UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The use of mixing-sensitive chemical reactions for the study of pulp fibre suspension mixing Thangavel, V. K. 1992

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

Item Metadata

Download

Media
831-ubc_1992_fall_thangavel_v_k.pdf [ 4.08MB ]
Metadata
JSON: 831-1.0058588.json
JSON-LD: 831-1.0058588-ld.json
RDF/XML (Pretty): 831-1.0058588-rdf.xml
RDF/JSON: 831-1.0058588-rdf.json
Turtle: 831-1.0058588-turtle.txt
N-Triples: 831-1.0058588-rdf-ntriples.txt
Original Record: 831-1.0058588-source.json
Full Text
831-1.0058588-fulltext.txt
Citation
831-1.0058588.ris

Full Text

T H E U S E OF MIXING-SENSITIVE C H E M I C A L REACTIONS FOR T H E S T U D Y OF PULP FIBRE SUSPENSION MIXING by V . K . T H A N G A V E L B . S c , The University of Madras, 1976 M . S c , The University of Madras, 1978 A THESIS SUBMITTED IN P A R T I A L F U L F I L L M E N T OF T H E REQUIREMENTS FOR T H E D E G R E E OF M A S T E R OF APPLIED SCIENCE in T H E F A C U L T Y OF G R A D U A T E STUDIES Department of Chemical Engineering We accept this thesis as conforming to the required standard T H E UNIVERSITY OF BRITISH C O L U M B I A June 1992 © V . K . Thangavel, 1992 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes 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 The University of British Columbia Vancouver, Canada Date DE-6 (2/88) Abstract The competitive, consecutive azo coupling between 1-naphthol and diazotized sulphanilic acid was used to investigate the mixing of a pulp fibre suspension in a stirred tank reactor operated in a semi-batch mode. The distribution between the mono and bis substituted product dyes depended upon the turbulent intensity in the reaction zone and was used to indicate the degree of mixing in the vessel. Coupling reactions were carried out under conditions where macromixing did not influence the product distribution and were used to indicate the turbulent dissipation and microscale mixing in the reaction zone. When fibres were present in the reactor some of the product dyes were adsorbed onto the fibres. Adsorption rate was not proportional to the equilibrium concentration of the dyes and it was necessary to correct the measured product distribution to account for the differential adsorption rates. Once this correction had been made the product distribution correctly indicated the degree of mixing. As the dyes were found to have different affinities for different pulps, separate correlations would have to be developed for each pulp tested. A l l results reported here were obtained with a commercially produced fully bleached kraft pulp. Tests were conducted in the stirred vessel using aqueous media and at fibre suspension concentrations up to 2.5% by mass. Test results in aqueous media compared favourably with past work and a predictive i i model developed by Baldyga and Bourne. In the presence of fibres a reduction in turbulent energy dissipation was detected under most operating conditions when compared with the aqueous tests. When the suspension mass concentration was increased to the point where the vessel contents ceased to be in complete motion the product distribution began to increase due to macroscale mixing effects. The technique could not be applied successfully to a high-shear mixer due to the high energy dissipation rates and dye adsorption, both of which reduced the product distribution below the analytical detection limit. While other mixing sensitive chemical reactions are available to permit mixing determination at these high dissipation rates, the extension of the reaction system to more than 3 adsorbable components would make the measurements difficult to conduct and interpret. i i i Contents Abstract i i LIST OF T A B L E S ix LIST OF FIGURES xiv A C K N O W L E D G E M E N T S xvii 1 Mixing Pulp Suspensions 1 1.1 Introduction 1 1.2 Mixing Concepts 3 1.3 Mixing in Pulp Suspensions 4 1.4 The Kinetics of Tracer Reaction 5 1.4.1 Reaction Used 5 1.4.2 Pre-equilibria and Salt Effects 9 1.4.3 Side Reactions 10 1.4.4 Product Distribution 12 1.5 Pulp Suspension Rheology 13 1.6 Mixing in Pulp Bleaching 15 1.7 Research Objectives 16 iv 2 Adsorption of Chemicals onto Cellulose Fibres 17 2.1 Introduction 17 2.2 Literature Search 18 2.2.1 Isotherms 18 2.2.2 Direct Dyes 19 2.2.3 Forces of Attraction 19 2.2.4 Dye Aggregation 21 2.2.5 Pulp Fibres 21 2.2.5.1 Surface Charge of Cellulose Fibres 22 2.2.5.2 Adsorption Dyes by Lignin in Fibre 22 2.3 Experimental 23 2.3.1 Fibre Used 23 2.3.2 Equipment and Procedure for 1-Naphthol Adsorption 24 2.3.3 Equipment and Procedure for Dye Adsorption 25 2.3.4 Experiment for Dye Aggregation 26 2.4 Results and Discussion 28 2.4.1 Adsorption of 1-Naphthol 28 2.4.2 Data Reproducibility 28 2.4.3 Adsorption of Mono and Bisazo Dyes 33 2.4.4 Diffusion of Chemicals onto Fibre 36 2.4.4.1 Diffusion of Diazotized Sulfanilic Acid 36 2.4.4.2 Diffusion of Monoazo Dye 38 v 2.4.5 Correction for Dye Adsorption using Xs 38 2.4.6 Correction for Dye Adsorption using Mass Balance 42 2.4.7 Dye Aggregation 44 2.4.8 Adsorption of Dyes by Various Types of Fibres 45 2.5 Conclusion 48 3 Mixing in a Semi-Batch Reactor 52 3.1 Introduction 52 3.2 Literature Search 52 3.2.1 Impeller Characteristics 52 3.2.2 Micromixing in a Semi-Batch Reactor 53 3.2.3 Macromixing 55 3.3 Equipment and Procedure 56 3.4 Aqueous Tests 62 3.4.1 Evaluation of Impeller Characteristics 62 3.4.2 Investigation of Macromixing and Micromixing 62 3.4.3 Effect of Feed Point 63 3.5 Tests in Pulp Suspension 63 3.6 Data Reproducibility of Power Measurement in the Stirred Tank 64 3.7 Results and Discussion for Aqueous Tests 65 3.7.1 Impeller characteristics in the Stirred Tank 65 3.7.2 Backmixing into Feed Pipe 65 3.7.3 Determination of Macromixing Time at Feedpoints A and B . . . . 67 vi 3.7.4 Comparison with Model - Aqueous Tests 67 3.7.5 Comparison with Similar Study 72 3.8 Results and Discussion for Pulp Suspension 74 3.8.1 Power Consumption by Pulp Suspension 74 3.8.2 Flow Pattern of Pulp Suspension in the Reactor 76 3.8.3 Macromixing Time 80 3.8.4 Variation of X s c o r r with C m at Various Stirrer Speeds 80 3.8.5 Relationship between X s c o r r and Average Energy Dissipation . . . . 85 3.9 Conclusions 85 4 Micromixing in a High-Shear Mixer 88 4.1 Introduction 88 4.2 Literature Search 88 4.3 Equipment and Procedure 89 4.3.1 Equipment 89 4.3.2 Assessment of Micromixing in the High-Shear Mixer 92 4.3.3 Critical Feed Time 93 4.4 Results and Discussion 94 4.5 Conclusion 98 5 Conclusion 99 6 Recommendations for Future Work 101 Nomenclature 102 References 108 vii Appendicies 114 A Fibre and Suspension Properties 114 A . 1 Fibre Dimensions 114 A.2 Water Retention Value 119 A.3 Concentration of Pulp Suspension 119 A.4 Half-Thickness of Fibre 120 B Data Reproducibility 121 C Preparation of Chemicals 126 C . l Preparation of 1-Naphthol at 0.52 mol.nr 3 127 C.2 Typical Adsorption Test 127 C.3 Preparation of Diazotized Sulfanilic Acid 128 D Computer Program for Xs Determination 132 E Adsorption Data 134 F Semi-Batch Reactor Test Data 147 G Measured Xs vs. C m in the Pulp Suspension at Various Stirrer Speeds . . . . 158 viii List of Tables 2.1 Composition of the softwood fibre used in this study 23 2.2 Concentration and volume of the reactants used for dye adsorption study . 25 2.3 Variables affecting the diffusion limitation of primary reaction 37 2.4 Variables affecting the diffusion limitation of secondary reaction 38 2.5 95% confidence interval of Xs/Xsmit at various fibre mass concentrations . 40 2.6 Range of mass balance in water tests and in pulp tests at various mass con-centrations and the 95% confidence intervals 42 3.1 Variation of mass balance with mass concentration in water and in pulp suspension 55 3.2 Experimental conditions for the azo coupling reaction 61 3.3 Data Reproducibility of average energy dissipation rates in the stirred tank (with water and with 2% C m fibre) 64 4.1 Chamber and rotor dimensions of the high-shear mixer 91 4.2 Experimental conditions for the azocoupling reaction 92 4.3 Product distributions (Xs) with addition time (tj) in the high-shear mixer . . 94 A . l Fibre length distribution by Kajaani FS-200 analyzer 118 B. l Data reproducibility for 1-naphthol adsorption 122 ix B.2 Data reproducibility for dye adsorption (analytical) 123 B.3 Data reproducibility for dye adsorption (experimental) 124 B. 4 Data reproducibility for power measurement in the stirred tank 125 C. l Concentrations of chemicals for azo coupling reaction 126 C.2 Conditions for the azo coupling reaction 126 E . l Adsorption of 1-naphthol (Initial concentration = 0.500 mol/m3) on pulp suspension (C m = 1.8%) with increasing time 134 E.2 Adsorption of 1-naphthol (Initial concentration = 0.998 mol/m3) on pulp suspension (C m = 1.8%) with increasing time 135 E.3 Adsorption of 1-naphthol (Initial concentration = 1.505 mol/m3) on pulp suspension (C m = 1.8%) with increasing time 135 E.4 Adsorption of 1-naphthol (Initial concentration = 2.523 mol/m3) on pulp suspension (C m = 1.8%) with increasing time 136 E.5 Adsorption of 1-naphthol (Initial concentration = 0.499 mol/m3) on pulp suspension (C m = 0.9%) with increasing time 136 E.6 Adsorption of 1-naphthol (Initial concentration = 0.498 mol/m3) on pulp suspension (C m = 2.7%) with increasing time 137 E.7 Adsorption of 1-naphthol on pulp suspension with increasing concentration of 1-naphthol 138 E.8 XSinjt and Xs after adsorption of dyes at various mass concentrations ( C J of pulp suspension. Time of adsorption = 30 minutes 139 E.9 XSini/Xs with increasing mass concentration ( Q J of pulp suspension . . . . 139 x E.10 Concentrations of dyes before and after adsorption at various mass concent-rations ( C J of pulp suspension. Time of adsorption = 30 minutes . . . . 140 E . 11 Mass balances before and after adsorption at various mass concentrations of pulp suspension 141 E . 12 Equilibrium concentration and adsorption of monoazo dyes with increasing mass concentrations of pulp suspension 142 E. 13 Equilibrium concentration and adsorption of bisazo dyes with increasing mass concentrations of pulp suspension 143 E . 14 Change in Xs and mass balance with increasing time in dye solution and with pulp suspension 144 E.15 Change in concentration of dyes with increasing time in dye solution and with pulp suspension 145 E . 16 Change in Xs and mass balance with mass concentrations of different pulp fibre suspensions 146 F. l Power numbers of water in the stirred tank with varying Reynolds number . 147 F.2 Power numbers of sucrose solution (viscosity = 15.4 mPa.s) in the stirred tank with varying Reynolds number 147 F.3 Power numbers of sucrose solution (viscosity = 218.6 mPa.s) in the stirred tank with varying Reynolds number 148 F.4 Product distributions (Xs) in solution with increasing addition time (tj) in the stirred tank. Rotational speed = 2 s'1 149 xi F.5 Product distributions (Xs) with varying rotational speeds. Addition time (tf)=30 minutes 149 F.6 Corrected product distributions (Xs c o r r) in pulp suspensions with increasing addition time (tj) 150 F.7 Motion of the fully bleached kraft pulp suspension (C m = 0.5%) in the stirred tank at various rotational speeds (N) 150 F.8 Motion of the fully bleached kraft pulp suspension (C m = 1.0%) in the stirred tank at various rotational speeds (N) 151 F.9 Motion of the fully bleached kraft pulp suspension (C m = 1.5%) in the stirred tank at various stirrer rotational speeds (N) 151 F.10 Motion of the fully bleached kraft pulp suspension (C m = 2.0%) in the stirred tank at various rotational speeds (N) 152 F . l l Motion of the fully bleached kraft pulp suspension (C m = 2.5%) in the stirred tank at various rotations speeds (N) 152 F.12 Motion of the fully bleached kraft pulp suspension (C m = 3.0%) in the stirred tank at various rotational speeds (N) 153 F.13 Corrected product distributions (Xs c o r r) with varying mass concentrations of pulp suspensions ( C J at different rotational speeds (N=10, 12, and 14 s"1) 154 F. 14 Measured product distributions, concentrations of dyes and mass balances at various C m of the pulp suspension (FBK) at N = 14 s 1 155 F.15 Measured product distributions, concentrations of dyes and mass balances at various C m of the pulp suspension (FBK) at N = 12 s"1 156 xii F. 16 Measured product distributions, concentrations of dyes and mass balances at various C m of the pulp suspension (FBK) at N = 10 s"1 xiii List of Figures 1.1 Chemical notations and Symbols of primary reaction 6 1.2 Chemical notations and symbols of secondary reaction 7 2.1 Adsorption of 1-naphthol on fibre at various mass concentrations vs. time 27 2.2 Equilibrium adsorption of 1-naphthol on fibre vs. concentration of 1-naphthol after adsorption. Mass concentration (C n i) = 1.8% 29 2.3 1/c vs. 1/q* for calculating the equilibrium constant ( K J for Langmuir adsorption isotherm 31 2.4 Equilibrium adsorption of 1-naphthol on fibre vs. mass concentration ( C ^ 32 2.5 Adsorption of monoazo dye by the fibres vs. equilibrium concentrations of monoazo dye 34 2.6 Adsorption of bisazo dye by the fibres vs. equilibrium concentrations of bisazo dye 35 2.7 Xsjnit and Xs after adsorption vs. mass concentration (Cm) of fibres at various Xs^jt values 40 2.8 Xs/Xs^n vs. mass concentration of fibre 41 2.9 Mass balances before and after adsorption vs. mass concentrations of fibre 43 2.10 Variation of monoazo dye concentration with time - (a) in dye solution and (b) with pulp fibres 46 xiv 2.11 Variation of bisazo dye concentration with time - (a) in dye solution and (b) with pulp fibres 46 2.12 Mass balance vs. time - (a) in dye solution and (b) with pulp fibres . . . . 47 2.13 Variation of Xs with time - (a) in dye solution and (b) with pulp fibres . . 47 2.14 Change in Xs with mass concentrations of various pulp suspensions at X s ^ = 0.101 49 2.15 Change in mass balance with mass concentrations of various pulp suspen-sions at Xs;,* = 0.101 50 3.1 Dimensions of the stirred tank and Rushton turbine 57 3.2 Flow diagram of the apparatus 59 3.3 Photograph of the apparatus 60 3.4 Impeller characteristics of the tank 66 3.5 Variation of Xs with addition time at impeller rotational speed = 2 s'1 . . 68 3.6 Xs vs. rotational speed (N) at addition time = 30 minutes. Dashed line show the values calculated using Baldyga and Bourne's model 69 3.7 Velocity distribution in the stirred tank reactor . , 70 3.8 Energy distribution in the stirred tank reactor „ 71 3.9 Xs vs. Reynolds number at addition time = 30 minutes. Dashed lines show the values obtained by Bennington and Bourne in a geometrically similar tank 73 3.10 Variation of power vs. water and pulp suspension (C m = 2%) 75 xv 3.11 Variation of power with impeller rotational speeds at various mass concent-rations ( C J of pulp suspension (FBK) 77 3.12 Photograph showing the motion of the fully bleached kraft pulp suspension (C m = 2%) in the tank at rotational speed (N) = 6 s"1 78 3.13 Photograph showing the motion of the fully bleached kraft pulp suspension (C m = 2%) in the stirred tank at rotational speed (N) = 8 s"1 79 3.14 Photograph showing the motion of the fully bleached kraft pulp suspension (C m = 2%) in the stirred tank at rotational speed (N) = 10 s"1 81 3.15 Photograph showing the motion of the fully bleached kraft pulp suspension (C r a = 2%) in the stirred tank at rotational speed (N) = 13 s"1 82 3.16 X s c o r r vs. addition time (t) with pulp suspension (C m = 2%) at rotational speed = 12 s'1 83 3.17 X s c o r r vs. mass concentrations of pulp suspensions at various rotational speeds (N = 10, 12 and 14 s"1) 84 3.18 X s c o r r vs. average energy dissipation rate with pulp suspensions at various mass concentrations (Cn i) 86 4.1 Dimensions of the high-shear mixer and feed points 90 4.2 Xs vs. addition time in the high-shear mixer 96 A . l Length weighted distribution of the F B K pulp fibres used in this study . . 116 A.2 Population distribution of the F B K pulp fibres used in this study 117 G. 1 Measured product distribution of the F B K pulp fibres at various mass con-centrations at stirrer speed (N) = 10, 12 and 14 s"1 158 xvi Acknowledgements M y sincere thanks to Dr. C.P.J . Bennington, for his guidance, encouragement and valuable discussions throughout the course of this thesis work. I appreciate the advice and suggestions of the supervisory committee, Dr. R.J . Kerekes, Dr. K . L . Pinder and Dr. P. Englezos. I thank the members of the Pulp and Paper Centre and the Chemical Engineering Department for their help in many aspects of this work. I thank the Ministry of Human Resource Development, Government of India, for sponsoring me as a commonwealth scholar, and my employer, Seshasayee Paper Mills , for providing me the required study leave. Financial support from the Canadian Commonwealth Scholarship and Fellowship Committee, Government of Canada, is greatfully acknowledged. Finally I would like to thank my wife, Damayanthi, and my son, Anush, for their support and patience. xvii Chapter 1 Mixing Pulp Suspensions 1.1 Introduction One of the most widely encountered process operations is mixing. Wohl (1968) comments in his review paper "... of all the unit operations, mixing is the one with the least theoretical background. Most of the design methods are based upon empirical correlations rather than on comprehensive analyses of the fluid flow regime". Inadequate mixing can be usually be detected by unsatisfactory product quality, however, over-design cannot be detected by product quality. Thus, lack of knowledge of mixing may lead to overdesign, and will result in higher capital and operating costs. Mixing is a central feature of many processes in the pulp and paper industry and the financial investment in both the capital and operating costs is considerable. In the pulp and paper industry mixing is very important in pulp bleaching operations. Improper mixing in the bleaching stage will lead to poor product quality, fibre damage and wastage of chemical. For example, improved mixing reduces chemical consumption and leads to a more uniform product (Atkinson and Partridge, 1966; Elliot and Fair, 1973; Robitaille, 1987). Since bleaching is a chemical reaction between the fibre and the bleaching chemical, mixing on 1 molecular dimensions is important. Few techniques are available to assess macroscale (Kolmodin, 1984; Bergnor, 1985) and microscale (Paterson and Kerekes, 1986) mixing in pulp suspensions. Techniques such as L D A measurement of the velocity distribution are limited to very dilute fibre concentrations (Molen and Maanen, 1978). So some other methods are required to assess mixing in pulp suspensions. In this thesis an attempt was made to assess mixing in pulp suspensions at the fibre-scale and microscale using mixing-sensitive chemical reactions. A laboratory technique using mixing sensitive chemical reactions was developed by Bourne and coworkers to assess mixing and turbulence in aqueous media and used to study a variety of mixers (Bourne and Toor 1977; Bourne, Kozichi and Rys,1981; Bourne and Rohani 1983; Bourne, Hilber and Tovstiga, 1985). This technique uses a set of competitive, consecutive chemical reactions which are mixing-sensitive. The distribution of reaction products is used to ascertain the mixing and turbulence. This technique was applied to fibre suspensions at low mass concentrations (C m < 1%) and found successful (Bennington and Bourne, 1990). In this case special attention to adsorption of reactants and products had to be made, i.e. fibre suspensions at low mass concentrations were utilized after saturation of adsorption sites so that differential adsorbtion of the product dyes was avoided. Thus the product distribution, used to indicate the mixing quality, did not change due to the presence of the fibres and could be used to indicate mixing quality irrespective of the presence of fibres. This present work attempts to account for adsorption so that special precautions like the saturating of the adsorption sites is not necessary. Also, by extending the test to higher mass concentrations (10%), a wider range of pulp mass concentrations can be assessed. 2 1.2 Mixing Concepts The term mixing is applied to operations which tend to reduce nonuniformities in the composition, properties, or temperature of material in bulk. Mixing is accomplished by movement of material between various parts of the whole volume. Mixing reduces the scales of inhomogeneities down to molecular dimensions. In turbulent single phase fluids it is very difficult to model the transport phenomena in full physical detail (Bourne, 1985). Beek and Miller (1959) proposed the following mechanism: (a) Distributive mixing: Relatively large eddies exchange positions and convect material so that, at a scale of observation which is coarser than the eddy size, macroscopic uniformity of concentration results. At a scale much smaller than the eddy size no significant mixing occurs. (b) Dispersive mixing: The larger eddies of (a) decay in size through the effects of turbulent shear and a finer-grained mixture is formed. At molecular scale the mixture remains, however, highly segregated. (c) Diffusive mixing: Diffusion within the finely dispersed structure operates over short distances and proceeds to randomize the mixture at the molecular scale. The result is termed a homogeneous mixture. The non-uniformity in a mixing process can be broadly divided into various scales. According to the definition given above distributive mixing belongs to the macroscale range. (approximate range is greater than 10 mm and less than the vessel dimensions). The diffusive mixing belongs to microscale (comparable to molecular dimensions). Turbulent mixing results in many fluid eddies of various sizes. It has the ability to mix different solutions of different compositions. Turbulent mixing occurs in the turbulent flow range. For stirred tanks flow is considered turbulent when the power number reaches a constant value. For stirred vessels this is reached at impeller Reynolds numbers greater than 3 10 000 (Edwards, 1985). Turbulence has no effect on a scale smaller than the Kolmogoroff velocity microscale (which is a function of kinematic viscosity and energy dissipation). In turbulent aqueous solutions this is on the order of 10-30 /xm — the diameter of the smallest eddy. Eddies smaller than this size are destroyed by viscous shear. Flow within an eddy is laminar. As the smallest eddies contain about 104 molecules, all eddies are of macroscopic size and turbulence is not a molecular phenomenon (McCabe and Smith, 1986). The main processes of micromixing are engulfment, deformation and molecular diffusion. In a turbulent regime diffusion occurs within vortices causing mass transfer between the slabs that build the vortex. Diffusion is accompanied by fine scale deformation leading to the disintegration of eddies caused by inertial and viscous forces (Bourne, 1985). Molecular diffusion allows the movement of the different reacting molecules across the boundaries of elements (for example small eddies). This reduction can occur with or without turbulence. However, turbulence increases the rate of the process by breaking the fluid elements or eddies into smaller ones, and thus allowing more area for molecular diffusion. 1.3 Mixing in Pulp Suspensions In pulp suspensions mixing is produced by creating relative motion within the pulp suspension. The complex rheology of pulp suspensions makes this difficult. Since bleaching is a chemical reaction between one or more bleaching agents and with the lignin located in fibre, the quality of mixing is responsible for the quality of the product. As fibre suspensions are discontinuous over short distances the fibre-scale mixing is important in pulp suspensions. The fibre-scale represents non-uniformities having a size comparable to fibres 4 and floes (an approximate size range is 0.05-10 mm, from a fibre diameter to a floe length). Fibre-scale mixing is attained by fluid shear and turbulence which breaks the floes up and leads to fibre-scale uniformity. 1.4 The Kinetics of The Tracer Reaction 1.4.1 The Reaction Used The reaction used to measure large scale as well as the fine scale concentration inhomogeneities is the well defined coupling of diazotized sulfanilic acid with 1-naphthol (Bourne et al. , 1981 ; Bourne et al., 1985). This is a competitive-consecutive reaction system and similar to reactions used by other workers (Paul and Treybal, 1971; Truong and Nethot, 1976; Bourne and Kozichi, 1977; Bourne, Rys and Suter, 1977). It is characterized by a fast primary reaction of 1-naphthol (A) with diazotized sulfanilic acid (B) to form 4-(4'-sulfophenylazo)-1-naphthol (R), and a slow secondary reaction of B with R to form 2,4-bis-(4 ' -sulfophenylazo)-1 -naphthol (S). A + B * R (1.1) k2 R + B » S (1.2) At T = 25°C: k, = 12 000 m 3. molds' 1, k 2 = 2.44 m 3 .molds' 1 (Bourne et al. , 1985). The chemical structures of the reactants and product dyes are given in Figures 1.1 and 1.2. In these reactions B is kept as the limiting reagent. The extinction coefficients for the product dyes are available (Bourne et al., 1985) and therefore the concentration of C R and 5 6 7 C s can be spectrophotometrically determined by measuring the extinction coefficients at several wavelengths in the range of 400-640 nm. Product mixtures were analyzed by assuming that the Lambert-Beer Law is valid at the concentration level used here (101—102 mol.m 3). According to Lambert-Beer Law: E = eCÔ (1.3) At a given wavelength (X), the absorption over a path length Ô is: E = eR[R]ô 4- es[S]5 (1.4) And upon rearranging: — = m + (i.5) eR5 sR From the measured values of absorption E at various wave lengths, a linear regression of E/eRô against es/eR was made. This was done using a computer program which gave the concentration of the dyes, mass balance and the goodness of the fit (r2). (See Appendix D). The azo coupling reactions are very rapid (half-life time < 1 s) and proceed during the mixing of the two reagent solutions (Bourne, Kozichi, Moergeli and Rys, 1981). Before the time the average concentration is established in the reactor the reactions have already taken place. Therefore the local reactant concentration in the reaction zone determines the rate of reaction. The observed reaction rate and in the case of a competitive-consecutive reaction the product distribution depend upon mixing. Thus the mixing process determines the product distribution which is defined as Xs = 2 C S / (2C S + CR) (1.6) 8 which is the fraction of the limiting reagent (B) present as bisazo dye (S). By measuring the product distribution the mixing is assessed. In the chemically controlled regime the reaction rate is slow compared to mixing. As the primary coupling rate (Eqn 1.1) is much faster than the secondary coupling, the product distribution Xs in a chemically controlled regime is zero (Xs -* 0) i.e. no S is formed (Bourne et al., 1985). In the mixing controlled regime the mixing is slow compared to reaction. So the R dye produced reacts with B forming S. The product distribution in this region is unity (Xs -» 1), i.e. all the R is converted into S. Any deviation from zero indicates the existence of inhomogeneities on either the macroscale or the microscale. So the Xs-values between these two limits (zero and one) indicate the degree of unmixedness (micromixing and macromixing). However, by adding B in a semi-batch mode with sufficient time for complete macromixing to occur, Xs is only a function of the micromixing. Hence Xs is a measure of micromixing and turbulence when the addition time is greater than the macromixing or bulk blending time. 1.4.2 Pre-Equilibria and Salt Effects 1-naphthol, diazotized sulfanilic acid and 4-(4'-sulfophenyl-azo) 1-naphthol (R) are subject to ionic equilibria. The reactive species in the system are in ionic form. The equilibria are strongly pH dependent and have great influence on the reaction rates (Bourne et al, 1985). The reactive species in reaction (1.1) are 1-naphtholate (AO") and the diazonium ions ( A r N 2 + ) . In reaction (1.2) the reactive species are the naphtholate form of R, viz. RO", and the diazonium ion A r N 2 + . The first substitution occurs in the p-, and not in the o-position. This was confimed by thin layer chromatography (Bourne et al. , 1985). In this paper the 4-9 (4'-sulfophenylazo)- 1-naphthol (R) is called the monoazo dyestuff and 2,4-bis(4'-sulfophenylazo)-1-naphthol (S) is called the bisazo dyestuff. In order to enhance mixing effects the reaction was carried out at pH 10, where the reagent molecules exist almost completely in their ionic form and thus the maximum chemical reaction rate is achieved (Zollinger, 1961). A solution of sodium carbonate and sodium bicarbonate was used as the buffer system. To avoid any local pH-gradients in the reaction zone, a buffer concentration of 75 times the 1-naphthol concentration was used for both buffer components. This concentration is higher than the minimum buffer concentration necessary to avoid pH-gradients in the reaction zone (Bourne et al., 1985). 1.4.3 Side Reactions The primary coupling of diazotized sulfanilic acid with 1-naphthol proceeds mainly in the para-position of the 1-naphthol. However, approximately two to three percent of the naphthol form ortho products (Bourne et al., 1985). By secondary coupling this then reacts at the more reactive para-position to form S. The rate constant for coupling at the para-position in a similar coupling was one order of magnitude higher than for the ortho-position. Thus the formation of S, starting from 1-naphthol, proceeds through two secondary couplings: K o-R + B » S (1.7) p-R + B *S (1.8) o-R means the monoazo dye, coupled in the c-position, so that secondary coupling occurs 10 in the /^-position with rate constant kp. p-R means the bisazo dye coupled in the p-position so the secondary coupling occurs in o-position with rate constant k0. In recent studies (Bourne, Kut, Leszner and Maire, 1990b) introduced a definition of the product distribution which includes the production of the ortho monoazo dye; Xs = 2C S / ( C U + C, , R + 2C S) (1.9) With Xs>0.15 both methods give identical results within experimental limits of spectrophotometric analysis ( ± 0.005 Xs). With Xs<0.15, the following empirical correlation between the 2 component-model and the three-component model is proposed by Bourne et al., (1990b). Xs ' = 0.884 Xs + 0.0203 (1.10) where Xs ' is the product distribution obtained using the three-component method. Studer (1989) proposed another correction using the H P L C (high pressure liquid chromatography) studies: Xss = -1.5548e-2 + 1.1427xXs(HPLC) (1.11) where Xss is the product distribution by two-component analysis (Eqn 1.6) and Xs (HPLC) is the product distribution obtained by H P L C which includes ortho monoazo dye (Eqn 1.9). The new three-component spectrophotometric analysis proposed by Bourne et al. (1990b) has been found to give inadequate resolution between the monoazo isomers because of their overlapping spectra. To avoid complications the small amount of o-R formed is neglected in the present work. Excess amount of diazonium ions degrade the bisazo dye (normally at high bisazo dye concentrations i.e. Xs>0.4). Since diazotized sulfanilic acid is the limiting reagent in the 11 micromixing study this degradation is avoided. Temperatures over 300 K degrade the bisazo dye (Bourne et al., 1985). Since the temperature used in this study was 298 K (close to 300 K) it was necessary to check i f dye degradation was a problem. 1.4.4 Product Distribution When B is maintained as the limiting reagent, Xs depends only upon the following factors: the ratio of the volume of 1-naphthol to diazotized sulfanilic acid, the ratio of the number of moles of 1-naphthol to number of moles of diazotized sulfanilic acid, the type of reactor, the ratio of the rate constants, the Schmidt number and mixing modulus. These can be expressed as: Xs = fWJNto, * A , reactor type, S c , VA/VB, M) (1.12) When the experimental conditions are fixed (choosing a reactor and selecting reactant concentrations, volumetric ratios and the operation temperature), Xs depends only on the mixing modulus, M , where M = * 2 C t o ô 0 / D (1.13) and as k2, C B o , and D are fixed by the experimental conditions, Xs depends only on ô0, where ô 0 = 0.5 X k = 0.5(vVe) I / 4 (1.14) Thus, Xs depends only on the local energy dissipation in the reaction zone. The turbulence in the mixing zone depends on the local energy dissipation (Baldyga and Bourne, 1984a, 1984b). More turbulence will lead to a large number of small size eddies with larger areas available for diffusion. Thus Xs is a measure of turbulence and micromixing in the mixing 12 zone. The local value of energy dissipation in a stirred tank varies with position over typically two to three orders of magnitude. The local value can be related to the average value (Baldyga and Bourne, 1988). The reaction zone in the azo coupling reaction spreads over various energy zones depending upon the average energy dissipation rate in the tank. 1.5 Pulp Suspension Rheology Pulp suspensions can consist of three phases: solid (the pulp fibres), liquid (usually water) and gas (air or bleaching chemical). However, in many applications, pulp suspensions can be considered to consist primarily of water and fibre. This occurs for low consistency pulp suspension ( C m < 6%). In these cases the mass concentration or consistency ( C J is sufficient to describe the suspensions. Pulp fibre suspensions are non-Newtonian fluids, possessing a yield stress and shear-thinning behaviour. Thus pulp suspensions are not only pseudoplastic but also have Bingham plastic property (Silvester, 1985; Bennington, Kerekes and Grace, 1989). This complicates the mixing process. Pulp suspensions have low apparent viscosities at high shear rates. In the high shear region surrounding a rotating impeller in a mixing vessel, the pulp suspensions will flow easily. In regions of low shear (remote from an impeller) high apparent viscosities are expected, which may result in low (or no) flow and hence poor mixing. Pulp suspensions form networks i f they have sufficient concentration. This concentration is called the "critical concentration", C c , and according to Mason (1954) this can be defined as follows: 13 When the concentration of a pulp suspension is increased, a critical concentration will be reached, above which the fibres cannot undergo unimpeded rotation, simply because there is insufficient space in which to move. C c = 1.5/r2 (1.15) where r = 1/d, 1 = fibre length and d = fibre diameter. The r value is 70 (1 = 2.41 mm, d = 35 ixm) for the F B K pulp used in this study and the critical concentration obtained by using Eqn 1.15 is 0.031%. i.e. above this concentration this pulp suspension forms networks. Therefore mixing complexities are expected at the concentration levels used in this study (0.5-3%). In a series of studies published in the mid-1960's, Wahren and co-workers showed that pulp networks formed at concentrations above the "critical" level had various properties of a solid: 'elasticity', rupture strength and viscoelasticity (Wahren, 1964). Meyer and Wahren (1964) proposed an explanation for this network strength, attributing it to the bending of fibres in the following manner: In shear flow, fibres bend and entangle, but once the shear stops, fibres cannot fully unbend because they are restrained by contact with other nearby fibres. Normal forces arise between the fibres. These in turn produce the frictional forces that provide the fibre network strength. As the consistency of the suspension increases, the number of fibre/fibre interactions increases, and this in turn increases network strength. The distribution of fibres in networks is never uniform, and local mass concentrations of fibres give rise to floes within the suspension. Since network strength depends on the number of fibre contacts, floes have a higher strength than the surrounding suspension. Thus, floes are not only regions of higher fibre mass concentration but they are also regions of greater strength than the surroundings. Thus, pulp network strength properties are very dependent upon fibre concentration, modulus of elasticity of the fibres, and the length to diameter ratio of individual fibres. The pulp network strength properties are also dependent upon the flow conditions which 14 cause fibre entanglement to take place. Hence a critical shear stress, the yield stress, is required to be applied to the suspension before flow occurs. This may lead to the formation of regions of stagnant fluid, where shear stresses are below the fluid's yield stress. This is important in pulp suspension mixing vessels where caverns of mixed suspension may form around a rotating impeller with the rest of the suspension stagnant. 1.6 Mixing in Pulp Bleaching The introduction and mixing of bleaching chemical with brown stock is one of the critical steps in the bleaching sequence. Whether the bleaching agent is chlorine, chlorine dioxide, or oxygen, it is important to have an efficient mixing system so that each fibre is contacted with the same strength of chemical for a constant period of time. The system has to be set up to prevent overbleaching which would cause drops in both pulp viscosity and strength and underbleaching which would cause a drop in pulp brightness, so that yield and pulp quality are maximized with minimum chemical consumption. In pulp suspension bleaching the removal of lignin from the fibres is the most important step. In pulp bleaching one of the reactants is lignin (on the fibre) and the other is chlorine or other bleaching agent. The lignin in unbleached pulp samples is about 5 % by weight. This is distributed all over the fibre in smaller dimensions, i.e. molecular dimensions. In order that bleaching is more effective each fibre should get an equal amount of bleaching agent or agents which reacts with the lignin in the fibre. So there should be fibre-scale mixing which is achieved by fluid shear and turbulence. Since Xs is a function of the energy dissipation (i.e. turbulence) in the mixing zone, it can be used to assess fibre-scale mixing and turbulence in the pulp 15 suspension. 1.7 Research Objectives In this thesis an attempt is made to assess mixing in pulp fibre suspensions using the mixing-sensitive chemical reaction developed by Bourne. The following were studied as part of the thesis: 1. Study of adsorption of chemicals (reactants and products) onto various types of fibres in order to correct the product distribution, Xs, for the presence of fibres. 2. Macroscale mixing time (aqueous medium and in pulp suspension) in a stirred tank reactor operating on a semi-batch mode. 3. Microscale mixing and turbulence in a stirred tank reactor in aqueous media to compare with results from previous studies and with results where pulp fibres are present. 4. Study of microscale mixing and turbulence in low consistency (0.5 to 3.0% C n i) in a stirred tank reactor and at medium consistency (10% C J in a high-shear mixer. 5. Assess the turbulence in the high-shear mixer with the measured values of energy dissipation and product distribution. 16 Chapter 2 Adsorption of Chemicals onto Cellulose Fibres 2.1 Introduction One important factor in applying the diazo-coupling reaction in the study of micromixing in the presence of pulp suspensions is the adsorption of 1-naphthol and the product dyes on the fibres and absorption through lumen and pits in the fibre walls. Since fibres are made up of cellulose and cellulose has attraction for ionic species, the effect of adsorption on the product distribution is important as it affects the reactions as well as the concentration of the products. Firstly, the adsorption of 1-naphthol on fibres may change the ratio of the amount of reactants available for reaction. This may slightly alter the product distribution from the case where no adsorption occurred (See Equation 1.12 which shows that Xs is a function of N A o / N B o ) . Secondly, i f the monoazo and bisazo dyes produced in the coupling reaction are equally adsorbed (with respect to their equilibrium concentrations), then the product distribution will be the same irrespective of the quantity of dye adsorbed on the fibres. On the otherhand if the dyes adsorb differentially then the product distribution will differ and wil l be greater or lesser depending upon which dye adsorbs preferentially. In this case, a correction is necessary to determine the product distribution in the absence of adsorption and 17 to relate it to the micromixing quality and energy dissipation within the vessel. 2.2 Literature Search 2.2.1 Isotherms The relation between the concentration of a substance that is distributed, at constant temperature, between two phases is termed the distribution isotherm. The distribution process is called adsorption if the substance which is distributed is retained by a surface but absorption i f this occurs in the interior of a body. When it is not known where the particles are retained the process is termed sorption (Mantell, 1951). Since the relationship between the different intermolecular interactions responsible for sorption in a true dyeing system are incompletely known, it is difficult to differentiate experimentally between the different interacting forces. The Langmuir isotherm and the Linear isotherm were encountered in this study. The Langmuir adsorption model assumes that adsorption occurs on localized sites with no interaction between adsorbate molecules, and that maximum adsorption occurs when the surface is covered by a monolayer of adsorbate (McCabe et al., 1986). The Langmuir adsorption coefficient ( K J is described by q* = Q K L c / ( l + K L c ) (2.1) where q* = Amount of adsorbate adsorbed per gram of adsorbent Q = Solid phase solute concentration corresponding to complete saturation of available sites 18 c = Residual liquid phase concentration at equilibrium. Values of K L and Q are determined by plotting 1/q* vs. 1/c. The Linear isotherm is the simplest model of all isotherms and is described by an equation of the form q e = K p c e (2.2) where Kp is the partition coefficient, c e is the equilibrium liquid phase concentration and q e is the equilibrium adsorption. This model describes a given set of adsorption data in terms of a single parameter, Kp. A l l adsorption isotherm models tend to predict linear sorption at low solution concentrations. The linear relationship is seldom valid over large ranges of concentration and it should not be used for extrapolation beyond the limits of a particular data set (Walter and Weber, 1985). 2.2.2 Direct Dyes Direct dyes are compounds which are able to dye cellulose fibres without any aid. This is due to of their affinity for cellulose. Anionic mono-azo dyes have some affinity for cellulose (Rys, 1972). This affinity can be increased by enlarging the planar conjugated system (alternate double bonds which favour resonance and stabilization). This is done by the introduction of further azo groups, resulting in bis-, tris-, and higher poly-azo dyes (Ryes, 1972). 2.2.3 Forces of Attraction The main forces responsible for adsorption in a dyeing system are Electrostatic forces, Hydrogen bonds and London forces. Ionic interactions are not possible when dyeing cellulose 19 with anionic dyes (diazo dyes) under basic conditions because cellulosic fibres are negatively charged. But dipole-dipole interactions are possible with polar molecules. These are considerably weaker than ionic or covalent bonds — they are only about 1 % as strong as ionic bonds. Their strength decreases very rapidly as the distance between the dipoles increases (Bradly, 1982). When a covalently bound hydrogen atom exists between two electronegative atoms, a hydrogen bond may be formed. Electronegative atoms like oxygen and nitrogen are fairly common in dye molecules and substrates. The hydrogen bond is a relatively weak one, and is about 5 to 10% as strong as an ordinary covalent bond (Bradly, 1982). Its formation also involves a low activation energy. Aromatic compounds, like 1-naphthol with electron withdrawing groups in the ring, form more powerful hydrogen bonds. Both dyestuffs and cellulose have many groups capable of entering into hydrogen bonding. Cellulose contains hydroxyl groups and dyes have azo groups which can make hydrogen bonds with the fibre. But the formation of hydrogen bonding between dyes and substrates in the presence of water is complicated. In many cases it makes no contribution to dye binding (Rattee, 1974). Nonpolar molecules and uncombined atoms experience weak attractions and these attractions are called London forces. When electrons move about in an atom or molecule, their motion is random, so that at any given instant there is a chance that more electrons will be on one side of the particle than on the other. At that particular instant the particle will be a dipole which is called instantaneous dipole as its existence is only momentary. Because of the presence of these dipoles, the molecules attract each other, and that attraction produces a momentary tug that helps hold them together (Bradly, 1982). The energy of attraction 20 diminished as the inverse sixth power of the distance between the molecules. So the London forces are effective only at very short distances. The stereochemistry of the dye molecules determines the closeness and the number of dye molecules that can approach the fibres. If more dye molecules can approach the fibre molecules the attraction is stronger. With dyes which have an affinity for cellulose this is favoured by the planar structure of the dye ions and the inner surface of cellulose. According to Dada and Wenzel (1991) the capacity of the adsorbent will depend on the orientation, packing density and the nature of molecules in the adsorbed phase. 2.2.4 Dye Aggregation Dye molecules can interact not only with adsorbing substrates such as cellulose, but also with one another. Dye aggregation, or association, is the interaction of dye molecules with one another. If there is aggregation between the dye molecules the concentration of the dye will change with time. By measuring the dye concentration over a period of time the state of aggregation can be determined. 2.2.5 Pulp Fibres The pulp fibre used in this study consist of softwood species or conifers. The density of the cell wall on an ovendry basis is about 1.5 g/cm3 (Stone and Scallan, 1966). The wall of a fibre composed of several layers. Each fibre has a primary wall and a 3-layered secondary wall with specific alignment of fibrils. Fibrils are aggregates of cellulose molecules, and their orientation can influence the characteristics of a pulp fibre. The fibre at the molecular level consists of microfibrils of cellulose embedded in a hemicellulose-lignin 21 matrix, arranged in an interrupted lamellar structure (Kerr, 1975). The hydrophillic properties of cellulose are due to the chemical constituents, namely the hydroxyl (OH) groups, located along cellulose chains. The OH-groups are capable of hydrogen bonding with water molecules, and thus impart specific properties to the water directly in contact with, or in the vicinity of them. These properties are different from those of water not directly in contact with cellulose and have been reviewed by Corte (1980). 2.2.5.1 Surface Charge of Cellulose Fibre Cellulose, as such, is negatively charged throughout the whole pH-regime (Clark, 1986). However, the fines fraction for many types of fibrous material has a higher content of noncellulosic material and thus a higher content of acidic groups (Kallmes, 1960). In the case of cellulosic surfaces, the present evidence suggests that the dissociation of the ionic groups is the main source of the charge (Strazdins, 1974). The major contributors to the fibre charge are the carboxylic groups present in the cellulose and hemicellulose components of pulp (Das and Lomas, 1973). An increase in pH or carboxyl group content increases the surface charge on the cellulose (Lindstrom, 1974). 2.2.5.2 Adsorption of Dyes by Lignin in Fibre Bleached pulp fibres consist of residual lignin. This is less than 0.5% for the fully bleached kraft pulp used in this study (Sjôstrom, 1989). Even though the amount of lignin is low it plays an important role in dye bonding. This is because of the sulphonic acid groups in the dyes. The sulphonic groups form lignin sulphonates as in the case of sulphite pulping. Since bisazo dye has two sulphonic groups it has more chances for bonding with 22 lignin in the fibres when compared to monoazo dye which has one sulphonic acid group. The /3-aryl ether bonds of lignin are cleaved by the sulphonic acid group to give styrene /3-sulphonic acid structure (Sjôstrom, 1989). The lignin is distributed all over the fibre. So only the lignin present on the surface is available for the bonding but the availability of the lignin inside the fibre wall may be diffusionally limited. 2.3 Experimental 2.3.1 Fibre Used The pulp fibre used in this study was obtained in sheet form (dried) from Western Pulp Partnership Ltd, Squamish Operations. It was a fully bleached kraft (FBK) pulp containing fibres from three different softwoods in the following ratio. Table 2.1: Composition of the softwood fibres used Fibre Type Weight % Douglas Fir 30 Cedar 30 Hemlock/Balsam 40 The coarseness of the fibre sample was = 0.214 mg/m. The length weighted average fibre length was 2.40 mm. The arithmetic average fibre length was 0.93 mm. The length weighted distribution and the population distribution are given in Appendix A .2 . 23 2.3.2 Equipment, Chemical Preparation and Procedures for Adsorption of 1-Naphthol The main equipment used for the adsorption study was a stirred tank, a Haake F3 temperature controller and a Varian Techtron UV-Visible Spectrophotometer Model Cary 1. The British Standard Disintegrator was used to disintegrate the pulp samples and a laboratory stirrer was used to mix the 1-naphthol with the pulp fibres. The adsorption of 1-naphthol was studied at mass concentrations (C,J of 0.9, 1.8 and 2.7% and at 1-naphthol concentrations 0.46, 0.92, 1.40 and 2.35 mol/m 3. The required amount of pulp for various consistencies was soaked in distilled water overnight and disintegrated in a British Standard Disintegrator for 15,000 revolutions. The pulp suspension was filtered over a filter paper on a Buchner funnel. 60 g of water was allowed to remain with the fibre in all the cases (Appendix C . l ) . The preparation of 1-naphthol solution is given in Appendix C . l . Pulp samples of a given mass concentration were prepared for study as follows. The required mass of pulp was added to the 1-naphthol solution after buffering to give the required suspension mass concentration. The total volume of the 1-naphthol solution after pulp addition was 1.06 L (60 grams water was added to 1 L of the 1-naphthol solution with the fibres). The water added with pulp reduced the concentration of 1-naphthol by about 5%, and was accounted for. The suspension was mixed using a stirrer and samples were taken at 3 minute intervals using a plastic syringe. Sampling was done in the top portion of the suspension (approximately half the distance from the stirrer shaft and vessel wall). The fibres were separated from the solution by filtering through a sintered glass funnel. Filter paper was not 24 used as it contains cellulose and would have adsorbed some of the 1-naphthol. The filtered 1-naphthol concentration was determined by measuring the U V absorbence in 1 cm quartz cuvettees on the Varian Techtron Cary 1 spectrometer at a wave length (X) of 332 nm. A sample of distilled water at the same pH equilibrated with the fibres in the same way as the sorbate solution was placed in the reference cuvette. From the concentration decrease in the solution, the amount adsorbed per gram of fibre at various time intervals was calculated. 2.3.3 Equipment, Chemical Preparation and Procedure for Dye Adsorption The volume of 1-naphthol used for each run in this study was 2 L (Appendix C. 1). 1 L of diazotized sulfanilic acid was prepared (Appendix C.2). Chemical concentrations and other details are given in table 2.2. Table 2.2: Details of concentration and volume used for dye adsorption study Volume of 1-naphthol (A) for each run 2 L Volume of diazotized sulfanilic acid (B) for each run 4 0 cm 3 Concentration of A 0 . 5 2 mol/m 3 Concentration of B 2 3 . 6 mol/m 3 N A O 1.04 mmol N B o 0 . 9 4 mmol v A / v B 5 0 N a 0 / N B O 1.106 Various concentrations of the dyes were prepared by varying the mixing conditions in a 25 4 L beaker using a laboratory stirrer (by altering the impeller rotational speed and the feed location). Pulp samples equivalent to 0.5, 1.0, 1.5, 2.0, 2.5, 6.0, and 10% mass concentrations (C i n) in a total volume of 200 cm 3 were used for each test at each product distribution (175 cm 3 dye was used for each case and the amount of water left in the pulp samples after filtering was 25 cm 3 in each case). The water added with the fibre reduced the initial concentration of the dye and was accounted for. Pulp samples of known mass corresponding to the various mass concentrations were immersed in the dye solution for 30 minutes (the sorption of 1-naphthol reached equilibrium after 30 minutes as shown in Figure 2.1). The samples were stirred at regular intervals. After 30 minutes the samples were filtered using a sintered glass funnel. The dye sample was diluted by a factor of 12 and the absorption at wavelengths between 400 and 600 nm measured in the Varian spectrophotometer. Seven measurements were taken in each case (at wavelengths 480, 510, 540, 570, 580, 590 and 600 nm). From the absorption measurements the concentration of monoazo and bisazo dyes were calculated by linear regression using a computer program (Appendix D). The concentration of the dye before the addition of fibres was also measured. From the difference in the concentration of dyes before and after adsorption by pulp fibres, the amount of monoazo and bisazo dye sorbed by the fibre was calculated. The mass balance and the product distribution were also calculated. 2.3.4 Experiment for Dye Aggregation For the study of dye aggregation the following experiment was carried out. The coupling 2 6 250 Si £ 200 £ O) E o O £ c o o "O < 150 100 50 0 */ / / ft ' 10 20 30 Time of Adsorption, minutes Figure 2.1: Adsorption of 1-naphthol on pulp suspensions at various equilibrium concentrations of 1-naphthol vs. time at 25+0.1 °C. Fully bleached kraft pulp fibres. Mass concentration = 1.8%. 27 reaction was carried out at moderate stirrer speed to obtain the dyes at a representative concentration (Xs = 0.16). To a portion of the dye a disintegrated pulp sample of known mass was added. Dye samples were taken at 30 minute intervals (filtered using sintered funnel) for a period of 3 hours. Dye samples without fibre were also taken in the sameintervals. The concentration of the dyes were measured using the spectrophotometer. If there was aggregation then the dye concentration would change with time. From the results the state of dye aggregation was determined. 2.4 Results and Discussion 2.4.1 Adsorption of 1-Naphthol The adsorption of 1-naphthol on cellulose at various time intervals with varying concentration of 1-naphthol is shown in Figure 2.1. The adsorption was very fast and more than 90% of the adsorption is over in the first 15 minutes. This shows that the surface adsorption was very fast due to agitation of the sample and that the bulk diffusion of 1-naphthol into the fibres was slow (Larson and Stenius, 1987). The adsorption increased with increasing concentration of 1-naphthol and followed the Langmuir isotherm in the concentration range as shown in Figure 2.2. Experiments for adsorption were conducted up to 1-naphthol concentrations 2.35 mol/m3 (5 Xconcentration used in this study for coupling reaction). 2.4.2 Data Reproducibility The error associated in this experiment consisted of two parts: analytical and 28 0.0 0.5 1.0 1.5 2.0 2.5 Concentration of 1-Naphthol, mol/m3 Figure 2.2: Effect of concentration of 1-naphthol on equilibrium adsorption of 1-naphthol on pulp suspensions at 25 ± 0 . 1 °C. Fully bleached kraft pulp fibres. Mass concentration = 1.8%. 29 experimental. The analytical error was due to the spectrometer and the regression analysis used for the study. The data reproducibility was ± 0.001 absorption units (i.e. less than 0.1% of the equilibrium concentration of 1-naphthol). The error associated with the experiment was higher. For 1-naphthol adsorption the 95% confidence limit ranged from+5.25 to ± 10.4% of the equilibrium adsorption of 1-naphthol per gm of fibre (Appendix B. 1). The error was high as the experiment involved many steps like weighing, filtering and diluting the sample before analysis in the spectrometer. The Langmuir adsorption coefficient (K L) was calculated from the slope and intercept of Figure 2.3 which was found to be 1.73 L/mmol of 1-naphthol. Similar results for the adsorption of chemicals on pulp fibre in aqueous medium is not available. From studies (Larson and Stenius, 1987) of adsorption using benzoic acid in microcrystalline cellulose in benzene gave the K L as 17 L/mmol. The adsorption decreases with increasing adsorption coefficient. The stronger adsorption in the present case may be due to the aqueous medium. According to Larson and Stenius (1987) there are two adsorption processes. The surface adsorption is a rapid process when the sample is agitated during the adsorption process. The bulk adsorption into the cellulose is much slower. They rationalized the affinities of the sorbates for cellulose using the acid-base theory. Strong acids and bases are strongly adsorbed and weak ones show much weaker adsorption. The adsorption of 1-naphthol at various mass concentrations of fibre are shown in Figure 2.4. The adsorption per gram of fibre was identical within experimental error at all fibre mass concentrations. The adsorption was same at all mass concentration levels. The water retention value or the fibre saturation point is the amount of water held in the 30 12 9 -\ 6 500 1000 1500 2000 2500 1 /c Figure 2.3: Relationship between 1/c and 1/q* for the calculation of equilibrium constant ( K J for Langmuir isotherm, c is the equilibrium concentration of 1-naphthol and q* is the equilibrium adsorption of 1-naphthol. 31 120 C m , % o 2.70 A 1.80 O 0.90 20 30 Time of Adsorption, minutes Figure 2.4: Effects of mass concentrations ( C J of pulp suspensions on equilibrium adsorption of 1-naphthol at 2 5 + 0 . 1 ° C . Fully bleached kraft pulp fibres. 32 fibre wall (Appendix A.2) . For the fully bleached kraft pulp used in this study the FSP is approximately 1.0 g/g of fibre (dried, fully bleached softwood kraft pulp). The fibre saturation point gives the amount of water in the cell walls. The fibre lumens can also contain water. For never dried fibres this value is about 3.5 g/g of fibre. However, in the case of dried fibres the cell walls are generally collapsed and the amount of water in the lumen is much lower. For this study the water in the lumen was approximately 1.0 g/g. If 1-naphthol solution in a 2% C m pulp suspension is considered then 2% of the total 1-naphthol solution is in the lumen, i.e., 2% by weight of the 1-naphthol is in the lumen. The amount of 1-naphthol adsorbed was less than 3% (Appendix C . l ) . Under the worst conditions this 5% of 1-naphthol may not be available for the reaction. In the worse case this will lower the N A / N B ratio from 1.10 to 1.05. The reaction can then be used to assess micromixing as long as the ratio is maintained constant and 1-naphthol is maintained in excess. 2.4.3 Adsorption of Dyes The adsorption of monoazo dye does not follow any definite pattern as shown in Figure 2.5. This is expected as the attractive forces between monoazo dye and cellulose are very weak (London force). It can be seen that the adsorption of monoazo dye is about 4 times greater than the adsorption of bisazo dye when compared on the basis of equal dye concentration (Figure 2.6). The adsorption of bisazo dye can be described by a Linear isotherm model as shown in Figure 2.6. The adsorption increased with increasing concentration of bisazo dye in solution. In this case the partition coefficient was 0.015 L/g of fibre. The points in the 33 Figure 2.5: Adsorption of monoazo dye by the fibre vs. equilibrium concentration of monoazo dye at 25±0 .1°C. Fully bleached kraft pulp fibres. Concentration of monoazo dye, micromole/litre Figure 2.6: Adsorption of bisazo dye by the fibre vs. the equilibrium concentration of the bisazo dye at 25+0.1 °C. Fully bleached kraft pulp fibres. Concentration of Bisazo Dye, micromole/litre figure are scattered at higher solution concentrations. 2.4.4 Diffusion of Chemicals onto Fibres The fibres were allowed to contact the 1-naphthol solution for 30 minutes before the coupling reaction was begun so that sorption of 1-naphthol was complete. Some 1-naphtholwas absorbed into the fibre. The reaction chemistry requires N ^ N ^ = constant. As described before (Section 2.4.2) approximately 5% may be unavailable for reaction. From diffusion limitation calculations the possibility of this adsorbed 1-naphthol available for coupling reaction can be estimated. Similarly the availability of adsorbed monoazo dye for the secondary coupling reaction also can be estimated. 2.4.4.1 Diffusion of Diazotized Sulfanilic Acid onto the Fibre In the case of fibres the Weisz Modulus is used to study the diffusional limitation (Levenspiel, 1984). The Weisz Modulus is given by the following equation: M w = (-r A ' " )W 2 /C b u l k D e f f (2.3) where -rA"' = Rate of reaction (Moles of A disappearing / m 3 of A.s) W = Fibre half width or radius Quik = Concentration of diazotized sulfanilic acid (B) D e f f = Effective diffusivity (effectiveness factor x bulk diffusivity) There is no diffusional limitation when M w < 0.15 (Levenspiel, 1984). The availability of the 1-naphthol in the fibre wall for the coupling reaction can be determined by estimating the Weisz Modulus. The rate constant for the first coupling reaction (Eqn 1.1) is very high 36 (kj = 12000 m'.mol^.s"1). The reactions proceed before the concentration of the diazotized sulfanilic acid reaches uniformity in the stirred tank. The concentration of diazotized sulfanilic acid in the reaction zone is taken as the bulk concentration for the estimation of Weisz modulus. The estimation of fibre half width is shown in Appendix A.4. Table 2.3: Variables affecting the diffusion limitation of 1-naphthol and estimation of the Weisz Modulus. Rate constant, k! 12 000 m l m o l ' . s 1 . Concentration of A , [A] 0.52 mol.nv3 Concentration of B, [B] 23.6 mol.nr 3 Fibre half width, W 27.9 jum (Appendix A.4) Rate of reaction, k,[A][B] 147 264 m o l . m l s 1 Diffusivity, D 7.8 X 10 1 0 mis" 1 Effectiveness factor, ef 0.5 (van Heiningen, 1992) Effective Diffusivity, D e f f 3.9 X 10 1 0 m i s 1 Weisz Modulus, M w 12 454 As M w > 0 . 1 5 there is likely diffusion limitation for the primary coupling reaction. However, the amount of 1-naphthol adsorbed was found to be less than 5% (2% in the lumen + 2 . 9 % due to adsorption (Appendix C. l ) ) . So the ratio of number of moles of A to B was « 1.05 even with the diffusion limitation. The product distribution (Xs), which is a function of number of moles of A to B should not be affected. The change in Xs for a change in N A O / N ^ from 1.10 to 1.05 was calculated using a computer model developed by Baldyga and Bourne (1988) . Xs was found to increase by 0 .001 when N A O / N B O decreased from 1.10 to 1.05. 37 2.4.4.2 Diffusion of Monoazo Dye If the monoazo dye that diffused into the fibre was not available for reaction the formation of S could be affected. The Weisz Modulus for the secondary coupling reaction (Eqn. 1.2) was calculated to find whether there was diffusional limitation for the secondary coupling reaction. Table 2.4 shows the variables affecting the diffusional limitation of monoazo dye. Table 2.4: Variables affecting the diffusion limitation of monoazo dye. Rate constant, k 2 2.44 m'.mol-'.s-1. Concentration of R 0.025 mol.ni-3 Concentration of B 23.6 mol.m"3. Rate of reaction, k2[R][B] 1.440 mol.nr 3.s _ 1. Diffusivity, D 7 . 8 X 1 0 1 0 m2.s-' Effectiveness Factor, ef 0.5 Effective Diffusivity, D e f f S ^ X l O ^ m l s 1 Fibre half width, W 2 . 7 9 x l O V m Weisz Modulus, M w 0.12 Since the value obtained is less than 0.15 there is no diffusion limitation for the secondary coupling reaction in the presence of fibres. Thus, the presence of fibres is judged to have a limited effect on the coupling reaction and the Xs can be used to assess mixing provided experimental conditions are kept identical between runs. 2.4.5 Correction for Dye Adsorption using Xs In the presence of fibres the adsorption of bisazo dye was greater than the adsorption of 38 monoazo dye when compared to their equilibrium concentrations. This is likely due to the attraction of sulphonic acid groups in the dyes by the residual lignin in the fibres. Bisazo dye has two sulphonic acid groups when compared to one in monoazo dye which resulted in more adsorption of the bisazo dye. So in the presence of fibres the measured concentration of bisazo dye was lower than monoazo when compared to their initial concentrations. This resulted in the measured Xs being reduced when fibres are present (Figure 2.7). By studying the adsorption at various mass concentrations of fibres and at various dye concentrations, correlations were made between product distribution measured after dye adsorption in the presence of fibres (Xs) and the product distribution measured in the absence of fibres (Xs^t). Figure 2.8 shows the relationship between X s / X s ^ and the mass concentration (C n i). Two equations that successfully correlated the results are shown below. Xs / X s ^ = 0.2exp(-CJ - 0.029C r a + 0.838 R 2 = 0.894 (2.5) Xs / Xs;,* = 0.0052Cm 2 - 0.089Cm + 1 R 2 = 0.903 (2.6) Confidence intervals for equvation (2.6) at various suspension mass concentrations are given in Table 2.5. 39 Fibre Mass Concentration, Mass Concentration! Table 2.5: 95 % confidence interval of Xs/Xs^ , at various mass concentrations of fibres for Eqn 2.6. 95 % confidence interval of Xs/XSinit 0.5 ±0.05 1.0 ±0.02 2.0 ±0.05 2.5 ±0.04 6 ±0 .06 10 ±0.08 As the Eqn 2.6 gave a better fit and correctly predicts the limit as C m -» 0 it can be used to correct the dye adsorption measurements. When applying this correlation for adsorption studies, the product distribution corrected for dye adsorption is termed as X s c o r r . Therefore, knowing the amount of fibre in the suspension and the measured value of Xs we can compute the value of Xs that represents the turbulence and micromixing in the suspension, X s c o r r . X s c o r r = Xs / (0.0052Cm 2 - 0.089Cm + 1 ± C.I.) (2.7) In this work the X s c o r r was obtained from Figure 2.8 which gave better results as the fits were better than that obtained from Equation 2.7. 2.4.6 Correction of Xs Using Mass Balance The mass balance on dye after dye adsorption decreases with increasing mass concentration as shown in Figure 2.9. Table 2.6 shows the variation in mass balance during different identical runs. 42 Fibre Mass Concentration, Table 2.6: Range of mass balance in water tests and in pulp tests (at various mass concentrations) and the 95% confidence intervals. c m , % Range of mass balance, % Mean of mass balance, % C . I . (95%) 0 92.4 - 95.8 94.3 ± 1.41 0.5 90.3 - 94.6 92.7 ± 2.22 1.0 88.1 - 93.1 90.7 ± 2.66 1.5 85.5 - 92.6 89.2 ± 2.89 2 81.4 - 87.7 84.9 + 2.39 2.5 79.9 - 85.2 83.1 + 2.13 In theory the value of X s c o r r could be determined from the known change in the mass balance and the measured equilibrium concentrations of the two product dyes (R and S). However two factors prevent this. First, the mass balance determined for identical runs was not constant. Even in the absence of fibres the mass balance was 94.3 + 1.41% (Table 2.6). Mass balances in this range are typical when using the 2-component analysis (Bourne et al., 1985; Bennington and Bourne, 1990). Second, this method cannot be used for Xs correction as there is no correlation between the monoazo dye adsorption and the equilibrium concentration of the monoazo dye (Figure 2.5). Because of these factors a correction to Xs based on measured mass balances could not be made. 2.4.7 Dye Aggregation The change in concentration of the dyes with time was observed for a period of 3 hours. This was done in dye solution alone and in the presence of pulp fibres. The concentration of the monoazo dye remained the same (within error limits) with change in time (Figure 44 2.10). This shows that there is no interaction between the monoazo dye molecules. If there was any interaction the concentration of the monoazo dye would decrease. Also there was no disintegration of the monoazo dye as the concentration remained the same. Similarly there was no interaction between the bisazo dye molecules as the concentration of this dye did not change with time (within error limits (Figure 2.11)). Also there cannot be any interaction between mono and bisazo dye molecules. The constant mass balance (Figure 2.12) shows that there was no dye disintegration in the test time of 3 hours which is seen from the constant mass balance (within error limits). The Xs remained the same (Figure 2.13) within the error limits. This concluded that there was no appreciable change in the variables with time (for 3 hours) in micromixing reaction carried out in aqueous medium. In the case of fibres there was an initial change in Xs, mass balance and concentration of monoazo and bisazo dye in the first 30 minutes due to sorption. After 30 minutes there was no appreciable change in these variables (Xs, mass balance, concentration of R and concentration of S). This shows that there is no interaction between the dye molecules for a period of three hours. Under normal conditions all tests and analysis were completed within this period. 2.4.8 Adsorption of Dyes by Various Fibres Adsorption studies were conducted with six different fibre types (fully-bleached kraft (FBK), semi-bleached kraft (SBK), unbleached kraft (UBK), stone groundwood (SGW), thermomechanical pulp (TMP) and nylon). The adsorption of the dyes at an Xs value of 0.110 and at 3 different mass concentrations (2, 4 and 6%) was examined. The results are 45 Time, minutes Time, minutes Figure 2 .11 : Concentration of bisazo Figure 2.10: Concentration of monoazo dye Time, minutes Figure 2.12: Mass balance vs. Time Time, minutes given in Figure 2.14. For all pulp fibres the value of Xs decreases. However, the adsorption is different for each pulp studied. The decrease in Xs for stone groundwood and thermomechanical pulp was very much higher than for the chemical pulps. This is due to the lignin content of the pulps which plays an important role in the adsorption process. The greater the amount of lignin the larger the adsorption of bisazo dye with the consequent reduction in Xs. The results with unbleached and semi-bleached kraft pulps also agree with this. The drop in Xs for unbleached pulp was more than that for semi-bleached pulp. The unbleached pulp has more lignin when compared to semi-bleached pulp. The fully bleached chemical pulp has very little lignin when compared to all the other pulps and it adsorbed less bisazo dye which resulted in comparatively smaller drop in Xs value. Nylon fibres showed the same Xs at various mass concentrations. This is expected as these fibres do not have any affinity for the dyes. When considering the mass balance the behaviour was the same for the nylon and chemical pulps. The mass balance of the nylon fibres remained the same at various fibre concentrations. This is expected as the nylon fibres have no attractive forces for the dyes. The mass balance in the case of wood fibres drops with increasing mass concentration (Figure 2.15). The mechanical pulps have more adsorption because of their high lignin content. The stone groundwood pulp showed the greatest adsorption. The larger surface area of the groundwood pulp may be another reason for this. 2.5 Conclusions 1. The adsorption of 1-naphthol on the fully bleached kraft pulp tested follows the Langmuir 48 0.12 Fibre Mass Concentration, % Figure 2.14: Product distribution (Xs) vs. mass concentrations ( C J of various pulp fibre suspensions. (a)nylon, (b) fully bleached kraft pulp, (c) semi-bleached kraft, (c) unbleached kraft, (d) thermomechanical pulp and (e) stone groundwood pulp. X s ^ = 0.110. Temperature = 2 5 ± 0 . 1 ° C . N ^ N * , = 1.10 and V A / V B = 50. C A o = 0.52 mol/m 3 , = 23.6 mol/m3. 49 adsorption isotherm. The Langmuir coefficient is found to be 1.73 L/mmol. 2. The adsorption of monoazo dye on a fully bleached kraft fibre does not give any correlation with its equilibrium concentration. The adsorption of bisazo dye follows the linear isotherm in the concentration range of interest. This may be due to the reaction between lignin and the dye. 3. The measured value of the product distribution can be corrected for the presence of fibres once a correlation for the adsorption is experimentally determined. 4. The adsorption of product dyes and hence the change in product distribution will vary for each pulp fibre used. Thus a separate correlation needs to be developed for each pulp studied. 5. The reduction in Xs with pulp mass concentration appears to depend on the lignin content of the fibres, with pulps having a higher lignin content adsorbing more dyes. 6. For the pulp studied here, at low mass concentrations (C m <3%), the adsorption of 1-naphthol on the fibre walls and the 1-naphthol present in the lumen solution is expected to reduce N A o / N B o At C m = 2% this is estimated to reduce N A o / N B o at most from 1.10 to 1.05. This is estimated to increase the value of Xs by 0.001 and is not expected to significantly affect the results reported here. Provided A is maintained in excess the reactions can be utilized. 51 Chapter 3 Mixing in a Semi-Batch Reactor 3.1 Introduction In this study a 22 L semi-batch reactor was used to study the micromixing in aqueous media and in pulp suspensions. The results obtained in aqueous media are compared with Bourne's model (1988) and with the results obtained by Bennington and Bourne (1990) in a geometrically similar tank. The corrected product distribution under various mixing conditions (mass concentration of pulp, rotational speed of the stirrer and reactant feed position) for a fully bleached chemical pulp in the stirred tank reactor were obtained. 3.2 Literature Search 3.2.1 Impeller Characteristics The turbine used in this study was a six-bladed Rush ton turbine. This is a flat-blade turbine with radial discharge. The blade widths are generally one-fifth to one-eighth of the diameter and the distance of the impeller from the tank bottom is equivalent to the diameter (Uhl and Gray, 1966) and these dimensions were used in this study. Four short baffles were 52 used to avoid swirling. The baffle width normally used in the industry is T/12 and is the same used in this study (T is the tank diameter). According to Nagata (1975) the power consumption of the impeller in the stirred tank increases by the insertion of the baffle plates, because the tangential flow of liquid is impeded. A l l turbine impellers have a constant value of N P (Power Number) above N R e » 104 and the common value for a flat-bladed turbine is about 5.0 (Oldshue, 1983). According to Nagata (1975) the correlation of power consumption is rather difficult as the apparent viscosity of pulp suspension depends upon the shear rate and will vary with rotational speed and distance from the impeller. He concluded that the point of incipient motion throughout the tank can be used as a criteria for evaluating the effects of various mixer variables. The power requirement to keep a 4% bleached sulfite pulp in motion with a propeller type impeller was found to be 200 to 400 W/m 3 (Uhl and Gray, 1966). 3.2.2 Micromixing in a Semi-Batch Reactor The semi-batch reactor was chosen for this work as a model for micromixing exists (Baldyga and Bourne, 1988) and aqueous phase reactions have been carried out in geometrically similar tanks. Previous study by Bennington and Bourne (1990) was carried out only at very low mass concentrations (<1% C J . In this study the micromixing was extended to pulp suspensions at higher mass concentrations. Although this type of reactor is rarely used today for mixing bleaching agents with pulp suspensions, this study will confirm the use of the technique and permit a comparison between mixing with and without the presence of fibres in a well studied mixer geometry. 53 Micromixing in liquids proceeds by molecular diffusion and laminar convection within shrinking laminae, which are embedded in energy dissipating vortices (Baldyga and Bourne, 1984a). A l l properties of fine scale turbulence depend on the kinematic viscosity, Kolmogoroff velocity microscales, vortex life times and rates of laminar deformation. From this general standpoint the time required for micromixing should be unchanged if i (the average rate of energy dissipation in the reaction zone) is held constant (Baldyga and Bourne, 1984b). With a given feed ratio of reagents, a given pair of reactions at a known temperature and a given volume of reagents, the same product distribution would result i f i was held constant (Baldyga and Bourne, 1984b). In this study the ratio of V A / V B was 50. The volume of A was 21.2 L and the flow patterns were not significantly changed by the small addition of B. The initial stoichiometric ratio of N A o / N B o was 1.10. The adsorption and solution held in the lumen may effectively reduce this ratio to at most 1.05 but the increase in Xs is expected to be less than 0.001 (section 2.4.4.1). The reactor temperature was measured before and after each run and the change was less than 0 . 2 ° C . At the end of a run, the limiting reagent B was fully consumed and concentrations of the mono- and bisazo dyes (R and S) were determined by spectrophotometry. A mass balance was carried out, whereby ( V A + V B ) ( C R + 2 C S ) was compared with the initial quantity of B added (V B CBO). Mass balances were usually 94.3 + 1.56% for aqueous tests in agreement with that found by other investigators and that obtained by adsorption study (Table 2.6). For pulp fibres the mass balance was lower and decreased with increasing pulp mass concentration. The mass balances at various suspension mass concentrations are shown in Table 3.1. 54 Table 3.1: Variation of mass balance with mass concentration in water and in pulp suspensions in the stirred tank reactor. c„„ % Mass balance range, % Mass balance (mean), % 95% confidence interval 0 91.8-96.1 94.3 + 1.56 0.5 86.1-90.9 88.1 ±2.42 1.0 84.3-89.3 86.4 ±1.79 1.5 83.1-88.6 85.5 ±2.54 2 79.9-87.6 84.5 ±2.99 2.5 81.2-84.5 83.0 ±1.23 Replicate tests were made at 2% C m (Table B.2 in Appendix B). The product distribution Xs, could be found within ± 0.005. 3.2.3 Bulk Mixing (macromixing) The micromixing model (Baldyga and Bourne, 1984b) requires the A-concentration in the neighbourhood of the reaction zone (which is near the feed point) to be known. The simplest formulation is to assume complete macromixing, so that the A-concentration is the average value for the whole tank (Baldyga and Bourne, 1984b). Since bulk blending requires a finite time (Q,the addition time (ff) should be greater than tm i.e. tf > tm. Failure to secure bulk blending will result in fresh B-solution being pumped to the feed point and reaction zone faster than A can arrive there by circulation from other tank regions. (This will be particularly true towards the end of the reaction as the A-concentration becomes progressively smaller). The minimum addition time (tc) for a given stirrer speed was obtained experimentally. This required measuring Xs at various addition times (t{) until Xs had 55 decreased to a constant value. This time was found to be 30 minutes for water at rotational speed (N) = 2 s 1 . It was observed that the liquid near the surface participated irregularly in the general circulation. A long addition time was chosen to ensure that the liquid in the stirred reactor was macroscopically homogeneous. Under these conditions a measured value of Xs exceeding that predicted for the chemical (slow) regime uniquely indicated inhomogeneity on the molecular scale (Bourne and Dell 'Ava, 1987). For pulp suspensions the macromixing time could be assessed only i f the suspension was in complete motion. If the motion was not complete then the zones of stagnant pulp would increase Xs as B would no longer be in excess in the mixed zone. In order to keep the pulp suspension in motion the stirrer speed and in turn the average energy dissipation rate is higher in pulp suspensions than in water. The energy required to keep the suspension in motion increases with increasing mass concentration of the suspension. 3.3 Equipment and Procedure A baffled stirrer tank reactor was used for this study. The dimensions of the tank are shown in Figure 3.1 and is geometrically scaled version (scale-up factor = 2 x ) of the vessel used by Bennington and Bourne (1990). The baffles were short and away from the walls to avoid the formation of dead zones by trapping of fibres at the baffles. A six-bladed Rushton turbine was used. The reactor was surrounded by a square outer plexiglass tank with provision for water circulation. This allowed maintenance of a set temperature and permitted photography without the lensing effect caused by curved vessel walls. Two feed points A , B were used which were in geometrically similar locations as those used by Bennington and 56 57 Bourne (1990). Feed point A was in the suction stream near the impeller (rA = 24 mm; zA = 30 mm) and feed point B was located near the top surface(rA = 58 mm; z B = 140 mm). The diameters of the feed pipes were 1.25 mm in both the feed points. A variable speed motor (250 W) was used to drive the impeller. A Sen so tec torque gauge (Model # QWFK-8M) coupled with a chart recorder was used to measure the shaft torque of the stirrer. In this method the axial torsion of the rotating shaft was measured by a strain gauge. The range of the torque meter was 0-1.9 N . m with a reproducibility of ± 0.05 N.m. The maximum stirrer speed was 1750 rpm and was measured to + 1 rpm by a Cole Parmer digital tachometer (Model # 08212). An Ismatec metering pump system (Model # 7617-70) with a capacity to pump between 1 and 600 cc/min was used to pump the diazotized sulfanilic acid to the mixer. The flow diagram is shown in Figure 3.2. Figure 3.3 shows a photograph of the experimental set up. The 1-naphthol was prepared in a Nalgene tank of 60 L capacity. The tank was provided with a stainless steel coil to regulate the temperature with the use of a thermostat. The temperature was maintained at 25 ± 0.1 °C. For each run in the stirred tank 21.2 L of 1-naphthol was used (concentration = 0.52 mol/m3) and the volume prepared each day was enough for two trials. Semi-batch operations were employed. In this mode of operation one of the reactant (A) was fully charged in the reactor and the other reactant (B) was slowly fed in. In order to inject the exact amount of diazotized sulfanilic acid a scale was used with the metering pump. In this method the beaker containing the diazotized sulfanilic acid was placed on the scale and the suction tube to the pump was placed in the solution in such a way that it did not touch the sides of the beaker. Thus the volume of the solution pumped was measured by the change in scale 58 D C A A Stirred tank B Rushton turbine C Metering pump D Weighing scale E Beaker for diazitized sulfanilic acid F Temperature controller G Motor Figure 3.2: Flow diagram of the apparatus. 59 60 reading as the density of the solution was close to unity. This is shown in Figure 3.2. Mixing was assessed by carrying out tests in which the feed time (tj), stirrer speed (N) and feed position were examined in aqueous media and in the presence of fibres at low mass concentrations (C m <3%). In the preparation of pulp suspension at various concentrations 1 L of water was transferred with the pulp in all cases. The concentration of 1-naphthol was made in such a way that after the addition of pulp to the 1-naphthol solution the total liquid volume was 21.2 L , with a 1-naphthol concentration of 0.52 mol/m3 and the desired suspension mass concentration. The experimental conditions are shown in Table 3.2. Table 3.2: Experimental conditions for the azo coupling reaction concentration of 1-naphthol, c A 0.52 mo/m3 concentration of diazotized sulfanilic acid, c B 23.6 mol/m 3 Volume of 1-naphthol, V A 21.2 x 10"3 m 3 Volume of diazotized sulfanilic acid, V B 4.24 X lO"4 m 3 v A / v B 50 c A /c B 0.022 N A o / N B o 1.10 Concentration of N a 2 C 0 3 39 mol/m3 Concentration of N a H C 0 3 39 mol/m3 Temperature 2 5 + o . r c kj/k2 4918 Samples were collected at the top portion of the solution or the pulp suspension with the help of a plastic syringe (approximately between the impeller shaft and the vessel wall). The fibres were separated using a sintered glass funnel and the dye solution diluted with buffer 61 for analysis. 3.4 Aqueous Tests 3.4.1 Evaluation of Impeller Characteristics Solutions of various viscosities were prepared using different sucrose concentrations (The viscosities were taken from the literature (Weast, 1986) by using the weight percentage of the sucrose). This was used to study the Reynolds number versus Power number relationship of the mixer. The Reynolds number over a range was created by changing the viscosity of the solution and by altering the stirrer speed. By measuring the power consumption at various stirrer speeds the relationship between the Reynolds number and Power number was obtained. The average energy dissipation of the pulp suspension at various mass concentrations (up to 3% C i n) and at various impeller rotational speeds (2 to 14 s"1) was also determined. The Reynolds number for pulp suspensions could not be calculated as it is not possible to measure the viscosity of the suspension. So the average energy dissipation at various stirrer speeds were measured and compared with water. The flow patterns in the pulp suspensions were also observed. The average energy dissipation rates were also correlated with the X s c o r r measured in these mass concentration ranges. 3.4.2 Investigation of Macromixing Time and Micromixing First the macromixing time was investigated by varying the feed time (tj). When the feed time exceeds a certain value, tc, the product distribution becomes independent of tf and depends on micromixing only. This was done at an impeller speed of 2 s"1 in aqueous 62 medium. If tf>tc at a low stirrer speed, it will be sufficient at higher stirrer speeds as the macromixing time decreases with increasing stirrer speed (N0 = constant), where 6 is the circulation time or tj. The variation of product distribution at various stirrer speeds (2-16 s"1) was carried out with 30 minutes feed time. As the feed time was greater than the critical feed time the product distribution was an indication of micromixing alone. 3.4.3 Effect of Feed Point The feed point has a large effect on product distribution. This is due to the variation in the rate of energy dissipation at various parts of the tank. Two feed points were used in this study. Feed point A was characterized by a high local energy dissipation rate and at a high local flow velocity. There is a steep gradient in the distribution of local energy dissipation rate and local flow velocity. At feed point A (rA = 24 mm ; z A = 3 0 mm (with reference to the impeller centre)) the main flow is radial and tangential while the axial flow is negligible. The energy dissipation rate as well as the local flow velocity at feed point B (rB = 58 mm; z B = 140 mm) is reduced compared to feed point A . The local energy dissipation rate is only a fraction of the average. The fluid movement occurs in all three directions and is not well defined. When the feed point is in the higher energy dissipation zone (feed point A) Xs is less (which indicates higher local rates of energy dissipation) when compared to other locations where the energy dissipation is less (feed point B). 3.5 Tests in Pulp Suspension The average energy dissipation of the pulp suspension at various mass concentrations (0.5 to 3% C n i) and at various impeller rotational speeds (2 to 12 s 1) was measured. In pulp 63 suspensions the mass concentration used to estimate the macromixing time was 2% and at a stirrer speed of 12 s 1 (higher speed was used when compared with aqueous media as it was required to keep the suspension in motion throughout the vessel). The coupling reaction was carried out in the presence of pulp suspensions at various mass concentrations (0.5 to 2.5 C J and at rotational speeds of 10, 12 and 14 s 1 . 3.6. Data Reproducibility of Power Measurements in the Stirred Tank The accuracy of the measurement of power improved with increasing rotational speed of the stirrer. To obtain the data reproducibility in aqueous as well as in pulp suspensions, ten measurements at each rotational speed were made in water and in a 2 % C m pulp suspension under similar conditions. The 95% confidence level of the measurements are shown in Table 3.3. Table 3.3: Data reproducibility of energy dissipation measurements in water and in F B K pulp suspension. Fully bleached kraft fibre. Temperature = 25+0.1 °C. R P M 2% C m Pulp Suspension Water Average energy dissipation rate, W/kg 95 % Confidence interval, W/kg Average energy dissipation rate, W/kg 95 % Confidence interval, W/kg 100 0.028 0.028±0.0065 0.008 0.008+0.0025 200 0.071 0.071+0.0165 0.058 0.058+0.0149 300 0.212 0.212+0.0228 0.265 0.265+0.0280 400 0.611 0.611±0.0175 0.670 0.670+0.0255 500 1.282 1.282+0.0212 1.340 1.340±0.0462 600 2.338 2.338±0.0868 2.330 2.330±0.0433 800 6.285 6.285+0.1100 4.890 4.890+3.300 64 For the pulp suspension the 95 % confidence limit was the mean value of average energy dissipation rate + 2 %. For water the 95% confidence limit was the mean value of average energy dissipation rate + 3% up to 700 rpm. At 800 rpm the 95% confidence limit was higher (mean ± 7 % ) . This may be due to the air entrainment in the impeller zone. 3.7 Results and Discussion for Aqueous Tests 3.7.1 Impeller Characteristics in the Stirred tank The power number in the stirred tank increased slightly in the Reynolds number range of 1000 to 10 000. After that it remained constant (Figure 3.4). The Power number remained around 5.2 + 0.2 in the turbulent range and is in agreement with the literature values for similar turbines operated under similar conditions (Uhl and Gray, 1966). When the Reynolds number was 105 there seemed to be a slight drop in power number. According to Uhl and Gray (1966) this was due to the aeration in the impeller zone. 3.7.2 Backmixing into the Feed Pipe The feed pipes checked in this work were of 1.25 mm and 3.25 mm in diameter. The feed rate was maintained at 14 cc/minute. The velocity at the larger nozzle was 2.8 cm.s 1 while it was 19 cm.s 1 with the smaller nozzle. With the larger nozzle the Xs increased with feed time while with the smaller nozzle the Xs decreased with increasing time. The decrease in Xs with increasing addition time was attributed to the backmixing of reactant B into the feed pipe as the velocity at which B pumped out was low. By using the smaller nozzle the velocity was sufficiently increased and the backmixing into the feed pipe avoided. So the 65 Figure 3.4: Impeller characteristics in the stirrer tank. Turbine used = Rushton turbine. The viscosity was varied using sucrose/water solutions of varying concentrations. Dotted line shows the literature value (Uhl and Gray, 1966). feed pipe with 1.25 mm was used in this study. 3.7.3 Determination of Macromixing Time at Feed points A & B Coupling reactions were conducted in aqueous media at N = 2 s"1 and the addition time were varied from 120 to 3000 seconds. As the macromixing time decreases with increasing rotational speed, by determining tc for N = 2, keeping tf > this value at all rotational speeds will ensure that Xs measures microscale mixing effects only. The location of feed point influenced the product distribution in macromixing as well as micromixing experiments. The difference was due to the inhomogeneous distribution of the local energy dissipation rates and local flow velocities in the reactor. At both feed points Xs values decrease with increased addition times (^ ) until they levelled off above tf = tc (Figure 3.5). The tc ~ 20-25 minutes and the addition time was kept at 30 minutes for all tests. By keeping tf >tc the contribution to Xs by macromixing was avoided. Thus the measured Xs was an indication of microscale mixing only. 3.7.4 Comparison with Model — Aqueous Tests The product distribution at feed points A and B were measured at various stirrer speeds. The results were compared with Baldyga and Bourne's model (Balyga and Bourne, 1988). The results are shown in Figure 3.6. The experimental results agree with the model only at medium stirrer speeds (5-10 s'1). This model is based on literature measurements of flow and energy distribution in typical stirred vessels (Figure 3.7 and Figure 3.8). Plug flow was assumed in the four circulation loops above and below the turbine, p and q were the centres 67 0.5 0.4 0.3 ( 0 X 0.2 0.1 0.0 0 10 20 30 40 50 60 Feed time, minutes Figure 3.5: Variation of Product distribution (Xs) with addition time at a constant rotational speed (2 s'1) in aqueous media at 2 5 ± 0 . 1 ° C . N ^ N ^ = 1.10 and V A / V B = 50. C A o =0.52 mol/m3, = 23.6 mol/m 3 . 68 0.3 0.2 ( 0 X 0.1 0.0 Feedpoint B Feedpoint A \ \ \ \ A \ \ \ \ \ \ \ 5 8 12 16 20 Rotational Speed, s-1 Figure 3.6: Product distribution (Xs) vs. rotational speed (N) with feed time ($) = 30 minutes (tf>te) in aqueous media at 25+0.1°C. NJN^ = 1.10 and V A / V B = 50. C A o = 0.52 mol/m 3, C B o = 23.6 mol/m 3. Dashed lines show the values obtained from Baldyga and Bourne's model (1988). 69 Figure 3.7: Flow pattern in stirred tank reactors after Baldyga and Bourne (1988). A, B = Feed points for azo coupling. Radial coordinates normalized to tank radius, axial coordinates normalized to tank diameter (height). Numbers on circulation loops are flow coefficients = Qj^/ND3- w = height of turbine blade, p, q = Centres of circulation. Figure 3.8: Distribution of rate of energy dissipation in the stirred tank after Baldyga Bourne (1988). <t> =e l o c a l /ë 4> = 15 8 5 1.6 1.5 1-0 0.6 0.5 0.3 0.25 0.2 P 1- -f + K X X X X X X X 71 of circulation. According to Baldyga and Bourne (1988) the literature cited was not consistent, nevertheless they can be used as a first comparison of such results with experiments. At very low stirrer speed the experimental product distribution was less than that predicted by the model. This was expected as it was difficult to keep the stirrer speed constant below 3 s 1 . At higher stirrer speeds the experimental results gave higher product distribution than that predicted by the model. This may be due to the deviation from the plug flow behaviour at higher stirrer speeds. 3.7.5 Comparison with Similar Study The results obtained were compared with Bennington and Bourne (1990) for a geometrically similar tank with geometrically similar feed points. The scale-up factor was two. The volume used by Bennington was 2.5 L while it was 21.2 L in this study. The results are shown in Figure 3.9. They agreed well for a small range of Reynolds numbers. At feed point A , the Xs value in both studies was similar up to N R e = 8 x l 0 4 . At higher Reynolds number the Xs value obtained in this study was higher. At feed point B the Xs value obtained in this study had lower values. The flow velocity in the larger tank is higher than in the small tank at a particular impeller speed. Thoma (1989) observed that in the small tank the reaction zone covered a greater distance than the reaction zone in the larger tank. While in geometrically similar tanks the distribution of the energy dissipation is independent of scale, the effective point of reaction and therefore the local energy dissipation rate responsible for the product distribution are not identical. The integrated energy dissipation rate at the point of reaction in a larger tank is greater than in a small tank 72 0.3 0.2 -CO 0.1 -0.0 \ ^ Feedpoint B • Feedpoint A \ \ A \ 0 4 x 1 0 * 8 x 1 04 12x10 4 16x104 2 0 x 1 0 4 N Re Figure 3.9: Product distribution vs. Reynolds number (N R e) with feed time = 30 minute ( > critical feed time) in aqueous medium. Dashed lines show the experimental values obtained in a geometrically similar tank by Bennington and Bourne (1990). Scale-up factor for the present work is two. NJN^ = 1.10 and V A / V B =50. C A o = 0.52 mol/m3, = 23.6 mol/m 3. 73 (Thoma, 1989). This indicates that the point of reaction is nearer the stirrer in the large tank. According to Bourne and Dell ' Ava (1987) the reaction becomes more localized with the increasing size of the tank. The average energy dissipation rate (when N R e = 10s) in the present study was 2.33 W/kg where as in the smaller tank it was about 0.6 W/kg. The higher energy dissipation rate in the present case was also a reason for the lower product distribution. The feed point B is at a longer distance from the impeller in the larger tank when compared to the small tank. So it may take more time for the reactant to reach the impeller in the larger tank when compared to the small tank. Considering all these variables the disagreements between these two geometrically similar tanks are reasonable. 3.8 Results and Discussion for Pulp Tests 3.8.1 Power Consumption by Pulp Suspension Figure 3.10 compares the energy dissipation rates in water and in 2% C mpulp suspension. The energy dissipation rate in the pulp suspension was proportional to its mass concentration only when there was complete motion in the vessel. For example, below a rotational speed of 10 s 1 the aqueous medium consumed more power than the pulp suspension (when the pulp suspension was not in complete motion). At 10 s 1 both had the same average energy dissipation (Table 3.3). Above 12 s_1 the pulp suspension had a higher average energy dissipation rate than water as the suspension was in complete motion. Below 12 s"1 there was no complete motion in the pulp suspension and the motion was spread only to a smaller volume which resulted in lower energy dissipation rates. When based on the entire reactor volume the average energy dissipation rate for complete motion in a 1 % C m suspension was 74 0 2 4 6 8 10 12 14 Rotational Speed, s -1 Figure 3.10: Variation of average energy dissipation rate with increasing rotational speed with water and in fully bleached kraft pulp suspension (C^ = 2%) at 25+0 .1°C. Fully bleached kraft pulp fibres. 75 500 W/m 3 and for a 2.5% C m suspension it was 7 kW/m 3 . Earlier studies show that the power requirement to keep a 4% bleached sulfite pulp in motion with a propeller type impeller was 200 to 400 W/m 3 (Uhl and Gray, 1966). With the increasing rotational speed of the stirrer, the average energy dissipation rate increased at all mass concentrations (up to 3 % C m as shown in Figure 3.11). But the energy dissipation rate could be correlated to the C m suspension mass concentration only when there was complete motion. At 13 s 1 the suspension having suspension mass concentration of 3% was not in complete motion which resulted in lower energy dissipation rate was lower than a suspension having 2% C m . When N = 13 s 1 the power consumption decreased due to the entrainment of air in the vessel. 3.8.2 Flow of Pulp Suspension in the Stirred Tank The flow conditions of the pulp suspension at 2% C m and at stirrer speeds 6, 8, 10 and 13 s"1 are shown in Figures 3.12 to 3.15. The colour of the azo dyes makes the observation of the flow easy by providing contrast between the fibre and liquid phases. At 6 s"1 there was motion only in the impeller radial plane. There was no motion above and below the impeller plane (Figure 3.12). Even in the impeller radial plane the motion hardly reached the wall. This indicated a well defined cavern formation with the cavern diameter just matching the vessel diameter. Below this stirrer speed no colour was visible due to the formation of a cavern having a diameter less than the vessel diameter. When the stirrer speed was increased to 8 s 1 the mixing zone increased in volume (Figure 3.13). More than 50% of the total volume of the pulp suspension was in motion. Still the top portion was not 76 TO 7 I 1 1 1 1 1 r Rotational Speed, s -1 Figure 3.11: Variation of average energy dissipation rate at various stirrer speeds with water and fully bleached kraft pulp fibre suspensions at C m = 1, 2 and 3%. Temperature = 25±0 .1°C 77 Figure 3.12: Motion of the fully bleached kraft pulp fibre suspension (C m = 2%) in the stirred tank at stirrer speed (N) = 6 s 1 78 79 in motion. Even though the pulp was in motion the floe size was large as seen from the photograph. At 10 s 1 the motion spread over a larger volume (Figure 3.14), except a small portion at the top. The floe size appeared to be smaller. When the stirrer speed was 13 s"1 there was complete motion and the turbulent motion in the impeller radial plane could be seen clearly (Figure 3.15). The floe size was small when compared to lower stirrer speed. There was plug flow only in the low stirrer speed range. At higher speed there was more tangential flow which was seen from the top. So the micromixing model which assumed plug flow might give different Xs than that obtained in this study at higher stirrer speeds. 3.8.3 Macromixing Time The critical addition time in pulp suspension was calculated by maintaining the suspension (C m = 2%) in motion at a rotational speed of 12 s 1 . The critical addition time was ~ 15 to 20 minutes for feed point A and 20 minutes for feed point B under these conditions. This is shown in Figure 3.16. The U was taken as 30 minutes (tf >tc) in all tests. 3.8.4 Variation of X s c o r r with Mass Concentration at Various Stirrer Speeds The variation of X s c o r r with increasing mass concentration and with the stirrer speed as a parameter are shown in Figure 3.17 (The variation of measured Xs with mass concentration is shown in Appendix E). First point B is considered. Here X s c o r r for the pulp suspension is greater than Xs for water. The X s c o r r seems fairly constant over a mass concentration range 0.5-2.0%. When the stirrer speed was 10 s 1, the 2.5% C m pulp suspension showed an increase in X s c o r r . This may be due to the cavern formation as the pulp suspension was not in full motion under these conditions. The macroscale mixing was 80 81 82 Feed Time, minutes 0.0 0.5 1.0 1.5 2.0 2.5 Mass concentration, % Figure 3.17: Corrected product distribution ( X s ^ ) vs. mass concentrations ( C J of pulp suspensions at various stirrer speeds (N) at 25+0.1 °C. Fully bleached kraft pulp fibre. Stirrer speeds used were 10, 12 and 14 s 1 . N ^ N ^ = 1.10 and V A / V B = 50. C A o = 0.52 mol/m 3 , Cg,, = 23.6 mol/m 3 . Open symbols indicate tests where the suspension was not in complete motion. 84 not complete and this contributed to increase in X s c o r r . At feed point A the X s c o r r was almost equal to the Xs for water in all cases. It increases slightly over the C m range 0.5-2.0%. A slight increase in Xs c o „was noticed when the stirrer speed was 10 s 1 and the mass concentration was 2.5%. As indicated previously the product distribution was dependent on the integrated energy dissipation over the reaction zone. At feed point A this was high compared to feed point B . There was a decrease in X s c o r r observed when the mass concentration was 0.5% and N = 14 s"1. X s c o r r is a measure of turbulence. This may be due to the higher integrated energy dissipation near the impeller in the presence of fibres. 3.8.5 Relationship Between X s ^ , . and Average Energy Dissipation The relation between X s c o r r and the average energy dissipation in the pulp suspensions is shown in Figure 3.18. At both feed points X s c o r r decreased with increasing average energy dissipation rate. With increasing energy dissipation rate the drop in X s c o r r was more at feed point B than in A . At feed point B the X s c o r r was very high when the stirrer speed was 10 s_1 and C m was 2.5%. At this condition the tf<tc and contribution by macromixing likely gave rise to an increase in X s c o r r 3.9 Conclusions 1. By keeping tf > tc, the contribution to Xs by macromixing can be avoided. 2. The product distribution (Xs) depends on the position of feed point. The Xs is low near the impeller vicinity where the turbulent energy dissipation rate is high. 3. In pulp suspensions X s c o r r is greater than Xs in water and increases with mass 85 0.25 0.00 1 1 ' ' 2 4 6 8 Energy Dissipation Rate, W / k g Figure 3 18' Corrected product distribution vs. average energy dissipation with varying mass concentrations ( C J and stirrer speeds (N). Fully bleached kraft pulp fibre Mass concentrations used were 0.5, 1.0 and 2%. Stirrer speeds used were 10, 12 and 14 s . 86 concentration. Therefore the turbulence seen by the reaction is less for fibre than water. This is applicable for both feed locations at N = 10 and 12 s"1, and at feed point B at N = 14 s 1 . 4. The Xs c o„ seems to decrease in the feed point A at 14 s"1 when the C m was 0.5%. This may be due to the higher integrated energy dissipation near the impeller in the presence of fibres. 5. The manner in which the fibres change the flow pattern is unknown. However, an increase in Xs would indicate a decrease in turbulence. Therefore the presence of fibres modifies the turbulence likely by changes both the flow through the zones of turbulence and the energy dissipation zones in the vessel (due to cavern formation). 87 Chapter 4 Assessment of Micromixing in a High-Shear Mixer 4.1 Introduction Mixing of medium consistency pulp suspensions requires imposition of intense shear on these suspensions. This is normally carried out in high-intensity or high-shear mixers. These type of mixers have high rate of energy dissipation in the mixing zone. The energy dissipation rates of the mixer used in this study were 46.3 W/kg when the impeller rotational speed was 16.7 s_1 and 5600 W/kg when the impeller rotational speed was 83.5 s 1. Bourne, Kut and Lenzner (1992) proposed that the diazo coupling reaction can be used to assess micromixing in mixers whose energy dissipation rate do not exceed 200-400 W/kg. According to this the diazo coupling reaction can be applied to the mixer in this study up to approximately 34 s"1 ( « 3 9 3 W/kg). The study was carried out to determine the possibility of using this coupling reaction in the conditions existing in the high-shear mixer. 4.2 Literature Search In a stirred tank reactor the energy dissipation rates of different zones differ widely. They vary over two orders of magnitude when comparing the impeller tip zone and the bulk 88 zone. In the high-shear mixers the energy dissipation rate may not vary as much in different zones as in the case of stirred tank (Bourne et al., 1992). This is due to the smaller volume and high energy dissipation rate applied in these type of mixers. In medium consistency pulp suspensions the high energy dissipation rate ensures floe disruption and good fibre scale mixing (Bennington et al., 1991). The turbulent fibre suspensions have hydrodynamic properties which resemble those of "Newtonian" fluids. 4.3 Equipment and Procedure 4.3.1 Equipment The high-shear mixer is a concentric device used to fluidize medium consistency (8-14%) pulp suspensions. The mixer is powered by a 22.4 kW variable speed DC motor which permits a rotational speed up to 83.3 s 1 . The 1-naphthol was added to the chamber formed between the rotor and a housing which provides the outer cylinder surface. The rotors have lugs to prevent slippage in case of pulp fibres and the housing have baffles to prevent fibre slippage at the outer wall. The lugs and baffles provide regions of high turbulence in the device. The chamber volume can be varied by using different rotor/housing combinations. The configuration used in this study is given in Figure 4.1. Detailed dimensions of the rotor and housing for the wide gap configuration are shown in table 4.1. The shaft torque of the mixer was measured by a strain gauge attached to the shaft. The accuracy of the strain gauge was + 0.3 N . m. The shaft rotational speed was measured by a remote optical sensor. The temperature measurement of fluids/suspensions inside the chamber was made with a platinum resistance thermometer. More details of the high shear 89 Figure 4.1: Dimensions of the high-shear mixer chamber and rotor (in millimetres). The position of the 6 feed points are shown. mixer are available in Bennington (1988). Table 4.1: Specifications of high-shear mixer housing and rotor (Bennington, 1988). Chamber Component Dimensions Housing diameter depth number of baffles baffle dimensions 220 mm 100 mm 6 10x10 mm Rotor diameter depth number of lugs lug dimensions 100 mm 100 mm 6 10x25 mm Gap width Volume of chamber 50 mm 3.41 xlO" 3 m 3 The number of feed nozzles used were 6. Multiple nozzles were used to reduce the critical addition time. The balance between the rate of arrival of fresh B in the reaction zone (V B c B o M z vr c ) and the rate of flow of A from various regions of the mixer to the reaction zone is an important aspect of macromixing (section 3.2.4). Bourne and Thoma (1991) verified that Xs is a function of nzt{ by using 4 nozzles in two geometrically similar stirred tank reactors. In this study the nozzle diameters were 1mm and they were equally spaced from each other. The positions of the nozzles were 74 mm from the centre of the shaft. The length of the nozzles (inside the chamber) were 15 mm. The nozzles had a conical shape with a base diameter of 6 mm. A l l the nozzles were connected to a common header which was attached to the outlet of the metering pump. The method of chemical addition used was 91 the same as described in section 3.3.3. 4.3.2 Assessment of Mixing in the High-Shear Mixer Table 4.2: Experimental conditions for the azo coupling reaction concentration of 1-naphthol, cA 0.52 mo/m3 concentration of diazotized sulfanilic acid, c B 23.6 mol/m 3 Volume of 1-naphthol, V A 3.3 x 10-3 m 3 Volume of diazotized sulfanilic acid, V B 6.6 x 10"5 m 3 v A / v B 50 c A /c B 0.022 N A o / N B o 1.10 Concentration of N a 2 C 0 3 39 mol/m3 Concentration of N a H C 0 3 39 mol/m3 Temperature 2 5 ± 2 ° C According to the literature search the diazo coupling reactions can be used to assess micromixing if the energy dissipation rate is less than 400 W/kg. For the high shear mixer used in this study this energy dissipation rate can fluidize a suspension with 4% C m (Bennington et al., 1990). The expected product distribution for a mixer can be calculated using Baldyga and Bourne's (1988) model. The product distribution in the high shear mixer under various conditions (rotational speed, ratio of chemicals and ratio of the volume of chemicals etc.) was computed using this model. The Xs value obtained at 16.7 s"1 in the high shear mixer with this model was 0.016 ( V A / V B = 50, N A / N B = 1.10 C A o = 0.52 mol/m 3, C B o = 23.6 mol/m3). The power number of the mixer was 3.4 and was used to 92 estimate the energy dissipation rate in the mixer. 4.3.3 Critical Feed Time The operation carried out in the high-shear mixer was a semi-batch process. In this mode of operation one of the reactant (A) was fully charged in the reactor and the other reactant (B) was slowly fed in. A short feed time is very important for high-shear mixers. This is because the high energy imparted to the fluid or pulp suspension is dissipated as viscous shear which increases the fluid temperature. For pulp suspensions, the more concentrated the suspension the higher will be the energy dissipation and in turn the rise in temperature. Bennington (1988) observed that in the thirty second duration of a standard test, the temperature of a 10% C m semi-bleached pulp increased by 6 °C in the wide gap configuration (the configuration used in this study). Bourne et al. (1985) observed that the bisazo dye disintegrates when the temperature was over 300 K . In order to avoid this the temperature should be kept below 300 K . The critical feed time (tc) is the shortest feed time (tj) at which only the microscale mixing contribute to Xs. When the tfis less than the critical feed time (tc) the macromixing also contributes to Xs. Bourne and Hilber (1990) concluded that the critical feed time (tc) can be reduced by increasing the number of nozzles. They showed that by employing a feed distributer with n z nozzles the tc can be reduced by the factor n z relative to tc for a single feed point provided that the reaction zones do not overlap. They also observed that the critical feed time increases with increasing reactor size. Dye samples were taken by opening the sampling plug at the top of the chamber. A 93 plastic syringe was used to take the samples. The dye solution was then diluted, buffered and analyzed in the spectrophotometer. The number of nozzles used in this study was 6. This would reduce the critical feed time by a factor of six. By reducing the addition time the temperature rise could be reduced. Previous study by Bennington et al. (1991) shows that the macromixing time was 1.4 seconds using red dyed nylon fibres as a tracer to a semi-bleached kraft pulp suspension in the wide gap configuration. 4.4 Results and Discussion Table 4.3: Xs vs. Average energy dissipation rate in the high-shear mixer at 25+0 .2°C. N A o / N B o = 1.10 and V A / V B = 50. Feed time, s [R], mol/m 3 [S], mol/m3 Xs Av. energy dissipation rate, W/kg Mass balance, % 7.5 0.405 0.0031 0.0151 58.35 88.92 7.5 0.445 0.0056 0.0243 58.35 98.59 10 0.406 0.0030 0.0145 55.23 89.01 10 0.418 0.0042 0.0197 58.35 92.10 15 0.397 0.0009 0.0044 55.23 86.07 20 0.413 0.0003 0.0014 58.35 89.30 20 0.398 0.0002 0.0001 58.35 86.03 30 0.403 -0.0004 -0.0019 58.35 87.00 40 0.413 -0.0009 -0.0042 58.35 88.92 60 0.415 -0.020 -0.0098 58.35 88.80 The product distribution is dependent on the energy dissipation in the mixing zone (Equation 94 1.14). As the high-shear mixers have high energy dissipation rates, the expected product distribution is low. Since the analytical error in determining the product distribution is ±0.001 (Appendix B) the measured Xs value should be over 0.01 for accurate determination. The approximate values of the maximum rates of energy dissipation range from 5 W.kg"1 in centrifugal pumps, 50 W . k g 1 in stirred tanks and 5000 W.kg"1 in rotor-stator devices (Bourne et al., 1992). They concluded that the 1-naphthol coupling reaction is too slow to characterize micromixing in high-intensity mixers (when ë >400 W.kg"1). This is because of the slow rate of the secondary reaction (Equation 1.2) when compared to the primary reaction. The primary reaction (Equation 1.1) is instantaneous (half-life in range of 0.2-0.6 ms) relative to the mixing (Bourne and Thoma, 1991). The second reaction has a rate constant which is about 5000 times smaller than that for the first reaction. The variation of Xs with increasing addition time is shown in Figure 4.2. The product distribution was low as predicted and it started to level off after 40 seconds. Thus the critical addition time (tc) was > 40 s. The macromixing time calculated for a pulp suspension at (10% C J in the same mixer configuration was 1.4 s (Bennington, 1989). This difference can be rationalized by comparing the scales of mixing measured by the two techniques. In the present case the mixing was of microscale (molecular dimensions) whereas in the Bennington study, mixing was assessed at the fibre-scale (~ 0.05-10 mm). With longer feed times (over 30 s) the Xs values obtained were negative. Negative Xs values were also obtained by Studer (1989). He observed that the two component analytical method gave too low and even negative Xs values with reactors having high energy dissipation rates. This is due to the linear regression used which does not include the ortho 95 Figure 4.2: Product distribution vs. addition time in the high-shear mixer in aqueous monoazo dye. By applying Studer's correlation (Eqn 1.11) the negative Xs value -0.0009 (Table 4.3) obtained by two component analysis in the high-shear mixer was found to be 0.013. The value calculated from Baldyga and Bourne model was 0.016 at the rotational speed used in the mixer (1000 rpm) and is in good agreement with the value of Xs once corrected. The experimental energy dissipation rate at 16.7 s"1 was 60 W . k g 1 (obtained by using the measured torque, angular velocity and the volume of the chamber). The energy dissipation rate calculated by power number was 46.3 W . k g 1 . The difference may be due to the accuracy of the torque meter ( ± 0 . 3 N.m) as the net torque measured at 16.7 s"1 was only 1.9 N .m. Bourne et al. (1992) proposed a new extended diazo coupling reaction involving a mixture of 1-naphthol and 2-naphthol with diazotized sulfanilic acid. The rate constant for the secondary coupling of 2-naphthol is 124.5 m3.mol"1.s"1 (50 times higher than the similar rate constant with 1-naphthol. This extended system is suitable for energy dissipation rates in aqueous solutions up to about 10s W.kg" 1. But this system is more complicated in the presence of fibres where making corrections for the five components is expected to be very difficult. As the Xs value was negative and tc was only just being approached (tc> 40-50 s) pulp suspensions were not tried as the sorption would further lower the Xs value, i.e. the bisazo dye concentration will be very low even without fibres and with the presence of fibres the bisazo dye concentration will be insignificant. So the mixing in the high shear mixer cannot be assessed by the two component method. Under these conditions the three component method may be used. In this study the Xs value leveled off at -0.005 when tf > 50s. By 97 applying the three component correction (Eqn 1.10) the value of Xs obtained was 0.016. 4.5 Conclusions 1. The two component method used in this study gave a very low or negative product distribution. By altering the stoichiometry and the volume of the reactants the reaction could by used to assess mixing in reactors where the energy dissipation rate is <400 W . k g 1 . 2. Mixers with higher energy dissipation rates need the new extended system to assess micromixing. But for pulp suspensions the extended system is unlikely to be suitable as with 5 components the adsorption effects would be complicated and it would be very difficult to make any correction for sorption. 3. Nylon fibres could be used as a model fibre suspension with the existing two component method. This avoids the sorption problem. However, negative Xs value still make assessment difficult. Also at very low Xs values test reproducibility is expected to be low. 98 Chapter 5 Conclusions 1. The consecutive, competitive azo coupling reaction between 1-naphthol and diazotized sulfanilic acid can be used to assess mixing in low consistency pulp suspensions by applying a correction for sorption of the dyes in presence of fibres (C m <3%). The average energy dissipation rate (?) in the stirred vessel is 0.1-7 W/kg. 2. There is good fibre-scale mixing with the chemicals when the addition is in the high energy dissipation range, i.e. near the impeller vicinity. The mixing at a constant impeller speed is not affected by change in mass concentration i f there is good macromixing and the addition of chemical is in the high energy dissipation zone. 3. In pulp suspensions the energy dissipation rate is an indication of mixing only when there is complete motion in the suspension without cavern formation. 5. With high-intensity mixers the two component system as well as the three component system (accounting for the ortho monoazo dye) is useful only to a certain local energy dissipation level ( < 400 W/kg). For higher energy dissipation rates the new extended system with five components with three reactants (1-naphthol, 2-naphthol and diazotized sulfanilic 99 acid) can be used up to 105 W/kg (Bourne et al., 1992). But the increase in the number of components make the analysis difficult. In the presence of fibres with lignin the adsorption with five components will be more complex and making the correction for adsorption effects very difficult. 100 Chapter 6 Recommendations for Future Work 1. In this work the fibre-scale mixing was studied upto 2.5% C m . It can be extended to higher mass concentrations of the pulp suspension provided the Xs is greater than 0.02 and A is maintained in excess. 2. The adsorption of product dyes on pulps with different lignin contents can be used to determine the relationship between lignin and dye adsorption. 4. The three component (considering the ortho monoazo dye) method can be used to improve the accuracy of the test. 101 Nomenclature SYMBOL DEFINITION UNITS A 1-naphthol B diazotized sulfanilic acid C concentration mol/m 3 cbulk bulk concentration mol/m 3 C c critical concentration fraction C m mass concentration of pulp suspension fraction C v volumetric concentration of pulp suspension fraction d fibre diameter m D diameter, impeller diameter m D diffusivity m2/s D e f f effective diffusivity m2/s D x vessel diameter m E total extinction at wave length X over path length b E modulus of elasticity N/m 2 FSP fibre saturation point kg water/kg fibre g acceleration due to gravity kg/m.s2 H liquid depth in vessel m k,kltk2 reaction rate constants m'mor's"1 102 K L Langmuir adsorption coefficient L/mmol K P partition coefficient cm3/gm of fibre / fibre length m < / > ; weighted fibre length by length m < / > „ , weighted fibre length by weight m L B length of impeller blade m mf mass of fibre g mw mass of water g M mixing modulus, Equation 1.11 M w Weisz modulus, ( -r A " ' )W 2 /C b u l k D e f f n number «! number of impeller blades n2 number of baffles N rotational speed s"1 N p power number, P/D 5 N 3 p N ^ , Reynolds number, D2Np/j« P power Watts q* fibre phase concentration mg/g of fibre q e equilibrium fibre phase concentration mg/g of fibre Q equilibrium fibre phase concentration mg/g of fibre Q volumetric flow rate m3.s_ 1 r rate of reaction mol.m^.s"1 103 r'" rate of reaction mol.m"3, s"1 rs inside fibre radius m r c outside fibre radius m R monoazo dye R rotor radius m o-R ortho monoazo dye p-R para monoazo dye S bisazo dye Se Schmidt number, v/D t time s tc critical feed time s tf feed time s t,„ macromixing time s T torque N.m v velocity m/s V volume m 3 w baffle width m W half thickness of fibre m W B width of impeller blade m W R V water retention value kg water/kg pulp Xs product distribution Xsjnit initial product distribution 104 corrected product distribution, corrected for adsorption in presence of fibres, Equation 2.7 105 GREEK LETTERS SYMBOL DEFINITION UNITS a volume ratio of A and B solutions before reaction /3 shear rate s"1 6 critical time or macromixing time s"1 ô path length cm ô 0 initial half-thickness of reagent layer m e energy dissipation rate W/kg ï average energy dissipation rate W/kg ef effectiveness factor X wave length m X k Kolmogoroff velocity microscale m /* viscosity N.s/m 2 /*„ apparent viscosity N.s/m 2 p density kg/m 3 a surface tension N/m TY shear stress N/m 2 T tortuosity TY network yield stress N/m 2 4> ratio of local to average rate of energy dissipation 106 107 References Angst, W. , J. R. Bourne and P. Dell 'Ava, "Mixing and Fast Chemical Reaction IX" , Chem. Eng. Sci. 39, 335-344 (1984). Atkinson, E . S. and H de V . Partridge, "Effects of Mixing and Degree of Chlorination on Quality and Bleaching Costs", Tappi 49(2), 66A-72A (1966). Baldyga, J. and J. R. Bourne, " A Fluid Mechanical Approach to Turbulent Mixing and Chemical Reaction II", Chem. Eng. Comm. 28, 243-258 (1984a). Baldyga, J. and J. R. Bourne, " A Fluid Mechanical Approach to Turbulent Mixing and Chemical Reaction III", Chem. Eng. Comm. 28, 259-264 (1984b). Baldyga, J. and J. R. Bourne, "Calculation of Micromixing in Inhomogeneous Stirrer Tank Reactors", Chem. Eng. Res. Des. 66, 33-38 (1988). Beek, J. and R. S. Miller, "Mechanisms of Mixing", Mixing in the Process Industries, N . Harnby, M . F . Edwards and A . W. Nienow, Eds., Butterworths Publishers, London, pp. 170-182 (1985). Bennington, C. P. J . , "Mixing Pulp Suspensions", PhD thesis, The University of British Columbia, Vancouver, B . C . (1988). Bennington, C. P. J . , R. J. Kerekes and J. R. Grace, "Mixing in Pulp Bleaching", J. Pulp Pap. Sci. 15(5), J186-J195 (1989). Bennington, C. P. J. and J. R. Bourne, "Effect of Suspended Fibres on Macro-mixing and Micro-mixing in a Stirred Tank Reactor", Chem. Eng. Comm. 92, 183-197 (1990). Bennington, C. P. J . , R. J. Kerekes and J. R. Grace, "The Yield Stress of Fibre Suspensions", Can. J. Chem. Eng., 68, 748-757 (1990). Bennington, C. P. J . , R. J. Kerekes and J. R. Grace, "Motion of Pulp Fibre Suspensions in Rotary Devices", Can. J. Chem. Eng., 69, 251-258 (1991). 108 Bergnor, E . , B. Backhand and A . Teder, "The Benefits of Better Mixing", Intl. Pulp Bleaching Conf., Quebec, PQ, CPPA/TAPPI , 233-241 (1985). Bourne, J. R. and F . Kozichi, "Mixing Effects During the Bromination of 1,3,5-trimethoxy Benzene", Chem. Eng. Sci. 32, 1538-1539 (1977). Bourne, J. R. , P. Ryes and K . Suter, "Mixing Effects in the Bromination of Resorcin", Chem. Eng. Sci. 32, 711-716 (1977). Bourne, J. R. and H . L . Toor, "Simple Criteria for Mixing Effects in Complex Reactions", A.I .Ch.E.J . 23, 603-604 (1977). Bourne, J. R. , F. Kozichi and P. Rys, "Test Reactions to Determine Segregation", Chem. Eng. Sci. 36, 1643-1654 (1981). Bourne, J. R. , F. Kozichi, U . Moergeli and P. Rys, "Mixing and Fast Chemical Reactions III", Chem. Eng. Sci. 36, 1655-1664 (1981). Bourne, J. R. and S. Rohani, "Micro-mixing and the Selective Iodination of 1-Tyrosine", Chem. Eng. Res. Des. 61, 297-302 (1983). Bourne, J. R., "Mixing in Single Phase Chemical Reactors", Mixing in the Process Industries, N . Harnby, M . F . Edwards and A . W. Nienow, Eds., Butterworths Publishers, London, pp. 170-185 (1985). Bourne, J. R. , C. Hilber and G. Tovstiga, "Kinetics of the Azo Coupling Reactions between 1-Naphthol and Diazotized Sulfanilic Acid", Chem. Eng. Comm. 37, 293-314 (1985). Bourne, J. R. and P. Dell 'Ava, "Micro-and Macro-mixing in Stirred Tank Reactors of Different Sizes", Chem. Eng. Res. Des. 65, 180-186 (1987). Bourne, J. R. and C. P. Hilber, "The Productivity of Micromixing-controlled reactions: Effect of Feed Distribution in Stirred Tanks", Trans. Instn. Chem. Engrs. 68, Part A , 51-56 (1990a). Bourne, J. R. , O. M . Kut, J. Lenzner and H . Maire, "Kinetics of the Diazo Coupling between 1-Naphthol and Diazotized Sulfanilic Acid", Ind. Eng. Chem. Res. 29(9), 1761-1765 (1990b). Bourne, J. R. and S. A . Thoma, "Some Factors Determining the Critical Feed Time of a Semi-Batch Reactor", Chem. Eng. Res. Des. 69, 321-323, (1991). Bourne, J. R. , O. M . Kut and J. Lenzner, "An Improved Reaction System To Investigate Micromixing in High-Intensity Mixers", Ind. Eng. Chem. Res. 31, 949-958 (1992). 109 Brady, J. E . and G. E . Humiston, "General Chemistry — Principles and Structures", John Wiley and Sons, New York, pp. 244-276 (1982). Clark, J. A . , "Surface Phenomena", Pulp Technology and Treatment of Paper, II Edition, Miller Freeman Publications Inc. San Francisco, pp. 79-99 (1986). Corte, H . , "The Raw Materials and Processing of Paper Making", Chapter 1 in Handbook of Paper Science, Vo l . 1, H . F . Ranee, Ed. , Elsevier, New York, pp. 1-42 (1980). Dada, A . E . and L . A . Wenzel, "Estimation of the Adsorbent Capacities from the Adsorption Isotherm of Binary Liquid Mixtures on Solids"., Ind. Eng. Chem. Res. 30, 396-402 (1991). Das, B. S. and H . Lomas, "Flocculation of Paper Fines. I. Adsorption of and Flocculation by Polyelectrolytes. II. Study of the Nature of the Solid Surface and Soluble Impurities", Pulp and Paper Magazine of Canada, 74(8), 95-102 (1973). Edwards, M . F . , "A Review of Liquid Mixing Equipment", Mixing in the Process Industries, N.Harnby, M . F. Edwards and A . W. Nienow, Eds., Butterworths Publishers, London, pp. 113-130 (1985). Elliot, R. G. and T. D . Farr, " M i l l Scale Evaluation of Chlorine Mixing", Tappi 56(11), 68-70 (1973). Ellis, R. G . , K . B. Jewett, K . A . Smith, W. H . Ceckler and E . V . Thompson, "Water Retention Ratio and its Use to Study the Mechanism of Water Retention in Paper", J. Pulp Paper Sci., 9(1), TR12-TR15 (1983). Gullichsen, J. and E. Hàrkônen, "Medium Consistency Technology", Tappi 64(6), 69-72 (1981). Kallmes, O., "Distribution of the Constituents Across the Wall of Unbleached Spruce Sulfite Fibres", Tappi 43(2), 143-146 (1960). Kerr, A . J. and D . A . I. Goring, "The Ultrastructural Arrangement of the Wood Cell Wall" , Cellulose Chemistry and Technology 9, 563-590 (1975). Kolmodin, H . , "How to Save Costs by Mixing Chlorine Dioxide and Pulp Homogeneously", Svensk Papperstidn. 18, 8-14 (1984). Larson. A . and P. Stenius, "Sorption of Small Organic Molecules by Cellulose from Hexane Solutions", Nordic Pulp and Paper Research Journal 3, 87-91 (1987). 110 Levenspiel, O., "The Porous Catalyst Pellet — Identifying the Rate Regime", Chemical Reactor Omnibook, Chap.22, OSU Book Stores, Oregon (1984). Lindstrom, T., Ch. Soremark, Ch. Heinegard and S. Martin-Lof, "The Importance of Electrokinetic Properties of Wood Fibre for Papermaking", Tappi 57(12), 94-98 (1974). MacDonald, R. G . , "Structural and Physical Properties of Pulpwood", Pulp and Paper Manufacture, Vol.1, McGraw H i l l , New York, pp. 1-32 (1965). Mantell, C. L . , "The Unit Operations of Adsorption", Adsorption, Mcgraw-Hill, New York pp. 1-20 (1951). Mark, R. E . , Ed. , "Handbook of Physical and Mechanical Testing off Paper and Paperboard", Vol . 2. Marcel Dekker Inc., New York, pp. 314-318 (1984). Mason, S. G . , "Fibre Motion and Flocculation", Tappi 37(11), 494-501 (1954). McCabe, L . , J. C. Smith and P. Harriott, "Adsorption", Unit Operations of Chemical Engineering, McGraw H i l l , New York, pp. 686-706 (1986). Meyer, R. and D. Wahren, "On the Elastic Properties of Three-Dimensional Fibre Networks", Svensk Papperstid. 67(10), 432-436 (1964). Molen, K . V . D . and H . R. V . Maanen, "Laser-Doppler Measurements of the Turbulent Flow in Stirred Vessels to Establish Scaling Rules", Chem. Eng. Sci. 33, 1161-1168 (1978). Nagata, S., "Power Consumption of Mixing Impellers", Mixing — Principles and Applications", John Wiley and Sons, New York, pp. 1-82 (1975). Oldshue, J. Y . , "Power Correlations and the Effects of Mixing Environment", Fluid Mixing Technology, McGraw-Hill , New York, pp. 43-71 (1983). Panshin, A . J. and C. de Zeenw, "Textbook of Wood Technology", 4th ed., McGraw H i l l , New York, pp. 131-135 & 166-172 (1980). Paterson, A . H . J. and R. J. Kerekes, "Fundamentals of Mixing in Pulp Suspension: Measurement of Microscale Mixing of Chlorine", J. Pulp and Paper Sci. 11(4), 108-113 (1985). Piirainen, R., "Optical Method Provides quick and Accurate Analysis of Fibre Length", Pulp and Paper 59(11), 69-71 (1985). Paul, E . L . and R. E. Treybal, "Mixing and Product Distribution for a Liquid-Phase, Second Order Competitive-Consecutive Reactions", A.I .Ch.E.J . 17, 603-608 (1977). I l l Rattee, I. D . and M . M . Breuer, "Finding Forces", The Physical Chemistry of Dye Adsorption", Academic Press, New York, pp. 1-25 (1974). Robitaille, M . A . , "High Intensity Mixing at the Chlorine Dioxide Stage", Pulp Paper Can. 88(4), T109-T111 (1987). Rydholm., "Species Anatomy and Physical Properties of Wood", Pulping Processes, Interscience Publishers, New York, pp. 42-67 (1965). Rys, P. and H . Zollinger, "Azo Dyes", Fundamentals of the Chemistry and Application of Dyes, Wiley-Interscience, New York, pp. 42-71 (1972). Silvester, R. S., "Classification of Non-Newtonian Fluids", Mixing of Non~Newtonian Media - A Technical Review, B H R A , Bedford, pp. 2-6 (1985). Sjôstrom, E . , "Lignin — Removing Bleaching", Wood Chemistry — Fundamentals and Applications, Academic Press, New York, pp. 146-164 (1989). Soszynski, R. M . , " The Formation and Properties of Coherent Floes in Fibre Suspensions", Ph. D . Thesis, University of British Columbia, Vancouver, B . C . (1987) Stone, J. E . and A . M . Scallan, "The effects of Component Removal Upon the Porous Structure of the Cell Wall of Wood. II. Swelling in Water and Fibre Saturation Point", Tappi 50(10), 496-501 (1967). Stone, J. E . , A . M . Scallan and G. M . A . Aberson, "The Wall Density of Native Cellulose Fibres", Pulp and Paper Mag. of Canada 67(5), T263-T268 (1966). Strazdins, E . , "Surface Chemical Aspects of Polymer Retention", Tappi 57(12), 76-80 (1974). Studer, M . , " Ablauf Schneller Reaktionen in Rotor-Stator - Mischern verschiedener Grosse", Ph.D. Thesis 9037, E T H Zurich, (1989). Thoma, S. A . , "Interactions Between Macro-and Micromixing in Stirred Tank Reactors", Diss. E T H Nr. 9012, Zurich, (1989). Truong, K . T. and J. Methot, "Segregation Effects on Consecutive Competing Reactions in a CSTR", Can. J. Chem. Engng. 54, 572-577 (1976). Uhl , V . W . and J. B. Gray, Eds., "Impeller Characteristics and Power", Mixing, Volume I, Academic Press, New York, pp. 111-175 (1966). van Heiningen, A . , Private Communications (1992). 112 Wahren, D . , "On the Visco—Elastic Properties of Fibre Networks", Svensk Papperstid. 67(9), 378-381 (1964). Walter, J. and J. R. Weber, "Adsorption Theory, Concepts, and Models", Chapter 1 in Adsorption Technology, Frank L . Slejko, Ed. , Marcel Dekker Publishers, New York, pp. 1-35 (1985). Weast, R. C , Ed. , "Concentration Properties of Aqueous Solutions: Conversion Tables", Handbook of Chemistry and Physics, 65th éd., CRC Press, Florida, pp. D222-D281 (1985). Wohl, M . H . , "Mixing of Non-Newtonian Fluids", Chem. Eng., 75(32), 113-118 (1968). Zollinger, H . , "The Mechanism of the Coupling Reaction", Azo and Diazo Chemistry, Interscience Publishers, New York, pp. 221-248 (1961). 113 Appendix A Fibre and Suspension Properties A . l Fibre Dimensions Fibre length is an important property that influences the yield stress of pulp suspensions. Higher the yield stress greater the energy required to achieve a required degree of mixing in the suspension. The average length of wood fibres is given in a number of publications (Rydholm, 1965; MacDonald, 1965; Panshin, 1980). Wood fibres are about 1-3 mm long, 20-45 ixm in diameter. For pulp fibres, a weighted average fibre length is a more useful measure of fibre length than simply the arithmetic average due to the disproportionate influence that the longer fibres have on most sheet properties (Mark, 1984). The average fibre length and its distribution for the fibres used in this study was determined by Kajaani FS-200 Fibre Length Analyzer (Piirainen, 1985) which counts fibres as well as measure their lengths. This instrument provides the population distribution which is described as the number of fibres in a certain fibre length range as a percentage of the total number of fibres analyzed. It also gives the length-weighted distribution which is defined as the percentage of fibres by length in each 114 range as a percentage of the total length of fibres. The two average fibre lengths are defined as follows: < 1 > = J ^ i z i (A . l ) < 1 (A.2) where < 1 >, = weighted fibre length by length < 1 > w = weighted fibre length by weight The length weighted distribution of the fibre used in this study is shown in figure A . 1 and the population distribution is shown in Figure A.2 . The standard printout from the Kajaani FS-100 gives the upper limit of each length range. This is reported in terms off range boundaries. The population as well as the length-weighted distribution are reported in the table A . l . Fibre diameters were not measured. Average fibre length was taken from the literature based upon the species of the pulp sample (Rydholm, 1965). 115 6 -^ 5 -Fibre Length, mm Figure A . 1 : Length weighted distribution of the fully bleached kraft pulp fibre (soft wood) used in this study using Kajaani FS-200 Analyzer. Length weighted average fibre length = 2.40 mm. 116 Figure A.2 : Population distribution of the fully bleached kraft pulp fibre (soft wood) used in this study using Kajaani FS-200 Analyzer. Arithmetic average fibre length = 0.93 mm. Fibre coarseness = 0.241 mg/m. 117 Table A . l : Length weighted distribution of the fully bleached softwood kraft pulp used in this study (by Kajaani FS-200 Analyzer). Length, mm Number of Fibres Weighted % Length, mm Number of Fibres Weighted % 0.00-0.20 9809 3.65 4.00-4.20 171 3.28 0.20-0.40 2307 2.87 4.20-4.40 133 2.68 0.40-0.60 1367 3.09 4.40-4.60 69 1.45 0.60-0.80 1044 3.33 4.60-4.80 60 1.32 0.80-1.00 950 3.94 4.80-5.00 39 0.89 1.00-1.20 739 3.75 5.00-5.20 28 0.67 1.20-1.40 629 3.79 5.20-5.40 16 0.40 1.40-1.60 645 4.50 5.40-5.60 12 0.31 1.60-1.80 625 4.94 5.60-5.80 10 0.27 1.80-2.00 532 4.71 5.80-6.00 7 0.19 2.00-2.20 496 4.87 6.00-6.20 4 0.11 2.20-2.40 541 5.81 6.20-6.40 1 0.03 2.40-2.60 506 5.92 6.40-6.60 4 0.12 2.60-2.80 413 5.22 6.60-6.80 4 0.13 2.80-3.00 404 5.48 6.80-7.00 1 0.03 3.00-3.20 368 5.34 7.00-7.20 0 0.00 3.20-3.40 300 4.63 3.40-3.60 266 4.34 3.60-3.80 242 4.18 3.80-4.00 206 3.76 Total number of fibres counted = 22 948 Arithmetic average fibre length = 0.93 mm Length weighted average fibre length = 2.40 mm Weight weighted average fibre length =3.11 mm Fibre coarseness = 0.241 mg/m 118 A.2 Water Retention Value The walls of fibres are made up of cellulose, hemicellulose and lignin that can take up water and swell (hydrophilic property). Wood fibres are tube-like with closed and pointed ends. The space enclosed by the fibre walls is called a "lumen". During pulping the hemicellulose and lignin are removed. This make the fibre porous. This pores can be enlarged by mechanical and chemical treatment. In the dry state the fibres are essentially non-porous (Stone and Scallan, 1966). So in the wet state the walls are swollen above their dry volume by an amount equal to the volume of water they contain. According to Stone and Scallan (1967) the amount of water contained within the water-saturated cell wall is the "fibre saturation point". The water held in the fibres is also expressed by WRR (water retention ratio) (Ellis et al., 1973). The lumens hold certain amount of water. This depends upon the morphology and the state of lumen collapse. The water in the cell walls and lumen is expressed by the term WRRk (Soszynski, 1987). This is a water retention ratio when lumen is full of water and fibre walls are at the saturation point. In beaten fibres, the microfibrils outside the surface of the fibre hold some water and the interface between a number of fibres form a capillary system which can hold some amount of water. A.3 Concentration of Pulp Suspension Pulp suspensions are usually characterized by their mass concentration or consistency, C m . The mass concentration is simply the mass of dry fibre divided by the total mass of the suspension: C m = mf I (mf + m J (A.3) 119 It is commonly determined by over drying a pulp sample, and standard procedures are specified for this test by the CPPA and TAPPI. A.4 Half-thickness of Fibre Half thickness of the softwood fully bleached kraft fibre used in the study can be calculated using the formula (van Heiningen, 1992): W = rf +(r) 1 / 2(r 0 -rj) (A.4) where r, and r 0 are inside and outside radii of the fibre and T is the tortuosity, when rf = 20 fim, r 0 = 25 /xm and tortuosity = 2.5 (van Heiningen, 1992) the half width of the fibre is 27.9 pm. 120 Appendix B Data Reproducibility The reproducibility of the results are important in any experiment. The error in a test result consists of two parts. They are (a) analytical and (b) experimental. Analytical error is due to the measurement in the spectrophotometer. Experimental error consists of various factors like weighing the sample, mixing the sample and filtering etc., The analytical error was determined by taking ten different measurements of the same dye sample. The experimental error was assessed by carrying out ten different experiments with identical experimental conditions and making measurements of the product distribution in the spectrophotometer. The data reproducibility was found out by applying the Gaussian distribution concepts to analytical and experimental data. First, the mean (m), the standard deviation (S) of the measurements and the standard error of the mean (S^ were calculated. It was necessary to find the confidence limits to a certain level of significance. The confidence limits (or confidence interval) defines the interval in which the "TRUE" mean of the overall population will lie, based on measurements of a finite sample. The confidence limits/interval is given by: 121 m ± t p . S E = m±tp(S/N 1 / 2) <R1) where tp, the t-value, accounts for the degree of uncertainty in the estimate. The t-value depends on: - Number of measurements which affect the degrees of freedom (N) - The confidence limits we wish to establish (95% confidence level in this case or the 95 % probability of making the right prediction) Table B . l shows the 95% confidence interval for the adsorption of 1-naphthol on fibre. Table B.2 shows the 95% confidence interval taking for analytical measurements of the dye concentration and Table B.3 shows the 95% confidence limits of the experimental measurements of the dye adsorption. The analytical error is low when compared to experimental error. This is expected as the experimental steps were more when compared to the analytical measurement. Table B.4 shows the 95% confidence limits for the power measurement in water and 2% C m pulp suspension. Table B . l : Adsorption measurements and data reproducibility. C m of pulp suspension = 1.8%. concentration of 1-naphthol, mol/m 3 iug of adsorbed 1-naphthol/g of fibre, (mean) Standard deviation, (S) ixg of adsorbed 1-naphthol/g of fibre 95% confidence interval Hg of adsorbed 1-naphthol/ g of fibre 0.46 108 3.3 108+5.25 0.92 145 4.3 145±6.78 1.40 166 6.6 166±10.4 2.35 199 6.2 199+9.92 122 Table B.2: Analytical measurements of the concentration of dyes for calculation of data reproducibility at 25±0 .1°C. N A o / N B o = 1.10 and V A / V B = 50. C A o = 0.52 mol/m 3, = 23.6 mol/m 3. [R], mol/m 3 [S], mol/m3 Xs Mass Balance, % 0.365 0.0261 0.1251 90.3 0.365 0.0264 0.1263 90.3 0.364 0.0265 0.1271 90.3 0.365 0.0264 0.1266 90.0 0.365 0.0265 0.1268 90.2 0.365 0.0260 0.1245 90.2 0.365 0.0272 0.1300 90.2 0.363 0.0263 0.1260 90.3 0.365 0.0264 0.1266 90.2 0.365 0.0260 0.1258 90.2 [R], mol/m 3 [S], mol/m 3 Xs Mass balance, % Mean 0.365 0.0264 0.1265 90.2 Standard deviation 0.000675 0.000346 0.00147 0.095 95% confidence interval 0.365± 0.000482 0.0264 + 0.000345 0.1265 + 0.00105 90.2 ± 0.068 123 Table B.3: Experimental measurements for calculation of data reproducibility of dye adsorption on pulp suspension (C m = 2%) a t 2 5 ± 0 . 1 ° C . N A o / N B o = 1.10 and V A / V B = 50. C A o = 0.52 mol/m 3, = 23.6 mol/m3. [R], mol/m 3 [S], mol/m3 Xs Mass Balance, % 0.304 0.0194 0.1131 84.7 0.301 0.0193 0.1137 84.0 0.297 0.0180 0.1080 82.2 0.290 0.0195 0.1186 81.3 0.306 0.0199 0.1153 85.3 0.269 0.0182 0.1137 75.5 00.296 0.0208 0.1233 83.5 0.296 0.0190 0.1140 82.4 0.281 0.0210 0.1245 83.4 0.296 0.0190 0.1193 78.8 [R], mol/m 3 [S], mol/m3 Xs Mass balance, % Mean 0.294 0.0194 0.1164 82.1 Standard deviation 0.011 167 0.000 972 0.005 05 2.97 95% confidence interval 0.294± 0.007 98 0.0194 + 0.000 695 0.1164 + 0.0036 82.1+2.12 124 Table B.4: Data reproducibility of energy dissipation measurements in water and in suspension (FBK). R P M 2% C m Pulp Suspension Water Energy dissipation rate, W/kg 95 % confidence interval, W/kg Energy dissipation, rate, W/kg 95 % confidence interval, W/kg 100 0.028 0.028± 6.46 xlO" 3 0.0078 0.0078+2.54 xlO" 3 200 0.071 0.071 + 1.65 xlO" 2 0.058 0.058 +1.49 xlO" 2 300 0.212 0.212± 2.28 xlO" 2 0.265 0.265+2.80 xlO" 2 400 0.611 0.611 + 1.75 xlO" 2 0.67 0.67+2.55 xlO" 2 500 1.282 1.282 + 2.12 xlO" 2 1.34 1.34+4.62X10"2 600 2.338 2.388± 8.68 xlO" 2 2.33 2.33+4.33 xlO" 2 700 4.032 4.032 + 1.40 xlO" 2 3.69 3 . 6 9 ± 6 . 6 5 x l 0 2 800 6.285 6.285+0.11 4.89 4.89+0.33 125 Appendix C Preparation of Chemicals Table C . l : Chemicals, formula weight and the concentration used Chemical Formula Weight Concentration, mol/m 3 1-naphthol (A), C 1 0 H 7 O H 144.17 0.52 Sodium carbonate, N a 2 C 0 3 105.99 39 (for buffer) Sodium bi-carbonate, N a H C 0 3 84.01 39 (for buffer) Hydrochlolic Acid, HC1 36.46 56 Sulfamic Acid, N H 2 - S 0 3 H 97.09 0.944 Sulfanilic acid (B) ,NH 2 -C 6 H 4 -S0 3 H 173.19 23.6 Sodium Carbonate 105.99 11.8 (for dissolving sulfanilic acid) Sodium Nitrite, N a N 0 2 69 24.5 XlO" 3 Table C.2: Conditions for coupling reaction Temperature of coupling reaction 25±0 .1 °C Volume of A/Volume of B, V A / V B 1000/20 = 50 Concentration of A/Concentration of B 0.022 Moles of A/Moles of B, N A o / N B o 1.10 PH 10 126 C . l Preparation of 1-Naphthol at 0.52 mol/m 3 The dissolution of 1-naphthol in water is slow. So the amount of 1-naphthol required was dissolved overnight in a volume of distilled water a little less than that required to make up a particular concentration. In the morning the stirrer and sides of the vessel were washed with the remaining water and stirred a further 15 minutes. The solution was then buffered to pH 10 by dissolving the N a 2 C 0 3 (4.13 g = 39 mol/m3) and N a H C 0 3 (3.28 g = 39 mol/m3), just before the reaction. To prepare a volume of 1 L of 1-naphthol solution having a concentration of 0.52 mol/m 3, the amount of 1-naphthol required was 0.075 g (equivalent to 0.52 mol/m3) and this amount was dissolved in 1 L of water. The solution was buffered to pH 10 just before coupling reacttion. C.2 Typical Adsorption Test - Weight of pulp - Mass concentration of pulp - O.D. weight of fibre - Amount of water allowed with the fibre - Total suspension weight - Volume of 1-naphthol solution after pulp addition - Concentration of 1-naphthol before pulp addition - Concentration of 1-naphthol after pulp addition 21.06 gm 94.499% 19.902 g 60 g 60 + 19.902 g 1.060 L 0.4978 mol/m 3 0.4978 x (1000/1060) 0.4696 mol/m 3 Mass concentration of the suspension = 19.90/(1060+19.90) 127 - Amount of 1-naphthol present in solution before adsorption - Amount of 1-naphthol present in solution after adsorption C.2 Preparation of Diazotized Sulfanilic Acid - Amount of sulfanilic acid required - Moles of N a 2 C 0 3 required = 23.6/2 mol/m 3 =11.8 mol/m3 = 11.8 molx 105.99 g.mol"1 - Amount of N a 2 C 0 3 required = 1.25 g 2. The sulfanilic acid was dissolved in 550 ml of distilled water. 100 ml was used for washing the sulfanilic acid from the side of the beaker wall during dissolution. 3. 350 g of ice was added to make the total volume to 1 L . By adding ice the temperature was brought down to 0°C (diazotization is normally carried out around 0°C). 4. The Sodium Nitrite (4% excess) was added to the above sulfanilic acid solution and dissolved. It was assumed that the reaction would be completed by adding N a N 0 2 in slight excess(4%) thus sulfanilic acid was the limiting reactant). - Moles of N a N 0 2 required = 23.6x1.04 mol/m 3 = 2 3 . 6 x l 0 3 X 1.04x69 g/L - Amount of NaN0 2 required = 1.69 g. 5. HC1 (37%) was added to the above solution to bring down the pH to 2. The acid was added as quickly as possible (within 2 minutes after adding NaN0 2 ) . The HC1 required was 2 moles per mole of sulfanilic acid and 1 mole per mole of sodium carbonate. This balanced the bases present. More acid was required to bring down the pH to 2. - HC1 requirement = 5.6 ml 6. After HC1 addition the reaction was allowed to proceed for 30 minutes. A magnetic stirrer was used to mix the reactants. 7. At the end of 30 minutes sulfamic acid was added to destroy the excess sodium nitrite and to stop the reaction. The sufamic acid required was equivalent to the excess sodium 129 nitrite present. So 4% of sulfamic acid was added to destroy the excess sodium nitrite. Sulfamic acid required = 23.6x0.04x97.09 mg/L = 90.17 mg For Analysis of the Dye: It was assumed that all A was converted to S. Therefore the maximum concentration = 0.498 mol/m 3. As the sample concentration should be -0 .05 mol/m 3 (for analysis) the dilution required was 1:12 to keep absorption in range, i.e. for 1 part of dye solution 11 parts of buffer required. Buffer was prepared to give ionic concentration I = 40 mol/m 3 N a 2 C 0 3 > 2 N a + + C 0 3 2 - (C.2) N a H C 0 3 * N a + + HC0 3 " (C.3) Concentration of N a 2 C 0 3 and N a H C 0 3 after dilution was 39 mol/m3 (75x0.52 mol/m 3). I = l/2(2[Na +] + [Co 3.]4+[Na+] + [ H C 0 3 ] + [monoazo]4) It was assumed that monoazo dye = 0.52 mol/m 3 Considering equation C.2 and C.3, if [Na 2 C0 3 ] and [NaHC0 3 ] = x I = 0.5(2x + 4x + x + x) = 4x Hence I = 40 = (1/12)4x39 + (1/12)4(0.52) + (11/12)4Xx 40 = 13.00 + 0.173 + 3.67x x - 7.316 mol/m 3 since the presence of monoazo dye was not significant it was ignored. Under this condition the concentration of N a 2 C 0 3 and N a H C 0 3 required was « 10 mol/m 3. 130 (1/12)39 + (ll/12)x = 10 x = 7.316 mol/m 3 To prepare 1 L of buffer with concentration 7.316 mol/m 3 the amount of N a 2 C 0 3 and NaHCO 3 required were 0.775 g and 0.615 g respectively. Spectrophotometer Analysis for 1-naphthol and R and S: The 1-naphthol was diluted 4 times before analysis. The absorbence was measured at 332 nm. Dye samples were diluted 12 times and the absorbence was measured at 7 wavelengths between 400 and 600 nm. 131 Appendix D Product Distribution Determination Computer Program The computer program used for the calculation o f product distribution, mass balance and concentration of dyes from the spectral data is given. 10 REM Calculate Xs from Spectral Data 2 0 REM (C) Chad P . J . Bennington 3 0 CLS 4 0 REM To compute data h i t " =F1 ^RETURN " 50 ON KEY (1) GOSUB 560 = ^ 60 KEY(1) ON 70 DEFINT I-K 80 LOCATE 5,1,1,0,7 9 0 INPUT "Test Identification";A$ 100 INPUT "Sample Dil u t i o n / /;DIL 110 INPUT "Total Moles of B i n i t i a l l y ";MOLEB 120 INPUT "Total Volume of Reactor =m**3 ";VTOT 130 LPRINT "Test ";A = 140 LPRINT " ": LPRINT " " 150 SUMXY=SUMX=SUMY=SUMX2=SUMY2=0! 160 LOCATE 10,1 170 INPUT "Enter Test Data Point ^WAVELENGTH,EXTINCTION ";LAMDA, EXT 180 REM Data from Bourne, H i l b e r and Tovstiga, Chem. Eng. Comm. 37: 190 REM 293-314 (1985). 200 IF LAMDA= 600 THEN ER= 142 : ES =1917 : GOTO 420 210 IF LAMDA= 590 THEN ER= 320 : ES =22 07 . : GOTO 420 220 IF LAMDA= 580 THEN ER= 638 : ES =2408 : GOTO 420 230 IF LAMDA= 570 THEN ER= 1102: ES =2531 : GOTO 420 240 IF LAMDA= 560 THEN ER= 1653: ES =2553 : : GOTO 420 250 IF LAMDA= 550 THEN ER= 2199: ES =2496 • : GOTO 420 260 IF LAMDA= 540 THEN ER= 2708: ES =2364 : : GOTO 420 270 IF LAMDA= 530 THEN ER= 3069: ES =2206 : : GOTO 420 280 IF LAMDA= 520 THEN ER= 3248: ES =2070 : • GOTO 420 290 IF LAMDA= 510 THEN ER= 5239: ES =2013 : : GOTO 420 300 IF LAMDA= 500 THEN ER= 3059: ES =2053 : • GOTO 420 310 IF LAMDA-490 THEN ER= 2757: ES =2140 : • GOTO 420 320 IF LAMDA= 480 THEN ER= 2403: ES =2211 : GOTO 420 330 IF LAMDA= 470 THEN ER— 2035: ES =2 219 : GOTO 420 340 IF LAMDA= 460 THEN ER= 1726: ES =2158 : GOTO 420 350 IF LAMDA= 450 THEN ER= 1463: ES =2000 : GOTO 420 132 133 Appendix E Adsorption Data A l l the experiments were conducted at 25+0 .1°C. Table E . l : Adsorption of 1-naphthol (Initial concentration = 0.50 mol/m3) on pulp suspension (C m = 1.8%) with increasing time. Time, min Absorption Concentration of 1-Naphthol, mol/m 3 Amount of 1-Naphthol present in solution/g of Fibre, mg 1-Naphthol Adsorbed /tg/g of Fibre 0 0.8090 0.4978 3.606 -3 0.7550 0.4646 3.568 38 6 0.7490 0.4609 3.539 67 9 0.7450 0.4585 3.520 86 12 0.7420 0.4566 3.506 100 15 0.7410 0.4560 3.501 105 30 0.7405 0.4557 3.499 107 134 Table E.2: Adsorption of 1-naphthol (Initial concentration = 0.998 mol/m3) on pulp suspension (Cm=1.8%) with increasing time. Time, min Absorption Concentration of 1-naphthol, mol/m 3 Amount of 1-naphthol present in solution/g of fibre, mg 1-naphthol adsorbed /xg/g of fibre 0 0.8110 0.9982 7.230 -3 0.7580 0.9323 7.159 71 6 0.7540 0.9280 7.126 104 9 0.7519 0.9254 7.106 124 12 0.7506 0.9238 7.094 136 15 0.7498 0.9228 7.086 144 30 0.7496 0.9226 7.084 146 Table E.3: Adsorption of 1-naphthol (Initial concentration = 1.505 mol/m3) on pulp suspension (C m = 1.8%) with increasing time. Time, min Absorption Concentration of 1-naphthol, mol/m3 Amount of 1-naphthol present in solution/g of fibre, mg 1-naphthol adsorbed, jug/g of fibre 0 0.8150 1.505 10.899 -3 0.7630 1.409 10.817 82 6 0.7600 1.403 10.774 125 9 0.7589 1.401 10.758 141 12 0.7577 1.399 10.741 158 15 0.7571 1.398 10.733 166 30 0.7571 1.398 10.733 166 135 Table E.4: Adsorption of 1-naphthol (Initial concentration = 2.523 mol/m3) on pulp suspension (C m = 1.8%) with increasing time. Time, min Absorption Concentration of 1-naphthol, mol/m 3 Amount of 1-naphthol present in solution/g of fibre, mg 1-naphthol adsorbed /xg/g of fibre 0 0.8201 2.523 18.278 -3 0.7699 2.369 18.191 87 6 0.7679 2.363 18.143 135 9 0.7666 2.359 18.113 165 12 0.7657 2.356 18.091 187 15 0.7652 2.354 18.079 199 30 0.7651 2.354 18.079 199 Table E.5: Adsorption of 1-naphthol (Initial concentration = 0.499 mol/m3) on pulp suspension (C m = 0.9%.) with increasing time. Time, min Absorption Concentration, mol/m 3 Amount of 1-naphthol/g of fibre, mg 1-naphthol adsorbed ttg/g of fibre 0 0.8110 0.4991 7.230 -3 0.7608 0.4682 7.190 40 6 0.7576 0.4662 7.160 70 9 0.7556 0.4650 7.141 89 12 0.7543 0.4642 7.128 102 15 0.7538 0.4639 7.124 106 30 0.7535 0.4637 7.121 109 136 Table E.6: Adsorption of 1-naphthol (Initial concentration = 0.498 mol/m3) on pulp suspension(Cm = 2.7%) with increasing time. Time, min Absorption Concentration, mol/m 3 Amount of 1-naphthol present in solution/g of fibre, mg 1-naphthol adsorbed, /xg/g of Fibre 0 0.8090 0.4978 2.404 -3 0.7517 0.4626 2.368 36 6 0.7428 0.4571 2.340 64 9 0.7365 0.4532 2.320 84 12 0.7320 0.4505 2.306 98 15 0.7304 0.4495 2.301 103 30 0.7298 0.4491 2.299 105 137 Table E.7: Adsorption of 1-naphthol on pulp suspension with increasing equilibrium concentration of 1-naphthol. Adsorption time = 30 minutes. Table E.8 X s ^ and Xs after adsorption of dyes at various mass concentrations ( C J of pulp suspension (FBK). Time of adsorption = 30 minutes. c m , % Sink (0.034) (0.070) (0.088) (0.146) (0.189) (0.195) 0.0 0.034 0.070 0.088 0.146 0.189 0.195 0.5 0.035 0.064 0.086 0.137 0.170 0.178 1.0 0.030 0.062 0.077 0.130 0.165 0.180 1.5 0.027 0.060 0.084 0.122 0.160 0.172 2.0 0.025 0.054 0.074 0.124 0.152 0.162 2.5 0.020 0.052 0.071 0.116 0.145 0.157 6.0 0.021 0.042 0.056 0.095 0.128 0.148 10.0 0.012 0.037 0.049 0.085 0.116 0.136 Table E.9: X s / X s ^ with various mass concentrations of pulp suspension (FBK). c m , % Xs/Xs^it 0 1.000 1.000 1.000 1.000 1.000 1.000 0.5 1.030 0.914 0.977 0.938 0.899 0.913 1.0 0.880 0.886 0.875 0.890 0.873 0.923 1.5 0.790 0.857 0.955 0.836 0.847 0.910 2.0 0.735 0.771 0.841 0.849 0.804 0.857 2.5 0.588 0.743 0.807 0.794 0.767 0.830 6.0 0.618 0.600 0.636 0.651 0.677 0.758 10.0 0.353 0.528 0.557 0.582 0.614 0.697 139 Table E.10: Concentration of dyes before and after adsorption at various mass concentrations of pulp suspension(FBK). Time of adsorption = 30 minutes. Table E. 11 : Mass balance before and after adsorption of dyes at various mass concentrations of pulp suspension (FBK). Time of adsorption = 30 minutes. c m , % (0.034) X S ^ (0.070) (0.088) X S in i t (0.146) X S ini t (0.189) XSinit (0.195) Mass Balance, % 0.0 95.4 95.8 92.4 93.4 93.7 95.2 0.5 93.4 93.6 90.3 92.3 92.1 94.6 1.0 92.3 92.7 88.1 90.9 87.1 93.1 1.5 90.0 91.8 85.5 87.1 88.3 92.6 2.0 83.4 86.8 84.9 85.2 81.4 87.7 2.5 85.2 84.8 82.0 82.3 79.9 84.2 6.0 78.8 76.5 71.0 74.9 77.7 80.4 10.0 71.6 71.5 65.2 68.1 64.8 76.6 141 Table E. 12: Equilibrium concentration and adsorption of monoazo dye on pulp suspensions (FBK) at various mass concentrations. c % equilibrium concentration, mol/m3 0.0 0.426 0.398 0.404 0.369 0.352 0.5 0.365 0.342 0.347 0.323 0.309 1.0 0.363 0.335 0.346 0.321 0.295 1.5 0.354 0.325 0.341 0.309 0.301 2.0 0.329 0.325 0.326 0.302 0.279 2.5 0.337 0.315 0.319 0.295 0.277 6.0 0.268 0.236 0.251 0.235 0.235 10 0.245 0.218 0.235 0.216 0.199 Adsorption, /xmol/g of fibre 0.0 1.592 1.284 1.299 -0.200 -0.875 1.0 0.991 1.312 0.743 0.198 1.287 1.5 1.248 1.527 0.821 0.919 0.525 2.0 2.156 1.139 1.348 1.029 1.421 2.5 1.404 1.297 1.346 1.092 1.209 6.0 1.645 1.759 1.606 1.379 1.14 10.0 1.152 1.172 1.067 0.963 0.981 142 Table E.13: Equilibrium concentration and adsorption of bisazo dye at various mass concentrations (C,J of pulp suspension(FBK). c % Equilibrium concentration, mol/m3 0.5 0.0067 0.0117 0.0163 0.0256 0.0318 0.0341 1.0 0.0050 0.0110 0.0145 0.0239 0.0290 0.0339 1.5 0.0050 0.0104 0.0155 0.0216 0.0286 0.0332 2.0 0.0043 0.0093 0.0129 0.0214 0.0251 0.0288 2.5 0.0038 0.0086 0.0122 0.0193 0.0234 0.0268 6.0 0.0028 0.0052 0.0074 0.0123 0.0172 0.0206 10.0 0.0015 0.0042 0.0061 0.0100 0.0130 0.0181 Adsorption, jumol/g of fibre 0.5 0.029 0.2910 0.3780 0.4380 0.7950 0.1390 1.0 0.099 0.2260 0.2430 0.3710 0.6270 0.4850 1.5 0.109 0.1950 0.1970 0.3330 0.5450 0.4850 2.0 0.105 0.1750 0.1710 0.3070 0.4920 0.4480 2.5 0.098 0.1610 0.1580 0.2880 0.4540 0.4090 6.0 0.061 0.1140 0.1170 0.2220 0.3270 0.2480 10.0 0.043 0.0910 0.1090 0.1850 0.2640 0.0170 143 Table E. 14: Change in Xs and mass balance with increasing time in dye solution and with pulp suspension (FBK). Mass concentation (Cn i) = 2%. Time, minute Xs Mass balance, % in solution with pulp suspension in solution with pulp suspension 0 0.157 0 92.0 0 30 0.163 0.138 92.1 77.3 60 0.164 0.167 0.139 0.137 91.9 94.4 77.6 77.7 90 0.167 0.141 95.8 78.0 120 0.164 0.163 0.139 0.140 91.8 94.9 78.3 79.0 150 0.163 0.141 93.1 77.4 180 0.165 0.168 0.137 0.138 93.7 94.1 79.0, 78.7 144 Table E.15: Change in concentration of dyes with time in dye solution and with pulp suspension (FBK). C m = 2%. Time, Concentration of monoazo dye, Concentration of bisazo dye, minute mol/m 3 mol/m 3 in solution with pulp in solution with pulp suspension suspension 0 0.359 0 0.033 0 30 0.357 0.307 0.035 0.025 60 0.355 0.308 0.035 0.025 0.363 0.309 0.036 0.025 90 0.369 0.309 0.037 0.024 120 0.355 0.312 0.035 0.025 0.365 0.313 0.036 0.026 150 0.360 0.308 0.035 0.025 180 0.362 0.316 0.036 0.024 0.361 0.314 0.036 0.025 145 Table E . 16: Product distributions and mass balances of various pulps at different mass concentrations after adsorption with dyes ( X s ^ =0.126) for 30 minutes. Fibre Type Mass Concentration, % Xs Mass Balance, % No fibre 0 0.110 99.2 2 0.113 101.9 Nylon 4 0.111 100.0 6 0.109 101.8 2 0.102 87.7 Fully bleached kraft 4 0.092 81.5 6 0.081 75.3 2 0.101 86.4 Semi-bleached kraft 4 0.083 80.0 6 0.079 75.4 2 0.099 91.1 Unbleached kraft 4 0.076 82.1 6 0.062 72.3 2 0.066 77.2 Themomechanical pulp 4 0.067 70.0 6 0.055 63.2 2 0.079 58.1 Stone groundwood 4 0.066 54.2 6 0.057 49.8 146 Table F . l : Power numbers of water in the stirred tank with varying Reynolds number. Impeller speed, s 1 Torque, N.m Angular velocity, CJ Power, W N P N R e 3.34 0.0095 20.94 1.95 5.3 33 300 5.00 0.0216 31.42 6.65 5.3 50 000 6.68 0.0374 41.89 15.37 5.2 66 700 8.33 0.0585 52.36 30.07 5.2 83 300 10.00 0.0816 62.83 50.27 5.0 100 000 11.67 0.1048 73.30 75.36 4.7 116 600 Table F.2: Power numbers of sucrose solution (viscosity = 15.4 mPa.s) in the stirred tank with increasing Reynolds number. Impeller speed, S"1 Torque, N.m Angular velocity, Power,W N P N R e 3.34 0.103 20.94 2.161 4.74 2 660 5.00 0.248 31.42 7.792 5.07 3 990 6.68 0.413 41.89 17.301 4.25 5 720 8.33 0.661 52.36 34.610 4.86 6 650 10.00 0.171 62.83 61.001 4.96 7 980 11.67 1.281 73.30 93.897 4.80 9 310 13.33 1.653 83.80 138.520 4.75 10 640 147 Table F.3: Power numbers of sucrose solution (viscosity of sucrose solution = 219 mPa.s) in the stirred tank with increasing Reynolds number. Impeller speed, R P M Torque, N.m Angular velocity, CJ Power, W N R e N P 1 0.010 6.28 0.065 61 4.21 2 0.031 12.56 0.389 121 4.12 3 0.072 18.84 1.358 182 4.03 4 0.145 25.12 3.642 243 4.28 5 0.227 31.40 7.128 304 4.29 6 0.310 37.68 11.680 364 4.07 7 0.434 43.96 19.079 425 4.19 8 0.558 50.24 28.034 486 4.13 9 0.723 56.52 40.864 546 4.42 10 0.889 62.80 55.829 607 4.21 11 1.116 69.08 77.093 668 4.36 12 1.343 75.36 101.208 728 4.81 13 1.550 81.64 126.542 789 4.34 Table F.4: Product distributions (Xs) in solution with increasing addition time in the stirred tank. Impeller rotational speed = 2 s"1 at 25±0 .1°C. N A o / N B o =1.10 and V A / V B = 50. C A o = 0.52 mol/m 3, C B o = 23.6 mol/m 3. Time, min Xs at feedpoint A Xs at feedpoint B 2 0.175 0.355 10 0.122 0.310 20 0.099 0.270 30 0.088 0.255 40 0.087 0.255 50 0.087 0.255 148 Table F.5: Product distributions (Xs) with varying impeller rotational speeds in the stirred tank in aqueous medium at 25±0 .1°C. Addition time = 30 minutes. N ^ N ^ =1.10 and V A / V B = 50. C A o = 0.52 mol/m 3, C B o = 23.6 mol/m 3. Feedpoint A N , s"1 [R], mol.m"3 [S], mol.m"3 Mass balance, % Xs 2 0.385 0.0202 92.32 0.0950, 0.0770 4 0.391 0.0161 91.84 0.0763, 0.0650 6 0.412 0.0136 95.30 0.0619 8 0.417 0.0117 95.58 0.0531 10 0.417 0.0110 95.27 0.0501 12 0.413 0.0106 94.23 0.0488 14 0.414 0.0105 94.40 0.0483 16 0.418 0.0115 95.71 0.0522 Feedpoint B 2 0.315 0.0545 92.02 0.257 4 0.348 0.0421 93.80 0.195, 0.193 6 0.371 0.0328 94.75 0.150, 0.153 8 0.378 0.0280 94.18 0.129 10 0.393 0.0259 96.08 0.117 12 0.390 0.0226 94.44 0.104 14 0.396 0.0216 95.31 0.098 16 0.395 0.0218 95.19 0.099 149 Table F.6: Corrected product distribution (Xs c o r r) in pulp suspensions (FBK) with increasing addition time at 25±0 .1°C. Impeller rotational speed (N) = 12 s"1, N A o / N B o = 1.10 and V A / V B = 50. C A o = 0.52 mol/m 3, C B o = 23.6 mol/m 3. Addition time, minute X s c o r r at feedpoint A X s c o r r at feedpoint B 10 0.071 0.178 20 0.065 0.159 30 0.065 0.154 40 0.065 0.154 Table F.7: Motion of the fully bleached kraft pulp suspension (C m = 0.5%) in the stirred tank at various impeller rotational speeds. R P M Torque, N.m Power, W Condition of suspension 120 0.078 0.98 No motion 240 0.163 4.09 Complete motion 480 0.527 26.47 Complete motion 600 0.759 47.67 Complete motion 720 1.193 89.90 Complete motion 150 Table F.8: Motion of the fully bleached kraft pulp suspension at 1% C m in the stirred tank reactor at various impeller rotational speeds. R P M Torque, N.m Power, W Flow condition of the pulp suspension 200 0.078 1.60 No motion. 400 0.310 12.98 Complete motion. 600 0.790 49.60 Complete motion. 800 1.616 135.00 Complete motion. Table F.9: Motion of the fully bleached pulp suspension at 1.5% C m in the stirred tank at various impeller speeds. R P M Torque, N.m Power, W Flow condition of the pulp suspension 600 0.89 56.15 Complete motion 720 1.27 95.78 Complete motion 840 1.61 142.16 Complete motion 151 Table F. 10: Motion of the fully bleached kraft pulp suspension (C m = 2%) in the stirred tank at various impeller rotational speeds. R P M Torque, N.m Power, W Flow condition of the pulp suspension 200 0.028 0.594 No motion. 300 0.143 4.494 Slight motion around impeller. 400 0.309 12.953 Axial movement of pulp around the impeller. Top layers stagnant. 500 0.519 27.178 Good motion around the impeller. Top portion moved slowly. 600 0.806 50.626 Good motion around impeller. Top portion moved moderately as a big lump. 700 1.168 86.648 Good motion all over the suspension. Table F. 11: Motion of the fully bleached kraft pulp suspension (C m = 2.5%) in the stirred tank at various impeller rotational speeds. R P M Torque, N.m Power, W Flow condition of the pulp suspension 600 0.81 50.80 Incomplete motion 720 1.26 95.18 Incomplete motion 840 1.73 152.37 Complete motion 152 Table F . 12: Motion of the fully bleached kraft pulp suspension (C m = 3%) in a stirred tank at various impeller rotational speeds. R P M Torque, N.m Power, W Flow condition of the pulp suspension 200 0.138 2.92 No motion 400 0.260 10.90 No motion 600 0.742 46.70 Slight motion 800 1.511 125.88 Good motion at the bottom and no motion at the top. 153 Table F . 13: Corrected product distributions (Xs c o r r ) with varying mass concentrations ( C J of pulp suspension (FBK) at impeller rotational speeds of 10, 12 and 14 s 1. Temperature = 2 5 + O . r C . N A n / N n o =1.10 and V A / V B = 50. C A o = 0.52 mol/m 3, C I l n = 23.6 mol/m 3 Pulp C m , % X s c o r r at feedpoint A X s c o r r at feedpoint B 0.5 0.061 0.152 1.0 0.065 0.161 1.5 0.069 0.169 2.0 0.071 0.194 2.5 0.092 0.231 0.50 0.051 0.135 1.00 0.055 0.140 1.50 0.058 0.145 2.00 0.065 0.154 2.50 0.072 0.181 0.5 0.042 0.119 1.0 0.046, 0.049, 0.048 0.121, 0.124, 0.122 1.5 0.053 0.128 2.0 0.055 0.135 2.5 0.056 0.150 154 Table F . 14: Measured product distributions, concentrations of mono azo and bis azo dyes and mass balances at various mass concentrations of the pulp suspension (FBK) at impeller speed = 14 s"1 at temperature = 25+0.1°C. N A o / N B o = 1.10 and V A / V B = 50. C A o = 0.52 mol/m 3, C B o = 23.6 mol/m 3 cm) % Measured Xs [R], mol/m 3 [S], mol/m3 Mass balance, % Feed point A 0.5 0.035 0.406 0.0074 90.9 1.0 0.041 0.396 0.0086 89.3 1.5 0.047 0.391 0.0096 88.6 2.0 0.048 0.380 0.0095 86.3 2.5 0.047 0.369 0.0092 83.7 Feed point B 0.5 0.114 0.352 0.0231 86.1 1.0 0.112 0.350 0.0221 85.2 1.5 0.113 0.345 0.0219 83.9 2.0 0.110 0.342 0.0220 83.4 2.5 0.120 0.335 0.0229 82.2 155 Table F . 15: Measured product distributions, concentrations of mono azo and bis azo dyes and mass balances at various mass concentrations of the pulp suspension (FBK) at impeller speed = 12 s° at temperature = 25±0 .1°C. N A o / N B o =1.10 and V A / V B = 50. C A o = 0.52 mol/m 3 and C B n = 23.6 mol/m 3. Measured Xs [R], mol/m3 [S], mol/m 3 Mass balance, % Feed point A 0.5 0.048 0.401 0.0100 90.9 1.0 0.065 0.365 0.0125 84.3 1.5 0.049 0.389 0.0101 88.4 2.0 0.051 0.388 0.0104 86.5 2.5 0.057 0.375 0.0120 84.5 Feed point B 0.5 0.127 0.348 0.0252 86.1 1.0 0.125 0.350 0.0250 86.4 1.5 0.126 0.345 0.0248 85.2 2.0 0.136 0.340 0.0268 83.3 2.5 0.151 0.326 0.0290 82.9 156 Table F . 16: Measured product distributions, concentrations of mono azo and bis azo dyes and mass balances at various mass concentrations of the pulp suspension (FBK) at impeller speed = 10 s"1 at temperature = 25±0 .1°C. N A o / N B o =1.10 and V A / V B = 50. C A o = 0.52 mol/m 3 and C H o = 23.6 mol/m 3.. c m , % Measured Xs [R], mol/m3 [S], mol/m3 Mass balance, % Feed point A 0.5 0.056 0.385 0.0115 88.1 1.0 0.052 0.382 0.0105 87.0 1.5 0.063 0.360 0.0122 83.1 2.0 0.062 0.380 0.0126 87.6 2.5 0.080 0.346 0.0156 81.2 Feed point B 0.5 0.143 0.343 0.0285 86.4 1.0 0.145 0.342 0.0289 86.4 1.5 0.150 0.330 0.0291 83.8 2.0 0.172 0.306 0.0320 79.9 2.5 0.200 0.309 0.0386 83.5 157 Appendix G: Figure G . l : Measured product distributions (uncorrected) with varying mass concentrations of pulp suspension (FBK) at impeller rotational speed (N) = 10, 12 and 14 s"1. Temperature = 25±0 .1°C . V A / V B = 50 and N A o / N D o =1.10. C A o = 0.52 mol/m 3, C B o = 23.6 mol/m 3. Open symbols indicate tests where the suspension was not in complete motion. (0 X 0.25 0.20 - • 0.15 -0.10 0.05 0.00 0.0 0.5 1.0 1.5 2.0 2.5 Mass concentration, 158 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0058588/manifest

Comment

Related Items