ALKALI STABILITY OF SOME URONIC ACIDS AND ITS IMPLICATIONS IN BOROHYDRIDE AND POLYSULFIDE COOKING N I L S H A R T L E R A N D
I N G A - L I S A SVENSSON
Central Laboratory of the Swedish Cellulose Industry, Stockholm, Sweden
2-0-(4-0-methyl-a-D-glucopyranosyluronic acid)-D-xylopyranose is unstable in alkaline solutions, the effect being more pronounced the higher the temperature. The experimental results make it most likely that the bond between the glucuronic acid and the xylose unit is split as a result of this degradation. One can conclude that when a substituent in the C-2 position is present in cellulose and hemicellulose, the peeling will not stop when the substituted glucose unit is reached. In alkaline cooking processes based upon softwoods, when measures are taken to lessen degradation resulting from peeling-e.g., addition of borohydride or polysulfide-the yield of glucomannans from wood to pulp should increase, whereas that of the xylan should remain more or less unchanged. When hardwoods are cooked, on the other hand, the yield of xylans 7 should increase. HE alkaline degradation of polysaccharides containing 1,4T g l p c o s i d i c bonds results mainly from the peeling reaction as well as from alkaline hydrolysis of glycosidic bonds followed by secondary peeling. A necessary condition for the initiation of the peeling reaction is the presence of a reducing end group. This is rearranged through epimerization with formation of a keto group at carbon 2 in the @-position to the glycosidically bonded carbon 4 and resultant lessening of the stability of this bond. The glycose units are split off one by one with formation of isosaccharinic acids. After an average of 50 peelings the reaction is terminated by formation of a stable metasaccharinic acid end group (4). If the saccharide is substituted a t carbon 2, there will be no epimerizarion and consequently no shortening of the polysaccharide chain so long as the substituent is left intact ( 8 ) . Birch xylan consists of p(1 -+ 4)-linked D-xylose units, substituted at approximately every tenth xylose unit in the C-2 position with 4-O-methyl-a-~-glucuronic acid residues. The number of the glucuronic acid substituents and the stability of their bonds to the xylan chain are consequently of great importance in determining the extent of peeling. In the present work the interest has been focused on the pure peeling by performing the experiments in alkali at temperatures of up to 100" C., in which case only primary peeling occurs and the uronic acid content remains unchanged. At higher temperatures, such as 170' C.. however, the glucuronic acid substituents are split off almost completely if the alkali concentration is 40 grams per liter or higher. At lower alkali concentrations the extent of splitting is considerably less. When alkaline solutions of birch xylan are kept at I O O ' C. for several hours, the amount of residue which can be isolated decreases to approximately 65% and the degree of polymerization decreases. T h a t the entire degradation is due to primary peeling is shown by the fact that birch xylan, when reduced with sodium borohydride and treated similarly, can be isolated in 100% yield and with almost the same degree of polymerization. If the uronic acid substituents impede the peeling, a random or even distribution of the substituents would result in between 2 and 5y0degradation, as the DP, of birch xylan is approximately 200. Fractionation experiments on xylan have shown that polysaccharides of moderately low glucuronic acid content can be isolated. but no fractions free from glucuronic acid could be found. Furthermore, Time11 (7) has
EO
I&EC PRODUCT RESEARCH A N D DEVELOPMENT
shown that the glucuronic acid substituents are not present on adjacent xylose residues and has concluded that acid residues are probably distributed at random along the polymer chain. In view of this, a decrease in yield of 35% seems extremely unlikely if it is assumed that the glucuronic acid substituents impede the peeling. If, on the other hand, it is assumed that the glucuronic acid bond has considerably lessened the stability when occurring in a terminal xylose unit, the glucuronic acid substituents would not impede the peeling and the observed extent of the peeling can be explained. This assumption differs from the common view that a substituent in the C-2 position stops the peeling. This conclusion has been based on experimental results obtained a t low alkali concentration and in particular a t low temperatures. However, in the present work only temperatures of 100' C. or higher are considered-Le., temperatures used in mill operations related to cellulose technology. I n an earlier communication part of this material was presented (7). Experimental Layout
Three model compounds were used to study the alkali stability of the glucuronic acid bond when it is situated at different positions in the xylan chain. T o simulate the case where the glucuronic acid is linked to a reducing end group the aldobiuronic acid 2-0- (4-0-methyla-o-glucopyranosyluronicacid)-D-xylopyranose was used.
The reducing end group in this acid probably has some influence on the alkali stability of the glucuronic acid bond and the same acid was therefore also studied &her reduction with sodium borohydride [2-0-(4-0-methy~-a-~-g~ucopyranosy~uronic acid)-D-xylitol]. COOH
To simulate the average situation in birch xylan, especially as regards the influence of adjacent bonds on the glucuronic
acid bond the aldotetrauronic acid (reduced with borohydride) was prepared and studied [ O-a-~-(4-O-methylglucopyranosyluronic acid)-(1 +. 2)-0-/3-~-xylopyranosyl-(l +. 4)O-P-n-xylopyranosyl-(I +. 4)-~-xylitol.
barium hydroxide on preparation of aldobiuronic acid. This epimerization proceeds to an equilibrium with one third of the "lyxose" and two thirds of the "xylose" acid, as shown by Roudier and Eberhard (6). After hydrolysis lyxose was identified by paper chromatography. Klemer, Lukowski, and Zerhusen (3) have shown that the degradation in alkali of 2-methyla-glucose proceeds via 2methyl-D-glucoseen(2, 3) : CH20H
This acid had the substituents positioned at the xylose unit having no glycosidic bond a t the C-4 position (7). T h e aldobiuronic acid was prepared by hydrolysis of birch xylan, followed by neutralization with barium hydroxide in combination with carbon dioxide. T h e acidic components, aldobiuronic acid and 4-O-methyl-~-glucuronicacid, were separated by chromatography on an ion exchanger using acetic acid of increasing concentration as eluent (2). T h e aldote trauronic acid was prepared by enzymatic hydrolysis of birch xylan using the technique of simultaneous dialysis (5). T h e enzyme used, pectinase, does not hydrolyze the xylosidic bonds closest to the glucuronic acid bond. Acid oligosaccharides were obtained having the glucuronic acid residue linked to the xylose unit farthest from the reducing end group. T h e acid oligosaccharides of different degrees of polymerization were separated by paper chromatography (7). The alkali treatment was performed under nitrogen in 4% sodium hydroxide solution between room temperature and 100' C . If complete stability was observed a t a certain temprrature, no further experiments were performed at lower temperatures. For most chemical reactions the necessary reaction time is doubled upon a 10' C. decrease in temperature. Due allowance was made for this fact and a t lower temperatures correspondingly longer reaction times were used. T h e progress of the degradation was followed by paper chromatography. Results
Aldobiuronic Acid. The aldobiuronic acid was treated in alkali over the whole temperature range. .4t 100" C. the reaction mixture became intensely yellow. O n the chromatogram no identifiable compounds were found and consequently the acid must have been completely degraded. The fact that different acid fragments are formed on treatment of aldobiuronic acid in alkali at 100' C . has been pointed out ( 9 ) . At lower temperatures there was much less degradation. At 60' C. only traces of aldobiuronic acid could be found on the chromatogram. At room temperature the reaction was allowed to proceed for a very long time and samples were \vi thdrawn a t regular intervals. The first sample Tvas t'aken after 48 hours. A new compound was found having a somewhat higher R, value and giving a color reaction with anisidine similar to that of the starting material. At room temperature, honever, no acid fragments are to be expected (9, 70). T h e faster moving acid proved to be 2-0- (4-0-methyl-a-~-glucopyranosyluronic acid)-D-lyxopyranose (methylglucuronolyxose) . COOH
This acid is formed so easily that it could be detected if the p H accidentally became too high during neutralization \vith
@
HO
'
H,OH
-
H,OH
OH-
dCH3
H
'
dCH3
I t is therefore most likely that the further degradation of one of the two aldobiuronic acids proceeds in a similar manner. Further degradation products were not isolated and the mechanism for the further degradation remains unexplained. Probably, however, the glucuronic acid bond is split a t a later stage of the degradation. Reduced Aldobiuronic Acid. T h e reduced aldobiuronic acid was practically stable a t 100 " C. Only traces of the reduced lyxose acid could be detected together with reduced xylose acid on the chromatogram. At lower temperatures it was completely stable. Reduced Aldotetrauronic Acid. T h e reduced aldotetrauronic acid was stable toward alkali a t 100" C. If alkaline hydrolysis of the uronic acid bond occurred through splitting of the glucuronic acid bond, glucuronic acid and xylotriol would be found on the chromatogram. This was not the case. Conclusions \\Then birch xylan is treated with alkali at 100' C. the glucuronic acid bond is stable so long as the glucuronic acid is not linked to a terminal reducing xylose unit. If, however, the substituent is positioned on a terminal reducing group, this end group is degraded even below 100' C. The reaction proceeds through some partly unknown intermediate stages and the most probable result is that the bond between the glucuronic acid and the xylose unit is split. This means that a 4-0methyl-D-glucuronic acid substituent in the C-2 position in xylose does not stop the peeling reaction occurring in alkaline medium either heterogeneously, as in most technical applications of interest. or homogeneously, as in the present model experiments. I n alkaline pulping the yield of bleached pulp is given, by and large, by the carbohydrate retention in the cooking. T h e retention of each individual carbohydrate is mainly influenced by its alkaline stability. When the alkaline stability is rather high, as in the case of softwood xylans which have arabinose substituents that stop the peeling reaction, any stabilization forced on the polymer ~ i l have l little effect on the retention. O n the other hand, when the alkaline stability is comparatively low, as in case of glucomannans having no substituents and hardwood xylans having only inefficient glucuronic acids. there is evidently much to gain by introducing some stabilizing measure. I n kraft pulping we know so far two means of stabilization-addition of sodium borohydride or sodium polysulfide. I n the case of softwood the retention of glucomannan is then increased, whereas that of xylan is actually slightly decreased, for reasons beyond the scope of the present discussions. Corresponding results for hardwood show that in the case of birch, contrary to what might have been expected but fully in line with the present results, the retention of r'le glucomannan as well as of the xylan increases drastically. VOL.
4 NO. 2 J U N E 1 9 6 5
81
When prehydrolyzed pine wood is cooked in a second stage with a kraft liquor containing borohydride, the retention of glucomannan and of xylan increases. This is also fully in accord with the present results because the pine xylan ought to have been modified as a result of prehydrolysis, with almost complete loss of arabinose, making the prehydrolyzed pine xylan rather similar to the birch xylan.
(4) Machek G., Richards, G. N., J . C'hem. s o d . 1957, P. 4500. ainter, T. J., Can. J . Chem. 37, 497 (1959). i i u d i e r , A. J., Eberhard, L., Bull. sot. Chirn. 1960, p. 2074. (7) Timell, T. E., S u m k 65, 435 (1962). (8) Whistler, R. L., BeMiller, J. N., Advan. Carbohydrate Chem. 13,289 (1958). (9) Whistler, R. L., Corbett, W. M., J . Am. Chem. Soc. 77, 3822 (1955). (10) Whistler, R. L., Richards, G. N., Zbid., 80, 4888 (1958).
Literature Cited (1) Xurell, R., Hartler, N., Persson, G., Acta Chern. Scand. 17, 545
RECEIVED for review September 4, 1964 ACCEPTEDFebruary 19, 1965
(1 967). \ - ~ - - I -
(2) Croon: I., Enstrom, B., Tappi 44, 870 (1961). ( 3 ) Klemer, A., Lukowski, H., Zerhusen, F., Cham. Bzr. 96, 1515 (1963).
Division of Cellulose, Fiber, and Wood Chemistry; 148th Meeting, ACS, Chicago, Ill., September 1964.
END O F SYMPOSIUM
PAD- BA KE CAR BOXY M ET HY LAT IO N
OF COTTON TEXTILE MATERIALS ROBERT M. R E I N H A R D T AND TERRENCE W. FENNER' Southern Regional Research Laboratory, Kew Orleans, La.
Cotton textile materials can b e carboxymethylated b y a new process in which the cotton is impregnated with an aqueous solution containing sodium chloroacetate and sodium hydroxide, padded or centrifuged to remove excess solution, and baked to etherify the cellulose. The effects of processing variables and the properties of the products have been determined. Cotton fabric has been carboxymethylated b y this new process on a semipilot plant scale. The treatment appears more suitable for use in continuous processing than the present conventional wet method for carboxymethylating cotton fabric. Standard textile finishing equipment may b e utilized. A single solution containing both reagents necessary for reaction i s employed. Much lower concentrations of sodium hydroxide can be used than the 40 to 50% solutions required in the conventional two-stage carboxymethylation process. Other a-halocarboxylic acid salts also were used to prepare alpha-substituted carboxymethylated cotton fabrics.
c
modification of cotton to impart new and improved properties with retention of the important fibrous nature of the starting material has long been a research objective of this laboratory. Many chemically modified cottons have been developed as a result of this research (7-9). Carboxymethylated cotton, a product which has been the subject of several previous publications (4, 5. 10-731, is now commercially produced for a captive use and on a commission basis. Cotton which has been carboxymethylated to a degree of substitution of 0.2 or less has properties considerably changed from those of the untreated material: a built-in starched effect, increased moisture regain, \rater absorbency, water permeability, changed dyeing characteristics, increased resistance to soiling from aqueous dispersions. greater ease of soil removal, cation exchange properties. high water swellability, and a greater receptivity to further chemical treatment than unmodified cotton. Products with degrees of substitution cf 0.3 or more disintegrate in water or dilute alkali solution. Textiles of this type are useful whenever a temporary, easily removable member is needed. More than 40,000,000 pounds per year of nontextile sodium carboxymethylcellulose (cellulose gum or CMC) are being produced in this country. Cellulose pulp is impregnated with sodium hydroxide to form alkali cellulose, which then reacts with chloroacetic acid or sodium chloroacetate (2). Sumerous variations of this technique have been described ( 6 ) . The over-all reaction may be written: 1
82
Cell-OH
HEMICAL
Present Address, U. S. Customs Laboratory, Yew Orleans, La. I&EC P R O D U C T RESEARCH A N D DEVELOPMENT
Cell-ONa
Lf K a O H + Cell-ONa
+ ClCHzCOONa
+
+ H2O
Cell-OCHzCOONa
+ NaCl
which is an example of the classical Williamson synthesis for the preparation of ethers. Recent mechanism studies on the reactions of a-halocarboxylic acids in aqueous alkaline solutions suggest that nucleophilic attack may not occur in one step, but that highly reactive a-lactones may be formed as transient intermediates (3).
No
ClCHzC -+
CHzC=O
\0 - v 0
+ C10
CHzCyO
\/ 0
+ -OH
+ HOCH2C
// \
0-
Such transient intermediates may also be involved in the carboxymethylation of cellulose. Fibrous cotton textile materials have been carboxymethylated by variations of the methods used on pulp. The most s u c c a f u l method 14) has been that in which cloth is padded with aqueous chloroacetic acid and then treated with a concentrated solution of sodium hydroxide in a wet, two-stage batch process. Reversing the order of application of the
