A Chemical Engineering Approach to Cellulose Substitution Kinetics

Transcription

1 A Chemial Engineering Approah to Cellulose Substitution Kinetis Tapio Salmi 1*, Pia Damlin 1, Jyri-Pekka Mikkola 1,2, Matias Kangas 1 1 Åbo Akademi, Proess Chemistry Centre, Industrial Chemistry and Reation Engineering, FI-25 Åbo/Turku Finland, 2 Umeå University, Tehnial Chemistry, Chemial-Biologial Center, Department of Chemistry, SE-9187 Umeå Sweden A general kineti model for the substitution of ellulose was developed. The model assesses the differenes in the reativity of the ellulose hydroxyl groups and a derease in the reativity as the substitution proeeds. The model predits the detailed distribution of the different mono (2, 3, 6-, di (23, 26, 36 and trisubstituted (236 units as a funtion of the reation time. The lassial Spurlin distribution is obtained as a speial ase of the general model. Numerial strategies were developed for the solution and omputer simulation of the model. Simulation results were shown to be valid for various reativity ratios of the hydroxyl groups. The modelling onept was verified with experimental data obtained for arboxymethylation of ellulose in a slurry reator. The detailed data, whih revealed the non-uniform reativities of the different hydroxyl groups and the deline of the substitution rate with time, were suessfully desribed by the mathematial model. 1. Introdution The researh of hemial omponents originating from renewable soures is obtaining a growing interest throughout the world. A very important maromoleule originating from biomass is ellulose, whih is the main omponent in wood. Cellulose is a rih soure of hemials, partiularly ellulose derivatives are manufatured in large quantities, and all their potentials have not been disovered yet. A lot of researh on ellulose derivatives has been arried out, but mainly qualitative work by hemists. In this work, we present the hemial engineering approah to the preparation of ellulose derivatives: detailed kineti modelling based on the hemial analysis of the substituted hydroxyl groups in ellulose. The most ommon ellulose derivatives are ellulose esters and ellulose ethers. Cellulose ethers are prepared by letting alkyl halogenides or allyl halogenides or alkene oxides to reat with alkali ellulose. The ellulose fibres are pretreated by alkali to promote swelling of the fibres and ionization of the hydroxyl groups. With alkyl halogenides, the reation proeeds as follows,

2 M + Cell-O - + RX Cell-OR + M + X - where ell-o- represents the ionized ellulose and RX is the alkyl halogenide. M + denotes an alkali metal ion, typially Na +. Depending on the halogenide, various produts are obtained: with methyl (RX = CH 3 Cl and ethyl hloride (R = CH 3 CH 2 Cl, methyl and ethyl ellulose are obtained as main reation produts, whereas with - monohloroarboxyli aids, arboxyalkylelluloses are obtained, suh as arboxymethyl and arboxyethyl elluloses. The most important of them is arboxymethylellulose (CMC, whih is obtained by letting alkaliellulose to reat with monohloroaeti aid: M + Cell-O - + CH 2 ClCOOH Cell-OCH 2 COOH + M + Cl -. The sodium salt of CMC is widely used, for instane, in detergents, alimentary produts, textiles, glues as well as emulgators in various produts suh as dyes, erami produts and osmetis. Pioneering models for the substitution kinetis of ellulose have been proposed by Spurlin (1939, who derived the lassial distribution between mono-, di- and trisubstituted anhydrogluose units for the ase that all of the hydroxyl groups (OH-2, OH-3, OH-6 have equal reativities. Timell (1952 presented differential equations for the substitution reations for ases that first order kinetis is obeyed and Salmi et al. (1994 have developed kineti models for arboxyalkylation of ellulose with - hloroarboxyli aids. Diffusion phenomena inside the fibers was treated (Salmi et al. 1994, Salmi et al. 23 with an empirial model and with a reation-diffusion model. A detailed study, inluding a omparison of three analytial methods, on the arboxymethylation of starh was arried out by Lazik et al. (22, who onluded that the reativities of the different hydroxyl groups in starh are different, but their ratio is onstant; thus the lassial model of Spurlin (1939 was appliable. The aim of the present paper is to develop a generalized simulation model for the substitution of ellulose, inluding both the reativity differenes and onentrations of the individual hydroxyl groups as well as the predition of the produt distribution of the substituted anhydrogluose units. 2. Experimental Setion The ommerial ellulose from birh hardwood was obtained from Metsä-Serla hemials; the ellulose obtained was sreened to a size of.35 mm. In the heterogeneous arboxymethylation experiments were performed in a glass laboratorysale autolave (5 ml with fixed molar ratios of ellulose-to-monohloroaeti aidto-naoh 1:4:8. 2-propanol was used as the solvent and the reation temperatures during arboxymethylation experiments were 3, 4 and 6 o C. The reation time was 12 min and samples were withdrawn from the reation mixture at pre-defined intervals (1 min, 2min, 4 min et. High-pH anion-exhange hromatography with pulsed amperometri detetion (HPAEC-PAD was applied for the hemial analysis. Before the analysis, the CMC samples were subjeted to aid hydrolysis with trifluoroaeti aid (TFA. The hydrolysis produts were analyzed by HPAEC with PAD on a Hewlett Pakard 11 Series LC system with a Dionex ED5 eletrohemial detetor, a Dionex

3 CarboPa PA1 anion-exhange olumn (semi-preparative 9 25 mm and analytial 4 25 mm ombined with guard 4 5 mm, olumn temperature 3 o C. The detetion was performed by the PAD with a gold working eletrode and triple pulse amperometry. The details of the analysis method are reported by Damlin et al. ( Mathematial Modelling Priniples The anhydrogluose unit of ellulose has three funtional hydroxyl groups (OH adjaent to arbon atoms 2, 3 and 6 (OH-2, OH-3, OH-6. The reativity of the hydroxyl group varies, depending on the substitution reation and the onditions. The most aidi hydroxyl group is OH-2, whih implies that it is laimed to be the most reative one, for instane, in hydroxymethylation. On the other hand, as the size of the substituent inreases, steri fators beome important, thus preferring the reativity of OH-6. The overall reation sheme is displayed in Fig δ 2 k δ3 23 ( δ 2 +δ3 k δ3 k 6 6 δ6 k3 k δ6 6 ( δ +δ k δ ( δ 2 +δ 6 k3 26 Figure 1. Reation sheme for the formation of substituted anhydrogluose units. Notation:, unsubstituted gluose unit; 2, 3, 6, monosubstituted gluose units (substituted at hydroxyl groups 2, 3 and 6; 23, 36, 26, disubstituted gluose units (substituted at hydroxyl groups 2 and 3, 3 and 6 and 2 and 6; 236, trisubstituted gluose units. The detailed produt distribution in the time sale an be simulated by using the mass balane (1 and the reation sheme. In the atual ase, the balane equations of the omponents in a onstant-volume bath reator are written as d ' i / dt = r i (1 where i =, 2,3, 6, 23, 236, R. The generation rates of different anhydrogluose units are listed below: r ( k2 + k3 + k6 R = (2 2 ( R r = k ( k k (3 3 ( R r = k ( k k (4

4 6 ( R r = k ( k k (5 23 ( R r = k + k k (6 26 ( R r = k + k k (7 36 ( R r = k + k k (8 R r 236 = ( k k k6 23 (9 The onsumption rate of the substitution reagent is obtained from R = (( k2 + k3 + k6 + ( k3 + k6 2+ ( k2 + k6 3+ ( k2 + k3 6+ k2 36+ k3 26+ k6 23 R r (1 If the substitution reagent (R undergoes side reations, their effets are added to the expression of r R. Typial side reations are substitution of hydroxyl groups and elimination reations in an alkaline environment, as disussed by Salmi et al (23. Furthermore, the reativity of a hydroxyl group diminishes due to the substitution to another group in the same anhydrogluose unit, the effet an be taken into aount in the rate onstant. The following semi-empirial funtion was used: Ea, j /( Rθ ' δ / DS δ k Pk k j = k j, ref e e e whih predits the deline of the reativity as the substitution progresses. The onentrations of mono (I-, di (II- and trisubstituted (III units are defined as (Fig. 2: I = + + ( ' II = + (13 III = 236 (14 The degree of substitution is expressed as the weighted sum of the mono-, di- and trisubstituted units (n=i, II, III, DS = I / + 2II / + 3III / (15 The differenes between the observed anhydrogluose units (, 2, 236 and those predited by the model (eqs 1-11 were minimized by nonlinear regression, after whih the total onentrations and DS were simulated (eqs Modeling Results and Disussion The arboxymethylation kinetis of ellulose in 2-propanol was reorded at 3, 4 and 6 o C in the bath reator. The overall degree of substitution (DS and the substitution rate inreased as the temperature was augmented. The onentrations of native (unsubstituted, mono-substituted (I, di-substituted (II and tri-substituted (III ellulose units at 6 o C are displayed in Fig. 2. At 3 o C and 4 o C, the reation rate was rather slow, and the onentrations of all of the substituted units (DS still inreased. At 6 o C, however, the rates were learly higher, and a maximum in the onentration of the mono-substituted units beomes visible. The onentration of the tri-substituted (11

5 units was very low at 3 o C and 4 o C, while it beomes learly visible at 6 o C (Fig. 2, left. Hydroxyl group OH-6 is the most reative one: the mono-substituted unit ( 6 is formed more rapidly, but delining more slowly than the other mono-substituted units ( 2 and 3. A harateristi feature obtained throughout the experimental domain was that the reation rate stagnates and a omplete substitution seems to be in pratie impossible. Preliminary parameter estimation efforts revealed that the reativities of the hydroxyl groups OH-2 and OH-3 are rather similar and the retardation proess is ompliated, depending both on the overall DS and the speifi effet aording to eq (11 (see Fig. 2. It was in pratie impossible to distinguish between the relative reativities of the groups OH-2 and OH-3; therefore the pre-exponential fators and ativation energies for the reativities of these groups were set equal. The regression analysis revealed that the hydroxyl group OH-6 has the highest retardation effet: evidently the substituted group in position 6 an by rotation shield the substitution to the OH-groups 2 and 3. One the substituted OH-6 has been formed, it reats further rather slowly to the disubstituted units ( 26 and 36 ; the disubstituted unit ( 23 is formed most rapidly of the di-substituted units, beause the shielding effet of substituted OH-6 is not present. To desribe the overall retardation effet, the approah given by eq. (11 is neessary. A summary of the parameter estimation statistis is provided here. The rate onstants obtained the following values:. k 2,ref = k 3,ref =.24 L/(mol min, k 6,ref =.33 L/(mol min, δ 2 =2., δ 3 =4.4, δ 6 =6.5, δ =4.8; E a,2 = E a,3 =14 kj/mol, E a,6 =127 kj/mol. The referene temperature (T ref is the average temperature of the experiments (in K. The transformed temperature (θ is defined as 1/θ =1/T-1/ T ref, in K. The overall estimation statistis was fairly good, taking into onsideration the omplexity of the system. The estimated ativation energies of the kineti onstants are reasonable, around 13 kj/mol for OH-6 and around 14 kj/mol for OH-2 and OH-3. The parameters listed in Table 1 represent the minimum set of parameters, whih is able to desribe the behavior of the system. It an be onluded that the modeling approah was suessful, taking the omplexity of the system into aount. 5. Conlusions An analytial method was developed whih enables the detailed determination of the arboxymethylation kinetis of ellulose. A mathematial model was developed for the substitution of the anhydrogluose units of ellulose and the kineti parameters inorporated in the model inluding the shielding effet were estimated with regression analysis. The model gave a reasonably good desription of the formation of arboxymethylellulose and has thus a perspetive for proess sale-up purposes.

6 Mole fration.6.4 Mole fration Time / [min] Time / [min] Figure 2. Measured (symbols and modeled (lines mole frations (x in arboxymethylation of ellulose; left figure:, unsubstituted gluose units;, monosubstituted;, di-substituted;, tri-substituted; right figure:, x 2 ;, x 6 ;, x 3;. Temperature: 6 o C Aknowledgement. This work is a part of the ativities at the Åbo Akademi Proess Chemistry Centre (PCC within the Finnish Centre of Exellene Programmes (2-25 and by the Aademy of Finland. Finanial support from the Knut and Alie Wallenberg Foundation, Bio4Energy Programme and Kempe Foundation is gratefully aknowledged. Referenes Damlin, P., Mikkola, J.-P., Salmi, T., 21, Charaterization of hardwood-derived arboxymethylellulose by high-ph anion-exhange hromatography using pulsed amperometri detetion, Cellulose Chem. Tehnol. 44 (1-3, Lazik, W., Heinze, Th., Pfeiffer, K., Albreht, G., Mishnik, P., 22, Starh derivatives of a high degree of funtionalization. VI. Multistep arboxymethylation, J. Appl. Polymer Si. 86, Salmi, T., Valtakari, D., Paatero, E., Holmbom, B., Sjöholm, R., 1994, Kineti study of the arboxymethylation of ellulose, Ind. Eng. Chem. Res. 33, Salmi, T., Tirronen, E., Lehtonen, J., Paatero, E., Valtakari, D., 23, Modeling of omplex organi solid-liquid reation systems in stirred tanks, Ind. Eng. Chem. Res 42, Spurlin, H.M., Arrangement of substituents in ellulose derivates, 1939, J. Am. Chem. So. 61, Timell, T.E., Spurlin, H.M., 1952, Svensk Papperstidning 55, 7-78