Download Article (PDF)

http://www.e-polymers.org e-Polymers 2005, no. 041. ISSN 1618-7229

A facile method to prepare methylcellulose from annual plants and wood using iodomethane

Daiyong Ye *, Xavier Farriol *

Chemical Engineering Department, Rovira i Virgili University, Avinguda dels Països Catalans, 26, 43007 Tarragona, Spain; Fax 0034977558544; [email protected], [email protected]

(Received: January 17, 2005; published: June 14, 2005)

Abstract: A novel facile method was developed to prepare methylcellulose from miscanthus, cardoon and eucalyptus. Pulps were obtained by the impregnation rapid steam pulping process (IRSP). A total-chlorine-free (TCF) method was used to bleach the pulps with hydrogen peroxide. Bleached pulps were mercerized in 40% NaOH solution for one hour at ambient temperature. Mercerized cellulose reacted with iodomethane in 2-propanol slurry at 60°C for 22 h. Mercerization and methylation were repeated. Intrinsic viscosities of methylcellulose were measured in 4% NaOH solution. Super-molecular substitution patterns of methylcellulose were determined by 13C NMR. Molecular weights of methylcellulose were measured in dimethyl sulfoxide by gel permeation chromatography. Pulping severity is a key factor influencing the properties of the methylcellulose prepared.

Introduction Because of increasing deforestation and the increasing consumption of cellulose products, the cellulose industry is investigating such new resources as overproduced crops, agricultural waste, unconventional plants and common wild plants to investi- gate whether it is feasible to use them to produce paper, paperboard and cellulose derivatives like tailor-designed methylcellulose as an additive for cement, food and drugs. The properties of methylcellulose are influenced by its molecular structure: e.g., its degree of substitution (DS), molecular weight, molecular weight distribution, degree of polymerization (DP), and distribution of methoxyl groups along the glucose unit and the polymer chain. These properties strongly depend on the methylation conditions and properties of cellulose, which, in turn, depend on bleaching conditions, pulping parameters, plant species, the time that the plant was harvested and even parts of the plant. Therefore, a method is required that can evaluate these parameters when methylcellulose is prepared directly from wood and annual plants. Suida first synthesized methylcellulose with dimethyl sulfate in 1905 [1]. Several years later, patents in which chloromethane was used as a methylation agent were published independently [2-4]. Cellulose was homogenously methylated in triethyl- benzylammonium hydroxide [5], trimethylbenzylammonium hydroxide [6], LiCl/N,N- dimethylacetamide [7], SO2-diethylamine-methyl sulfoxide [8] and LiCl/dimethyl sulfoxide [9]. The activation of cellulose was the most important step in homoge- 1 neous methylation. Sodium hydroxide powder [8] and dimsyl sodium [9] were used as activation agents in homogeneous methylation. In heterogeneous methylation, caustic alkaline solutions were used to mercerize cellulose and form alkali cellulose; methylcellulose was synthesized in organic slurry with toluene [10-11] or isopropanol [12]. Timell and Purves found that the heterogeneous methylation was intermicellar with dimethyl sulfate while the uniform substituent distribution of commercial methylcellulose indicated that it was intramicellar with chloromethane [13]. Philipp et al. synthesized methylcellulose with chloromethane to study methylation on the laboratory scale [14]. Taipa et al. synthesized methylcellulose from commercial pulps with dimethyl sulfate and iodomethane in 2-propanol slurry at 30°C [12]. Methylcellulose is now produced by methylation of alkali cellulose with chloromethane under high pressure in industry. Traditionally methylcellulose has been synthesized from cotton or commercial wood dissolving pulps. In the present investigation, it was synthesized from bleached pulps, which were prepared from plant stalks in the laboratory. The impregnation rapid steam pulping process (IRSP) took place in our laboratory [15]. The stalks were impregnated in concentrated sodium hydroxide solutions under high pressure. This chemical impregnation improved the swelling, penetration and diffusion of the pulping chemicals and facilitated the subsequent cooking. Pulps were obtained from rapid steam explosion pulping. Short steam cooking time and steam explosion facilitated the delignification and defibration. The main characteristics of this steam pulping were high screened yield, low lignin content or kappa number a, high hemicellulose content, high intrinsic viscosity and degree of polymerization. The accessibility and reactivity of cellulose were improved by steam explosion pulping [16]. A TCF bleaching method was used to obtain pulps with low lignin contents. It was easier to compare and analyze the experimental results of methylcellulose synthesized from iodomethane, which had amplified different properties. The main properties of methylcellulose (e.g., viscosity, intrinsic viscosity, rheological data, degree of substitution and molecular weight) can be measured in the laboratory. This method, therefore, was appropriate for evaluating the methylcellulose production of plants.

Experimental part

Materials and chemical reagents The experiments used homogeneous batches of miscanthus, cardoon and eucalyptus stalks harvested in Spain. All chemicals were bought from Sigma-Aldrich, and were used without any further purification.

Equipment The stalks were chipped and stored in a refrigerator below 0°C. The impregnation was carried out in a 2-l batch reactor made of ANSI 304-L and 316-stainless steel. Steam explosion pulping was carried out in a steam explosion-pulping reactor, which had two vessels. One vessel was used for direct cooking with saturated steam at high temperature and pressure. Another vessel was used for receiving cooked

a The kappa number is a measure of lignin content in pulp. Higher kappa numbers indicate higher lignin content. 2 materials after sudden decompression. Methylation was carried out in a reaction flask over a heat plate connected to a coiled reflux condenser. Tap water was used as coolant. Intrinsic viscosities were measured in a capillary Ubbelohde viscometer and calculated by plotting a series of reduced viscosities vs. corresponding concentrations and extrapolating to zero concentration. 13C NMR spectra were measured in a spectro- meter operating at 300 MHz.

Experimental process The chipped stalks were impregnated in 20% or 30% sodium hydroxide solutions under 15 bar nitrogen pressure for different times. After the impregnation, the stalks and liquors were collected and weighed. The stalks were stored in a refrigerator below 0°C. A sample of liquor was collected so that neutralization titration could analyze the weight of residual NaOH in the soaked stalks. The impregnated stalks were directly heated by saturated steam in the steam explosion reactor at various retention times and temperatures. These cooking stalks were suddenly decompressed into the receiving vessel. Suspected solid was collected by filtration and washed several times by distilled water until the pH was close to 7. Unbleached pulps were dried in an oven at 60°C to constant weights. The bleaching methods were PP, EPP and PEP (E stands for alkaline extraction and P stands for hydrogen peroxide bleaching). During alkaline extraction, pulps were extracted by 10% NaOH solution at ambient temperature for 1 h with 3 - 4% consistency. After this extraction, pulps were collected by vacuum filtration and washed with distilled water. The hydrogen peroxide bleaching was performed with 3 - 4% consistency in 0.2 M NaOH and 0.15 M H2O2 solution for 1 h at 60°C. At the end of bleaching, pulps were washed by distilled water and collected by filtration. Bleached pulps were dried in an oven at 60°C to constant weights. A total of 5 g of bleached pulp (oven-dried weight) was mercerized in 200 g 40% NaOH solution at ambient temperature. After 1 h, the mercerized cellulose was filtered through a vacuum filter and pressed until the weight ratio of pulp and NaOH solution was 1/5. After filtration, 150 ml 2-propanol and various volumes of iodomethane were placed in a flask. The methylation lasted for 22 h. Since a single methylation could not provide a sufficient degree of substitution, mercerization and methylation were repeated. After methylation, the methylcellulose was cooled, neutralized with acetic acid, collected by filtration and washed three times by acetone and ethanol, respectively. The methylcellulose was dried in an oven at 60°C over night and stored in a refrigerator at 4°C for further analysis.

Characterization The kappa numbers of pulps were determined by TAPPI T 236 om-99. Pulp viscosities were determined by TAPPI T 230 om-99 (capillary viscometer method). The intrinsic viscosities of pulps were estimated by the Schula-Blaschke equation. The degree of substitution (DS) of methylcellulose was determined by 13C NMR in dimethyl sulfoxide at 80°C. The spectra were measured under conditions suggested by the research of Takahashi et al. [17]. DS values were calculated from integrated NMR signals. A 13C NMR spectrum of eucalyptus methylcellulose is shown in Fig. 1.

3 The water-soluble content was measured by distilled water extraction. Intrinsic viscosities were measured in 4% NaOH solution. The molecular weights of methylcellulose were measured in dimethyl sulfoxide (DMSO) by GPC. The eluent was DMSO of HPLC grade. The column was HP PL-GPC purchased from HP. The flow rate of the eluent was 0.7 ml/min. Dextrans of different molecular weights were used as the standard GPC materials.

3 9

7 6

105 100 95 90 85 80 75 70 65 60 ppm 13.66 10.98 24.60 8.47 42.29

200 180 160 140 120 100 80 60 40 20 ppm Fig. 1. 13C NMR spectrum of methylcellulose prepared from eucalyptus

Results and discussion

Impregnation The aim of impregnation was to obtain a uniform distribution of pulping chemicals in stalks. The uniform distribution led to more uniform pulp, better quality, fewer rejects and shorter cooking time [18-19]. The reactive pulping chemicals were mass transferred into the stalk voids by penetration (which was governed by the pressure gradient) and by diffusion (which was controlled by the concentration gradient of the penetrating chemicals) [20]. NaOH and anthraquinone (AQ) were used as pulping chemicals under 15 bar pressure. Chemicals penetrated and diffused into the capillaries and the stalk voids. The stalk fibres swelled until maximum absorption. Water, NaOH, AQ and alkaline soluble chemicals transferred between the fibre and the bulk solution until an equilibrium stage was reached. The delignification, some softening of fibres and defibration occurred during swelling and penetration. Some lignin that reacted with NaOH degraded and dissolved in the alkaline solution. The initial white colour of the alkaline solution became darker and blacker. Cardoon stalks absorbed 47.6% alkali, miscanthus stalks absorbed 44.6%, eucalyptus stalks absorbed 43.7% [21]. As time and initial NaOH concentration increased, the weight of the chemicals absorbed by cardoon fibres also increased [21].

4 Steam explosion pulping Explosion pulping was invented by Mason [22]. Vit and Kokta modified, improved and developed the process to produce pulps suitable for papermaking. They used such techniques as chemical impregnation of chips, short-duration saturated steam cooking and sudden pressure release [23-25]. Steam explosion pulping can be divided into two stages: rapid steam cooking and steam explosion. In the rapid steam cooking stage, the typical cooking time is several minutes and the typical cooking temperature is over 180°C. The short cooking time prevents side reactions and improves the selectivity and yield of pulps. Water has a plasticizing effect on lignin and hemicellulose, and their softening temperature is reduced to about 100°C. Steam cooking at temperatures above their glass transition temperature leads to additional permanent fibre softening because of internal structural changes [25]. The structure softening results in defibration. Lignin reacts with residual NaOH absorbed in the fibrils and degrades. AQ protects cellulose and hemicellulose and increases the yield of pulping during steam cooking at high temperature [26-27]. The more uniformly distributed are NaOH and AQ in the fibrils, the better is the quality of the pulps. Cellulose and hemicellulose are degraded and some of it is converted to polysaccharides. Degraded lignin and low-molecular-weight polysaccharides dissolve in alkaline aqueous solutions. These degraded substances diffuse into the bulk cooking liquid. Therefore, the places they occupied are vacant, and gradually develop into capillaries and voids. During the cooking process, more and more voids appear. The increasing number of voids facilitates and improves the effect of the subsequent steam explosion pulping.

Fig. 2. Steam explosion effect of a fibril

During steam cooking, interior capillaries and fibril voids are gradually filled with high- pressure liquid. When the cooking pressure is suddenly released, the high-pressure liquid evaporates, which subjects the fibres to high impact forces. The fibres are lacerated, as shown in Fig. 2. The mechanical explosion tears and breaks the fibres from interior capillaries and voids, and produces smaller fibres, fibrils and micro- fibrils. The surface areas of pulps increase significantly. This is the defibration, defibrillation and laceration of the steam explosion process. The chemical pretreatment during impregnation swells and softens the fibres, and also probably dissolves some lignin or has some effects of disintegration and defibration. The physical pretreatment of the chipped stalks (i.e., compression and decompression during impregnation under 15 bar pressure and further compression before steam cooking over 180°C) leads to considerable fibre deformation and partial fibre separation [23-25]. The steam cooking leads to softening, probably as a result of defibration and defibrillation of the interior fibre structure [24]. The explosion promotes defibration, and probably some internal fibrillation. As a result, the steam explosion pulping produces pulps

5 with higher yields, lower lignin contents, higher porosities, higher specific surface areas, and higher hydrophilicity than traditional pulping [25]. These pulps, therefore, have higher accessibility and reactivity, which facilitates pulp dissolution and cellulose derivatization. The pulping conditions of miscanthus, cardoon and eucalyptus stalks are listed in Tab. 1. The cooking temperature and time were combined into a single parameter for pulping severity, the P factor, which is calculated by the following equation [28]: t T −100 P = log R = log( exp( )⋅ dt) (1) o ∫0 14.75 where Ro is the severity of steam explosion, T the reaction temperature in °C, and t the retention time in min.

Tab. 1. Pulping conditions of cardoon, eucalyptus and miscanthus Plant Cooking time in min Cooking temperature in °C P factor cardoon 4 170 2.66 cardoon 4 180 2.96 eucalyptus 8 180 3.26 eucalyptus 16 180 3.56 eucalyptus 24 180 3.74 eucalyptus 16 190 3.85 eucalyptus 24 190 4.03 miscanthus 4 180 2.96 miscanthus 8 180 3.26 miscanthus 16 180 3.56 miscanthus 24 180 3.74

Fig. 3. Intrinsic viscosities of unbleached pulps

6 When the P factors increased, intrinsic viscosities and kappa numbers decreased for the cardoon, miscanthus and eucalyptus pulps. Cardoon fibres had lower P factors, so cardoon pulps had higher intrinsic viscosities and kappa numbers than miscanthus and eucalyptus pulps. On the other hand, eucalyptus pulps exhibited lower values because they were subjected to higher pulping severities. These data are shown in Figs. 4 and 5 below [21]. The stalks of cardoon and miscanthus had looser fibre structures than those of eucalyptus. Hence the pulping severities of stalks of cardoon and miscanthus were lower than of stalks of eucalyptus. High quality pulps of cardoon and miscanthus could be prepared. The stalks of eucalyptus could easily be pulped to lower kappa numbers. Cardoon and miscanthus are potential sources of cellulose like traditional eucalyptus.

Fig. 4. Kappa numbers of unbleached pulps

TCF bleaching TCF bleaching does not produce any organochlorines, which are hazardous substances such as dioxin, an endocrine disrupter and human carcinogen. In the present investigation, the TCF method only used cool alkaline extraction and hydrogen peroxide oxidation. Some of the residual lignin that reacted with NaOH was degraded and dissolved in the alkaline solution. Hydrogen peroxide breaks down conjugated side chains in lignin and other chromophores to yield colourless aliphatic substances by the nucleophilic attack of the perhydroxyl anion in alkaline solutions. The resulting chemicals become more hydrophilic and are at least partly removed by washing. Bleached pulps with low lignin contents were obtained by this TCF bleaching method. As the P factors increased, intrinsic viscosities and kappa numbers of the bleached pulps decreased for cardoon, miscanthus and eucalyptus pulps. Though cardoon pulps had lower P factors, they could be bleached to pulps of highest intrinsic viscosities and lowest kappa numbers among pulps of the three plants. This might be because cardoon fibres have more voids in their stalks and pulps. Eucalyptus pulps had lower intrinsic viscosities and kappa numbers than miscanthus pulps because they underwent higher pulping severities. These data are shown in Figs. 6 and 7 [21]. 7 Comparison of the results of pulping and bleaching shows that excellent bleached cardoon pulps can be produced at low temperatures and short cooking times, which saves a lot of energy. Miscanthus can also be used to prepare very good bleached pulps. Miscanthus and cardoon could substitute wood and become new sources of industrial cellulose.

Fig. 5. Intrinsic viscosities of bleached pulps. Pulps of miscanthus were bleached by an EPP sequence (E stands for alkaline extraction and P stands for hydrogen peroxide bleaching). The bleaching sequences for the cardoon pulps were PP or PEP. The bleaching sequence for the eucalyptus pulps was PP

Fig. 6. Kappa numbers of bleached pulps

Mercerization of bleached pulps Mercer found that a strong solution of sodium hydroxide led cellulose to shrink in width and length and become denser [29]. The NaOH solution penetrates and diffuses into the voids of bleached pulps. This process is shown in Fig. 7. After 8 absorbing some of the NaOH solution, the original fibre I develops into the larger fibre II. The diameter of fibre III is largest because the capacity of absorption is maximum. When the diffusion and mass transfer reach an equilibrium state, most fibres adjust to become fibre IV.

Fig. 7. Swelling and diffusion during mercerization

Bleached pulp is impure cellulose. The bleached pulp usually contains more than 90% cellulose, which is a complex composite with both crystalline and amorphous structures [30-31]. During mercerization, the crystalline area is attacked by the strong NaOH solution [32], which leads to the cellulose chain gradually being debonded from the crystal [32]. Thus, amorphous regions are formed and developed by alkali attacking the outer surface of the crystal [32]. The crystalline cellulose gradually develops into amorphous cellulose. During mercerization, most of the cellulose gradually transfers its crystal lattice from cellulose I to Na-cellulose II; some of the cellulose changes its crystal lattice from cellulose I to Na-cellulose I and the amorphous cellulose II develops to Na-cellulose II [32]. Excess alkali was pressed out at a press factor of 5. This press factor was higher than the traditional press factor, 3, which means that the steam-exploded pulps have a greater capacity of absorption and that the pulps have more surfaces and voids. These steam-exploded pulps, then, were more accessible and reactive, which favoured the following methylation.

Fig. 8. Water-soluble contents of methylcellulose

Fig. 9. Intrinsic viscosities of methylcellulose

Fig. 10. Molecular weights of methylcellulose

Methylation of bleached pulps The methylation of alkali cellulose with a methyl halide is a nucleophilic substitution, which yields a rate-controlled distribution of methoxyl on the methylcellulose mole- cules. In a heterogeneous organic reaction medium, the methylation is a rate-deter- mining transport process with chloromethane as the methylation agent under high pressure [13]. Under ambient pressure and with dimethyl sulfate and iodomethane as methylation agents, however, the methylation can be diffusion-controlled [12]. From the data [21] shown in Figs. 8, 9, 10 and 11, the P factor is the key factor that influences the synthesis of methylcellulose. As the P factor increases, the water- soluble content seems to increase for all samples of cardoon and eucalyptus.

Fig. 11. Degrees of substitution of methylcellulose

Tab. 2. Distribution of methoxyl groups along the glucose unit

Plant P factor DS at C2* DS at C3** DS at C6*** cardoon 2.66 0.41 0.25 0.34 cardoon 2.96 0.47 0.44 0.42 eucalyptus 3.26 0.38 0.22 0.16 eucalyptus 3.56 0.41 0.20 0.20 eucalyptus 3.74 0.44 0.21 0.21 eucalyptus 3.85 0.40 0.19 0.23 eucalyptus 4.03 0.53 0.28 0.26 miscanthus 2.96 0.20 0.09 0.17 miscanthus 3.26 0.33 0.16 0.21 miscanthus 3.56 0.40 0.34 0.32 miscanthus 3.74 0.67 0.38 0.47 *: Second hydroxyl group. **: Third hydroxyl group. ***: Sixth hydroxyl group.

Intrinsic viscosities and molecular weights decreased while P factors increased. The degrees of substitution increased perfectly with increasing P factors. Of the three plants, cardoon was the one that produced methylcellulose of the highest molecular weight. Pulping severities were higher for eucalyptus stalks, so the methylcellulose they produced had lower intrinsic viscosities. Eucalyptus was suitable for synthesizing methylcellulose of high water-soluble content. Miscanthus methylcellulose had medium intrinsic viscosities and water-soluble contents [21]. Miscanthus was suitable for producing methylcellulose with high DS. From Tab. 2, substitution values at the second OH group were bigger than at the third and sixth OH groups. Substitutions at the third and the sixth OH groups were more difficult. With increasing P factor, not only the total DS values but also the DS values at the second, third and sixth OH groups increased. 11 Conclusions Pulping severity was the factor that most influenced the properties of synthesized methylcellulose. Cardoon easily provided high quality pulps and was suitable for synthesizing high-molecular-weight methylcellulose. Miscanthus was suitable for synthesizing methylcellulose with a high degree of substitution. Eucalyptus was suitable for synthesizing methylcellulose with high purity and degree of substitution. Rapid steam explosion pulping led to pulps with low lignin contents, high intrinsic viscosities, high accessibility and reactivity. Methylcellulose, which was synthesized in an ordinary apparatus, had sufficient and different properties to compare plant species, and process parameters of pulping and bleaching. Therefore, this method was appropriate for evaluating whether certain plants could be used to produce methylcellulose.

[1] Suida, W.; Monatsh. 1905, 26, 413. [2] Lilienfeld, L.; British Patent 1912, 12854. [3] Leuchs, O.; German Patent 1912, 322586. [4] Dreyfus, H.; French Patent 1912, 462274. [5] Bock, L. H.; Ind. Eng. Chem. 1937, 29, 985. [6] Stuchlík, J.; Tichý, M.; Procházk, V.; Chem. Listy. 1956, 50, 662. [7] McCormick, C. L.; Lichatowich, D. K.; J. Polym. Sci., Polym. Lett. 1979, 17, 479. [8] Isogai, A.; Ishizu, A.; Nakano, J.; Eda, S.; Kato, K.; Carbohydr. Res. 1985, 138, 99. [9] Petrus, L.; Gray, D. G.; BeMiller, J. N.; Carbohydr. Res. 1995, 268, 319. [10] Denhan, W.; Woodhouse, H.; J. Chem. Soc. 1913, 103, 1755. [11] Steele, R.; Pacsu, E.; Textile Res. J. 1949, 19, 771. [12] Tapia, C.; Hagar, J. S.; Siches, P.; Valenzuela, F.; Basulato; Latin American Applied Research 1996, 26, 221. [13] Timell, T.; Purves, C. B.; Svensk Papperstidning 1951, 54, 303. [14] Philipp, B.; Bischoff, K. H.; Loth, F.; Cellul. Chem. Technol. 1979, 13, 23. [15] Montané, D.; Jollez, P.; Salvadó, J.; Farriol, X.; Chornet, E.; TAPPI 1996, 79, 253. [16] Cuculo, J. A.; Aminuddin, N.; Hudson, S. M.; Wilson, A. F.; ‘Cellulose (Direct Dissolution)’, Polymeric Materials Encyclopedia, CRC Press, Boca Raton 1996. [17] Takahashi, S.; Fujimoto, T.; Miyamoto, T.; Inagaki, H.; J. Polym. Sci., Part A: Polym. Chem. 1987, 25, 987. [18] Gustafsson, R. R.; Jimenez, G.; McKean, W.; Chian, D. S.; Proceedings of the 1988 Pulping Conference, Atlanta 1988, 685. [19] Gullichsen, J.; Sundqvist, H.; Proceedings of the 1995 Pulping Conference, Chicago 1995, 227. [20] Stone, J. E.; Förderreuther, C.; Tappi 1956, 39, 679. [21] Ye, D.; Farriol, X.; Abstract in ‘The First Workshop on Regenerated Cellulose and Cellulose Derivatives’, Karlstad 2003.

12 [22] Mason, W. H.; US Patent 1928, 1655618. [23] Vit, R.; Kokta, B. V.; Canadian Patent 1986, 1.212.505. [24] Kokta, B. V.; Vit, R.; Proc. CPPA Annual Meeting, Technical Section, Montreal 1987. [25] Kokta, B. V.; Ahmed, A.; in “Environmentally friendly technologies for the pulp and paper industry”, Young, R. A.; Akhtar, M.; editors; Wiley Inc., 1998, 191. [26] Abott, J.; Bolker, J. I.; Tappi 1982, 65, 127. [27] Blain, T. J.; Tappi 1983, 76, 137. [28] Chornet, E.; Overend, R. P.; in “Steam explosion techniques – fundamentals and industrial applications, Focher, B.; Marzeti, A.; Crescenzi, V.; editors; Gordon and Breach Science Publishers, 1988, 21. [29] Mercer, J.; British Patent 1850, 13296. [30] Krässig, H. A.; “Cellulose-structure, accessibility and reactivity”, Gordon & Breach, Amsterdam 1993, 1. [31] Lai, Y. Z.; in “Chemical modification of lignocellulosic materials”, Hon, D. N.-S.; editor; Marcel Dekker, Inc., New York 1996, 35. [32] Takai, M.; Colvin, J. R.; J. Polym. Sci., Polym. Chem. 1977, 16, 1335.