solution was sparged with oxygen at atmospheric pressure, mixed homogeneously and subjected to the desired current for the reaction duration. Samples of the electrolyte were analyzed for hydrogen peroxide concentration at regular intervals as detailed in Appendix IV. Following the in-situ generation of peroxide, the brightening reaction commenced by introducing 10 g (O.D.) of TMP pulp at 1% consistency into this peroxide enriched brightening liquor. Upon the termination of a brightening run, the pH of the electrolyte, containing brightened pulp, was adjusted to 5.5 - 6 by the addition of 0.01M sulfuric acid. Afterwards, the brightened pulp slurry mixture was removed from the reactor and transferred into a handsheet maker. Repeated washing processes were performed in a handsheet maker, in which brightened pulp slurry was subjected to thorough washing by distilled water. The cleaned pulp slurry, which may still contain some carbon fibres was detwatered, compressed into a handsheet and then air dried overnight in the constant humidity room. Ex-situ brightening differs from in-situ brightening in that the current and oxygen were discontinued and the actual brightening of the pulp was carried out with the filtered liquid in a separate 1000 ml beaker (Refer to page 31). A pressed handsheet was made as per CPPA standard C.5 [31] and brightness of the sheet was measured as per CPPA standard E.l [25]. 39 Chapter 5.0. -Experimental Results and Discussion 5.0 Introduction The objective of this work was to investigate anthraquinone-2-sulfonic acid (denoted as AQS) as a potential electrochemical mediator for the reduction of oxygen to hydrogen peroxide in the in-situ brightening of TMP pulp, at atmospheric pressure and in alkaline media. To achieve the proposed aim, the fundamental electrochemical behavior of the selected redox couple had to be scrutinized. The cyclic voltammetry and the I,E. vs. S C E - (current, potential vs. standard calomel electrode) measurements were performed and the results obtained for this purpose. Furthermore, a series of experiments on the electroreduction of oxygen mediated by AQS was performed and an experimental design was employed to evaluate the effects of key variables. A flowsheet of the overall experimental layout is summarized in Table 5.0.. 5.1 5.2 5.3 5.4 5.5 Examine the electrochemical behavior of AQS Study the electrochemical generation of hydrogen peroxide via electro-reduction of oxygen mediated by AQS Determine the brightening effect on TMP with the proposed redox cycle Study the factorial design on key parameters of the proposed redox cycle Study the aging of the carbon cloth electrode Table 5.0 Flowsheet of the Overall Experimental Plan 40 5.1 Electrochemistry of Anthraquinone-2-sulfonic acid: This section was to explore methods for improving the electro-reduction of oxygen mediated by the AQS redox couple at TMP brightening conditions. The H-cell electrochemical reactor (Figure 4.3.1), was employed to examine the cyclic voltammogram (CV) and current, electrode potential vs. SCE curves (I, E vs. SCE curves) of the proposed AQS system, as outlined below. Study of Cyclic Voltammetry(CV) A.) N2/5mMAQS B) Q2/5mMAQS Study of I,E vs. SCE Curve A) O2/0iriMAQS B) N2/5mMAQS C) Q2/5mMAQS Study of Bulk Electrolysis Table 5.1. Outline of Section 5.1 5.1.1 Cyclic Voltammogram (CV) Cyclic voltammetry is an effective tool for the investigation of electrochemical systems. Before proceeding with experimentation, CV is utilized to examine the ability of the GC-12 carbon cloth electrode employed in the brightening trials, to promote the reduction of oxygen as well as AQS. The ability of the specified electrodes to promote these reductions is important for the proposed cycle to engage in the continuous in-situ generation of hydrogen peroxide. CV can also verify the appearance of new species in the bulk solution ~ indicating the production of hydrogen peroxide. 5.1.1 5.1.2 5.1.3 41 A) Study of Cyclic Voltammetry fN7/5mMAOS) The CV obtained from the use of GC-12 carbon cloth in a N2/5mMAQS solution indicated a well defined redox couple, where couple I has Ei/ 2 equal to -0.66 V vs. SCE (Figure 5.1.1). < X-Axis P o t e n t i a l ( V vs. SCE) Figure 5.1.1.Cyclic Voltammogram for the AQS/HiAQS Redox Couple Formed in the Presence ofN? (Scale of the X-axis: 1 cm = -0.1 V; Y-axis: lcm = 1 * 10\"4 A) Operating Conditions:(5mM AOS. 60\u00C2\u00B0C, pH.10.5, Scan Rate lOOmV/s; Cathode: 2 layers of 23* 8 cm2 GC-12 Carbon Cloth (Working Electrode) = 368 cm2; Anode: 5.7 * 1.4 cm2 Platinized Titanium = 8 cm2, Additives: 4.7mM Na2Si03; 0.32mM MgS04; 0.1M Na2S04 0.22mM DTPA;) The CV shown in Figure 5.1.1, illustrates the well-known redox reaction that is typical of anthraquinone derivatives (Reaction 6 of Section 2.8). The two distinct peaks indicate that AQS is first reduced (reduction-upper peak), then the H2AQS is being oxidized (oxidation-lower peak). 42 B) Study of Cyclic Voltammetry rO?/5mMAOS) The CV obtained in an CVSmMAQS solution indicates that two reduction reactions have occurred (Figure 5.1.2.) .0 X-Axis: P o t e n t i a l (V vs. SCE) Figure 5.1.2.Cyclic Voltammogram for the AOS/HgAQS Redox Couple Formed in the Presence of 0 7. (Scale of the X-axis: 1 cm = -0.1 V: Y-axis: 1cm = 5*1(T6 A ) Operating Conditions:(5mM AQS, 60\u00C2\u00B0C, pH,10.5, Scan Rate 150mV/s, Cathode: 2 layers of 23*8 GC-12 Carbon Cloth (Working Electrode) =368 cm2; Anode: 5.7*1.4 cm2 platinized titanium = 8cm2 ; Additives: 4.7mM Na2Si03; 0.32mM MgS04; 0.1M Na2S04 0.22mM DTPA;) 4 3 The CV shown in Figure 5.1.2, illustrates two distinct cathodic waves (upper peaks). The first cathodic wave, shown at -0.46 vs. SCE is believed to correspond to the irreversible reduction of oxygen to hydrogen peroxide that is commonly observed in an oxygen-saturated solution. The second cathodic wave, shown at -0.78V vs. SCE, approximates that of the cathodic wave (AQS reduction wave) shown in Figure 5.1.1, which occurred at approximately -0.73 V vs. SCE. The difference, observed in the positions of cathodic waves for AQS reduction between Figure 5.1.1 and Figure 5.1.2. is -0.06V vs. SCE., and is most likely attributed to the use of different scan rates. The scan rates used for the N2/5mMAQS CV and 02/5mMAQS CV were lOOmV/s and 150 mV/s respectively. Since CV is a function of the scan rate, the CV may often shift slightly depending on the scan rates used in the CV tests. In comparison to the anodic waves (lower peak) presented in the nitrogen-saturated conditions (Figure 5.1.1.), the anodic curve for the oxygen-saturated condition was very small. This exceedingly small anodic peaks re-confirmed the reaction between oxygen and H 2AQS, (refer to reaction 7 of section 2.8.1). In other word, a large percentage of oxygen, in the bulk solution, is engaged in the continuous generation of hydrogen peroxide through its reaction with H 2AQS; therefore, lesser amount of H2AQS is available to becoming oxidized, resulting in a smaller anodic wave. Subsequently, the Ei/ 2 value for the 02/5mMAQS conditions cannot be accurately evaluated based on such a small anodic wave. 44 5.1.2. Current (I) vs. Electrode Potential (E) The importance of the current (I), vs. Electrode Potential (E vs. SCE) curve is that it provides a better understanding of the ability of a given electrode to promote the necessary reduction of the active species. It indicates the optimal range of current density that can be employed in the proposed system at specified operating conditions. The I, E vs. SCE measurements for O 2/0mMAQS, N 2/5mM AQS and 0 2 /5mM AQS were obtained and plotted in Figure 5.1.3.. 400 potentials in Volts, (V) Potentials in Volts, E Figure 5.1.3 I, E vs. SCE ;Current vs. the Potential of the Working electrode Operating conditions: (60 \u00C2\u00B0C, pH,10.5) (Cathode: 2 layers of 23*8 cm GC-12 Carbon Cloth = 368 cm2; Anode: 5.7 * 1.4 cm 2 - 8 cm 2 (Additives: 4.7mM Na 2 Si0 3 ; 0.32mM MgS0 4 ; 0.1M Na 2 S0 4 ; 0.22mM DTP A) 45 A) (VOmMAOS As illustrated Figure 5.1.3., the slope of the I-E curve in the 02/OmMAQS solution changed drastically around -1.0V vs. SCE. In the range of 0.0V to -1.0V, the potential decreases gradually with respect to the increasing current. When the potential is less than -1.0V, the slope is significantly reduced. This indicates that, within range, incrementing the current density continuously will result in a more rapid decrease in the electrode overpotential. This phenomena is attributed to the limit imposed by the mass transfer of oxygen to the electrode surface under the specified operating conditions. B) Ni/ 5mMAOS In the potential range from -0.6V to -1.5 V vs. SCE, the I-E curve for the N2/5mMAQS indicates that the electrode potential is directly proportional to the applied current. The potential, E , decreases with respect to the rising current. C) Q2/5mMAOS The I-E curve of the C>2/5mMAQS solution has characteristics that strongly resemble those seen for both the 02/OmMAQS and N2/5mM AQS solutions. As illustrated in Figure 5.1.3., the first part of the I-E curve, (i.e. approximately in the range from 0 to -.8V vs. SCE), the curves of 02/5mMAQS and 02/OmMAQS resemble each other, thereby suggesting the occurrence of oxygen reduction. As the electrode becomes more negatively polarized and its potential becomes more negative than -.8V vs. SCE, similar trends are observed for the I-E curves of the 02/5mMAQS and the N2/5mMAQS solutions rather than that of O2/0mMAQs. This similarity implies the reduction reaction ~AQS/H 2AQS predominates at potential below -0.8V vs. SCE. 46 5.1.3 Bulk Electrolysis Some interesting phenomena were observed during the electrolysis of the bulk solution. In these experiments, the H- cell Reactor (Figure 4.3.1.) was employed at similar operating conditions to those used in section 5.1.2 -- 60\u00C2\u00B0C; pH=10.5, 5mM AQS, 2 layers of 23 * 8 GC-12 carbon cloth working electrode = 368 cm2; 5.7 * 1.4 cm 2 platinized titanium anode = 8 cm 2 Additives: 4.7mM Na 2Si0 3; 0.32mM MgS0 4 ; 0.1M Na2SO4,0.22mM DTP A. A continuously increasing current was applied to the working electrode in the N 2/5mMAQS solution until the potential of the working electrode reached the potential of couple I ~ approximately -300 mV vs. SCE. A prominent color change in the bulk solution from pale yellow to burgundy red was observed after 15 minutes. This color change confirms the observations from the CV tests by indicating the electrochemical reduction of the anthraquinone (yellow) to hydroanthraquinone (red) at the cathode. The current and the nitrogen were then shut off followed by purging with oxygen. The color of the bulk solution returned to its original pale, yellowish color in approximately 50 minutes. The color restoration suggested that hydroanthraquinone has reacted with oxygen to form anthraquinone and hydrogen peroxide. More importantly, the rate of color change implied that the reaction of H 2 AQS with oxygen is the rate determining step for the in-situ generation of peroxide from oxygen mediated by the AQS/H 2 AQS redox couple. A small part of the solution was pipetted for the determination of the hydrogen peroxide concentration. The presence of hydrogen peroxide was proven. This corresponds to the results from CV tests and the current vs. the electrode potential curve, confirming, in accordance with the proposed redox theory, the reaction of the H 2 AQS with oxygen forming hydrogen peroxide and AQS.(Refer to Figure 5.1.1.) 47 5.2 Electrochemical Generation of Hydrogen Peroxide 5.2.0. Introduction The results presented in section 5.1. supported the proposed redox reaction. In section 5.2, experiments were designed to further examine the different methods for improving the electrochemical generation of hydrogen peroxide mediated by the AQS complex. Important variables such as the current density, effect of gas flow and its composition, reactor design, area of the cathode, material of cathode, and the aging of the carbon cloth were investigated in the absence of pulp, by trial and error, in six progressive stages. A flowsheet of the experimental plan is summarized in Table 5.2.O.. 48 Table 5.2.0. Outline Of The Six Progressive Stages Stage Areas of Studies Reactor Descriptions 5.2.1. Stage I A) Effect of AQS H - Cell Reactor Cathode: 2 layers of 23*8cm2 carbon cloth Anode: 5.7 * 1.4 cm platinized titanium B) Effect of doubling current density H -Cell Reactor Cathode: 2 layers of 23*8cm2 carbon cloth Anode: 5.7* 1.4 cm2 platinized titanium C) Effect of AQS vs. AQ2S H - Cell Reactor Cathode: 2 layers of 23*8cm carbon cloth Anode: 5.7 *1.4 cm platinized titanium 5.2.2 Stage II Preliminary Experiments Reactor Configuration #1 with cation membrane Cathode: 3 layers of 23*8cm2 carbon cloth Anode: 5.7*1.4 cm platinized titanium 5.2.3. Stage III Current Density Reactor Configuration #1 without cation membrane Cathode: 3 layers of 23*8cm2 carbon cloth Anode: 2.5*1.4 cm2 platinized titanium 5.2.4. Stage IV A) Effect of fritted glass separation Reactor Configuration #1 Cathode: 4 layers of 23 *8cm2 carbon cloth Anode: 2.5* 1.4 cm2 platinized titanium mixer B) Effect of carbon felt cathode Reactor Configuration #2 Cathode: 2 layers of 40*10 cm carbon felt Anode: 2.5* 1.4 cm2 platinized titanium mixer 5.2.5 Stage V Number of carbon cloth layers vs. rate of hydrogen peroxide production Reactor Configuration#2 Cathode: layers of 41*10 cm carbon cloth Anode: 2.5* 1.4 cm2 platinized titanium Cathode support: polyethylene 5.2.6. Stage IV Effect of AQS with Reactor Configuration #3 Reactor Configuration#3 Cathode: 3 layers of 41*10 cm2 carbon cloth Anode: 2.5*1.4 cm platinized titanium cathode support: polyethylene mixer inserted at 30\u00C2\u00B0 centering anode 4 9 5.2.1 Stage I Introduction The H - Cell reactor (Figure 4.3.1.) was employed in Stage I experiments to further validate the continuous in-situ generation of peroxide in the proposed electrochemical system with improved accuracy and increased current efficiency. The experimental plan in Stage I is summarized in Table 5.2.1. A) B) Q Determine the effect of AQS with a mixer and an additional layer of carbon cloth Determine the effect of doubling the current density Determine the effect of using AQ2S (Anthraquinone 2,6 disulfonic acid). Table 5.2.1. Flowsheet of Stage I 50 0 60 120 180 240 300 Time (min) Figure 5.2.1. Effect of 5mMAOS vs. OmM AOS on the rate of EbO? generation Operating Conditions: (60\u00C2\u00B0C; pH=10.5) (Additives:4.7 mM Na2Si03; 0.32 mM MgS04; 0.1 M Na2S04, 0.22 mM DTP A) (H - Cell Reactor - - Cathode: 2 layers of 23*8 cm2of GC-12 carbon cloth = 368 cm2; Anode: 5.7*1.4 cm2 platinized titanium = 8 cm2) 51 A)Effect of AQS As illustrated in Figure 5.2.1, the generation of hydrogen peroxide at the carbon cloth cathode, in the absence of AQS was evident, but the amount of hydrogen peroxide generated was small under the conditions employed. The hydrogen peroxide concentration obtained at 300 minutes for the O2/0mMAQS/0.15Amp run (20mM - - Table 5.2.2. of Appendix V), which was approximately 30% that of the O2/5mMAQS/0.15Amp run (67mM- - refer to Table 5.2.1.of Appendix V). This concurs with the findings of the CV and the I, E vs. SCE curves - - in addition to electroreduction of 0 2 to H 2 0 2 , AQS is also engaged in generating hydrogen peroxide. In accordance to the proposed theory, AQS is first reduced to H 2AQS, which then reacts with oxygen, yielding both peroxide and AQS. In this way, AQS is recycled and re-engaged in the generation of peroxide. 52 B) Effect of Current Density In this section, the importance of current density was investigated. 0 60 120 180 240 300 Time (min) Figure 5.2.2 Effect of Current Density on the Rate of Hydrogen Peroxide Generation Operating Conditions: (60\u00C2\u00B0C; pH=10.5; 5mM AQS) (Additives: 4.7mM Na2Si03; 0.32 mM M MgS04; 0.1 M Na2S04, 0.22mM DTP A) (H - Cell Reactor -- Cathode: 2* layers of 23*8cm2of GC-12 carbon cloth = 368 cm2; Anode: 5.7* 1.4 cm platinized titanium = 8 cm ;) 53 From Figure 5.2.2., the importance of current density was confirmed. The results, as expected, indicated that the rate of hydrogen peroxide production was significantly reduced when the current was halved. In 300 minutes, the concentration of hydrogen peroxide obtained in O2/5mMAQS/0.08Amp run was 19% less than that of O2/5mMAQS/0.15Amp.( Refer to Table 5.2.3. and Table 5.2.1. of Appendix V). In contrast to the rate of peroxide production, the current efficiencies at higher current density are consistently lower than those obtained at a lower current density. Therefore, a compromise must be made between the production rate and the current efficiency of hydrogen peroxide generation when choosing an optimal current density. O The Effect of Anthraquinone 2,6 disulfonic acid Table 5.2.4 of Appendix V and Figure 5.2.3. presents the results showing the effect of anthraquinone2,6 - disulfonic acid - - denoted as AQ2S, on the rate of hydrogen peroxide generation. Note that the concentration of anthraquinone 2,6 disulfonic acid used was half that of the usual AQS concentration due to its low solubility at the operating conditions. The maximum anthraquinone 2,6 disulfonic acid that could be completely dissolved in the bulk solution was 2.5 mM. 54 0 60 120 180 240 300 Time (min) Figure 5.2.3. Effect of AQ2S vs. AOS on the Rate of Hydrogen Peroxide Generation Operating Conditions: (60\u00C2\u00B0C; pH=10.5; 2.5mM AQ2S) (Additives: 4.7 mM Na2Si03; 0.32 mM MgS04; 0.1 M Na2S04, 0.22mM DTP A) (H - Cell Reactor - - Cathode: 2*23*8cm2of GC-12 carbon cloth =368 cm2; Anode: 5.7*1.4 cm2 platinized titanium =8 cm2) 55 In comparison to runs with no anthraquinone complex (refer to O2/0mMAQS/0.15A run of Figure 5.2.1), the results from Figure 5.2.3. indicate that the rate of hydrogen peroxide generation was enhanced by the addition of an alternative anthraquinone derivative ~ i.e. anthraquinone 2,6 disulfonic acid. At 300 minutes, a 49% increase in the concentration of hydrogen peroxide was observed in the presence of the anthraquinone 2,6 disulfonic acid than in the absence of anthraquinone compounds. Although the results confirmed that anthraquinone 2-6 disulfonic acid promotes the hydrogen peroxide generation, its use in the proposed system is limited. In addition to its low solubility, the use of anthraquinone 2,6 sulfonated complex in the brightening liquor will dye the pulp red. Because of these two disadvantages, the possibility of using anthraquinone 2,6 disulfonic acid as an alternative hydrogen peroxide promoter was ruled out. 56 5.2.2 Stage II Preliminary Investigation The employment of the 250ml H - Cell reactor of Section 5.2.1. yielded promising peroxide generation results. However, this H - Cell Reactor was not suitable for pulp brightening trials due to the its size constraint. In Stage II, experiments were conducted to examine the rate of hydrogen peroxide generation in a scaled-up O2/AQS electrochemical reactor - - Reactor Config. #1 . (Refer to Figure 4.3.2.1.) Increasing the volume of the reactor was an essential step before attempting any brightening trials. If the volume of the reactor is too small, the brightened pulp retrieved from the reactor will not be sufficient for standard paper testing. It is important to note that the only source of mixing in reactor configuration #1 was provided via the purging of oxygen gas through a glass gas sparger. In addition, a cation membrane was employed to substitute the anodic/cathodic separator of the H - Cell reactor, so that the transport of peroxide onto the anode surface was minimized. Finally, an additional layer of carbon cloth , i.e. (23*8 cm2) was used on the carbon cloth cathode in an attempt to increase the electro-active area of the carbon cloth electrode. 57 20 Current efficiency (%1 vs. time 0 60 120 180 240 300 Time (min) Figure 5.2.4. The Rate of H7O7 Generation in a Scaled-up Reactor Operating Conditions:(60\u00C2\u00B0C; pH=10.5, 5mM AQS, Additives: 4.7mM Na 2 Si0 3 ; 0.32 mM MgS0 4 ; 0.1 M Na 2 S0 4 ; 0.22mM DTPA)( Reactor Config.tfl- - Cathode: 3 layers of 23*8 cm 2 GC-12 carbon cloth = 552 cm2; Anode: 5.7*1.4 cm 2 Platinized titanium = 8 cm2; Separator = cation membrane) Due to the IR drop induced by the presence of the cation membrane, the highest current that could be applied in the Stage II experiments, without exceeding the desirable range, of potential, was 0.2 A. Figure 5.2.4. indicates that the concentration of the H 2 0 2 produced increased over time in the solution that contained 0 2/5mMAQS at 0.2A. The current efficiency remained above 88% towards the end of the 300 minutes run, suggesting that the cation membrane may have successfully minimized the destruction of peroxide on the anode. 58 5.2.2. Stage III Effect of Cation Membrane and Current Density The cation membrane, used in the Stage II experiments, may have successfully suppressed the oxidation reaction of hydrogen peroxide by the anode (Reaction 10 of Section 2.8). However, it did so at the cost of a lower production rate of hydrogen peroxide. Due to the inherently high IR drop of the cation membrane separator, the maximum current that could be applied to Reactor Config.#l without exceeding the desirable potential range - - (between 0 V and -IV vs. SCE), was 0.2 A.. This low current density translated to a low rate of peroxide generation. In Stage III, the importance of current density on the rate of peroxide production was examined with Reactor Configuration #1, without cation membrane. To counterbalance the effect of the cation membrane, the size of the platinized titanium anode was reduced from 5.7 * 1.4 cm 2 to 2.5 * 1.4.cm2. As expected, upon discarding the cation membrane, the applicable current, without exceeding the desirable range of the potential, rose from 0.2 Amp to 0.42A and subsequently, from 0.2 A to 0.7Amp. The results of the O2/5mMAQS/0.42A. and the O 2/5mMAQS/0.7A. experiments are illustrated in Figure 5.2.5 and summarized in Tables 5.2.6.and 5.2.7. of Appendix V respectively. 59 30 0 60 120 180 240 300 Time (min) Figure 5.2.5. Effect of Current Density on the Rate of Peroxide Generation Operating Conditions: (60\u00C2\u00B0C; pH=10.5, 5mM AQS, Current: .0.42A vs. 0.7 A) (Additives:4.7 mM Na2Si03; 0.32 raM MgS04; 0.1 M Na2S04, 0.22mM DTPA); (Reactor Config.#l - - Cathode: 3 layers of 23 *8 cm2 GC-12 carbon cloth = 552 cm2; Anode: 2.5*1.4 cm2 platinized titanium; in the absence of cationic membrane) Figure 5.2.5. indicate that increasing the current density raised the rate of hydrogen peroxide generation, but lowered the current efficiencies. For example, the concentration of hydrogen peroxide obtained for the 0.7A. run was 16% higher than that of the 0.42 A. run, at 300 minutes. On the contrary, the current efficiency obtained for 0.7A was 8% lower than that of the 0.42 A. run at 300 minutes. (Refer to Table 5.2.6. and Table 5.2.7. of Appendix V.) 6 0 5.2.4. Stage IV Introduction In Stage IV, a mixer was incorporated into reactor configuration #1 (Figure 4.3.2.1.) to aid both the gas-liquid and liquid/electrode mass transfer, the current distribution of the electrochemical cell, and later, the dispersion of pulp. In addition to the mixer, a cathode with 4 layers of (23*8 cm2) carbon cloth was employed in Stage IV. The flowsheet of section 5.2.4 is summarized as follows: A) Determine the effect of medium sized fritted glass B) Investigate another cathode material: carbon felt Table 5.2.2. The Flowsheet of Experimental Plan, Stage IV A) The Effect of Fine & Medium Sized Fritted Glass Separator Experiments were conducted to investigate the effect of using both fine and medium sized fritted glass separators. Similar to the function of the cationic membrane, the fritted glass was used to inhibit the oxidation of hydrogen peroxide on the anode. The fritted glass separation would hopefully create an anodic/cathodic barrier to impede the transport of peroxide onto the anode surface and thereby increase the net production of hydrogen peroxide. For the fine pored separator run, AQS seemed to have crystallized and blocked the fine pores of the separator within the first half hour. Because of this blockage of the pores, the IR drop was significantly increased so that the maximum current that could be applied, plunged to approximately 0.02Amp within the first half hour A similar, but not so severe trend was observed in the medium sized fritted-glass separator. As shown in Table 5.2.8, of Appendix V, the maximum current that could be applied to the system in the presence of the medium sized fritted-glass separator was 0.2 Amp. 61 0 60 120 180 240 300 Time (min) Figure 5.2.6. The Effect of a Medium Sized Fritted Glass on the Rate of Hydrogen Peroxide Generation Operating Conditions: (60\u00C2\u00B0C; pH=10.5; 5mM AQS) (Additives: 4.7mM Na2Si03; 0.32 mM MgS04; 0.1 M Na2S04; 0.22mM DTPA) (Reactor ConfigJl - - Cathode: 4 layers of 23* 8 cm2of GC-12 carbon cloth =552 cm2; Anode: 2.5*1.4 cm2 platinized titanium =3.5 cm2) From Figure 5.2.6. it is evident that the presence of the fritted-glass separator improved both the rate and the current efficiency of peroxide production . Similar to the theory of the cation 62 membrane, the fritted-glass separator was used to impede the transport of H2O2 to the anode surface, which , in turn, allows a higher yield of H2O2 and current efficiency. Unfortunately, the medium sized fritted-glass separator was also found to be unsuitable for the proposed cell. Apart from current limitation problems, fouling posed an important concern in the use of the medium sized fritted-glass separator. The AQS crystallized and blocked the pores, even in the medium sized separator after a few runs, despite several attempts to thoroughly clean the pores, and as a result, the IR drop and the current limitation problems drastically worsened. Improvements and modifications, other than the use of the cation membrane or the fritted glass separator must be implemented in the existing apparatus used in this section (Figure 4.3.2.1.) before commencing pulp brightening trials. 63 B) The Effect of Carbon Felt To improve the reactor design, an alternative source of material was tested for the working electrode. Carbon felt has long been used as an ideal electrode for ex-situ electroreduction of oxygen to hydrogen peroxide in a strong alkaline solution. It is a cheap and stable cathode on which the intrinsic kinetics of oxygen reduction allow high selectivity for the formation of perhydroxyl anions, i.e. HO2\". It was used in experimental Stage IV- section B, instead of the carbon cloth electrode. The anode remained at a reduced size (2.5*1.4 cmz). One of the most important features that may make carbon felt a more attractive source of cathode material than carbon cloth is its high porosity. Carbon felt has approximately 95% porosity, whereas carbon cloth has only approximately 70% porosity. Because of its high porosity, carbon felt yields more accessible reactive area. In addition, the sturdiness of the carbon felt enabled it to be positioned along the wall of the beaker, yielding a much larger exposed area. As illustrated in the undivided reactor configuration #2 of Figure 4.3.2.2., approximately 376cm of exposed area was obtained through this new arrangement of the cathode, which is nearly five times that of the tube carbon cloth cathode As a result of this increase in the exposed cathode area, the maximum current was increased from 0.3A (Table 5.2.9) to 1 A. (Table 5.2.10). 64 0 30 60 90 120 Time (min) Figure 5.2.7. Effect of Carbon Felt on the Rate of Peroxide Production Operating Conditions: (60\u00C2\u00B0C; pH=10.5; 5mM AQS ) Additives: (4.7mM Na2Si03; 0.32mM MgS04; 0.1 M Na2S04, 0.22mM DTPA); ( Undivided Cell - - Reactor Config.#2; Cathode: 2 layers of 40*10 cm2 carbon felt = 800 cm2 Anode: 2.5 *1.4 cm2 platinized titanium = 3.5 cm2) 6 5 As can be seen in Figure 5.2.7., increasing the exposed electro-active area of cathode has enabled the electrochemical cell to be operated at a higher current and thereby increasing the rate of hydrogen peroxide generation. Thus, a significant reduction in the H2O2 production time was achieved through this reactor design modification. Unfortunately, the increased current also caused a more rapid decline in current efficiency. An in-situ, three hour TMP brightening experiment was performed with the brightening liquor obtained at 120 minutes from the carbon felt run shown in Figure 5.2.7. - (Refer to Table 5.2.10 of Appendix V). Through the course of the three hour brightening run, while continuously supplying current and gas, a significant part of the carbon felt disintegrated into fine fibers which then attached and integrated with the pulp matrix. These dark fibers certainly discolored and darkened the pulp. As a result of the darkening problem, the measured pulp brightness and yellowness were not accurate and therefore not reported in this study. Due to the fibre darkening problems, it can be concluded that the carbon felt is not suitable to be employed in the in-situ brightening of pulp. Although the use of carbon felt in the in-situ brightening of TMP proved to be impractical, these experiments introduced a novel concept for increasing the exposed electro-reductive area of a carbon cloth cathode. 66 5.2.5. Stage V The Stage V experiments were carried out to determine the number of carbon cloth layers (41 * 10 cm2 ) required to raise the rate of the hydrogen peroxide production. The reactor design for Stage V experiments arose from the concept of the carbon felt cathode arrangement of Stage IV-B. As illustrated in Figure 4.3.2.2., the carbon cloth cathode was placed along the wall of the reactor to allow its maximum exposure to the bulk electrolytes. An inert, perforated, polypropylene cathode support was employed to provide carbon cloth with a skeletal backbone and to prevent the carbon cloths from being entangled in the mixer. This cathode support also minimized the contact between the pulp and the carbon cloth and thereby reducing the dark blemishes in the brightened pulp. It is important to note that availability, cost and pore-sizes were considered in choosing the cathode support. If the pores are too small, as in the case of the fritted glass, it will act as a current flow barrier, resulting in a significant increase in the IR drop of the electrochemical cell. The cathode support used for the Stage V experiments was obtained from the plastic (polypropylene) wrapping used around conventional gas cylinders, which is widely available in any small chemical laboratory. The pores for the chosen plastic support are moderate in size ~ 1.2 * 2 cm diamond shaped and therefore would not substantially raise the IR drop of the electrochemical cell. 67 0 2 4 6 number of carbon cloth layers Figure 5.2.8. Number of Carbon Cloth Layers vs. the Rate of H2O2 Production Operating Conditions: (60\u00C2\u00B0C; pH=10.5; 5mM AQS; 3 hours; in the absence of pulp) Additives: (4.7mM Na2Si03; 0.32mM MgS04; 0.1 M Na2S04; 0.22mM DTPA) (Undivided Cell- - Reactor Configuration #2; Cathode: 1,2,3,5 or 6 layers of 41* 10cm2 carbon cloth; Anode:2.5*1.4 cm2 platinized titanium = 3.5 cm2) 68 Figure 5.2.8. indicates that using three layers of carbon cloth (41*10cm2) yielded the best results in an 180 minute run without pulp. Beyond or below three layers, the production of hydrogen peroxide waned. The results implies that the electro-active area of electrodes can be increased up to 3 layers of carbon cloths. Beyond three layers, increasing the layers of carbon cloths seemed to adversely affect the rate of peroxide generation. This negative effect suggests that the peroxide generated may have being trapped between the carbon cloth fibers, and was therefore subjected to further electro-reductive reactions inside of the carbon cloth ~ i.e. H2O2/H2O. Increasing the layers of carbon cloth electrode may have also suppressed the convection and the mass transport of AQS and oxygen to the active face of the electrode, resulting in a reduced rate of hydrogen peroxide production. 69 5.2,6. Stage VI Final Modifications in Reactor Design As can be seen in the reactor #3 - - Figure 4.3.2.3., both the positions of mixer and the anode were placed at the center of the reactor. This rearrangement in the reactor design would hopefully improve mixing, and therefore the current distribution and the mass transfer in the electrochemical cell. 60 n 1 0 20 40 60 80 100 120 140 Time (min) Figure 5.2.9. The Rate of H 7 Q 7 Generation vs. Time (Reactor Config.#3) Operating Conditions: (6Q\u00C2\u00B0C; pH=10.5; 5mM AQS; Current: 1.5 A); Additives: (4.7mM Na2Si03; 0.32mM MgS04; 0.1 MNa 2S0 4; 0.22mM DTP A); (Undivided Cell - - Reactor Config.#3; Cathode: 3 layers of41*10*cm2 carbon cloth = 1280 cm2; Anode: 2.5*1.4 cm2 platinized titanium =3.5 cm2) As a result of centering the anode and modifying the mixing position, the current distribution for Figure 4.3.2.3. has been improved. The IR drop is much lower in this experimental design. Hence, a higher current ~ 1.5 Amp ~ could be applied. 7 0 As shown in Figure 5.2.9., the rate of hydrogen peroxide generation was dramatically improved. Up to 49.5(+/-)2.5 mM of peroxide was produced in 2 hours by electroreduction of oxygen in a pulp brightening liquor (free of pulp) containing 02/5mMAQS at 1.5 Amp. The current efficiencies were well above 85%. (Refer to Table 5.2.12.of Appendix V) These results showed a significant improvement over those shown in Table 5.2.11. of Appendix V for Stage V, in which the maximum amount of peroxide obtained in 180 minutes was 35 (+/-)2.5 mM.. The amounts of hydrogen peroxide generated in Stages V and VI are ample for brightening 1 % consistency pulp at a 15wt % hydrogen peroxide charge. In the following sections of this thesis, the brightening of TMP will be examined. 71 5.3 Electrochemical Brightening of Pulp 5.3.0 Introduction The results of the cyclic voltammetry test, and I-E curve analysis indicate the ability of the proposed system to promote the reduction of AQS to H2AQS. The experimental results showed that H2O2 would be continuously generated in such a system via the proposed AQS-complex/oxygen redox couple. In particular, as much as 49.5mM(+/-)2.5mM of hydrogen peroxide concentration was obtained in two hours in the experiment of section 5.2.6. A brightening liquor containing 49.5mM(+/-)2.5mM of hydrogen peroxide is equivalent to brightening of TMP at 1% consistency at 15 % H 2 0 2 charge. The following section describes the ability of the proposed AQS complex/oxygen redox cycle to brighten TMP. Some key variables with respect to this process will also be further investigated. The experimental plan for this section is illustrated in Table 5.3.0. Section 5.3.1. Section 5.3.2. Section 5.3.3. Section 5.3.4. Section 5.3.5. Hydrogen peroxide charges vs. brightening effects Time vs. brightening effects Effect of carbon residuals from carbon cloth cathode Ex-situ brightening vs. in-situ brightening Brightening effects of the proposed system vs. merchant H2C>2 Table 5.3.0. Flowsheet of the Overall Experimental Plan, Section 5.3 Note: In all the experiments of section 5.3., brightening conditions were standardized and summarized as follows: Operating Conditions: (Temp=60\u00C2\u00B0C; pH=10.5; Current: 1.5 A.; Q2 purged at 90ml/min @ STP) Additives: (5mM AQS, 4.7mM Na2Si03; 0.32mM MgSQ4; 0.1 M Na2SQ4, 0.22mM DTPA) Cathode: 3 layers of (41*10 cm2)carbon cloth; Anode: 2.5*1.4 cm2 platinized titanium) 72 5.3.1. Hydrogen peroxide charges vs. brightening effects Hydrogen peroxide is a lignin preserving, oxidative brightening agent. It will brighten mechanical pulp without significantly decreasing the pulp yield. In-situ brightening of mechanical pulp involves a continuous generation of hydrogen peroxide which reacts with the lignin of pulp in the electrochemical cell. In the preliminary investigations, the brightening effects vs. H2O2 were scrutinized. From the results of the preliminary investigation shown in Table 5.3.1 (Refer to Appendix VI), the significance of the initial H2O2 concentration in the brightening process is clearly demonstrated. Higher brightness gains and yellowness reductions were obtained through employing a higher initial concentration of H2O2 brightening liquor. Results from Table 5.3.1. also suggests that alkaline darkening would occur in the case of insufficient initial hydrogen peroxide concentration. For example, in run Dl , the brightness of pulp increased by 2.3 ISO(%), whereas the yellowness worsened by 0.8 %(R457), i.e. the yellowness of the pulp was 0.8% (R457) higher than the original pulp. The results from run Dl imply that, at a low initial hydrogen peroxide concentration, the brightening effect of hydrogen peroxide is suppressed by the alkaline darkening effect. In contrast to run Dl , the brightening results from run CI are much superior. It is therefore important to generate ample hydrogen peroxide in the absence of pulp before commencing the brightening process. Table 5.3.1. Initial Hydrogen Peroxide Concentration (prior to addition of pulp) vs. In-situ Brightening Effects in 3 hours KI N Initial Concentration of H A ( m M ) Brightness Gain %(ISO) Yellow ness Changes %(R45\"7) CI 35 +13.7 -12.1 1)1 14 +2.3 +0.82 Additional Operating Conditions: 1) (Generation of peroxide in the absence of pulp: 3 Hours; In-situ Brightening of Pulp:3 Hours ) 2) The Brightness/Yellowness for the original pulp is 44.7 (+/-)1 %(ISO) and 33.3 (+/-) .6% (R457). 3) Note: The yellowness achieved in Dl was worse than that of the original pulp; the yellowness measurement increased by 0.82% (R457) from that of the original pulp. 73 5.3.2. Ex-situ Brightening vs. In-situ Brightening The following experiments examine the effect of reaction time upon brightening for both in-situ and ex-situ brightening. Experiments were performed separately at 1,2, 3 and 5 hour for ex-situ brightening method and at 1, 2 and 3 hour for in-situ brightening method. All the experiments present in section 5.3.2. were repeated once to ensure the accuracy of the experimental data. The procedures for both the in-situ and the ex-situ brightening process are illustrated Table 5.3.2.. A) A 02/5mMAQS/l .5Amp run was carried out for 2 hours in the absence of pulp, till the hydrogen peroxide concentration reached approximately 49.5 (+/-) 2.5 mM. B) This 49.5(+/-)2.5mM H 2 0 2 brightening liquor was filtered before the addition of pulp. C) During the three hour 1% pulp brightening process, current was off. Oxygen and mixer were on. In-situ B) The pulp was added into this 49.5(+/-)2.5 mM H 2 0 2 brightening liquor without any filtration. C) During the three hour 1% pulp brightening process, current, oxygen and mixer were kept on. D) After 3 hours, the brightened pulp was retrieved and washed in 0.01 M sulfuric acid. E) Pulp was made into a handsheet and air dried. Table 5.3.2. Procedures for Both In-situ and Ex-situ Brightening 74 A) Ex-situ Brightening The results of time vs. ex-situ brightening effects are summarized in Table 5.3.2.of Appendix VI and illustrated in Figure 5.3.1.. Figure 5.3.1. Brightening Gains (% ISO) and Yellowness Reduction (%) of Ex-situ Brightening vs. Time Operating Conditions: \u00E2\u0080\u00A2 Ex-situ Brightening of Pulp: 1, 2 , 3 & 5 Hours \u00E2\u0080\u00A2 The Brightness/Yellowness for the original pulp is 44.7 (+/-)1%(IS0) and 33.3 (+/-) .6%(R457). \u00E2\u0080\u00A2 Please refer to section 5.3.0. for details of the standard operating conditions. 7 5 Figure 5.3.1. or Table 5.3.2. (Appendix VI) presents the changes in brightness gains and yellowness reductions over a 5 hour period for ex-situ brightening. These results indicate an initial sharp rise in both brightness gains and yellowness reductions in the first and the second hour reaction followed by a moderate leveling off effect, in the 2-3 hour interval. Beyond 3 hours, however, the brightening effect of hydrogen peroxide declined sharply. As a result, there was little incentive in continuing the reaction beyond 3-5 hours. There are several reasons that may explain the retardation in the brightening effect during 3-5 hour period. First, hydrogen peroxide is not very stable. It decomposes easily in alkaline conditions and beyond a certain time, little is left. Another factor that may explain the decreasing brightening effect is the consumption of hydrogen peroxide upon its reaction with the pulp. Since hydrogen peroxide is not being generated continuously, i.e. ex-situ brightening, the active hydrogen peroxide may have been depleted over time. Finally, similar to conventional hydrogen peroxide brightening, it may be that hydrogen peroxide is incapable of increasing the yellowness reduction beyond a certain value. Those specific values represent the limit of reaction under the conditions described. 76 B) In-Situ Brightening Figure 5.3.2. and Table 5.3.3 (see Appendix VI) indicate the increases in brightness gains and yellowness reductions for in-situ brightening runs at 1, 2 and 3 hours. In a three hour run, the brightness of pulp was increased by 13.9(+/-) 1%IS0 while the yellowness was lowered by 12.3(+/-)0.6 %. Figure 5.3.2. Brightening Gains (% ISO) and Yellowness Reduction (%) of In-situ Brightening vs. Time Operating Conditions: \u00E2\u0080\u00A2 In-situ Brightening of Pulp: 1, 2 & 3 Hours \u00E2\u0080\u00A2 The Brightness/Yellowness for the original pulp is 44.7 (+/-)1 % (ISO)and 33.3 (+/-) .6 % (R457). \u00E2\u0080\u00A2 Please refer to section 5.3.0. for details of the standard operating conditions. 77 Similar to the ex-situ brightening runs, the most significant brightening effect occurred during the first hour. The results show an initial sharp rise in both brightness gains and yellowness reductions during the first hour. Nevertheless, in contrast to what was expected, the in-situ brightening effect diminished much faster than that of the ex-situ brightening. Figure 5.3.2 indicates that the slope of the in-situ brightening effect reduces drastically beyond one hour and thus, there was little incentive to continue the reaction beyond 1 hour. For this reason, a five hour run was not employed for in-situ brightening. The rapid decline of the in-situ brightening effect beyond the first hour period can be attributed to several factors. Again as mentioned earlier, hydrogen peroxide is not very stable. It decomposes easily in alkaline conditions and can be depleted easily upon its reaction with the pulp. But more importantly, hydrogen peroxide, may not have been generated in-situ continuously, as expected, due to poor mixing. The mixing system employed in this part of investigation was primitive, and most likely - -inadequate. In fact, since the surface of the carbon cloth would be much coarser and rougher than that of the glass beaker, a much more powerful mixing system is required to overcome the friction force created between the carbon cloth surface and the swirling pulp slurry. Often, some pulp was found to be caught in between the carbon cloth cathode and the cathode support, thus creating mixing dead zones. As a result, the electro-active area of the carbon cloth may have been blocked by stagnant pulp chunks, which in turn, hindered the continuous, in-situ generation of hydrogen peroxide. Meanwhile, the already generated peroxide may have also been trapped inside the carbon cloth, and thus is subjected to further reduction to water, inside the cathode. For these reasons, the in-situ brightening method was found to yield inferior brightening results to those obtained through ex-situ brightening method. 78 In summary, the results from section 5.3.2. indicate that both in-situ and ex-situ brightening give promising brightening effects. Although, the difference in the yellowness reductions of the two methods were insignificant, the ex-situ brightening method seems to bring about more brightening gains than the in-situ brightening method. The brightness gain obtained for ex-situ brightening was approximately 4 points higher than that of the in-situ brightness gains. More importantly, the physical appearance of the pulp was dramatically improved in the ex-situ brightening runs due to the filtration process, which removed the residual carbon fibers from the brightening liquor. From the brightness gains as well as the physical appearance of the resultant brightened pulp, it can be concluded that the ex-situ brightening with filtration method is superior to the in-situ brightening method . 7 9 5.3.3 Effect of Residual Carbon Fibers Introduction Experiments were carried out to determine whether or not filtration of the carbon fibre residuals from the brightening liquor prior to adding the pulp would improve the quality of ex-situ brightened pulp. Effect of Filtration First, it is important to note that both runs were repeated once to ensure the accuracy of the findings. The results shown in Table 5.3.3. indicate that filtering the residual carbon fibre did not dramatically improve either the yellowness reduction nor the brightness gains. However, although the difference of the brightening results observed between both filtered and non-filtered runs were found to be within experimental error, the physical appearance of the brightened pulp was improved with the filtration process. The pulp retrieved from a filtered brightening solution, contained less visible fine carbon cloth fibre than pulp from non-filtered brightening solution. Table 5.3.3. Effect of Carbon Fibres Residuals on Brightening Process (Ex-situ) CARBON FIBER BRIGH ri-NING TIME (Min) BRIGHTNESS GAINS % (ISO) YELLOWNESS CHANGES %(R457) filtered 180 +16.3 -7.5 not filtered 180 +14.3 -8.3 Operating Conditions: 1) The in-situ electrochemical generation of hydrogen peroxide was carried out for three hours in the absence of pulp. The solution, which contained 35 (+/- 2.5.mM) of peroxide was then used as the brightening liquor without filtration. Both current and oxygen were employed during brightening of pulp. 2) Ex-Situ Brightening of Pulp:3 Hours 3) The Brightness/Yellowness for the original pulp is 44.7 (+/-)! % (ISO)and 33.3 (+/-) .6 % (R457). 80 5.3.4 Brightening Effects of In-situ Generated H^Oi vs. Merchant Peroxide The following experiments examine the effect of reaction time upon the brightening results of in-situ generated peroxide mediated by A Q S , merchant peroxide and merchant peroxide with 5 m M A Q S Figure 5.3.3. and Figure 5.3.4. respectively illustrate the brightness and yellowness responses over a 5 hour period for 49.5mM merchant peroxide and 49.5mM merchant peroxide with 5 m M A Q S and 49.5 m M H b O i from the ex-situ electrochemically generation mediated by A Q S , followed by filtration. (Refer to Tables 5.3.5, 5.3.6 and 5.3.7 of Appendix V I for experimental data.) 0 2 4 6 Time (hours) Figure 5.3.3. Brightening Effects of Merchant Peroxide(49.5mM) vs. Merchant Peroxide (49.5mM) with 5mM AQS vs. Ex-situ Generated Peroxide (49.5mM) 81 15 0 2 4 6 Time (hours) Figure 5.3.4. Yellowness Reduction Effect of Merchant Peroxide(49.5mM) vs. Merchant Peroxide (49.5mM) with 5mM AQS vs. Ex-situ Generated Peroxide 49.5mM) The results shown in Figure 5.3.3. and Figure 5.3.4. suggest that the presence of 5mM AQS inhibited the brightening effect of merchant hydrogen peroxide. The results obtained for merchant peroxide with 5mMAQS were consistently inferior to those for merchant peroxide alone. This suggests that AQS without the proper application of oxygen/current will reduce the brightening effect of hydrogen peroxide. In comparing the results of all three cases over 5 hours, the ex-situ brightening method of the proposed AQS/H2AQS/oxygen system (Table 5.3.7. of Appendix VI) consistently yielded better results than the merchant peroxide/5mMAQS run and gave similar results to those of the merchant peroxide only run. 82 5.4 Factorial Experiments The objective of this investigation was to determine the effects (main and interaction) of three process variables ~ oxygen, catalyst and current on the rate of hydrogen peroxide generation in the absence of pulp. The temperature and p H were held constant at 60 \u00C2\u00B0C and 10.5 respectively. The levels of the variables are summarized in Table 5.4.1. Table 5.4.1.Variables and their level in the 2 3 factorial run Variables LEVEL -('LOW') LEVEL +CHIGH) Catalyst (mM AQS) I m M 5 m M Current (Amp) 0.3 1.5 Gas (4.88 litres/min) AIR O X Y G E N Eight runs (2 ) were performed in random order according to the design created by the J A S S 2.2 software (Table 5.4.2.). A l l runs were replicated once to estimate the experimental error in the hydrogen peroxide measurements. From the replicated runs, Appendix VI, the response errors for hydrogen peroxide measurements are: s H 202= (+/-)2.5 mM. The experimental results, where all values are averages , are presented in Figure 5.4.1. 83 14 / / 12 V J 7\u00E2\u0080\u0094 1mM AQS catalyst ,effect_ A 49.5 \ \ V > \u00E2\u0080\u0094 / 4 v\u00E2\u0080\u0094-/ 4.5 Oxygen 4 J3 1.5 Amp / A. 3 / J 4 / J./0.3 Amp 5mM AQS Figure 5.4.1. Cube Plot of H?.Q? Concentration (mM) as a Function of Oxygen Catalyst and Current Density Levels. 84 Standard Operating conditions: see section 5.4. and Part II of Appendix VII Figure 5.4.1. indicates that the highest hydrogen peroxide concentration \u00E2\u0080\u0094 49.(+/-)2.5mM, was obtained at the highest level of all three factors. Also, the presence of all three factors, i.e. current/gas/catalyst seems to be beneficial for the in-situ generation of H2O2 in the proposed system. In the catalyst(+)gas(-)current(-), current(+)catalyst(-)gas(-) or even catalyst(+)/current(+)gas(-) experiments, very little hydrogen peroxide was obtained. The J A S S 2.1 software calculated the main and interaction effects, as well (Table 5.4.2.) Table 5.4.2. Main and Interaction Effects in 23 Factorial Experiments F.ffecis Effects on the Changes of H2O2 Concentration (mM) \ k i i n Effect - - Oxygen (0) 16.1 Main Effect - - Catahst (C) 7.4 Main Effect - - Current (I) 7.4 O C Interaction Effect 10.6 CI Interaction Ef fcu 10.9 OI Interaction 12.1 0C1 Overall Interaction 6.6 85 Main Effects A l l the main effects presented in Table 5.4.2. were statistically significant for the rate of electrochemical generation of H2O2 . This is consistent with the previous results of this thesis. A s can be seen in Table 5.4.2. the effect of oxygen was found to be the most critical variable. This finding corresponds to the color-change observation made in the bulk electrolysis study, which implies that the limiting step of the proposed AQS/oxygen system is the reaction between oxygen and H2AQS (Reaction 7). It is interesting to note that the other two main effects \u00E2\u0080\u0094 the catalyst effect and current effect \u00E2\u0080\u0094 were calculated to have the same values indicating they were equally important. This implies that the current and catalyst exert the same impact on the proposed system and can only be of an importance when they co-exist. Interaction Effects Similar results were observed for interaction effects. The oxygen/current interaction effect was determined to be the most prominent, whereas the interaction effect of catalyst/current closely followed that of catalyst/oxygen interaction effect. Finally, the catalyst/current/gas interaction effect was the weakest of all ~ nevertheless, still significant. From this 2 3 factorial design, the oxygen effect was determined to be most crucial. However, the presence of two key factors ~ oxygen and sufficient current ~ are essential and the presence of catalyst factor - A Q S is beneficial in the in-situ electrochemical generation of hydrogen peroxide in the proposed AQS/oxygen system. 86 5.5 Aging of the Electrode For all runs reported above, the cathode- carbon cloths used in these thesis project were never re-used. The results were found to be irreproducible when the carbon cloth cathode was re-used. Table 5.5. Aging of the Carbon Cloth Cathode fl of Repetith e use Final [I2O2 concentration (mM) 0 49.5 1 36.5 i 23.0 ' \u00E2\u0080\u00A2 3 \" \" 15.5 4 8.0 5 7.5 Operating Conditions (60\u00C2\u00B0C; pH=10.5; 5mM AQS; 2 hours; in the absence of pulp) Additives: (4.7mM Na2Si03; 0.32mM MgS04; 0.1 M Na2S03) (Cathode: 3 layers of 41 * 10cm2 carbon cloth; Anode:2.5* 1.4 cm2 platinized titanium) Similar to the fouling problems observed with the use of the fritted-glass separator described in section 5.2.4, the AQS coupled with magnesium sulphate may have, again, contributed to the fouling problems. AQS crystallizes and may then block the pores of the carbon cloth. Despite careful washing and cleansing of the carbon cloth after each run, it did not prevent the aging of the electrode. Little red crystals of AQS could be easily detected on the surface of the carbon cloth after approximately three runs. These crystals reduced the electro-active area of the carbon cloth and prevented the carbon cloth from electrochemically reducing oxygen and AQS. To ensure the accuracy of the experimental results reported above, a new carbon cloth was required for each run. Although the brightening results using the proposed current/AQS/oxygen system seem promising, it can not be realized in industry due to the high operating cost associated with having to replace the carbon cloth for each run. 87 Chapter 6.0. - General Discussion The proposed anthraquinone complex/oxygen redox couple involves the use of theAQS/JrbAQS complex to brighten thermo-mechanical pulp through a continuous in-situ generation of hydrogen peroxide from oxygen in an alkaline environment. The ability of this redox couple to operate under the TMP brightening conditions was investigated and the following observations were made. 1) The cyclic voltammetry studies revealed the presence of the AQS/H2AQS couple. It also indicated that the AQS/H2AQS reaction would take place on the carbon cathode surface. The experiments employing cyclic voltammetry also indicated the generation of hydrogen peroxide from oxygen in the solution. 2) The current-potential curves indicated that adding the AQS complex to the proposed system helps to depolarise the cathode, thereby increasing the range of current applicable to the proposed system. The ability of the cathode to promote the reduction of AQS/H2AQS was made evident by this experiment. 3) The rate of the color change observed in oxygenating the bulk solution suggested that the reaction of H2AQS with oxygen is the rate determining step for the in-situ generation of peroxide from oxygen mediated by the AQS/H2AQS redox couple. 4) The preliminary investigations indicated that the rate of hydrogen peroxide generation can be raised by increasing the ratio of cathode area to electrolyte volume and by improving mixing. 5) Anthraquinone 2,6-disulfonic acid was found to be an unsuitable redox couple for the proposed system due to low solubility and its ability to dye the pulp red. 88 6) Carbon felt was found to be an unsuitable electrode for the in-situ brightening of pulp due to pulp darkening problems. 7) Modifications of reactor design, which included increasing electro-active area of the cathode and centering the mixer and the anode, improved the current distribution and the mass transfer of the electrochemical cell, and thereby increased the rate of hydrogen peroxide generation. 8) Both the cation membrane and the fritted-glass separators were found to suppress the oxidation of peroxide on the anode, and therefore, higher current efficiencies in the in-situ generation of peroxide were obtained. However, the increased IR drop, induced by the above two separators, restricted the current that could be applied to the electrochemical cell, resulting in a significantly reduced rate of hydrogen peroxide generation . 9) Due in part to the residual carbon fibres in the in-situ product, the brightening results of ex-situ brightened pulp were much superior to that of the in-situ brightened pulp. The results also implied that the continuous in-situ generation of hydrogen peroxide in the in-situ brightening method may have been hindered and interrupted due to inadequate mixing. 10) The ex-situ electrochemical brightening with filtration of the brightening liquor generated comparable results to those of the merchant peroxide of the same concentration. 11) In accordance with a well-known fact of the conventional peroxide brightening of pulp, it was found that brightening effect is a function of hydrogen peroxide charges 89 and time. Similar to the course of a conventional peroxide brightening reaction, initial sharp increases in brightening effect followed by a leveling effect are observed in both in-situ and ex-situ brightening methods. 12) Factorial experiments confirmed that oxygen and sufficient current were pre-requisites and AQS was beneficial for the electrochemical generation of hydrogen peroxide in the absence of pulp. The evidence presented in this thesis study suggests that the proposed AQS complex/oxygen redox cycle is, indeed, operating to produce hydrogen peroxide at high current efficiencies. This redox cycle is also successful in engaging in pulp brightening. It appears that all three ingredients, i.e. current, gas and catalyst, are responsible for the generation of hydrogen peroxide and thus for the brightening effects of thermomechanical pulp. 90 Chapter 7.0. - Conclusion and Recommendations Conclusion. The combination of the anthraquinone 2-sulfonic acid, oxygen and sufficient current in an alkaline environment was found to bring about significant brightness gains and yellowness reduction in the brightening of thermomechanical pulp. TMP of 1% consistency was successfully brightened in-situ and ex-situ by the AQS redox mediated electroreduction of oxygen to 62.5 (%ISO)/20.2(%R457) and 68.4 (%ISO)/21.2 (%R457), respectively. This brightening effect is attributed to the action of hydrogen peroxide generation via the AQS/H2AQS redox cycle. The O2/AQS/H2AQS redox cycle proved to be a capable reducing system, which was effective in brightening pulp. However, the use of this redox couple was not practical in comparison to the conventional pulp brightening methods due to the rapid-fouling of the costly carbon cloth electrode, i.e. at approximately $507m2. Further research may be conducted to prevent the aging process of the electrode as well as to increase the effectiveness of the 0 2/AQS/H2AQS redox cycle. 91 Recommendations The following recommendations are suggested for future work in the brightening of TMP utilizing the AQS/oxygen redox couple: 1) Further investigation is needed on the effect of factors such as pulp consistency, size/type of the reactor, a high pressure reactor system, agitation/mixing, size of carbon electrode, temperature, current, pH, concentration of anthraquinone complex, etc.. 2) Improve cost efficiency by developing a recovery system in which the electrolyte and the catalyst can be recycled. 3) Design a mechanical separation system based on centrifugal force (hydrocyclone), for example, to separate the brightened pulp and the fine carbon cloth fibres. 4) Implement an on-line monitoring and measuring device so that the optimum operating conditions can be determined. 5) Search for alternative electro-catalysts for oxygen reduction to hydrogen peroxide which do not age rapidly nor produce fines that contaminate the pulps. 6) Investigate the reaction between H2AQS and oxygen to further the understanding of the proposed system and thereby increase the peroxide production. 7) Suppress the aging and fouling problems associated with the carbon fiber electrode, by scavengers or in-situ filters. 92 Bibliography Bibliography 1. Alder, E., \"Lignin Chemistry-Past, Present and Future\", Wood Science, Technology 11, 169-218 (1977). 2. Berl E., \" A New Cathodic Process for Production of H2O2\", Trans Electrochem. Soc, 76 (1939)359. 3. Britt, K.W. ,\" Handbook of Pulp and paper Technology\", 2 n d Edn., 317-319 (1970). 4. Clifford A . et al, \"Electrosynthesis of Alkaline Peroxide Solutions\", Electrochem. Soc. Spring Meeting May 1990, Montreal, Canada. 5. Colodette J.L. and Dence, C.W., \"Factors Affecting Hydrogen Peroxide Stability in the Brightening of Mechanical and Chemimechanical Pulps, Part IV: The Effect of Transition Metals in Norway Spruce TMP on Hydrogen Peroxide Stability\", Journal of Pulp and Paper Science, Vol . 15(3), 79-80 (1989). 6. Economu, D.J. and Alkire, R.C., \"Two Phase Mass Transfer in Channel Electrolyzers with Gas-Liquid Flow\", Journal of Electrochemical Society, 132, 601 (1985). 7. Fleury, R.A. and Rapson, W.H., \"Characterization of Chromophoric Groups in Groundwood and Lignin Model Compounds by Reaction with Specific Reducing Agents\", Pulp Paper Mag. Can., Mar., 1968, T l 54-60. 8. Foller P.C. and Bombard R., \"Review of Processes for the production of Mixtures of Caustic Soda and Hydrogen Peroxide via the Electroreduction of Oxygen\", J. Appl. Electrochem., 25 (1995) 613-627. 9. Foller P.C. et.al., \"The Use of Gas Diffusion Electrodes in the On-Site Generation of Oxidants and Reductants \", Fifth Int. Forum on Electrolysis in the Chemical Industry, Nov. 1991, Fort Lauderdale, USA.. 10. Gilead, E., Kirowa-Eisner, E., Penciner, J., \"Interfacial Electrochemistry. A n Experimental Approach\", Addison-Wesley Publishing Company Inc., Chap.9 368-396 (1975). 11. Joyce, P. and Mackie, D., \"Brightening of Mechanical Pulps\", Int. Pulp Bleaching Conference, 107-118 (Toronto, 1979). 12. Joyce, P., and Mackie, D.W., \"Brightening of Mechanical Pulp\", International Pulp Bleaching Conference Reprints, CPPA (Toronto), 107-110 (1979). 13. Kerekes, R.J. and Grace J.R., \"Mixing in Pulp Bleaching\", Journal of Pulp and Paper Science, 15(5), J186-195 (1984). 93 Bibliography 14. Kirk, R.E., Othmer, D.F., Grayson, M . and Ekroth, D., \"Concise Encyclopedia of Chemical Technology\",3rd, Edn., 365-366 (1985). 15. Kirk-Othmer Encyclopedia of Chemical Technology, vol.11, 391-403 (1965). 16. Kollman, F.F.P. and Cote, W.A., \"Principles of Wood Science and Technology\". Springer, Chap 1-2(1968). 17. Kraft, \"Pulp and Paper Manufacture\", Vol . 1, Edition McDonald R & Frank, J.H. 18. Masao, S., Hisatsugu, K., and Kozo, K. , \"Electrochemical Production of Hydrogen Peroxide by Reduction of Oxygen\", Journal of Chemical Engineering of Japan, 18(5), 409 (1985). 19. Mclntyre J. and Phillips R. \"Method of Operating a Liquid-Gas Electro-chemical Cell\", US Patent 4,406,758 (1983) [To: Dow Chemical Co.]. 20. Murphy, T.D., \"Design and Analysis of Industrial Experiments\", Chem. Engineering\", 84(12), T168-182 (1977). 21. Oloman C. and A.P. Watkinson, \"Electrolytic Production of Alkaline Peroxide Solutions\", US Patent, 2,969,210 (1976) [To CPDL]. 22. Oloman C , \"Trickle Bed Electrochemical Reactors\", J. Electrochem. Soc. Soc, 126 (1979) 1885-1893. 23. Oloman, C. and Watkinson, A.P., \"The Electroreduction of Oxygen to Hydrogen Peroxide on Fluidized Beds\", Canadian Journal of Chemical Engineering, 53, 268 (1975). 24. Oloman, C. and Watkinson, A.P., \"The Electroreduction of Oxygen to Hydrogen Peroxide in Fixed Bed Cathodes\", Canadian Journal of Chemical Engineering, 54, 312 (1976). 25. Oloman, Colin \"Electrochemical Processing For the Pulp & Paper Industry\", The Electrochemical Consultancy Pub., Romsey 1996. 26. Pattyson, G.W., \"Kamyr M C Mixer for C10 2 Mixing at Great Lakes Forest Products, Thunder Bay, Ontario\", Pre-Print, 70 t h Annual Meeting for the Technical Section of CPPA, A63-A68 (1984). 27. Polcin, J. and Rapson, W.H., \" Effects of Bleaching Agents on the Absorption Spectra of Lignin in Groundwood Pulps: Part I and Part II\", Pulp Paper Mag. Can.72 no.3: 69-91 (T103-24), (1971). 28. Reeve, D.W., 'Introduction to the Principles and Practice of Pulp Bleaching' in 'Pulp Bleaching: Principles and Practice\", edited by Dence, C.W. and Reeve, D.W., Tappi Press, Atlanta, P.3. (1996). 94 Bibliography 29. Regner, A . , \"Electrochemical Processes in the Chemical Industry\", Artia, Prague, (1957). 30. Reichert, J.S. and Pete, R., \"Peroxide Bleaching of Groundwood - Recent Developments and Commercial Status\", Tappi Journal, 32(2),96-107(1949). 31. Sawyer, D.T., Heineman, W.R., and Beebe, J.M., Chemistry Experiments for Instrumental Methods, pp. 79-95, Wiley, 1984. 32. Singh, R.P., \"The Bleaching of Pulp\", Tappi Press, 3 r d Edn., 229 (1979). 33. Slove, M.L. , \" The Role of Alkali in Pulp Bleaching\", Tappi Journal 48(9),533-535 (1965). 34. Smook, G.A. \"Handbook for Pulp & Paper Technologists\", 2 n d Ed., Angus Wilde Pub., Vancouver, 1992. 35. Strunk, W.G., \"Factors Affecting Hydrogen Peroxide Bleaching for High-Brightness TMP\". Pulp & Paper, 54(6), 156-160 (1980). 36. Takahashi, K. and Alkire, R.C., \"Mass Transfer in Gas-Sparged Porous Electrodes\", Chemical Engineering Communications, 38,209 (1985). 37. Terregrossa, L.O., \"Effect of Mixing on CIO2 Bleaching\", Tappi Pulping Conference, Houston, Texas, 635-641(1983). 38. Vogel A.I., \"Textbook of Quantitative Inorganic Analysis\", 4 Edn., Longman (1978). 39. Volkman, Y., \"Optimization of the Effectiveness of a Three-Dimensional Electrode with Respect to its Ohmic Variables\"), Electrochemica Acta, 24, 1145 (1979). 95 Appendix I Analysis of Factorial Design Factorial design of experiments is a mathematical method that will determine the experimental response surface. It is an organized approach toward a collection of information. Through this analysis, maximum information can be gathered per experiment . It also exhibit information reliability by taking the experimental variations into consideration. More importantly, it will also illustrate important interactions among the experimental variables [20]. Part 1-2 Factorial Design In a two level factorial design, for example, the low level and high level of each variable are coded as -1' and '+1' in a two variable (namely XI and X2) factorial design. The factorial design for this 2 by 2 analysis is listed below: Table I Two Variable Factorial Design (2 )Factorial Design Run XI X2 XI *X2 Responses 1 - - + Y l 2 + - - Y2 3 - + - Y3 4 + + + Y4 The factorial experimental design also allows for the calculation of main and interaction effects. A) Main Effects Consider n variables (Xi..n) at two levels, i.e. 'low' and 'high'. The main effect is calculated as the difference between the average 'high' and 'low' level responses. Main Effect of Xi = \u00C2\u00A3 (responses at 'high' Xi - responses at 'low' Xi) (half the number of factorial runs) 96 B) Interaction effects The interaction effect is the average response difference between one half of the factorial runs and the other half. To illustrate the calculation of the interaction effects as 22 design is considered. Table I includes the interaction column as well, formed from the columns of the two factors that comprise the interaction, by multiplying the entries in the factorial column. C) Assessment of Significance of Effects The effects calculated from Factorial Design are point estimates. They do not indicate the reliability or precision of these estimates. The precision is generally given in the form of a confidence interval, i.e., the interval which includes the \"true\" effect at a stated confidence level. The commonly used confidence levels are 90, 95, 99%. The confidence interval width is a function of the reponse error estimate, the number of data in the estimate, the number of degrees of freedom in response error estimate for an effect. The confidence interval (CI) for effects are calculated as: For main effects and interaction effects (CI) ^Effect Estimate +1 s/ (N/4) A (0.5) Part II - 23 Factorial Design In a three level factorial design\u00E2\u0080\u009Ethe low level and high level of each variable are coded as '-1' and '+1' in a three variable (namely XI, X2 and X3) factorial design. The factorial design for this 2 by 3 analysis is listed below: Table II. Three variable Factorial Design (23 Factorial Design) Run XI X2 X3 Responses 1 -1 -1 -1 Yl 2 +1 -1 -1 Y2 3 -1 +1 -1 Y3 4 +1 +1 -1 Y4 5 -1 -1 +1 Y5 6 +1 -1 +1 Y6 7 -1 +1 +1 Y7 8 +1 +1 +1 Y8 97 Appendix II Current Efficiency of Hydrogen Peroxide Generation Reaction 5 from Section 2.8 02+H20+2e\" ==>H02 +OFT Average Current efficiency of the process over time is calculated from Current efficiency = nFR/I Where n = Electron stoichiometry for reaction 5 F = Faraday number (96480 coulombs/equivalent) R = Production rate of hydrogen peroxide (kmole/s) I = Total current (kAmp) For example: At 1.5 A current, after 120 minutes run (Table 5.2.12. of Appendix V), hydrogen peroxide concentration inlOOO ml is 49.5 M Note: Electron Stoichiometry for Reaction 5 (Section 2.8) = 2 Step I Theoretical amount = 120 min * 60 s/min * 1.5 A. = 56.0 mM 96500 *2* 1000ml Step II The actual amount of hydrogen peroxide (mM) obtained from titration = 49.5 mM Therefore: The rate of H2O2 Generation @120 minutes = ((49.5) mM/(120 min*60 sec/min) =6.9 mM/sec Step III Current Efficiency = Actual amount of peroxide obtained Theoretical amount of peroxide Current Efficiency = (49.5mM/56mM)= 88 % 98 Appendix III Cyclic Voltammetry c V Cyclic voltammetry (CV) is a useful tool for identification of steps in the overall reaction and of new species which appear in solution during electrolysis. CV consists of cycling the potential of an electrode, which is immersed in an unstirred solution, and measuring the resulting current. The potential of this working electrode is controlled vs. a reference electrode such as a saturated calomel electrode (SCE) or Ag/AgCl electrode. In cyclic voltammetry, the potential applied to an electrode is a linear potential scan with a triangular waveform as shown in Figure III. The sweeping potential in Figure III scans negatively from +0.8 (a) to -0.2V (SCE) (c) and then back to +0.80 V (SCE) (e). The scan rate is reflected by the slope of the triangular waveform.. Figure III. Typical scan potential triangular waveform for cyclic voltammetry [311 Switching potentials at 0.8 and -0.2 V (SCE) 99 The basic feature of a voltammogram (i.e. a plot of current vs. potential during cyclic voltammetry) is illustrated in Figure IV. [10]. In a typical cyclic voltammogram (Figure IV), a substance being reduced during the cathodic scan (a -> b c), emerges as a positive peak. The peak current (Ipc) represents the point (b) at which the reduction rate is equal to the rate of diffusion of the reducible species to the electrode surface. When the scan is reversed, (c -> d -> e), the reduced species, which is still at the surface of the electrode, is re-oxidized, yielding an anodic peak. The significance of a cyclic voltammogram include the magnitudes of the anodic peak current (Ipa), the cathodic peak current (Ipc), the anodic peak potential (Epa), and the cathodic peak potential (Epc). POTENTIAL, V vs SCE Figure IV Cyclic Voltammogram [101 100 For an electrochemically reversible couple, the formal reduction potential (E) is centered between its anodic and cathodic peak potentials. Slow electron transfer at the electrode surface results in an increase in the peak separation. E= (Epa + Epc)/2 The peak current for a reversible system is described by the Randies- Sevcik equation for the forward sweep of the first cycle [31]: Ip=2.69 * 1 0 5 z 3 / 2 A D 1 / 2 C v ' / 2 where Ip: Peak Current (A). z: Electron Stoichiometric coefficient. A: Electrode area (cm2) D: Diffusion coefficient (cm2/s) C: Concentration (mole/ cm3) v: Scan rate (V/s) According to the above equation, the position and shape of a given peak are dependent on several factors, which may include the sweep rate, electrode material, solution composition and the concentration of reactants. The Ip, for example, increases with v Vi and is directly proportional to concentration. The values of Ipa and Ipc should be approximate each other for a simple reversible couple. 101 Appendix IV Hydrogen Peroxide Titration Calculation Hydrogen Peroxide reacts with potassium iodine via reaction shown below. H 2 0 2 +2KI + H 2 S0 4 = = > K 2 S0 4 +2H20+I2 The titration procedure described below is based on the above equation and a method presented in literature [30]. A 5 ml sample of bulk solution was collected for peroxide analysis using a pipette The 5 ml sample was analyzed for hydrogen peroxide as the following standard procedure. \u00E2\u0080\u00A2 A 5ml sample of fibre-free liquor was drawn from the reactor and added to 200ml of ice-cooled 20%H2SO4 in a 250ml flask. \u00E2\u0080\u00A2 5ml of the KI solution (approximately 165g KI per liter) was added. \u00E2\u0080\u00A2 3 drop of (NH4)2Mn04 (saturated, freshly prepared) and three drop of starch indicator (0.5 tol %) were then added. \u00E2\u0080\u00A2 The sample was then titrated with 0.01N Na 2S 20 3, where 1ml of 0. IN Na 2S 20 3 titrated corresponds to a sample concentration of lOmM H 2 0 2 . 102 Stage IA Table 5.2.1. 5mM AQS/Oxy Current: 0.15Amp H-Shaped CELL Time Actual H202 Theo H202 Current Efficiency (min) (mM) (mM) (%) 0 0 0.0 100 30 4.8 5.6 86 90 13.0 16.8 77 145 19.0 27.0 70 210 25.0 39.2 64 300 30.0 56.0 54 Stage IB Table 5.2.2. OmM AQS/Oxy gen Current: 0.15 Amp H-Shaped CELL Time Actual H202 Theo H202 Current Efficiency (min) (mM) (mM) (%) 0 0 0.0 100 30 4.2 5.6 75 60 5 11.2 45 180 8.5 33.6 25 210 9 39.2 23 300 11.0 56.0 20 Stage IB Table 5.2.3. 5mM AQS/Oxy gen Current: 0.08 Amp H-Shaped CELL Time Actual H202 Theo H202 Current Efficiency (min) (mM) (mM) (%) 0 0 0.0 100 30 4.5 3.0 100 90 8.5 9.0 95 210 15.5 20.9 74 300 20.0 29.8 67 103 Stage IC Table 5.2.4. 2.5mM AQ2S/Ox ygen Current: 0.15 Amp H-Shaped CELL Time Actual H202 Theo H202 Current Efficiency (min) (mM) (mM) (%) 0 0 0.0 100 15 3 2.8 100 120 18 22.4 80 210 20 39.2 51 300 22 56.0 39 Stage II Table 5.2.5. 5mM AQS/Oxy gen Current: 0.2 Amp cation membrane Reactor Conf.#1 Time Actual H202 Theo H202 Current Efficiency (min) (mM) (mM) (%) 0 0 0.0 100 30 1.86 1.9 100 75 4.7 4.7 100 96 6 6.0 100 115 7 7.2 98 135 8 8.4 95 150 8.7 9.3 93 180 10.2 11.2 91 210 11.8 13.1 90 300 16.4 18.7 88 104 Stage III Table 5.2.6. 5mM no cation AQS/Oxy membrane gen Current: 0.42 Amp Reactor Conf.#1 Time Actual H202 Theo H202 Current Efficiency (min) (mM) (mM) (%) 0 0 0.0 100 15 3 2.0 100 30 4.1 3.9 100 60 6 7.8 100 120 10 15.7 64 180 14 23.5 60 300 20 39.2 51 Stage III Table 5:2.7. 5mM AQS/Oxy gen Current: 0.7 Amp no cation membrane Reactor Conf.#1 Time Actual H202 Theo H202 Current Efficiency (min) (mM) (mM) (%) 0 0 0.0 100 27 3.5 5.9 100 132 12.5 28.7 100 270 23 58.8 100 300 28 65.3 43 105 StaaelVA Table 5.2.8. 5m M AQS /Oxygen Current: 0.2 Amp Reactor Config.#1 Fritted Glass Separation Time Actual H202 Theo H202 Current Efficiency (min) (mM) (mM) (%) 0 0 0.0 100 20 5 1.2 100 107 8.5 6.7 100 180 10 11.2 89 210 11 13.1 84 240 12.5 14.9 84 300 14 18.7 75 StaqelVA Table 5.2.9. 5mMAQS /Oxygen Current: 0.2 Amp Reactor Config.#1 no fritted glass Time Actual H202 Theo H202 Current Efficiency (min) (mM) (mM) (%) 0 0 0.0 100 30 2 1.9 100 67 3.5 4.2 84 90 4.5 5.6 80 120 5.5 7.5 74 150 6.5 9.3 70 174 7.5 10.8 69 210 8.5 13.1 65 300 9.0 18.7 60 106 Stage IV- Table 5.2.10 Reactor B Config.#2 5mMAQS Carbon Felt /Oxygen Electrode Current 1.0 Amp Time Actual Theo Current H202 H202 Efficiency (min) (mM) (mM) (%) 0 0 0.0 100 15 5 4.7 100 40 10 12.4 80 80 16.5 24.9 66 120 20 37.3 54 Stage V Table 5.2.11 5mMAQS /Oxygen Current: 1.0 Amp Reactor Config.#2 Carbon Cloth Electrode H202 # of carbon cloth (mM) 9 1 14 2 35 3 32 5 13 6 Stage VI Table 5.2.12 5m M AQS /Oxygen Current: 1.5 Amp Reactor Config. #3 Carbon Cloth Electrode Time Actual H202 Theo H202 Current Efficiency (min) (mM) (mM) (%) 0 0 0.0 100 15 7 7.0 100 30 14 14.0 100 60 28 28.0 100 70 36 32.6 110 100 45 46.6 97 110 47.5 51.3 93 120 49.5 56.0 88 107 A p p e n d i x VI P a r t I - E x p e r i m e n t a l Resul ts i n Section 5.3 Note: In all the experiments of section 5.3.2., some brightening conditions were standardized and summarized as follows: \u00E2\u0080\u00A2 Operating Conditions: (Temp=60\u00C2\u00B0C; pH=10.5; Current:1.5 A.; O2 purged at 4.88 litres /min @STP) \u00E2\u0080\u00A2 Additives: (5mM AQS, lg Na2Si03; 0.08g MgS04; 0.1 M Na2S04, 0.22mM DTPA) \u00E2\u0080\u00A2 Reactor Configuration #3: Cathode: 2 layers of 41*10cm2 carbon cloth; Anode:2.5*1.4 cm2 platinized titanium) \u00E2\u0080\u00A2 The electrochemical in-situ generation of peroxide was carried out for 2 hours till the H 2 0 2 concentration reached 49.5(+/-)2.5.mM. This solution is filtered and used as the brightening liquor for an unbrightened pulp, which has brightness and yellowness at 44.7( +/-)1 %ISO and 33.3(+/-) .6 % R457 respectively. Experiments From Section 5.3.2 Table 5.3.2. Ex-situ brightening Effect vs. Time Brightening Time (min) Brightness Gain \"..(ISO) \ cllow ncvs t huiiues,,..iR457) 60 +10.0 -6.4 120 +16.7 -10.6 180 +19.5 -12.2 300 +21.1 -12.8 Operating Conditions \u00E2\u0080\u00A2 Brightening of Pulp Ex-Situ: 3Hours, i.e. current was not used while brightening of pulp is taking place. Table 5.3.3 In-situ brightening Effects vs. Time Timc(min) ISO brightness Yellowness Changes Gains % (ISO) \"..(R457) 60 +11.5 -7.3 120 +12.5 -9.3 180 +13.9 -12.3 Operating Conditions \u00E2\u0080\u00A2 Brightening of Pulp In-Situ: 3Hours, i.e. current was used while brightening of pulp is taking place. 108 Part II - From Experimental Section 5.3.4. Note: In all the experiments of section 5.3.4., some brightening conditions were standardized and summarized as follows: \u00E2\u0080\u00A2 The brightness and yellowness for the origianl pulp are at 44.7( +/-)1 %ISO and 33.3(+/-) .6 % R457 respectively. \u00E2\u0080\u00A2 Operating Conditions: (Temp=60\u00C2\u00B0C; pH=10.5; 0 2 purged at 4.88 litres /min @ STP) \u00E2\u0080\u00A2 Additives: (5mM AQS, lg Na2Si03; 0.08g MgS04; 0.1 M Na2S04, 0.22mM DTP A) Table 5.3.5. Brightening Effects of Merchant Peroxide(49.5mM) Time (hour) . - <;_Brightness-Gains ISO(%) Yellowness Changes (%) R457 1 +13 -8 3 +19 -11 \" 5 +20 -13 Table 5.3.6. Brightening effects of merchant hydrogen peroxide(49.5mM) +5mM AQS Time Brightness Gains Yellowness Changes (hour) . ISO(%) < (%) R457 1 +9 -6 3 +14 -8 5 +18 -11 Table 5.3.7. Brightening effects of the Ex-situ brightening method (49.5mMH202+5mMAQS) Time Brightness Gains Yellowness Changes (hour) ISO (%) (%) R457 . 1 +11.0 -6.4 : 3 \u00E2\u0080\u00A2 \u00E2\u0080\u00A2: +19.5 -12.2 \" 5 +21.1 -12.8 109 Appendix VII Experimental results in section 5.4 Note: In all the experiments of section 5.4., the standard brightening conditions are summarized as follows: \u00E2\u0080\u00A2 Operating Conditions: (Temp=60\u00C2\u00B0C; pH=10.5; 0 2 purged at 4.88 litres /min @ STP) \u00E2\u0080\u00A2 Additives: (lg Na2Si03; 0.08g MgS04; 0.1 M Na2S04, 0.22mM DTPA) \u00E2\u0080\u00A2 Reactor employed: (Reactor Config.#3; Cathode: 2 layers of 41 cm by 10 cm carbon cloth; Anode: 2.5 cm by 1.4 cm platinized tianium) Table 5.4.1. Variables an their level in the 23 factorial run Variables L E V E L -(\u00E2\u0080\u00A2LOW) L E V E L +('fflGH') Catalyst (mM A Q S ) I m M 5 m M Current (Amp) 0.3 1.5 Gas (4.88 litres /min) A I R O X Y G E N Table 5.4.2.: Design of the 23 Full Factorial Experiment XI Catalyst X2 Gas X3 Current *Hydrogen Peroxide Concentration (mM) Y l -1 -1 -1 12.0 Y2 1 -1 -1 4.5 Y3 -1 1 -1 12.0 Y4 1 1 -1 12.5 Y5 -1 -1 1 3.0 Y6 1 -1 1 4.0 Y7 -1 1 1 14.0 Y8 1 1 1 49.5 Note: The Hydrogen Peroxide concentration given in Table 5.4.2. are average values. 110 Part II of Appendix VH-factorial design Table 5.4.1 5mMAQS/Oxygen/1.5 ';Time Amount of Current 1 (min) H202 Efficiency (mM) (%) 0 0 100 15 7 100 30 14 100 60 28 100 70 36 100 100 45 96 110 47.5 93 120 49.5 88 Table 5.4.2 5mMAQS/Oxygen/0.75 Amp lime Amount of Cm run \u00E2\u0080\u00A2 mm i H202 1 fficiuio (mM) (%) 0 0 100 45 13.5 100 90 24 100 120 27 86 150 30 77 180 35 75 230 40 67.5 240 40 64.6 Table 5.4.3 Q-l-1 5mMAQS/Oxygen/0.30 Amp lime Amount nl\" C intuit (mini 11202 Efficients (mM) (%) 0 0 100 20 2.5 100 105 10.5 100 120 12.5 100 210 19 97 240 21 94 111 Table 5.4.4 Q-2 lmMAQS/Qxygen/1.5 Amp 1 lime \mnimt of ( UIlL'lU (mm) H2(>: Efficiency (mM) <%) 0 0 100 20 5.5 74 60 10.5 45 90 13.5 40 120 14 31 Table 5.4.5 Q-3 5mMAQS/Air/ .5 Amp Time Amount of Current (min) 11202 Efficient (mM) (\u00C2\u00BB..) 0 0 100 20 5.5 73 60 5.5 25 90 3 9 120 4 9 Table 5.4.6 Q-4 5mMAQS/N itrogen/1.5 Amp Time Amount of Current (min) 11202 1 filcienc\ (mM) (%) 0 0 100 15 2.5 45 60 3.5 16 120 4.5 10 240 5 6 Table 5.4.7 Q-5 ImMAQS/ Air/1.5 Amp Time Amount of Current (min) H202 Efficiency im\l) <\"..> 0 0 100 15 2.5 45 50 5 27 90 5 15 180 2 3 112 Table 5.4.8 5mMAQS/Air/0.3 Amp I ime (min) Aminint ol JI202 (mM) Current iifficiencv (%) 0 0 100 20 4 100 60 5 100 90 4.5 67 120 4.5 50 Table 5.4.9 lmMAQS/Air/0.3 Amp Time #.** (min) Amount of H202 (mM) Current - Efficiency (%) 0 0 100 15 2 100 45 4 94 120 12 100 135 13.5 100 160 13.5 94 180 14.5 90 240 20 93 Table 5.2.10 ImM AQS/Oxy gen/0.3 Amp Time Amount of Current iniiiu II202 Efficient^ (m\l) (%) 0 0 100 60 6 100 120 12 100 150 14 100 240 22 98 300 26 93 113 Appendix VIII Glossary Brightness: Reflectivity of paper sample using light of specified wave length (457nm), commonly used an index of brightness. Cellulose: Material that forms the solid frame work or cell walls of all plants. It is the fibrous substance that remains after the non-fibrous portions, such as lignin, have been removed from the pulp during the cooking and bleaching operations. Chromophore: A covalently unsaturated group responsible for electronic absorption Consistency: The concentration of fiber in a pulp suspension expressed as the weight percentage in a pulp and water mixture. CPPA: Canadian Pulp and Paper Association Current density: current/electrode area Current efficiency: Actual rate of desired reaction/ rate of desired reaction in stoichiometric equivalence with total current Divided reactor: Reactor divided by ionically conductive separators into two chambers holding anolyte and catholyte separately. DTPA: Diethylene-triamine-pentaacetic-acid. Extractives: Substance in wood, not an integral part of the cellular structure, that can be dissolved out by a solvent that does not react chemically with wood components. Fiber: An elongated tapering, thick walled cellular unit which is the structural component of woody plants. Handsheet: A sheet made by depositing the fiber from a suspension of fiber and water on a laboratory sheet mold, followed by pressing and drying under carefully standardized conditions. Hemicellulose: Short-chain polysaccharides having a DP of 15 or less. Holocellulose: Total carbohydrate content of fibers. Lignin: Natural binding constituent of the cells of wood. It is removed along with other organic materials during the pulping and bleaching stages. Mechanical pulp: Pulp produced by reducing wooden logs and chips into their fiber components by the use of mechanical energy. Middle lamella: Lignin-rich cementing layer between cell walls in plants. SHE: standard hydrogen electrode 114 SCE: Saturated Calomel electrode V(vs SCE) = 0.24 +V(vs.SHE) at 25 \u00C2\u00B0C Redox Cycle: Chemical/electrochemical process involving oxidation and reduction reactions. Stone Groundwood: Process in which fibers are loosened and peeled out from wooden logs by pressing against grindstones. Superficial Current density: Current density based on the area of the separator or the projected area of the electrode TAPPI: Technical Association of Pulp and Paper industry TMP: Thermomechanical pulp, produced from pre-steamed wood chips by passing them into a refiner. 115 "@en . "Thesis/Dissertation"@en . "1998-05"@en . "10.14288/1.0058939"@en . "eng"@en . "Chemical and Biological Engineering"@en . "Vancouver : University of British Columbia Library"@en . "University of British Columbia"@en . "For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use."@en . "Graduate"@en . "The in-situ brightening of pulp by the electro-reduction of oxygen, mediated with anthraquinone-2-sulfonic acid"@en . "Text"@en . "http://hdl.handle.net/2429/7834"@en .