Degradation of cellulosic materials by heating in DMAc/LiCl

Article abstract of DOI:10.1016/S0040-4039(02)01767-7

N,N-Dimethylacetamide (DMAc, 1)/lithium chloride is a frequently used solvent system for cellulosic materials. Heating or refluxing insoluble samples in DMAc/LiCl represents a common protocol for facilitating the subsequent dissolution. It is shown by kinetics using gel permeation chromatography (GPC) that the improved solubility is not caused by an alleged 'activation' of the pulp, but by progressing degradation of the cellulosic material. Ketene aminal 2 and N,N-dimethylketeniminium ions (3), the latter being known as reagent for cleavage of ethers and acetals, have been demonstrated to be present in DMAc and DMAc/LiCl at elevated temperatures by specific trapping of these intermediates in a Claisen rearrangement and in a thermal [2+2]-cycloaddition, respectively.

Full text of DOI:10.1016/S0040-4039(02)01767-7

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 7757–7759
Degradation of cellulosic materials by heating in DMAc/LiCl
Antje Potthast,a Thomas Rosenau,a Herbert Sixtab and Paul Kosmaa,*
aChristian-Doppler-Laboratory, University of Agricultural Sciences, Muthgasse 18, A-1190 Vienna, Austria
bR & D Department, Lenzing AG, A-4860 Lenzing, Austria
Received 2 August 2002; revised 22 August 2002; accepted 23 August 2002
Abstract—N,N-Dimethylacetamide (DMAc, 1)/lithium chloride is a frequently used solvent system for cellulosic materials.
Heating or refluxing insoluble samples in DMAc/LiCl represents a common protocol for facilitating the subsequent dissolution.
It is shown by kinetics using gel permeation chromatography (GPC) that the improved solubility is not caused by an alleged
‘activation’ of the pulp, but by progressing degradation of the cellulosic material. Ketene aminal 2 and N,N-dimethylketeni-
minium ions (3), the latter being known as reagent for cleavage of ethers and acetals, have been demonstrated to be present in
DMAc and DMAc/LiCl at elevated temperatures by specific trapping of these intermediates in a Claisen rearrangement and in
a thermal [2+2]-cycloaddition, respectively. © 2002 Elsevier Science Ltd. All rights reserved.
N,N-Dimethylacetamide (DMAc, 1) containing lithium
chloride is a very common solvent system in cellulose
chemistry,1,2 where it is frequently used for derivatiza-
tion under homogeneous conditions.3,4 Moreover, it has
become the standard solvent for gel permeation chro-
matography (GPC) measurements of cellulosics.5–7 To
achieve dissolution in DMAc/LiCl, cellulosic pulps are
subjected to so-called ‘activation procedures’, e.g. by
treatment with liquid ammonia,8,9 swelling in water
followed by solvent exchange to DMAc,5,10 or freeze-
drying.11 A very common protocol is also heating or
refluxing the cellulose samples in DMAc, or in DMAc
containing LiCl.12,13 This treatment is thought to cause
intra- and intercrystallite swelling, breaking of hydro-
gen bonds and increased accessibility. In the present
work, two cellulose samples of different provenience
(cf. Fig. 1) which are readily soluble in DMAc/LiCl
have been employed to study the effects of the high
temperature activation method in more detail.
LiCl content increased, and became rather severe at
temperatures near the boiling point of DMAc (164°C).
Cellulose degradation in the absence of LiCl is less
pronounced than in its presence.
The cellulose degradation was observed for both pulps,
the degradation rate being different. It appeared thus
plausible to assume that a highly reactive, hitherto
unidentified intermediate was present, which effected
this cleavage, since merely a thermal decay could by no
means account for the magnitude of the observed
effects.
As demonstrated by GPC measurements, heating in
DMAc or DMAc/LiCl definitely causes cellulose degra-
dation, which also explains the observed positive effect
on dissolution (see Fig. 1). Already at temperatures as
low as 85°C, a clear decrease in molecular weight was
observed by GPC. The loss in DP (degree of polymer-
ization) became more pronounced as temperature and
Figure 1. Degradation of two different pulps upon heating in
DMAc/LiCl (9% w/v) at different temperatures. Pulp A:
beech sulfite pulp; pulp B: eucalyptus prehydrolysis kraft
pulp, error bars 5%.
Keywords: DMAc; LiCl; cellulose; degradation; keteniminium ions;
trapping; cycloaddition; kinetics; gel permeation chromatography.
* Corresponding author. Tel.: +43-1-36006 6071; fax: +43-1-36006
6059; e-mail: pkosma@edv2.boku.ac.at
0040-4039/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)01767-7

7758
A. Potthast et al. / Tetrahedron Letters 43 (2002) 7757–7759
Addition of methanol to heated DMAc/LiCl produced
small amounts of 1-dimethylamino-ethane-1,1-diol (4),
which was indicative of the intermediacy of N,N-
dimethylketene semiaminal (2)—the enol form of
DMAc—and/or N,N-dimethylketeniminium ions (3).
The occurrence of both intermediates in DMAc/LiCl at
temperatures above 85°C was finally confirmed by trap-
ping (see Scheme 1). For intermediate 2, the reaction
with allyl alcohol to an ketene aminal, which undergoes
immediate Claisen-type rearrangement to 4-pentenoic
acid N,N-dimethylamide (5),14 was employed. To trap
3, 1-decene was used, which reacted in a thermal,
suprafacial–antarafacial [2+2]-cycloaddition to produce
from secondary amides in polar aprotic solutions in the
presence of several salts has been reported,18,19 but has
never been described to occur in DMAc/LiCl mixtures.
Keteniminium salts are known as reagents to affect
cleavage of ethers, acetals and ketals.20–22 In organic
solutions containing electrolytes, hydroxyl groups and
amino groups do not interfere with the reaction, as they
are surrounded by a solvent shell and are thus shielded
from attack by the strongly electrophilic keteniminium
reagent.
The same general mechanism as in Scheme 2 must be
assumed also for the cleavage of glycosidic bonds in
cellulosics. The electrophilic transient species 3 attacks
the glycosidic oxygen, followed by cleavage of the
glycosidic bond and secondary hydrolytic reactions.
a
cyclobutaniminium salt, finally giving 3-octyl-
cyclobutanone (6)15 upon hydrolysis. Trapping prod-
ucts 5 and 6 were found in heated DMAc and in heated
DMAc/LiCl independent of the presence of dissolved
or suspended pulp, lower amounts of trapping product
being obtained in the presence of cellulose.
The detrimental effect of keteniminium ions on the
integrity of cellulosic pulps was confirmed by addition
of an external degrading agent of the keteniminium
type (Scheme 3): when 4-(1-chloro-2-methylpropenyl)
morpholine 7 (0.1% relative to pulp), which immedi-
ately releases keteniminium ion 8 upon thermal treat-
ment, was added to a solution of pulp in DMAc/LiCl
(9%, w/v) heated to 85°C, the cellulose was completely
degraded to oligomers within 30 min.
N,N-Dimethylketeniminium cations (3) are formed
from DMAc via enol 2 in a thermal elimination. The
water thereupon released is incorporated into the Li+
ligand shell, which prevents the immediate reverse reac-
tion. The chloride anions act as counterions to the
keteniminium cation. Thus, a dynamic equilibrium
between 3, lithium cations, hydroxyl anions and chlo-
ride anions is established. The equilibrium concentra-
tion of 2 is determined by the endothermicity of the
elimination process and the sum of strong ionic interac-
tions in the solution; it is usually very low.16,17 How-
ever, if keteniminium ions are removed from the
reaction system, e.g. by trapping or by a competitive
reaction with cellulose, they are continuously regener-
ated. The thermal formation of keteniminium salts
Scheme 2. General mechanism of the cleavage of ether and
acetal bonds by N,N-dimethylketeniminium ions (3).
Scheme 1. Formation of intermediates 2 and 3 from DMAc
(1) and proof of their intermediacy by specific trapping reac-
tions.
Scheme 3. Degradation of cellulosics by cleavage of glyco-
sidic bonds induced by keteniminium ions (3 or 8).

A. Potthast et al. / Tetrahedron Letters 43 (2002) 7757–7759
7759
In summary, DMAc/LiCl at temperatures above 85°C
contains 3 as highly reactive intermediate, which is able
to cleave glycosidic bonds. Heating or refluxing cellu-
losics in DMAc or DMAc/LiCl should no longer be
used as an ‘activation procedure’: it causes pulp degra-
dation, and it thus influences the molecular weight
distribution of the material, so that subsequent GPC
analysis will not report the data of the original sample.
12. Klemm, D.; Philipp, B.; Heinze, T.; Heinze, U.;
Wagenknecht, W. Comprehensive Cellulose Chemistry;
Wiley-VCH: Weinheim, 1998; Vol. 2, p. 331.
13. Tosh, B.; Saikia, C. N.; Dass, N. N. Carb. Res. 2000, 327,
345–352.
14. 4-Pentenoic acid N,N-dimethylamide (5). 1H NMR
(CDCl3): l 2.36 (m, 4H, CH2
3.01 (s, 3H, N-CH3), 4.99 (dd, 1H, CHꢀCH2
Hz, 3J=1.5 Hz), 5.06 (dd, 1H, CHꢀCH2 3J=17.5 Hz,
2J=1.5 Hz), 5.88 (m, 1H, CHꢀCH2). 13C NMR (CDCl3):
6
-CH2
6
), 2.95 (s, 3H, N-CH3
6 ),
6
6 ,
3J=10.0
6
,
6
l 29.1 (3CH2), 32.5 (2CH2), 35.3 (N-CH3), 37.1 (N-CH3),
115.0 (ꢀ5CH2), 137.6 (4CHꢀ), 172.2 (1CO). The two N-
methyl groups are not magnetically equivalent due to the
hindered rotation around the C-N amide bond.
Acknowledgements
The authors would like to thank Dr. A. Hofinger and
Dr. J. Ro¨hrling for NMR measurements, as well as Ms.
G. Linsberger for practical assistance.
15. Characterized as 3-octylcyclobutanone-2,4-dinitrophenyl-
hydrazone: yellow crystals; mp=144–146°C. 1H NMR
(CDCl3): l 0.89 (t, 3H, 8%CH3-), 1.28 (m, 10H, 2%CH2,
3%CH2, 4%CH2, 5%CH2, 6%CH2), 1.54 (m, 4H, 1%CH2, 7%CH2),
3
2
4
2.56 (m, 1H, CH), 2.65 (m, 2H, CHA, CHA), 3.14 (m,
2H, 2CHB, 4CHB), 7.86 (d, 1H, 6%%CH), 8.28 (dd, 1H,
5%%CH), 9.11 (s, 1H, 3%%CH), 10.72 (s, 1H, NH). 13C NMR
(CDCl3): l 14.1 (8%CH3), 22.6 (7%CH2); 27.4 (2%CH2); 28.0
(3CH); 29.2; 29.4; 29.5 (3%CH2, 4%CH2, 5%CH2); 31.8;
(6%CH2), 36.2 (1%CH2), 37.1 (2CH2), 39.5 (4CH2), 116.0
(6%%CH2); 123.5 (3%%CH2); 128.7 (4%%CH2); 129.9 (5%%CH2);
137.5 (2%%CH2); 144.8 (1%%CH2); 160.0 (1CO). As cis/trans-
isomerism occurs at the CꢀN double bond of the hydra-
zone, C-2 (cis) and C-4 (trans) of the cyclobutane lose
their magnetic equivalence.
References
1. Dawsey, T. R.; McCormick, C. L. J. Macromol. Sci.-Rev.
Macromol. Chem. Phys. 1990, C30, 405–440.
2. Morgenstern, B.; Kammer, H. W. TRIP 1996, 4, 87–92.
3. Heinze, T. Prog. Polym. Sci. 2001, 26, 1689–1762.
4. Rahn, K.; Diamantaglou, M.; Klemm, D.; Berghmans,
H.; Heinze, T. Angew. Makromol. Chem. 1996, 238, 143–
163.
5. Kennedy, J. F.; Rivera, Z. S.; White, C. A.; Lloyd, L. L.;
Warner, F. P. Cellulose Chem. Technol. 1990, 24, 319–
325.
16. Barthel, J.; Krienke, H.; Kunz, W. Physical Chemistry of
Electrolyte Solutions; Steinkopff: Darmstadt, 1998.
17. Varela, L. M.; Rega, C.; Sarmiento, F. J. Chem. Phys.
1999, 110, 4483.
6. Striegel, A. M. Carb. Polym. 1997, 34, 267–274.
7. Schelosky, N.; Ro¨der, T.; Baldinger, T. Das Papier 1999,
53, 728–738.
18. Stewart, R. The Proton: Applications to Organic Chem-
istry; Academic Press: New York, 1985.
19. Zhang, W. M.; Collins, R.; Mahmood, K.; Dowd, P.
Tetrahedron Lett. 1995, 36, 2729–2732.
8. Morgenstern, B.; Berger, W. Acta Polymerica 1993, 44,
100.
9. Ro¨der, T.; Morgenstern, B.; Schelosky, N.; Glatter, O.
20. Bo¨hme, J.; Viehe, H. J. Iminium Salts in Organic Chem-
Polymer 2002, 42, 6765–6773.
10. Kvernheim, A. L.; Lystad, E. Acta Chem. Scand. 1989,
43, 209–211.
11. Ro¨hrling, J.; Potthast, A.; Rosenau, T.; Lange, T.; Ebner,
G.; Sixta, H.; Kosma, P. Biomacromolecules 2000, 3,
969–975.
istry Part I; Wiley: New York, 1976.
21. Arcelli, A.; Cecchi, R.; Porzi, G.; Rinaldi, S.; Sandri, S.
Tetrahedron 2001, 57, 6843–6846.
22. Penn, J. H.; Deng, D. L. Tetrahedron 1992, 48, 4823–
4830.