Biomacromolecules 2005, 6, 3467-3473
3467
Structural Differences between the Lignin-Carbohydrate Complexes Present in Wood and in Chemical Pulps Martin Lawoko, Gunnar Henriksson, and Go¨ran Gellerstedt* Department of Fibre and Polymer Technology, Royal Institute of Technology, KTH, SE-100 44 Stockholm, Sweden Received August 17, 2005
Lignin-carbohydrate complexes (LCCs) were prepared in quantitative yield from spruce wood and from the corresponding kraft and oxygen-delignified pulps and were separated into different fractions on the basis of their carbohydrate composition. To obtain an understanding of the differences in lignin structure and reactivity within the various LCC fractions, thioacidolysis in combination with gas chromatography was used to quantify the content of β-O-4 structures in the lignin. Periodate oxidation followed by determination of methanol was used to quantify the phenolic hydroxyl groups. Furthermore, size exclusion chromatography (SEC) of the thioacidolysis fractions was used to monitor any differences between the original molecular size distribution and that after the delignification processes. Characteristic differences between the various LCC fractions were observed, clearly indicating that two different forms of lignin are present in the wood fiber wall. These forms are linked to glucomannan and xylan, respectively. On pulping, the different LCCs have different reactivities. The xylan-linked lignin is to a large extent degraded, whereas the glucomannan-linked lignin undergoes a partial condensation to form more high molecular mass material. The latter seems to be rather unchanged during a subsequent oxygen-delignification stage. On the basis of these findings, a modified arrangement of the fiber wall polymers is suggested. Introduction Covalent linkages between lignin and carbohydrates (LCbonds) have been proposed to exist in wood1 and in chemical pulp2-5 although ambiguity in the types, frequencies, and quantity exist. Three main types of native LC bonds have been suggested in the literature viz. benzyl esters, benzyl ethers and phenyl glycosides.6 The former are labile under alkaline pulping conditions, whereas the latter two should be stable under these conditions. Indeed, the prevalence of alkaline-stable LC bonds in pulps has been shown to contribute in part to the slow delignification in the final phase.7 However, the possibility that LC bonds may be formed during alkaline pulping has also been investigated using model compounds.8,9 It has been suggested that lignin condensation reactions of different types may be the cause of the low reactivity of residual lignin during the final delignification phase.10,11 Recently, we have developed methods for the quantitative isolation and characterization of LCC from spruce wood12 and chemical pulps.7 It was proposed that no pure lignin fraction was present in wood since all of the lignin was found to exist chemically linked to polysaccharides. From kraft pulps with different lignin contents, it was shown that the lignin was degraded and/or dissolved at a rate dependent on the polysaccharide type to which it was bound.13 Furthermore, only about 10% of the residual lignin in these pulps was “free” and not linked to carbohydrates. In the present work, we have further investigated the reasons for the differences in delignification rate of the various LCCs by * To whom correspondence should be addressed.
studying the chemical structure of the lignin present in the different LCC types both in wood and in the corresponding kraft pulps. The results obtained have been related to the present knowledge about the ultrastructure of wood and to known features of the kraft pulping and subsequent oxygendelignification stages for making chemical pulp. Materials and Methods Isolation of LCCs. Acetone-extracted spruce wood as well as the corresponding unbleached and oxygen-delignified kraft pulps were treated with an endoglucanase (Novozyme 476, Novozyme, Denmark) and, after swelling in urea, further fractionated in aqueous alkali to give lignin carbohydrate complexes (LCCs) in a quantitative yield based on the starting material. The fractionation and workup procedure were carried out according to previously described methods.7,12 In the case of the wood, a pre-milling in a laboratory mill (Retsch mixer mill, Type MM2) for 3 h at 100 oscillations per second was necessary in order to make the material accessible to enzymatic attack. The complete analytical protocol is shown in Figure 1. Lignin and Carbohydrate Analyses. The Klason lignin was determined according to TAPPI test Method T222 om83 with a slight modification in that, instead of boiling to complete hydrolysis of polysaccharides, autoclaving at an elevated temperature (125 °C) and pressure (1.4 bar) was adopted. Carbohydrate analysis was performed as described previously.14 The GC-FID analysis was performed using a Hewlett-Packard 6890 instrument equipped with a BPX 70 column (12 m, 0.32 mm, 0.25 µm film thickness). Split injection was used. The injector temperature was set at 230
10.1021/bm058014q CCC: $30.25 © 2005 American Chemical Society Published on Web 10/21/2005
3468
Biomacromolecules, Vol. 6, No. 6, 2005
Lawoko et al.
Figure 1. General scheme for the isolation of LCCs from wood and pulps. Fractions P1-P4, S4, S7 from wood; P3-P, P4-P, S7-P from unbleached spruce kraft pulp; P4-PO from oxygen-delignified kraft pulp.
Figure 2. Lignin model compounds of the β-O-4 (guaiacylglycerolβ-guaiacyl ether) and β-β (pinoresinol) types used in this work.
°C, the detector at 250 °C, and the oven at 215 °C, with helium as carrier gas at a flow rate of 0.9 mL/min. Phenolic Hydroxyl Groups. The content of phenolic hydroxyl groups in the lignin was determined by periodate oxidation and determination of the liberated methanol.15 The methanol determination was performed by GC FID using a Hewlett-Packard 5790 gas chromatograph. The injector and detector temperatures were kept at 150 and 250 °C, respectively. The column was an Agilent technologies packed column (Porapak S 80/100 mesh, 1.2 m length, 0.32 cm outer diameter) with helium as carrier gas at a flow rate of 60 mL/min. Thioacidolysis. Thioacidolysis16 was performed on two lignin model compounds, viz. guaiacylglycerol-β-guaiacyl ether17 and pinoresinol18 (Figure 2), on wood, on LCCs from wood, and on the LCCs isolated from the kraft and oxygendelignified pulps. Portions of the respective product mixtures (except from model compounds) were silylated and then analyzed by gas chromatography to quantify the uncondensed β-O-4 structures present in the lignin portion. The analysis was performed with a Hewlett-Packard 6890 instrument. A DB 5MS column was used (30 m, 0.32 µm i.d., 0.25 µm film
thickness). The detector and injector temperatures were maintained at 250 °C. A temperature program was set for the oven as follows; from 120 to 200 °C at a rate of 15 °C per min, from 200 to 260 °C at a rate of 5 °C per min, at 260 °C for 15 min, from 260 to 300 °C at a rate of 20 °C per min, and at 300 °C for 5 min. Split injection was used. Quantitative calculations were done as described before.16,19 On further portions of the thioacidolysis products, a second reaction, involving a reduction with Raney nickel, was performed in order to study the distribution of dimeric lignin products. The conditions and analytical data were the same as those previously used.19,20 Acetylation. Acetylation of thioacidolysis products was performed as described earlier for lignin samples, in pyridine: acetic anhydride (1:1, v/v) and allowing the reaction to proceed overnight.21 The excess acetic anhydride was eliminated by adding methanol and cooling the mixture in an ice bath. The pyridine was removed by addition of toluene followed by rotary evaporation. Products were dissolved in tetrahydrofuran (THF) prior to SEC analysis. Size Exclusion Chromatography. SEC analysis was performed on a system of three Ultra styragel columns (Waters, Milford, MA, 100, 500, and 1000 Å, respectively) connected in series, with a Waters 2487 UV-light detector set at 280 nm. Acetylated thioacidolysis products were dissolved in tetrahydrofuran (THF) of analytical grade and analyzed at a flow rate of 0.8 mL/min. Results and Discussion LCCs in Wood and Kraft Pulps. All of the lignin present in the wood could be solubilized after a short period of ball
Biomacromolecules, Vol. 6, No. 6, 2005 3469
Differences in Lignin-Carbohydrate Complexes
Table 1. Sugar and Lignin Analyses of Spruce Ball Milled Wood (BMW), of Enzyme Solution, of Hydrolysate after Enzyme Treatment and of the Various LCC Fractions Isolated from Wood and Pulps fractiona BMW enzyme hydrolysate fractiond P1/5.0 P3/15 P2/9.3 P4/18 S4/6.0 S7/22 Kraft pulp P3-P/38f P4-P/5.8 S7-P/5.3 O-delign. pulp P4-PO/18
arabinan (g)b
xylan (g)b
mannan (g)b
galactan (g)b
glucan (g)b
klason lignin (g)b
mass balance (g)b
lignin yield (%)
1.3 0 0
5.9 0 0
12.2 0.2c 13.7
1.8 0.2c 7.7
46.7 0 78.6
26.7
94.6
100
3.2 1.0 1.2 0.7 3.1 4.1 0.7 0.5 1.2 4.4 0.5 0.9
2.4 0.9 3.1 3.2 18.7 34.9 7.3 3.5 11.2 56.8 6.7 7.6
33.5 2.5 19.2 33.6 4.3 14.2 5.7 4.0 20.2 6.3 5.0 41.6
5.2 85.2 9.3 12.5 3.6 1.7 75 88 7.4 7.7 80 13.3
39 7 56 41 65e 29 4.5 2.4 35 23 0.9 19.6
95.9 96.8 92.8 92.8 97.0 88.1 93.2 98.9 76.8 98 93 86
8 4 20 28 15 25 100 12 45 27 100 80
12.6 0.2 4.0 1.8 2.3 4.2 0.4 0 1.7 0.4 0.3 2.8
a Abbreviations according to Figure 1. b Amount in 100 g of sample. c Amount found in the volume of enzyme solution used for 100 g of sample. Designation/amount of fraction (g) obtained from 100 g of wood meal or pulp. e May include a small amount of enzyme. f Contains cellulose in addition to the glucane-lignin complex.
d
milling followed by endoglucanase treatment and urea swelling. Employing the dissolution-precipitation scheme shown in Figure 1, a total of six LCC fractions were obtained. These were classified into four main types based on their carbohydrate compositions; a galactoglucomannan-ligninpectin complex (GalGlcMan-L-P) containing 8% of the lignin in wood (P1), a glucan-lignin complex (Glc-L) with 4% lignin (P3); and two network structures. One of these, a glucomannan-lignin-xylan complex (GlcMan-L-Xyl), with a predominance of glucomannan over xylan was found in two fractions, viz P2 and P4, which together contained 48% of the lignin. Finally, the S4 and S7 fractions both contained a xylan-rich xylan-lignin-glucomannan (XylL-GlcMan) complex and these accounted for about 40% of the lignin. The detailed composition of all of the fractions obtained from the wood is further shown in Table 1. For reference purposes, a complete sugar and Klason lignin analysis was performed on the starting material, the ball milled wood, whereas the enzyme itself as well as the hydrolysate after enzyme treatment of the wood was analyzed for the presence of sugars (Table 1). On enzyme treatment, a portion of the galacto-glucomannan hemicellulose goes into solution together with the glucan, which can be due either to some enzymatic degradation of the hemicellulose or to simple dissolution in the aqueous solution. For the two pulp samples, unbleached kraft pulp and oxygen delignified kraft pulp, the protocol for isolation of the LCCs was similar to that for the wood but with the following differences. (1) After acetone extraction, the pulp was first swollen in water and then directly treated with enzyme without prior milling. (2) The urea treatment was done for 12 h (48 h in the case of wood), and after the alkaline borate treatment an undissolved residue was obtained. This was washed and again treated with enzyme as previously described.7 (3) The products dissolved during the enzymatic hydrolysis and during the urea treatment were not studied in detail,
since, in previous work, no lignin was detected in the hydrolysate and only an estimated 10% of free lignin had been found in the urea solution.7 From the unbleached pulp sample, three major LCC fractions were obtained, viz. a glucan-lignin (Glc-L) fraction (P3-P), a glucomannan-lignin-xylan (GlcManL-Xyl) fraction (P4-P), and a xylan-lignin-glucomannan (Xyl-L-GlcMan) fraction (S7-P) (Figure 1). Again, the latter two network fractions differ in their relative amounts of glucomannan and xylan as shown in Table 1. After a subsequent (harsh) oxygen delignification with a decrease in kappa number from 35 to 10, the procedure shown in Figure 1 afforded one predominant LCC fraction, a GlcManL-Xyl complex (P4-PO) containing 80% of the total pulp lignin. Lignin Structure in Wood and Pulp LCCs. Most of the different LCCs obtained from wood and pulps were subjected to thioacidolysis in order to determine the content of uncondensed β-O-4 structures present in the lignin. The method provides a selective and quantitative cleavage of all β-O-4 structures present in a lignin-containing sample, and from lignin units without any further lignin linkage in the aromatic C5 or C6 position, simple phenylpropane structures having ethylthio groups on the side-chain carbons are formed.16 These can be quantified thus providing an indirect measure of the extent of lignin reactions, e.g., during a kraft cook.22 Any cleavage of a β-O-4 structure in a lignin structure will result in the formation of a new phenolic hydroxyl group. Such groups were quantified using the periodate oxidation method.15 The analytical data based on the thioacidolysis and periodate oxidations are collected in Table 2. The results from thioacidolysis gave lower values for the LCC samples obtained from wood than for the reference wood. At the same time, the total content of phenolic hydroxyl groups was higher. This supports the assumption that even a very short time of ball milling (3 h) results in a certain (mechanochemical) cleavage of β-O-4 structures in lignin. It is also well-known that isolated milled wood lignin
3470
Biomacromolecules, Vol. 6, No. 6, 2005
Lawoko et al.
Table 2. Content of Uncondensed β-O-4 Structures and Phenolic Hydroxyl Groups Present in the Lignin Fractions of Wood and LCCs
fraction wood wood LCC P1 P3 P2 P4 S4 S7 unbl. pulp LCC P3-P P4-P S7-P O-del. pulp LCC P4-PO
β-O-4 structures, µmol/g lignin
phenolic OH, µmol/g lignin
phenolic OH/ 100 C9, contribution to total lignina
1313
730b
13b
1169 658 1247 1300 NAc 974
NAc 1558 1257 717 NAc 1300
50 292 37
2129 1450 1230
5.8 13.2 4.5
90
680
10
1.1 4.6 3.7 5.9
a Based on the yield of lignin in the various fractions as given in Table 1. b Value for spruce wood.23 c NA ) not analyzed.
from spruce has a much higher content of phenolic hydroxyl groups than wood lignin, approximately 20 per 100 phenylpropane units,19 due to the extensive milling that is required in the isolation procedure. Among the various LCC fractions from wood, those fractions rich in galactoglucomannan were found to have higher contents of β-O-4 structures than the glucane-lignin and xylan-rich fractions. The large difference in the phenolic hydroxyl group content between fractions P2 and P4 despite their similar contents of β-O-4 structures indicates, however, that their morphological origin or their location relative to cellulose may be different. The lowest content of β-O-4 structures was found in the glucane-lignin (P3) fraction which also had the highest content of phenolic hydroxyl groups. It can thus be assumed that the milling energy is to a great extent absorbed by the crystalline cellulose microfibrils and that any lignin located in the immediate vicinity should consequently be preferentially degraded. After a kraft cook, only three different LCC fractions could be isolated by adopting the scheme shown in Figure 1. The well-known comprehensive degradation of galactoglucomannan during the initial phase of a kraft cook together with the changes in lignin and xylan content during the cook thus seems to affect the resulting LCCs so that the lignin-rich P2 and S4 fractions and the galactose-rich P1 fraction are completely absent. Thioacidolysis showed that the remaining content of β-O-4 structures was low in all three LCC fractions whereas analyses of methanol after periodate oxidation showed that they had a high content of phenolic hydroxyl groups. This is in good agreement with other analytical data on kraft pulp lignins.22,23 The fact that the predominant LCC fraction, P4-P, still contained an appreciable amount of hydrolyzable β-O-4 linkages indicates, however, that, e.g., condensation reactions11 may to some extent prevent the desirable lignin fragmentation encountered in a kraft cook. In previous work, the delignification rates of the different LCCs during the final part of a kraft cook were found to be
Figure 3. Size exclusion chromatography of a mixture of acetylated thioacidolysis products from pinoresinol and guaiacylglycerol-βguaiacyl ether. The dimeric and monomeric products are denoted “a” and “b” respectively.
in the order xylan-lignin-glucomannan > glucan-lignin > glucomannan-lignin-xylan when the cook was allowed to proceed from kappa number 55 to kappa number 25. The same trend in delignification rates was observed on subsequent oxygen delignification.13 In the present work, the comparatively high stability of the glucomannan-ligninxylan complex (P4-P) toward delignification was further confirmed by isolation of the P4-PO fraction as the sole remaining LCC after the oxygen-delignification stage. The fraction still contained a considerable amount of β-O-4 structures, whereas the amount of phenolic end-groups had decreased due to the oxidative character of the stage.24 SEC of Thioacidolysis Products. The molecular size (mass) distribution of the acetylated thioacidolysis products obtained from the various LCCs was studied in tetrahydrofuran (THF) using the corresponding products obtained from spruce wood as a reference. Such distributions have previously been studied for wood19,25 and kraft pulp11,19 using either THF or dioxane-water as solvent. In those studies, peaks representative of trimeric, dimeric, and monomeric products resulting from the selective cleavage of β-O-4 linkages in lignin have been observed in wood, whereas pulp samples have been shown to also contain higher molecular mass material. To obtain reliable retention times for monomeric and dimeric lignin degradation products in the chromatographic system used, the model compounds guaiacylglycerol-βguaiacyl ether and pinoresinol (Figure 2) were subjected to thioacidolysis and the acetylated products were then analyzed in the SEC system. These compounds yield major dimeric and monomeric reaction products respectively, and in addition, guaiacol is liberated from the aryl ether structure. On admixture of the reaction products and SEC analysis, the chromatogram shown in Figure 3 was obtained. In this, the expected dimeric (a) and monomeric (b) products are clearly seen together with two further peaks of lower molecular weight. The latter were not identified, but they were assumed to belong to guaiacol and e.g. pyridine, the latter originating from the acetylation reaction. The peaks originating from dimers and monomers in the reference wood sample were clearly discernible, as shown in Figure 4A. The monomer peak was, however, more complex than that obtained for the model compound and contained at least three separate peaks in close proximity to
Differences in Lignin-Carbohydrate Complexes
Biomacromolecules, Vol. 6, No. 6, 2005 3471
Figure 5. Schematic view of a fiber wall segment in spruce wood based on a previously published model.26
Figure 6. Major dimeric (acetylated) products found after thioacidolysis and Raney-nickel reduction of the fraction S7. The origin of 1 ) 5-5, 2 ) β-1, and 3 ) β-5 structure in lignin.
Figure 4. Size exclusion chromatography of acetylated thioacidolysis products from (A) spruce wood, (B) glucomannan-lignin-xylan LCCs (mixture P2 + P4) from wood, (C) xylan-lignin-glucomannan LCCs (S4, dotted line, and S7, full line) from wood.
each other. This may be due to the high-resolution power of the chromatographic system employed and to the formation of monomeric structures differing from each other in, e.g., the number of ethylthio groups attached to the side chain.16 In addition to the dimer peak, unresolved peaks appearing at shorter retention times (i.e., higher molecular mass) were also visible. The general appearance of the wood sample chromatogram is in good agreement with previous data for wood and shows that the thioacidolysis of spruce wood results in a very extensive degradation of the polymeric lignin to form monomers and dimers as predominant products together with smaller amounts of trimers-oligomers.19,25 The major LCC fractions from wood, viz. the GlcManLig-Xyl complexes, P2 + P4, gave on SEC analysis a chromatogram almost identical to that of the wood sample (Figure 4B). Similar chromatograms were also obtained from the GalGlcMan-L-P (P1) and the Glc-L (P3) complexes. The Xyl-L-GlcMan (S4 and S7) complexes, on the other hand, gave almost exclusively a monomer peak indicating a lignin having a very high percentage of uncondensed β-O-4 structures (Figure 4C). The fact that the two major types of
LCC from wood, viz. P2 + P4 and S4, S7, contain glucomannan and xylan respectively as the predominant polysaccharide, in both cases combined with substantial amounts of lignin (Table 1), is in line with the ultrastructural arrangement of the polysaccharides and lignin in the cell wall as previously suggested.26 The large difference in the lignin structure in these two types of LCC demonstrates, however, that the fiber wall must have a more complex pattern than previously thought with one lignin type surrounded only by xylan and another only by glucomannan. A modified schematic illustration of the fiber wall encompassing these structural differences is shown in Figure 5. The presence of two different types of lignin in the fiber wall was further supported by the analysis of the thioacidolysis products from P2+P4 and S4, S7 after a subsequent second reaction stage, a reduction with Raney nickel.20,27 The latter converts all lignin side chains into reduced hydrocarbon structures thus facilitating a subsequent GC-MS analysis. The results obtained for the P2+P4 fraction showed a pattern of dimeric and monomeric lignin degradation products that was very similar to that obtained earlier from spruce wood. The S4 and S7 fractions, on the other hand, gave a large amount of the predominant monomer, 2-methoxy-4-n-propylphenol (acetate), originating from β-O-4 structures but only very small amounts of dimeric material. In the latter portion of the chromatogram, the three major peaks were identified as products originating from 5-5, β-1 and β-5 structures (Figure 6) by GC-MS and comparison of the fragmentation patterns.27 The possibility that two types of lignin are formed in the fiber wall may be due to the fact that xylan, being an acidic hemicellulose, can have an influence on the mode of monolignol polymerization different from that exerted by (galacto)-glucomannan or cellulose. In previous work, it has been shown that the structure of synthetic lignin (DHP), prepared by metal salt catalyzed polymerization of coniferyl
3472
Biomacromolecules, Vol. 6, No. 6, 2005
Lawoko et al.
Figure 8. Size exclusion chromatography of acetylated thioacidolysis products from glucomannan-lignin-xylan LCC (P4-PO) from oxygendelignified kraft pulp.
Figure 7. Size exclusion chromatography of acetylated thioacidolysis products from (A) glucomannan-lignin-xylan LCC (P4-P) from unbleached kraft pulp and (B) glucan-lignin LCC (P3-P, dotted line) and xylan-lignin-glucomannan LCC (S7-P, full line) from unbleached kraft pulp.
alcohol, can differ depending on the exact reaction conditions. When manganese (III) is used, either DHP devoid of phenolic hydroxyl groups or a “normal” DHP can be prepared by slight changes in reaction conditions.28,29 Furthermore, by an artificial lignification of primary maize cell walls with coniferyl alcohol, it could be shown that a slow introduction of coniferyl alcohol at a low pH (pH ) 4) resulted in a lignin with a greater proportion of β-O-4 linkages as compared to lignification at pH 5.5.30 A detailed understanding of the prerequisites determining the mode of monolignol polymerization in the wood cell must, however, await further experimental information. The predominant LCC fraction from the unbleached kraft pulp, viz. the GlcMan-L-Xyl (P4-P) fraction, showed, in comparison to the corresponding wood P2 + P4 fraction (Figure 4B), a larger amount of material having a short retention time (Figure 7A). This is in line with previous work showing that the kraft pulping process induces a condensation in the pulp lignin and that this leads to an increase in the molecular mass distribution following thioacidolysis.11 In the case of the Glc-L (P3-P) and the Xyl-L-GlcMan (S7-P) fractions, on the other hand, completely different molecular mass distribution patterns were obtained, and very little lignin material with a high molecular mass was found (Figure 7B). In both of these fractions, the most prominent peaks were found at long retention times, demonstrating the presence of components having a very low molecular mass. Thus, the chainlike lignin structure present in the S4 and S7 fractions from wood (Figure 4C) seems to be efficiently degraded during the course of a kraft cook and only a very small amount of residual β-O-4 structures can be found (cf Table 2). After a further delignification of the kraft pulp using an alkaline oxygen stage, the resulting pulp afforded one major LCC fraction (Table 1, Table 2, P4-PO) as a result of the
enzyme treatment and further workup according to the scheme in Figure 1. On thioacidolysis and GPC of the resulting mixture after acetylation, the chromatogram shown in Figure 8 was obtained. Obviously, despite the comprehensive lignin dissolution in the oxygen stage, most of the residual lignin in the pulp has a structure which (after thioacidolysis) seems to be similar to that of the major LCC fraction in the unbleached pulp (Figure 7A, P4-P). Conclusions LCCs were prepared in quantitative yield from spruce wood as well as from the corresponding unbleached and oxygen-delignified kraft pulps. Four types of LCC were isolated from wood, viz. GalGlcMan-L-P, Glc-L, GlcMan-L-Xyl, and Xyl-L-GlcMan. The latter two differed in the relative amounts of glucomannan and xylan attached to the lignin. After kraft pulping, three types of LCC were found with the predominant portion of the lignin present in the GlcMan-L-Xyl fraction. This fraction was completely dominant after a subsequent oxygen stage. Thioacidolysis of the various LCCs showed that the predominant inter-lignin linkage, the β-O-4 linkage, was always present in a large amount although, as expected, a large decrease in the absolute amount was found when going from wood to kraft pulp and further to oxygen-delignified pulp. Thioacidolysis in combination with size exclusion chromatography of the various LCCs from wood revealed pronounced differences in the lignin structure between the GlcMan-L-Xyl and the Xyl-L-GlcMan fractions. The latter type of lignin was found to have a rather linear coupling mode of β-O-4 structures whereas the former gave a chromatographic pattern similar to that obtained from wood. A substantial portion of high molecular mass material was formed from the major LCC obtained from kraft pulp and oxygen-delignified pulp, the GlcMan-L-Xyl fraction, clearly showing the presence of lignin condensation reactions occurring during the kraft cook. Acknowledgment. The Swedish Research Council, Contract No. 621-2001-2323, is acknowledged for financial support. Andrea Majtnerova´ of this Department is thanked for advice on the thioacidolysis experimental work. References and Notes (1) Bjo¨rkman, A. SVensk Papperstidn. 1957, 60, 243-251.
Differences in Lignin-Carbohydrate Complexes (2) Yamasaki, T.; Hosoya, S.; Chen, C.-L.; Gratzl, J. S.; Chang, H.-m. 1st International Symposium on Wood and Pulping Chemistry; Stockholm, Sweden, 1981; Proceedings, 2, 34-42. (3) Minor J. L. J. Wood Chem. Technol. 1986, 6, 185-201. (4) Gellerstedt, G.; Lindfors, E.-L. Holzforschung 1984, 38, 151-158. (5) Karlsson, O.; Westermark, U. J. Pulp Paper Sci. 1996, 22, J397J401. (6) Fengel, D.; Wegener, G. In Wood Chemistry, Ultrastructure, Reactions; Walter De Gruyter: Berlin, Germany, 1984; p 167. (7) Lawoko, M.; Henriksson, G.; Gellerstedt, G. Holzforschung 2003, 57, 69-74. (8) Gierer, J.; Wa¨nnstro¨m, S. Holzforschung 1986, 40, 347-352. (9) Iversen, T.; Wa¨nnstro¨m, S. Holzforschung 1986, 40, 19-22. (10) Gierer, J.; Imsgard F.; Pettersson, I. Appl. Polym. Symp. 1976, 28, 1195-1211. (11) Gellerstedt G.; Majtnerova, A.; Zhang L. C. R. Biologies 2004, 327, 817-826. (12) Lawoko, M.; Henriksson, G.; Gellerstedt, G. Holzforschung submitted for publication. (13) Lawoko, M.; Berggren, R.; Berthold, F.; Henriksson, G.; Gellerstedt, G. Holzforschung 2004, 58, 603-610. (14) Theander, O.; Westerlund, E. A. J. Agric. Food Chem. 1986, 34, 330-336. (15) Lai, Y.-Z. In Methods in Lignin Chemistry; Lin, S. Y., Dence, C. W., Eds.; Springer-Verlag: Heidelberg, Germany, 1992; pp 423-434. (16) Rolando, C.; Monties, B.; Lapierre, C. In Methods in Lignin Chemistry; Lin, S. Y., Dence, C. W., Eds.; Springer-Verlag: Heidelberg, Germany, 1992; pp 334-349.
Biomacromolecules, Vol. 6, No. 6, 2005 3473 (17) Miksche, G. E.; Gratzl, J.; Fried-Matzka, M. Acta Chem. Scand. 1966, 20, 1038-1043. (18) Erdtman, H. SVensk Kemisk Tidskrift 1934, 46, 229-233. (19) O ¨ nnerud, H.; Gellerstedt, G. Holzforschung 2003, 57, 165-169. (20) Lapierre, C.; Pollet, B.; Monties, B.; Rolando, C. Holzforschung 1991, 45, 61-68. (21) Gellerstedt, G. In Methods in Lignin Chemistry; Lin, S. Y., Dence, C. W., Eds.; Springer-Verlag: Heidelberg, Germany, 1992; pp 487497. (22) Gellerstedt, G.; Lindfors, E.-L.; Lapierre, C.; Monties, B. SVensk Papperstidn. 1984, 87, R61-R67. (23) Gellerstedt, G.; Lindfors, E.-L. SVensk Papperstidn. 1984, 87, R115R118. (24) Gellerstedt, G.; Lindfors, E.-L. Tappi J. 1987, 70 (6), 119-122. (25) Suckling, I.; Pasco, M.; Hortling, B.; Sundquist, J. Holzforschung 1994, 48, 501-503. (26) Salme´n, L.; Olsson, A.-M. J. Pulp Paper Sci. 1998, 24, 99-103. (27) O ¨ nnerud, H. Holzforschung 2003, 57, 377-384. (28) Landucci, L. L. J. Wood Chem. Technol. 2000, 20, 243-264. (29) O ¨ nnerud, H.; Zhang, L.; Gellerstedt, G.; Henriksson, G. Plant Cell 2002, 14, 1953-1962. (30) Grabber, J. H.; Hatfield, R. D.; Ralph, J. J. Agric. Food Chem. 2003, 51, 4984-4989.
BM058014Q
