Isolation, modification, and NMR assignments of a series of cellulose oligomers

Article abstract of DOI:10.1021/ja990561u

A homologous series of cellulose oligomers from two to eight repeating subunits have been isolated and size-fractionated from the hydrolysis products of microcrystalline cellulose. Chemical modification of cellotriose (1), cellotetraose (2), cellopentaose (3), and cellohexaose (4) to the corresponding β-methyl glycosides 13-16 proceeded in three steps in overall yields of 16-46%. Peracetylation produced oligomers 5-8 in 70-75% yield, and subsequent formation of the β-methyl glycosides gave 9-12 in 42-89% yield. Removal of the acetate-protecting groups employing guanidine provided 13-16 in 73-79% yield. This modification eliminated anomeric equilibration and permitted a detailed NMR solution study of these oligomers. The complete ¹H and ¹³C chemical shift assignments of each peracetylated and deprotected oligomer were obtained through a combination of DQF-COSY, HMQC, HMBC, and HMQC-TOCSY experiments. All the resonances in methyl cellotriose (13) and methyl cellotetraose (14) were readily distinguishable from one another and directly assignable. Severe overlap of the resonances for the inner pyranose rings of methyl cellopentaose (15) and methyl cellohexaose (16) was observed and could only be resolved and assigned using a comprehensive battery of 3D pulse sequences. These results demonstrate the utility of multidimensional NMR experiments in assigning the signals from a repeating polysaccharide and represent the first necessary step in a comprehensive, systematic study of cellulose oligomers in solution.

Full text of DOI:10.1021/ja990561u

7228
J. Am. Chem. Soc. 1999, 121, 7228-7238
Isolation, Modification, and NMR Assignments of a Series of
Cellulose Oligomers
Lisa A. Flugge, Jarred T. Blank, and Peter A. Petillo*
Contribution from the Roger Adams Laboratory, Department of Chemistry, UniVersity of Illinois,
Urbana, Illinois 61801
ReceiVed February 22, 1999. ReVised Manuscript ReceiVed June 14, 1999
Abstract: A homologous series of cellulose oligomers from two to eight repeating subunits have been isolated
and size-fractionated from the hydrolysis products of microcrystalline cellulose. Chemical modification of
cellotriose (1), cellotetraose (2), cellopentaose (3), and cellohexaose (4) to the corresponding â-methyl glycosides
13-16 proceeded in three steps in overall yields of 16-46%. Peracetylation produced oligomers 5-8 in 70-
75% yield, and subsequent formation of the â-methyl glycosides gave 9-12 in 42-89% yield. Removal of
the acetate-protecting groups employing guanidine provided 13-16 in 73-79% yield. This modification
eliminated anomeric equilibration and permitted a detailed NMR solution study of these oligomers. The complete
1H and 13C chemical shift assignments of each peracetylated and deprotected oligomer were obtained through
a combination of DQF-COSY, HMQC, HMBC, and HMQC-TOCSY experiments. All the resonances in
methyl cellotriose (13) and methyl cellotetraose (14) were readily distinguishable from one another and directly
assignable. Severe overlap of the resonances for the inner pyranose rings of methyl cellopentaose (15) and
methyl cellohexaose (16) was observed and could only be resolved and assigned using a comprehensive battery
of 3D pulse sequences. These results demonstrate the utility of multidimensional NMR experiments in assigning
the signals from a repeating polysaccharide and represent the first necessary step in a comprehensive, systematic
study of cellulose oligomers in solution.
Introduction
simple repeating structures is typically limited to 3 or 4
residues.7 Consequently, most studies of carbohydrate dynamics
in solution have focused on dimers or trimers as models, which
may not provide a complete picture of the more biologically
relevant longer oligomers.8 Recently, advances in solution NMR
techniques and molecular modeling methods have progressed
to where it is now feasible to undertake the systematic study of
more complex carbohydrates including linear oligomers with
known biological activity.8,9
Of all the possible glycoforms available for a systematic
structural study, cellulose and its oligomers possess one of the
simplest repeating motifs. Composed solely of repeating â-1,4-
glucose monomers, this near-ubiquitous carbohydrate, by virtue
of its simple composition, may serve as a model for other more
complex carbohydrates in solution. Cellulose, found in most
plants and some algae and bacteria, is the most abundant
polymer found in nature6,10 and has found use in a large variety
of applications due to its solubility profiles, its rigidity, and its
inherent lack of toxicity.11 Cellulose and cellulose-derived
compounds are found in a diverse number of disparate applica-
tions including pharmaceuticals, food stuffs, cosmetics, textiles,
paper, and other material-based applications.6,10 Despite exten-
Carbohydrates represent an important class of biopolymers
and are possibly the least understood contributors to biological
structure and function.1 Polysaccharides play essential roles in
a myriad of biological processes such as cell-cell recognition,2
signaling,3 protein folding,4 embryogenesis and development,5
and cellular and extra-cellular structure.6 Despite the obvious
importance of polysaccharides, development of a comprehen-
sive, detailed structural description of carbohydrates in solution
is substantially less advanced compared to other important
biopolymers such as proteins and nucleic acids.1 Although a
more complete understanding of the structure and dynamics of
carbohydrates would provide insight into how these polymers
function in vivo, their spectral overlap and flexibility hinder
their comprehensive solution study by modern NMR methods.1
The assignment of H and 13C NMR chemical shifts for an
1
arbitrary carbohydrate remains a challenge, and unlike other
biopolymers, enriching the NMR active isotopes in an oligosac-
charide is a problem that has yet to be solved in a general sense.
Presently, the assignment of linear carbohydrate polymers with
* To whom correspondence should be addressed at 261 Roger Adams
Laboratory, Box 43, Department of Chemistry, University of Illinois, 600
South Matthews Avenue, Urbana, IL 61801. Tel: (217) 333-0695. Fax:
(217) 244-8559. E-mail: alchmist@alchmist.scs.uiuc.edu.
(1) Dwek, R. A. Chem. ReV. 1996, 96, 683-720.
(2) Drickhammer, K. J. Biol. Chem. 1988, 263, 9557-9560.
(3) (a) Sharon, N. Trends Biochem. Sci 1993, 18, 221-226. (b) Lasky,
L. A. Annu. ReV. Biochem. 1995, 64, 113-139.
(4) Hoffmann, D.; Florke, H. Folding Des. 1998, 3, 337-343.
(5) (a) Ivatt, R. J. In The Biology of Glycoproteins; Ivatt, R. J., Ed.;
Plenum: New York, 1984. (b) Mann, P. L.; Waterman, R. E. Acta Anat.
1998, 161, 153-161.
(6) Kennedy, J. F.; White, C. A. In Carbohydrate Chemistry; Kennedy,
J. F., Ed.; Clarenden, Oxford, U.K., 1988.
(7) (a) Toffanin, R.; Kvam, B. J.; Flaibani, A.; Atzori, M.; Biviano, F.;
Paoletti, S. Carbohydr. Res. 1993, 245, 113-128. (b) Masoud, H.; Perry,
M. B.; Brisson, J.-R.; Uhr´ın, D.; Richards, J. C. Can. J. Chem. 1994, 72,
1466-1477. (c) Taguchi, T.; Kitajima, K.; Muto, Y.; Yokoyama, S.; Inoue,
S.; Inoue, Y. Eur. J. Biochem. 1995, 228, 822-829.
(8) van Halbeek, H. Curr. Opin. Struct. Biol. 1994, 4, 697-709.
(9) Imberty, A. Curr. Opin. Struct. Biol. 1997, 7, 617-623.
(10) (a) Brown, R. M. J. Macromol. Sci., Chem. 1996, A33, 1345-1373.
(b) Brown, R. M.; Saxena, I. M.; Kudlicka, K. Trends Plant Sci. 1996, 1,
149-156.
(11) Hon, D. N.-S. In Polysaccharides in Medicinal Applications;
Dumitriu, S., Ed.; Marcel Dekker: New York, 1996; pp 87-106
10.1021/ja990561u CCC: $18.00 © 1999 American Chemical Society
Published on Web 07/21/1999

Cellulose Oligomers
J. Am. Chem. Soc., Vol. 121, No. 31, 1999 7229
sive ongoing research on the value-added uses of cellulose,
questions remain regarding the solution properties and dynamics
of the polymer. A clearer picture of cellulose in solution is a
necessary prerequisite for studies on cellulose chelation behavior
and would have a broad impact in the areas of carbohydrate
solution dynamics and protein-carbohydrate recognition events.
Remarkably, little is known about the structure of cellulose
oligomers in solution, although considerable information is
available on the solid-state properties of high-molecular-weight
cellulose. Solid-state characterizations by fiber X-ray diffrac-
tion,12 atomic force microscopy,13 solid-state NMR,14 IR
spectroscopy,15 Raman spectroscopy,16 and small-angle X-ray
scattering17 have been reported. These studies have elucidated
the gross morphology of the two main forms of cellulose fibers,
cellulose-I and cellulose-II, which differ in the orientation of
the individual polysaccharide chains with respect to one another.
In cellulose-I, the ribbonlike chains are laid down in staggered
parallel sheets and are held in place by interstrand hydrogen
bonding. Cellulose-II is a modified form of cellulose, which is
formed when cellulose-I is treated with base or when cellulose
derivatives are saponified. These cellulose chains are oriented
in antiparallel sheets that are no longer staggered.10 Due to the
insolubility of native cellulose in both forms and the lack of
signal dispersion in the proton spectra, NMR studies on cellulose
have generally been limited to solid-state 13C experiments.
Although these spectra have broad line widths that are useful
in determining the allomorphs present in the solid state, they
reveal little about specific interactions in solution.18
mers indicate that all C6-O6 bonds are predominately in one
conformation, resulting in a single intramolecular hydrogen bond
between OH3 and O5, and leaving all other hydroxyl groups
free to form intermolecular hydrogen bonds.
Despite the obvious importance of cellulose, even short
cellulose oligomers have defied a detailed description of their
behavior in aqueous media. High-resolution NMR spectroscopy
offers a powerful means to probe the properties of water-soluble
species, including carbohydrates composed of simple repeating
motifs.22 In the case of high-molecular-weight cellulose frag-
ments, insolubility has hindered NMR studies. Solubilizing
agents that disrupt the hydrogen-bonding network of high-
molecular-weight cellulose have been employed in some NMR
studies but have revealed little about cellulose in its native
state.18,23These studies sought to determine in which solvent
systems cellulose could be dissolved, what, if any, covalent
modification occurred in solution, and the degree of substitution
in cellulose after partial derivatization. The resultant spectra were
of poor quality, and broad line widths of at least 50 Hz were
typically observed. Additionally, some of the agents that
contained metal ions exhibited evidence of ion complexation
to the cellulose, which does not give a true picture of naturally
occurring cellulose.
Smaller water-soluble cellulose fragments have yet to be
studied systematically because of the inherent resonance overlap
problem that plagues most solution NMR investigations of
carbohydrates. Small oligomers of cellulose ranging from
cellobiose to celloheptaose are water-soluble, and details of their
solution structure can serve as important models for the structure
of linear oligosaccharides in aqueous media. A more complete
understanding of the structure of cellulose oligomers in solution
would also have a broad impact in a number of disciplines. For
example, the interaction of cellulose with exogenous species
such as the cellulose-binding proteins is of great interest.24
Small, water-soluble cellulose fragments could serve as models
for cellulose in NMR studies of cellulose-binding proteins such
as CBH1. Current studies with CBH1 and cellulose fragments
are plagued by a dual equilibria problem, where in addition to
the equilibrium of cellulose binding to CBH1, the cellulose is
also in anomeric equilibrium.25-33Additionally, the flexibility
and dynamics of the glycosidic linkage continues to be an area
of active research. The dynamic behavior of cellulose oligomers
The short cellulose oligomers cellotetraose,19 methyl â-cel-
lobiose,20 and methyl â-cellotriose,21 have been crystallized and
analyzed by X-ray diffraction, revealing additional details of
molecular structure which have been used to infer details of
the longer chains. Cellotetraose exists as a hemihydrate in the
crystalline state with packing analogous to that of cellulose-II.
The primary difference between the structure of high-molecular-
weight cellulose-II and the short oligomers is the conformation
about the C6-O6 bond. Diffraction studies on cellulose-II
suggest that the C6-O6 conformation alternates between two
different states.18 By contrast, diffraction studies on the oligo-
(12) French, A. D.; Roughead, W. A.; Miller, D. P. In The Structures of
Cellulose: Characterization of the Solid State; Atilla, R. H., Ed.; ACS
Symposium Series 350; American Chemical Society: Washington, DC,
1987; pp 15-37.
(13) (a) Kuutti, L.; Peltonen, J.; Pere, J.; Teleman, O. J. Microsc. 1995,
178, 1-6. (b) Baker, A. A.; Helbert, W.; Sugiyama, J.; Miles, M. J. J.
Struct. Biol. 1997, 119, 129.
(22) (a) Peters, T.; Pinto, B. M. Curr. Opin. Struct. Biol. 1996, 6, 710-
720. (b) Evans, J. N. S. Biomolecular NMR Spectroscopy; Oxford University
Press: Oxford, U.K., 1995.
(14) Nagavi, B. G.; Mithal, B. M.; Chawla, J. S. Tappi J. 1986, 69, 132-
(23) Nehls, I.; Wagenknecht, W.; Philipp, B.; Stscherbina, D. Prog.
Polym. Sci. 1994, 19, 29-78.
133.
(15) (a) Yang, C. Q. Text. Res. 1991, 61, 433-440. (b) Kataoka, Y.;
Kondo, T. Macromolecules 1998, 31, 760-64.
(16) (a) Atalla, R. H.; Ranua, J.; Malcom, E. W. Tappi J. 1984, 67, 96-
99. (b) Hamad, W. Y.; Eichhorn, S. J. Eng. Mater. Technol. 1997, 119,
309-313. (c) Liu, Y. L.; Kokot, S.; Sambi, T. J. Analyst 1998, 123, 633-
636.
(17) (a) Saito, M. Polym. J. 1983, 15, 213. (b) Furuta, T.; Yamahara,
E.; Konishi, T.; Ise, N. Macromolecules 1996, 29, 8994-8995. (c) Muller,
M.; Chzihak, C.; Vogl, G.; Fratzl, P.; Schober, H.; Rickel, C. Macromol-
ecules 1998, 31, 3953-3957.
(18) (a) Atilla, R. H., Ed. The Structures of Cellulose: Characterization
of the Solid State; ACS Symposium Series 350; American Chemical
Society: Washington, DC, 1987. (b) Kamide, K.; Saito, M. Macromol.
Symp. 1994, 83, 233-271.
(19) (a) Poppleton, B. J.; Mathieson, A. McL. Nature (London) 1968,
219, 1046-1048. (b) Gessler, K.; Krauss, N.; Steiner, T.; Betzel, C.;
Sandmann, C.; Saenger, W. Science 1994, 266, 1027-1029. (c) Gessler,
K.; Krauss, N.; Steiner, T.; Betzel, C.; Sarko, A.; Saenger, W. J. Am. Chem.
Soc. 1995, 117, 11397-11406.
(20) Ham, J. T.; Williams, D. G. Acta Crystallogr., Sect. B 1970, 26,
1373-1383.
(21) Raymond, S.; Henrissat, B.; Qui, D. T.; Kvick, A.; Chanzy, H.
Carbohydr. Res. 1995, 277, 209-229.
(24) (a) Be´guin, P.; Aubert, J.-P. FEMS Microbiol. ReV. 1994, 13, 25-
58. (b) Tomme, P.; Warren, A. J.; Miller, R. C.; Kilburn, D. G.; Gilkes, N.
R. In Enzymatic Degradation of Insoluble Carbohydrates; Saddler, J. N.,
Penner, M. H., Eds.; American Chemical Society: Washington, D.C., 1995.
(25) Reinikainen, T.; Ruohonen, L.; Nevanen, T.; Laaksonen, L.; Kraulis,
P.,; Jones, A.; Knowles, J. K. C.; Teeri, T. Proteins 1992, 14, 475-482.
(26) Poole, D. M.; Hazelwood, G. P.; Huskisson, N. S.; Virden, R.;
Gilbert, H. J. FEMS Microbiol. Lett. 1993, 106, 77-84.
(27) Din, N.; Forsythe, I. J.; Burtnick, L. D.; Gilked, N. R.; Miller, R.
C.; Warren, R. A. J.; Kilburn, D. G. Mol. Microbiol. 1994, 11, 747-755.
(28) Linder, M.; Mattinen, M.-L.; Kontteli, M.; Lindeberg, G.; Ståhlberg,
J.; Drakenberg, T.; Reinkainen, T.; Pettersson, G.; Annila, A. Protein Sci.
1995, 4, 1056-1064.
(29) Linder, M.; Lindeberg, G.; Reinkainen, T.; Teeri, T. Pettersson, G.
FEBS Lett. 1995, 372, 96-98.
(30) Johansson, G.; Ståhlberg, J.; Lindberg, G.; Engstro¨m, A° .; Pettersson,
G. FEBS Lett. 1989, 243, 389.
(31) Kraulis, P. J.; Clore, G. M.; Nilges, M.; Jones, T. A.; Pettersson,
G.; Knowles, J.; Gronenbern, A. M. Biochemistry 1989, 28, 7241-7257.
(32) Linder, M.; Terri, T. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 12251-
12255.
(33) Mattinen, M.-L.; Linder, M.; Teleman, A.; Annila, A. FEBS Lett.
1997, 407, 291-296.

7230 J. Am. Chem. Soc., Vol. 121, No. 31, 1999
Flugge et al.
Figure 1. Absorption of cellulose oligomers versus fraction number during chromatographic run: peak A, Glucose; peak B, Cellobiose; peak C,
Cellotriose; peak D, Cellotetraose; peak E, Cellopentaose; peak F, Cellohexaose; peak G, Celloheptaose.
in solution would serve to address this issue. Finally, cellulose
can complex metal ions, but the conformational effects of this
are largely unknown. The prerequisite for all of these afore-
mentioned studies is direct knowledge of the 1H and 13C
chemical shifts of each center in each oligomer to be studied.
The systematic study of cellulose oligomers presents several
obstacles: the availability of the oligomers, the anomeric
equilibrium of the reducing end that occurs in aqueous solution,
and the inherent difficulty of NMR on carbohydrates.34,35
Cellulose itself can be obtained from natural sources or can be
synthesized chemically or biosynthetically.36 However, there is
presently no direct means to control polydispersity. Cellulose
oligomers of uniform length must either be made synthetically,
a complicated, time-consuming procedure,37 or be isolated as
the hydrolysis products of high-molecular-weight cellulose.
This report describes the isolation of cellulose oligomers from
the acid-catalyzed hydrolysis of high-molecular-weight cellulose,
the chemical modification of these fragments to eliminate their
the peracetylated intermediates have also been determined by
a battery of NMR experiments including 2D experiments such
as COSY and HMQC, and 3D HMQC-COSY and HMQC-
TOCSY experiments. This represents the initial step toward a
comprehensive study of the structure and dynamics of cellulose
in solution.
Results and Discussion
Isolation. Isolation of the oligomers from the acid-catalyzed
hydrolysis of microcrystalline cellulose was achieved using the
column chromatography method of Miller.38 This proved
effective for obtaining near-gram quantities of cellulose oligo-
mers, representing a significantly higher yield than other
reported methods.39 Over a five-day period, oligomers ranging
from glucose to celloheptaose were separated (Figure 1). This
method yielded oligosaccharides ranging from 700 mg of
cellotriose to 200 mg of celloheptaose from 10 g of high-
molecular-weight cellulose. Previous work on the solubility of
cellulose oligosaccharides has shown that celloheptaose repre-
sents the largest water-soluble oligocellulose that can be isolated
by chromatographic separation.40,41
1
anomeric equilibration, and a complete assignment of all H
and 13C resonances for each fragment. The NMR resonance
assignments of the modified â-methyl cellulose oligomers and
(34) Heyraud, A.; Rinaudo, M.; Vignon, M. Biopolymers 1979, 18, 167-
185.
(35) Gast, J. C.; Atalla, R. H.; McKelvey, R. D. Carbohydr. Res. 1980,
84, 137-146.
(36) (a) Lee, J. H.; Brown, R. M.; Kuga, S. Proc. Natl. Acad. Sci. U.S.A.
1994, 91, 7425-7429. (b) Kudlicka, K.; Brown, R. M.; Li, L. K.; Shin, H.;
Kuga, S. Plant Physiol. 1995, 1995, 111-123. (c) Kudlicka, K.; Lee, J.
H.; Brown, R. M. Am. J. Bot. 1996, 83, 274-284. (d) Nakatsubo, F.;
Kamitakahara, H.; Hori, M. J. Am. Chem. Soc. 1996, 118, 1677-1681.
(37) Takeo, K.; Yasato, T.; Kuge, T. Carbohydr. Res. 1981, 93, 148-
156.
(38) (a) Miller, G. L. Anal. Biochem. 1960, 2, 133-140. (b) Miller, G.
L.; Dean, J.; Blum, R. Arch. Biochem. Biophys. 1960, 91, 21-26.
(39) (a) Gum, E. K.; Brown, R. D. Anal. Biochem. 1977, 82, 372-375.
(b) El Rassi, Z.; Mechref, Y. Electrophoresis 1996, 17, 275-301. (c)
Nguyen, D. T.; Lerch, H.; Zemann, A.; Bonn, G. Chromatographia 1997,
46, 113-121. (d) Guttman, A.; Ulfelder, K. W. J. Chromatogr., A 1997,
781, 547-554. (e) El Rassi, Z.; Postlewait, J.; Mechref, Y.; Ostrander, G.
K. Anal. Biochem. 1997, 244, 283-290.
(40) Wolfrom, M. L, Dacons, J. C. J. Am. Chem. Soc. 1952, 74, 5331-
5333.

Cellulose Oligomers
J. Am. Chem. Soc., Vol. 121, No. 31, 1999 7231
Scheme 1
Modification of the Isolated Oligomers. In aqueous solution
reducing carbohydrates typically equilibrate between the R and
â anomers.10 The existence of both anomers further complicates
the already severe spectral overlap of these oligomers. Modi-
fication to the corresponding â-methyl glycosides eliminates
this process by producing a single anomerically pure compound.
Furthermore, the â-methyl glycosides 13-16 would correspond
more generally to the structure of high-molecular-weight
cellulose than the corresponding R-anomers. Modification of
the cellulose oligomers was achieved through a four-step
reaction sequence to methylate the reducing end of cellotriose
(1), cellotetraose (2), cellopentaose (3), and cellohexaose (4)
(Scheme 1). These â-methyl oligomers 13-16 have previously
been synthesized by enzymatic means to afford, starting from
methyl â-cellobioside, a series of oligomers up to methyl
â-cellopentaoside.42 The yields for the longer oligomers were
poor (14% 13, 36% 14, 6% 15), and cellohexaose was not
isolated. The chemical modification reported herein is moderate-
yielding (from 16% to 46% overall yield in 4 steps) and should
be applicable to other carbohydrate oligomers. Celloheptaose
and cellooctaose, although available from our separation, were
not modified as they are sparingly soluble in aqueous solution,
which would hinder NMR studies in the absence of isotopic
enrichment. Additionally, on the basis of our assignment data
for 13-16, we believe that the severe spectral overlap of these
longer oligomers would make their complete assignments
impossible.
Formation of the glycosyl bromides proceeded smoothly using
30% HBr in glacial acetic acid.44 This reaction proceeded faster
with the higher oligomers; 5 reacted in 1 h, while the reaction
was complete in 30 min for 6, 20 min for 7, and 10 min for 8.
We believe this is due in part to the differing solubilities of the
two anomers present in the reaction mixture, which becomes
less important with increasing oligomer size. The â anomer,
being more polar, is more likely to dissolve in the highly polar
glacial acetic acid than the less polar a anomer. Thus, 5 with a
1:1 anomeric ratio reacts relatively slowly compared to the other
oligomers with higher â anomer content and greater molecular
weight. The reaction proceeds quantitatively, and the product
was used without further purification. Methylation was per-
formed under standard Koenigs-Knorr conditions using methanol
and methylene chloride in the presence of CaSO4‚1/2H2O, HgO,
and HgBr2.44 On average the reactions were completed in 3 days.
The resulting products were purified by flash column chroma-
tography on silica gel when necessary. Column chromatography
resulted in product decomposition and much lower yields. Crude
yields ranged from 89% for 9 to 42% for 12. Column
chromatography decreased these yields to approximately 20%.
Where chromatography was not performed, the products were
carried on and purified following the subsequent step.
Glycoside 9 was deacetylated under Zemple´n conditions using
methanol and NaOMe.44 These conditions were found to be
ineffective for the higher oligomers due to their insolubility in
methanol. Deprotection of the higher oligomers was instead
accomplished using guanidine in ethanol and methylene chlo-
ride45 and purified by either recrystallization or trituration in
ethanol, yielding 73-79% of the deprotected products.
NMR Assignments. Complete NMR assignment of unpro-
tected carbohydrates can be difficult due to severe overlap of
1H and 13C resonances. Cellulose oligomers, by nature of their
simple repeating structure, have an especially severe spectral
overlap problem. The bulk of 1H resonances occur between 3.0
and 4.5 ppm, while most of the corresponding 13C resonances
occur between 55.0 and 80.0 ppm. The use of modern high field
The cellulose oligomers were peracetylated with acetic
anhydride in the presence of pyridine and DMAP.43 Reaction
times ranged from 12 h for 1 to 3-4 days for the longer
oligomers. We hypothesize that the longer reaction times are a
function of the additional hydroxyl groups and the reduced initial
solubility of the longer oligomers. Acetylation products were
predominately â, with anomeric ratios ranging from 1:1 â/R
(5) to 4:1 â/R (8) by NMR. The anomers were not separated,
as the mixture could be elaborated to the corresponding glycosyl
bromides.
(41) Longer oligomers have been separated. See the following: (a) Wells,
G. B.; Lester, R. L. Anal. Biochem. 1979, 97, 184-190. (b) White, C. A.;
Corran, P. H.; Kennedy, J. F. Carbohydr. Res. 1980, 87, 165-173.
(42) Samain, E.; Lancelon-Pin, C.; Fe´rigo, F.; Moreau, V.; Chanzy, H.;
Heyraud, A.; Driguez, H. Carbohydr. Res. 1995, 271, 217-226.
(43) Bonner, W. A. J. Am. Chem. Soc. 1958, 80, 3376.
(44) (a) Vogel, A. I. Vogel’s Textbook of Practical Organic Chemistry,
5th ed.; Longman Scientific and Technical, New York, 1989. (b) Wulff,
G.; Ro¨hle, G. Angew. Chem., Int. Ed. Engl. 1974, 13, 157-216. (c) Paulsen,
H. Angew. Chem., Int. Ed. Engl. 1982, 21, 155-224.
(45) Kunesch, N.; Miet, C.; Poisson, J. Tetrahedron Lett. 1987, 28, 3569-
3572.

7232 J. Am. Chem. Soc., Vol. 121, No. 31, 1999
Flugge et al.
Table 1. Proton and Carbon Assignments for 13
spectrometers coupled with multidimensional pulse sequences
can disperse these resonances and allow for the complete
assignment of proton and carbon signals.46
In this work, most proton spectra were recorded at a resonance
frequency of 750 MHz and 13C spectra were recorded at a
resonance frequency of 188 MHz. A range of multidimensional
spectra were employed to aid in the resonance assignments,
including DQFCOSY,47 HMQCPS,48 COSYPS,49 and HMBC50
and the 3D experiments HMQC-COSY and HMQC-TOCSY.51
Assignments were made for the â-methyl peracetylated oligo-
mers 9-12 as well as the â-methyl oligocelluloses 13-16.
Assignment of Protected Oligomers. The proton and carbon
resonances of the fully protected oligomers 9-12 were assigned
using a combination of COSY, HMQC, and HMBC experi-
ments. Methyl cellotriose decaacetate (9) was assigned using
spectra taken at a proton resonance frequency of 500 MHz, while
the longer oligomers were assigned with data taken at a proton
resonance frequency of 750 MHz. Coupling networks were
elucidated using COSY data. One-bond 13C-1H correlations were
obtained from HMQC data, and the position of the glycosidic
linkages was determined using HMBC data.
The â-methyl peracetylated oligomers (9-12) exhibited well-
resolved signals, with overlap appearing in 11 and becoming
problematic in 12. As expected, the acetate-protecting groups
caused chemical shift dispersion of the resonances relative to
the free oligomer and consequently the peracetylated oligomers
were much easier to completely assign. In each oligomer all of
the resonances for each ring position clustered together, with
the H5s resonating near 3.5 ppm, the H4s near 3.7 ppm, one
set of H6s at 4.0 ppm, the H1s and remaining H6s together at
4.5 ppm, the H2s around 4.8 ppm, and the H3s near 5.0 ppm.
The exception to this trend is the H4 of the terminal ring that
is not part of a glycosidic linkage. This resonance occurs with
the H3s near 5.0 ppm. In all of the oligomers, the ring next to
the terminal ring on the nonreducing end (n - 1 ring) exhibits
the most upfield resonances at the majority of the ring positions.
The carbon resonances also group together by ring position.
The C6s occur the most upfield at around 62 ppm, with the
C2s at 71.8 ppm, the C3s at 72.8 ppm, the C5s at 72.9 ppm,
the C4s around 76 ppm, and the C1s around 100 ppm. Again,
the exception is the C4 on the n - 1 ring, occurring at 67.8
ppm in all of the oligomers. The carbon resonances show no
general trend as seen in the proton resonances where one ring
is consistently the most upfield or downfield. In most of the
positions the order in which the resonances occur is the same
from oligomer to oligomer.
1H
multiplicity
13C
ring A
H1
H2
H3
H4
H5
H6
H6′
ring B
H1
H2
H3
H4
H5
4.403
3.305
3.633
3.628
3.597
3.818
3.993
d J ) 8.0
dd J ) 9.4, 7.6
102.94
72.99
74.13
78.36
74.62
59.81
m
m
ddd J ) 10.0, 5.1, 2.2
dd J ) 12.6, 5.4
dd J ) 12.3, 2.2
4.533
3.361
3.648
3.677
3.618
3.826
3.980
d J ) 8.0
dd J ) 9.3, 8.0
102.21
72.79
73.90
78.21
74.67
59.69
m
t J ) 8.8
ddd J ) 9.7, 5.2, 2.2
dd J ) 12.6, 5.4
dd J ) 12.3, 2.3
H6
H6′
ring C
H1
H2
H3
H4
H5
H6
H6′
4.509
3.315
3.507
3.418
3.487
3.736
3.917
d J ) 7.9
dd J ) 9.5, 7.9
t J ) 9.3
dd J ) 9.9, 9.1
ddd J ) 9.8, 5.9, 2.3
dd J ) 12.6, 6.0
dd J ) 12.4, 2.3
102.43
72.74
75.32
69.29
75.82
60.42
workers reported the 13C chemical shifts for a series of
R-peracetylated cellulose oligomers. These studies were per-
formed at a proton resonance frequency of 100 MHz, which
proved inadequate to assign the proton resonances now available
from this study. Although the carbon resonances were assigned,
the resonances of the interior residues all exhibited overlap.53
Assignment of Deprotected Oligomers. The deprotected
oligomers 13-16 were assigned using the same strategy as for
9-12, with the addition of HMQC-COSY and HMQC-
TOCSY data to assign the more crowded regions of the 1H and
13C spectra. As expected the chemical shift range of the
unprotected oligomers was substantially smaller compared to
their peracetylated counterparts. The optimal mixing times for
the 3D experiments were determined by a series of TOCSY
experiments utilizing mixing times of 5, 10, 20, 30, 40, 50, 60,
70, 80, 90, and 100 ms. These data were used to determine the
mixing times for TOCSY- and COSY-type data, which were
slightly different for each oligomer. Methyl cellohexaose was
determined to require only TOCSY-type data, as the signals
could be assigned by analogy to the shorter oligomers. COSY,
HMQC, and HMBC experiments were used to assign as much
of the backbone as possible and, in the case of methyl cellotriose
(13), were sufficient to completely assign the molecule. (Table
1) Higher oligomers could not be assigned directly and required
the extra dispersion only possible with a 3D experiment.
The assignment strategy for methyl cellotetraose (14) started
with assigning as many resonances as possible from the 2D data
sets. This information was then used in the interpretation of
the 3D data. The starting point for assigning the rings using
HMQC-TOCSY and HMQC-COSY data was the anomeric
Previous work has been performed on the fully peracetylated
oligomers with no terminal methyl group.52 Most of the NMR
experiments were done to determine the degree of acetate
substitution, and not NMR assignments. Thomson and co-
(46) (a) Chan, T.-M.; Osterman, R. M.; Morton, J. B.; Ganguly, A. K.
Magn. Reson. Chem. 1997, 35, 529-532. (b) Over, D. E.; Bardet, M.;
Marchon, J.-C.; Ramasseul, R. Magn. Reson. Chem. 1995, 33, 224-227.
(c) Huckerby, T. N.; Brown, G. M.; Nieduszynski, I. A. Eur. J. Biochem.
1995, 231, 779-783. (d) Reddy, G. P.; Chang, C.-C.; Bush, C. A. Anal.
Chem. 1993, 65, 913-921.
(47) Piantini, U.; Sørenson, O. W.; Ernst, R. R. J. Am. Chem. Soc. 1982,
104, 6800-6801.
(48) (a) Mueller, L. J. Am. Chem. Soc. 1979, 101, 4481. (b) Bax, A.;
Griffey, R. G.; Hawkins, B. L. J. Am. Chem. Soc. 1983, 105, 7188. (c)
Bax, A.; Subramanian, S. J. Magn. Reson. 1986, 67, 565.
(49) Marion, D.; Wuthrich, K. Biochem. Biophys. Res. Commun. 1983,
113, 967-974.
(50) Bax, A.; Summers, M. J. Am. Chem. Soc. 1986, 108, 2093.
(51) Lerner, L.; Bax, A. J. Magn. Reson. 1986, 69, 375.
(52) (a) Goodlett, V. W.; Dougherty, J. T.; Patton, H. W. J. Polym. Sci.
Part A: Polym. Chem. 1971, 9, 155-161. (b) Kamide, K.; Okajima, O.
Polymer J. 1981, 13, 127-133.
(53) Capon, B.; Rycroft, D. S.; Thomson, J. W. Carbohydr. Res. 1979,
70, 145-149.

Cellulose Oligomers
J. Am. Chem. Soc., Vol. 121, No. 31, 1999 7233
Table 2. Proton and Carbon Assignments for 14
1H
multiplicity
13C
1H
multiplicity
13C
ring A
H1
H2
H3
H4
H5
H6
H6′
ring C
H1
H2
H3
H4
H5
H6
H6′
4.404
3.304
3.636
3.634
3.597
3.818
3.991
d J ) 7.97
102.92
72.99
74.12
78.36
74.62
59.82
4.532
3.361
3.646
3.667
3.611
3.827
d J ) 8.08
102.18 or 102.21
72.77 or 72.79
73.85
78.21
74.66
t J ) 8.03
t J ) 8.55
m
m
m
m
m
m
dd J ) 12.03, 6.40
dd J ) 12.15, 2.01
dd J ) 12.22, 5.45
dd J ) 12.28, 1.93 or
dd J ) 12.32, 1.79
59.70 or 59.72
3.977 or 3.980
ring B
H1
H2
H3
H4
H5
H6
H6′
ring D
H1
H2
H3
H4
H5
H6
H6′
4.532
3.361
3.642
3.680
3.616
3.827
d J ) 8.08
t J ) 8.55
m
m
m
102.18 or 102.21
72.77 or 72.79
73.85
78.09
74.66
4.506
3.316
3.509
3.417
3.488
3.736
3.915
d J ) 7.83
102.40
72.73
75.32
69.30
75.83
60.41
t J ) 8.73
t J ) 9.25
t J ) 9.47
ddd J ) 9.99, 5.92, 2.15
dd J ) 12.36, 5.88
dd J ) 12.41, 2.16
dd J ) 12.22, 5.45
59.70 or 59.72
3.977 or dd J ) 12.28, 1.93 or
3.980
dd J ) 12.32, 1.79
Table 3. Proton and Carbon Assignments for 15
1H
multiplicity
13C
1H
multiplicity
13C
ring A
H1
H2
H3
H4
H5
H6
H6′
ring B
H1
H2
H3
H4
H5
ring D
H1
H2
3
H4
H5
H6
H6′
ring E
H1
H2
H3
H4
H5
H6
H6′
4.403
3.304
3.64
3.63
3.59
3.820
3.991
d J ) 8.01
102.87
72.97
74.06
78.28
74.59
4.533
3.360
3.65
3.69
3.62
3.829
3.977
d J ) 7.95
102.13
72.75
73.80
78.16
74.64
59.67
dd J ) 9.53, 8.29
t J ) 8.55
m
m
m
m
m
m
dd J ) 12.20, 6.53
dd J ) 12.26, 2.21
59.75
dd ) 12.06, 5.41
br d J ) 11.92
4.533
3.360
3.65
3.69
3.62
3.829
3.977
d J ) 7.95
t J ) 8.55
m
m
m
dd ) 12.06, 5.41
br d J ) 11.92
102.13
72.75
73.80
78.04
74.64
59.67
4.510
3.315
3.507
3.416
3.486
3.737
3.916
d J ) 7.95
102.35
72.70
75.27
69.23
75.83
60.35
dd J ) 9.47, 8.03
t J ) 9.25
t J ) 9.49
ddd J ) 9.70, 5.70, 2.46
dd J ) 12.42, 5.84
dd J ) 12.37, 2.25
H6
H6′
Ring C
H1
H2
H3
H4
H5
H6
H6′
4.533
3.360
3.65
3.69
3.62
3.829
3.977
d J ) 7.95
t J ) 8.55
m
m
m
dd ) 12.06, 5.41
br d J ) 11.92
102.13
72.75
73.80
78.04
74.64
59.67
protons, enabling assignment of each ring. (Table 2) Methyl
cellopentaose (15) and methyl cellohexaose (16) were assigned
in a similar fashion. (Tables 3 and 4)
The proton and carbon assignments were accomplished for
the oligomers through methyl cellohexaose (16). Most of the
protons and carbons for 13 and 14 were distinguishable as
separate signals. By contrast, the inner ring protons and carbons
of 15 and 16 exhibited overlap. The exception to this trend was
the signal for C4 of the n - 1 ring. This signal was clearly
separate in all oligomers. Presumably, proton and carbon signals
for the inner rings of longer oligomers such as celloheptaose
would simply overlap with the other interior ring signals and

7234 J. Am. Chem. Soc., Vol. 121, No. 31, 1999
Table 4. Proton and Carbon Assignments for 16
Flugge et al.
1H
multiplicity
13C
1H
multiplicity
13C
ring A
H1
H2
H3
H4
H5
H6
H6′
ring B
H1
H2
H3
H4
H5
ring D
H1
H2
H3
H4
H5
H6
H6′
ring E
H1
H2
H3
H4
H5
4.405
3.304
3.64
3.63
3.59
3.819
3.991
d J ) 8.1
102.91
72.97
74.10
78.32
74.60
59.70
4.533
3.358
3.65
3.69
3.62
3.828
3.976
d J ) 8.0
102.18
72.75
73.82
78.05
74.65
59.65
dd J ) 9.53, 8.29
br t J ) 8.5
m
m
m
m
m
m
dd J ) 12.3, 5.1
dd J ) 11.9, 2.0
dd J ) 12.0, 5.0
br d J ) 11.5
4.533
3.358
3.65
3.69
3.62
3.828
3.976
d J ) 8.0
br t J ) 8.5
m
m
m
dd J ) 12.0, 5.0
br d J ) 11.5
102.18
72.75
73.82
78.05
74.65
59.65
4.533
3.358
3.65
3.69
3.62
3.828
3.976
d J ) 8.0
br t J ) 8.5
m
m
m
dd J ) 12.0, 5.0
br d J ) 11.5
102.18
72.75
73.82
78.18
74.65
59.65
H6
H6
H6′
ring C
H1
H2
H3
H4
H5
H6
H6′
H6′
ring F
H1
H2
H3
H4
H5
H6
H6′
4.533
3.358
3.65
3.69
3.62
3.828
3.976
d J ) 8.0
br t J ) 8.5
m
m
m
dd J ) 12.0, 5.0
br d J ) 11.5
102.18
72.75
73.82
78.05
74.65
59.65
4.509
3.313
3.506
3.416
3.486
3.736
3.915
d J ) 7.8
102.39
72.71
75.29
69.27
75.81
60.38
dd J ) 9.4, 8.0
t J ) 9.3
t J ) 9.5
ddd J ) 9.6, 5.7, 2.4
dd J ) 12.5, 5.9
dd J ) 12.6, 2.1
yield no additional information. All of the rings in the oligomers
except the terminal, nonreducing ring exhibited roughly similar
chemical shifts, with the H2s the most upfield near 3.4 ppm,
the H5s at 3.7 ppm, the H4s and H3s near 3.8 ppm, the H6s at
3.9 and 4.0 ppm, and the H1s at 4.4 ppm. The signals from the
interior rings are the most downfield signal for most ring
positions. The carbon resonances exhibit this same clustering,
with the same order as in the peracetylated oligomers, where
C6 is the most upfield, followed by C2, C3, C5, C4, and C1.
However, the nonreducing terminal ring signals do not cluster
with the signals for the other rings. Presumably, this is because
it is the only ring without a glycosidic linkage at C4. There is
no general trend for the carbon signals as there is for the proton
signals, where the signals for one ring are noticeably up- or
downfield from the others, but the order of signals for each
position in any given ring remains the same through the
oligomer series. In addition, the proton and carbon chemical
shifts did not markedly change with the size of the oligomer
changed, indicating no dependence on oligomer length.
A limited amount of work has previously been done on the
unmodified cellulose oligomers. The assignments for the oli-
gomers 13-16 are similar to those for cellulose in NaOH/D2O23
and for cellohexaose (4),34 although more peaks are discernible
in the methyl cellohexaose (16). In addition, the methyl group
shifts C1 downfield relative to its position in the unprotected
cellulose oligomers. Gast and co-workers performed experiments
on cellobiose, cellotriose (1), and cellotetraose (2).35 As their
experiments were performed at a proton resonance frequency
of 100 MHz, little or no detail from the proton spectra could
be obtained from their data. The carbon spectra only differenti-
ated between terminal and interior residue signals. However,
their assignments are comparable to ours, although they seem
to be consistently 2 ppm downfield from the current findings.
Heyraud and co-workers performed experiments on a series of
cellulose oligomers up to cellohexaose (4) at a proton resonance
frequency of 250 MHz.34 They did not attempt proton assign-
ments but were able to differentiate some of the signals in the
carbon spectra; for instance, in cellotetraose all four C4 signals
were different. They did identify the different signal for the C4
on the n - 1 ring but were unable to assign it to the n - 1 ring.
However, most signals overlap, even in cellotetraose (2). Again,
their results were comparable to the current results, although
their assignments were consistently 3 ppm downfield from those
in this study.
Conclusion
The isolation, modification, and NMR assignments for a series
of cellulose oligomers have been completed. The modification
procedure is a simple way of obtaining anomerically pure
carbohydrates that are useful NMR substrates for many ap-
plications, including the study of protein-carbohydrate interac-
tions and of metal chelates of carbohydrates. The NMR
assignments represent the most complete NMR characterization
to date of cellulose oligomers. This provides another method
to characterize cellulose and its derivatives in addition to the
methods currently available and enables the eventual elucidation
of the solution dynamics of the cellulose oligomers. In addition,
this represents an important step toward understanding carbo-
hydrate solution behavior.
Experimental Section
General Information. Flash column chromatography following the
method of Still employed EM Science silica gel 60 (230-400 mesh).
Analytical thin-layer chromatography was performed with 0.25 mm
coated commercial silica gel plates (E. Merck, DC-fertigplatten
Kieselgel F254 or the corresponding Altech plate). Melting points were

Cellulose Oligomers
J. Am. Chem. Soc., Vol. 121, No. 31, 1999 7235
obtained on a Buchi melting point apparatus in an open capillary tube
and are uncorrected. High-resolution EI and FAB mass spectral data
were obtained on either a 70-VSE or 70-SE-4F mass spectrometer. FAB
spectra were obtained using either Na° or K° as the ionization source
(as indicated).
calculated ) 527.159, found ) 527.159 (HRFABMS, Na° source).
Cellotetraose (C24H43O21): mass calculated ) 667.230, found ) 667.230
(HRFABMS, M+1). Cellopentaose (C30H53O26): mass calculated )
829.283, found ) 829.283. (HRFABMS, M + 1). Cellohexaose
(C36H63O31): mass calculated ) 991.335, found ) 991.335. (HR-
FABMS, M + 1). Celloheptaose (C42H72O36Na): mass calculated )
1176.0, found ) 1175.3. (LRFABMS, Na° source).
NMR Experiments. All spectral data were obtained on either a
Varian Unity-Plus spectrometer (1H resonance frequency of 400 MHz),
a Varian Unity-Plus spectrometer (1H resonance frequency of 500
MHz), a Varian INOVA spectrometer (1H resonance frequency of 500
MHz), or a Varian INOVA spectrometer (1H resonance frequency of
750 MHz). Proton (1H) chemical shifts are reported in delta (δ) units,
parts per million (ppm) downfield from tetramethylsilane (TMS).
Splitting patterns are designated as s, singlet; d, doublet; t, triplet; q,
quartet; quin, quintet; sex, sextet; m, multiplet; and br, broad. Coupling
constants are reported in Hertz (Hz). 13C NMR chemical shifts are
reported in parts per million relative to the center line at 77.23 ppm of
the deuteriochloroform triplet. Spectra taken in D2O are referenced to
an internal benzene standard at 7.44 ppm. All spectra were acquired at
20 °C. All experiments were acquired with standard pulse sequences
supplied with the software. The 3D data were processed with the Varian
processing program VNMR 6.2. Two different mixing times were used
in some of the HMQC-TOCSY experiments to give both TOCSY and
COSY data. The number of data points taken for the 3D HMQC-
TOCSY of methyl cellotetraose, with a mixing time of 100 ms, and
with 4 scans per increment was 1024 × 48 × 24, which was linear
predicted to 1024 × 128 × 128 and zero-filled to 2048 × 512 × 512.
The data were processed with line broadening (lb ) 0.82, lb1 ) -13.6,
lb2 ) 15.6), Gaussian multiplication and shift (gf ) 0.35, gfs )
-0.0015, gf1 ) 0.01, gfs1 ) -0.0006, gf2 ) 0.16), and sine bell
multiplication and shift (sb ) 0.56, sbs ) -0.21, sb2 ) 0.18, sbs2 )
-0.02) as determined by the interactive weighting routine. The 3D
HMQC-COSY of methyl cellotetraose was acquired with a mixing
time of 10 ms and with 2 scans per increment and had 1024 × 48 ×
16 data points, subsequently linear predicted to 1024 × 128 × 128
data points and zero-filled to 2048 × 512 × 512 data points. The data
were processed with line broadening (lb ) 0.84, lb1 ) -49.4, lb2 )
2.3), Gaussian multiplication (gf ) 0.38, gf1 ) 0.009, gf2 ) 0.18),
and sine bell multiplication and shift (sb ) 0.49, sbs ) -0.21) as
determined by the interactive weighting routine. The HMQC-TOCSY
of methyl cellopentaose was acquired with mixing times of 60 ms and
10 ms and initial dimensions of 1024 × 64 × 24. The data were linear
predicted to 1024 × 192 × 128 and zero-filled to 2028 × 512 × 512
data points. The data were processed with line broadening (lb ) 1.3,
lb1 ) -20, lb2 ) 17.5), Gaussian multiplication and shift (gf ) 0.24,
gfs ) 0.042, gf1 ) 0.006, gfs1 ) 0.0003, gf2 ) 0.05, gfs2 ) 0.04),
and sine bell multiplication and shift (sb ) 0.35, sbs ) -0.12, sb1 )
0.009, sbs1 ) -0.009) as determined by the interactive weighting
routine. The HMQC-TOCSY of methyl cellohexaose was taken with
a mixing time of 80 ms and a data set size of 1024 × 64 × 24. The
data were linear predicted to 1024 × 192 × 48 and zero-filled to 2048
× 512 × 512 data points. The data were processed with line broadening
(lb ) 3, lb1 ) -20, lb2 ) 5), Gaussian multiplication and shift (gf )
0.20, gf1 ) 0.0059, gfs1 ) 0.0002, gf2 ) 0.013), and sine bell
multiplication and shift (sb ) 0.20, sbs ) -0.047, sb1 ) 0.0096, sbs1
) -0.009).
Oligomer Modification. (A) Cellotriose Undecaacetate (5). Cel-
lotriose (1) (73.0 mg, 0.145 mmol, 1 equiv) and DMAP (1 mg, 0.008
mmol, 0.06 equiv) were placed in a flask which was purged with N2,
and pyridine (3.0 mL, 37 mmol, 260 equiv) and acetic anhydride (0.5
mL, 5.3 mmol, 37 equiv) were added via syringe. The solution was
stirred 19 h at room temperature, then transferred to a flask containing
15 mL of ice water slurry and stirred for an additional 90 min. The
mixture was extracted with CHCl3 (5 × 10 mL). The organic layers
were combined and washed with saturated CuSO4 (4 × 10 mL),
saturated NaHCO3 (4 × 10 mL), and brine (10 mL). The organic layer
was then dried (MgSO4) and concentrated under reduced pressure. The
residue was dried to yield 98.8 mg (70%) of an off-white solid that
was a 1:1 mixture the R and â anomers of 5 by NMR: mp 105-109
1
°C. H NMR (500 MHz, CDCl3) 1.947 (s, 3H), 1.949 (s, 3H), 1.957
(s, 3H), 1.962 (s, 3H), 1.964 (s, 3H), 1.969 (s, 6H), 1.975 (s, 3H),
1.976 (s, 3H), 1.988 (s, 3H), 1.990 (s, 3H), 2.001 (s, 3H), 2.003 (s,
6H), 2.055 (s, 3H), 2.059 (s, 6H), 2.092 (s, 3H), 2.103 (s, 3H), 2.113
(s, 3H), 2.115 (s, 3H), 2.139 (s, 3H), 3.565 (ddd, J ) 9.8, 5.2, 2.4 Hz,
1H), 3.570 (ddd, J ) 9.8, 5.0, 2.6 Hz, 1H), 3.608 (ddd, J ) 9.9, 4.4,
2.3 Hz, 1H), 3.610 (ddd, J ) 9.9, 4.4, 2.3 Hz, 1H), 3.712 (ddd, J )
10.0, 4.6, 2.1 Hz, 1H), 3.737-3.793 (m, 4H), 3.950 (ddd, J ) 10.1,
4.1, 2.1 Hz, 1H), 4.004 (dd, J ) 12.5, 2.4 Hz, 1H), 4.009 (dd, J )
12.5, 2.5 Hz, 1H), 4.062 (dd, J ) 12.4, 4.3 Hz, 1H), 4.066 (dd, J )
12.2, 4.6 Hz, 1H), 4.080 (dd, J ) 12.2, 5.2 Hz, 1H), 4.096 (dd, J )
12.3, 5.1 Hz, 1H), 4.319 (dd, J ) 12.5, 4.4 Hz, 1H), 4.324 (dd, J )
12.5, 4.4 Hz, 1H), 4.369 (dd, J ) 12.1, 2.1 Hz, 1H), 4.375 (dd, J )
12.1, 2.0 Hz, 1H), 4.446 (d, J ) 7.9 Hz, 1H), 4.452 (d, J ) 7.9 Hz,
1H), 4.454 (dd, J ) 12.1, 2.4 Hz, 2H), 4.458 (d, J ) 7.9 Hz, 1H),
4.459 (d, J ) 7.9 Hz, 1H), 4.815 (dd, J ) 9.4, 7.9 Hz, 1H), 4.841 (dd,
J ) 9.4, 7.9 Hz, 1H), 4.868 (dd, J ) 9.3, 7.9 Hz, 1H), 4.869 (dd, J )
9.2, 8.0 Hz, 1H), 4.962 (dd, J ) 10.2, 3.4 Hz, 1H), 5.002 (t, J ) 9.0
Hz, 1H), 5.018 (t, J ) 9.8 Hz, 1H), 5.023 (t, J ) 9.7 Hz, 1H), 5.085
(t, J ) 9.2 Hz, 1H), 5.092 (t, J ) 9.3 Hz, 2H), 5.099 (t, J ) 9.2 Hz,
1H), 5.184 (t, J ) 9.2 Hz, 1H), 5.385 (dd, J ) 10.3, 9.3 Hz, 1H),
5.621 (d, J ) 8.3 Hz, 1H), 6.211 (d, J ) 3.8 Hz, 1H). 13C NMR (100
MHz, CDCl3) 20.408, 20.478, 20.558, 20.611, 20.712, 20.732, 20.805,
20.867, 61.095, 61.342, 61.458, 62.097, 67.583, 69.253, 70.369, 70.641,
71.449, 71.585, 71.675, 71.931, 72.173, 72.526, 72.587, 72.648, 72.753,
72.788, 73.439, 75.840, 75.995, 76.083, 91.439, 100.308, 100.710,
100.751, 168.796, 168.824, 169.189, 169.228, 169.268, 169.420,
169.538, 169.695, 169.807, 170.088, 170.137, 170.469, 170.433. Exact
mass calculated for C40H54O27K ) 1005.249. Found ) 1005.249
(HRFABMS, K° source).
(B) Cellotetraose Tetradecaacetate (6). Cellotetraose (2) (49.5 mg,
0.0743 mmol, 1 equiv) and DMAP (1 mg, 0.008 mmol, 0.1 equiv)
were placed in a flask which was purged with N2, and pyridine (3.0
mL, 37 mmol, 500 equiv) and acetic anhydride (0.5 mL, 5.3 mmol, 71
equiv) were added via syringe. The solution was stirred for 22 h at
room temperature, then transferred to a flask containing 20 mL of ice
water slurry and stirred for an additional 90 min. The mixture was
extracted with CHCl3 (5 × 10 mL). The organic layers were combined
and washed with saturated CuSO4 (4 × 10 mL), saturated NaHCO3 (4
× 10 mL), and brine (10 mL). The organic layer was then dried
(MgSO4) and concentrated under reduced pressure. The residue was
dried to yield 69.0 mg (74%) of an off-white solid which was a mixture
Isolation of Cellulose Oligomers. Microcrystalline cellulose (Avicel,
10 g) was hydrolyzed by stirring with fuming HCl solution (200 mL)
for 2 h. The solution was diluted with cold Milli-Q water (600 mL)
and neutralized with Na2CO3, then centrifuged. The supernatant was
loaded onto a column of radius 5 cm, length 95 cm, and a reservoir
volume of 1 L. The column resin consisted of 1:1 Darco G-60/Celite
545, successively equilibrated with solutions of 0.2% stearic acid in
absolute ethanol, and 50% aqueous ethanol saturated with stearic acid.
The resin was packed in the column and equilibrated with water prior
to introduction of the oligosaccharides. The oligosaccharides were eluted
with a water/ethanol gradient (0-45% ethanol over 5 days) and
monitored by analyzing each fraction with an orcinol cocktail.7,54 The
fractions corresponding to each oligosaccharide were combined, dried,
and verified by mass spectrometry. Cellotriose (C18H32O16Na): mass
1
of the R and â anomers of 6 by NMR. H NMR (500 MHz, CDCl3):
1.938 (s, 3H), 1.959 (s, 3H), 1.966 (s, 3H), 1.973 (s, 3H), 1.986 (s,
3H), 1.996 (s, 3H), 1.999 (s, 3H), 2.005 (s, 3H), 2.014 (s, 3H), 2.066
(s, 3H), 2.071 (s, 3H), 2.100 (s, 3H), 2.125 (s, 6H), 3.542 (ddd, J )
10.1, 5.1, 2.1 Hz; 1H), 3.562 (ddd, J ) 10.0, 5.1, 2.1 Hz; 1H), 3.613
(ddd, J ) 9.9, 4.4, 2.3 Hz; 1H), 3.716 (ddd, J ) 9.9, 4.3, 2.1 Hz; 1H),
3.726 (t, J ) 9.4 Hz, 1H), 3.733 (t, J ) 9.4 Hz, 1H), 3.777 (dd, J )
9.9, 8.9 Hz; 1H), 4.010 (dd, J ) 12.5, 2.3 Hz, 1H), 4.069 (dd, J )
12.2, 5.1 Hz, 1H), 4.073 (dd, J ) 12.2, 5.2 Hz, 1H), 4.075 (dd, J )
(54) Sorenson, M.; Haugaard, G. Biochem. Z. 1933, 260, 247-277.

7236 J. Am. Chem. Soc., Vol. 121, No. 31, 1999
Flugge et al.
12.1, 4.6 Hz, 1H), 4.334 (dd, J ) 12.5, 4.4 Hz, 1H), 4.376 (dd, J )
12.1, 2.1 Hz, 1H), 4.383 (dd, J ) 12.2, 2.1 Hz, 1H), 4.422 (d, J )
7.79 Hz, 1H), 4.452 (d, J ) 7.9 Hz, 2H), 4.457 (dd, J ) 12.2, 2.0 Hz,
1H), 4.806 (dd, J ) 9.4, 7.8 Hz, 1H), 4.812 (dd, J ) 9.3, 7.8 Hz, 1H),
4.881 (dd, J ) 9.2, 8.0 Hz, 1H), 5.013 (dd, J ) 9.4, 8.5 Hz, 1H),
5.031 (t, J ) 9.8 Hz, 1H), 5.078 (t, J ) 9.2 Hz, 1H), 5.083 (t, J ) 9.3
Hz, 1H), 5.101 (t, J ) 9.3 Hz, 1H), 5.191 (dd, J ) 9.5, 8.9 Hz, 1H),
5.630 (d, J ) 8.3 Hz, 1H). 13C NMR (125 MHz, CDCl3) 20.368, 20.412,
20.460, 20.449, 20.552, 20.668, 20.724, 61.342, 61.434, 61.925, 62.015,
67.561, 70.320, 71.409, 71.589, 71.638, 71.883, 72.113, 72.408, 72.498,
72.606, 72.646, 72.741, 73.390, 75.795, 75.943, 76.066, 76.070, 91.403,
100.252, 100.399, 100.684, 168.730, 168.969, 169.080, 169.172,
169.199, 169.346, 169.504, 169.597, 169.612, 170.057, 170.078,
170.126, 170.363. Exact mass calculated for C52H70O35Na ) 1277.360.
Found ) 1277.360 (HRFABMS, Na° source).
1H), 4.881 (dd, J ) 9.3, 8.0 Hz, 1H), 5.015 (dd, J ) 9.5, 8.4 Hz, 1H),
5.033 (t, J ) 9.7 Hz, 1H), 5.066 (t, J ) 8.9 Hz, 2H), 5.080 (t, J ) 9.2
Hz, 1H), 5.086 (t, J ) 9.7 Hz, 1H), 5.104 (t, J ) 9.4 Hz, 1H), 5.194
(dd, J ) 9.5, 8.9 Hz, 1H), 5.633 (d, J ) 8.3 Hz, 1H). 13C NMR (125
MHz. CDCl3) 20.419, 20.469, 20.493, 20.515, 20.597, 20.748, 20.784,
61.457, 61.537, 61.962, 62.010, 62.131, 67.689, 70.448, 71.532, 71.688,
71.786, 71.999, 72.218, 72.464, 72.499, 72.582, 72.770, 72.834, 73.516,
73.570, 73.803, 75.864, 76.045, 76.115, 100.290, 100.420, 100.463,
100.516, 100.766, 168.863, 169.104, 169.227, 169.305, 169.355,
169.484, 169.653, 169.729, 169.769, 170.209, 170.234, 170.500. Exact
mass calculated for C76H102O51Na ) 1853.529. Found ) 1853.528
(HRFABMS, Na° source).
(E) Methyl â-D-Cellotriose Decaacetate (9). Cellotriose undecaac-
etate (5) (97.0 mg, 0.100 mmol, 1 equiv) was dried by azeotropic
distillation with anhydrous benzene distilled from sodium (3 × 10 mL).
Residual benzene was removed in vacuo. A solution of 30% HBr in
HOAc (0.2 mL, 1.0 mmol, 10 equiv) was added to the flask via syringe,
and the mixture was stirred under N2 for 1 h. CH2Cl2 (3 mL) was added
to the flask, and the organic layer was washed with ice water (2 × 10
mL), ice/saturated NaHCO3 (2 × 10 mL), and ice water (10 mL). The
organic layer was then dried (MgSO4) and concentrated under reduced
pressure. The product was dissolved in CHCl3 and transferred to a flask
containing CaSO4‚1/2 H2O (76.2 mg, 0.525 mmol, 5.25 equiv), HgO
(96.9 mg, 0.447 mmol, 4.47 equiv), HgBr2 (1 mg, 0.003 mmol, 0.03
equiv), CH2Cl2 (10 mL), and CH3OH (1.5 mL, 37 mmol, 370 equiv),
which had been stirring under N2 for 90 min. The flask was wrapped
in foil and allowed to stir under N2 for 36 h. The reaction mixture was
filtered through a pad of Celite 545. The solvent was removed under
reduced pressure, and the resulting solid was redissolved in CHCl3.
This was filtered into a separatory funnel and washed with ice water
(2 × 10 mL), ice/saturated NaHCO3 (2 × 10 mL), and again with ice
water (10 mL). The combined aqueous washes were back-extracted
with CHCl3 (10 mL), the organic layers were combined and dried
(MgSO4), and the solvent was removed under reduced pressure to yield
72.1 mg (76%) of 9 as an off-white solid: mp 94 (dec) 1H NMR (500
MHz, CDCl3) 1.960 (s, 3H), 1.964 (s, 3H), 1.980 (s, 3H), 1.990 (s,
3H), 2.007 (s, 3H), 2.019 (s, 3H), 2.020 (s, 3H), 2.073 (s, 3H), 2.115
(s, 3H), 2.123 (s, 3H), 3.452 (s, 3H), 3.565 (ddd, J ) 9.9, 4.8, 2.1 Hz,
1H), 3.569 (ddd, J ) 9.9, 4.9, 1.8 Hz, 1H), 3.619 (ddd, J ) 9.9, 4.4,
2.3 Hz, 1H), 3.748 (t, J ) 9.6 Hz, 1H), 3.752 (t, J ) 9.5 Hz, 1H),
4.018 (dd, J ) 12.5, 2.3 Hz, 1H), 4.069 (dd, J ) 11.9, 4.7 Hz, 1H),
4.100 (dd, J ) 12.1, 5.1 Hz, 1H), 4.338 (dd, J ) 12.3, 4.5 Hz, 1H),
4.354 (d, J ) 7.9 Hz, 1H), 4.384 (dd, J ) 12.1, 2.1 Hz, 1H), 4.457 (d,
J ) 7.9 Hz, 1H), 4.468 (d, J ) 7.9 Hz, 1H), 4.519 (dd, J ) 12.0, 2.1
Hz, 1H), 4.835 (dd, J ) 9.4, 7.9 Hz, 1H), 4. 862 (dd, J ) 9.6, 7.9 Hz,
1H), 4.887 (dd, J ) 9.3, 7.9 Hz, 1H), 5.037 (t, J ) 9.7 Hz, 1H), 5.103
(t, J ) 9.2 Hz, 1H), 5.106 (t, J ) 9.3 Hz, 1H), 5.145 (t, J ) 9.4 Hz,
1H). 13C NMR (125 MHz, CDCl3) 20.434, 20.493, 20.536, 20.625,
20.672, 20.725, 20.835, 56.967, 61.467, 61.693, 62.107, 67.690, 71.531,
71.549, 71.746, 71.982, 72.426, 72.647, 72.677, 72.860, 76.098, 76.443,
100.538, 100.772, 101.392, 169.074, 169.297, 169.638, 169.774,
170.188, 170.317, 170.502. Exact mass calculated for C39H54O26K )
977.254. Found ) 977.255 (HRFABMS, K° source).
(F) Methyl â-D-Cellotetraose Tridecaacetate (10). Cellotetraose
tetradecaacetate (6) (68.0 mg, 0.0542 mmol, 1 equiv) was dried by
azeotropic distillation with anhydrous benzene distilled from sodium
(3 × 10 mL). Residual benzene was removed in vacuo. A solution of
30% HBr in HOAc (0.2 mL, 1.0 mmol, 19 equiv) was added to the
flask via syringe, and the mixture was stirred under N2 for 30 min.
CH2Cl2 (3 mL) was added to the flask, and the organic layer was washed
with ice water (2 × 10 mL), ice/saturated NaHCO3 (2 × 10 mL), and
ice water (10 mL). The organic layer was then dried (MgSO4) and
concentrated under reduced pressure. The product was dissolved in
CHCl3 and transferred to a flask containing CaSO4‚1/2 H2O (66.8 mg,
0.460 mmol, 8.49 equiv), HgO (32.1 mg, 0.148 mmol, 2.73 equiv),
HgBr2 (1 mg, 0.003 mmol, 0.05 equiv), CH2Cl2 (10 mL), and CH3OH
(1.5 mL, 37 mmol, 680 equiv), which had been stirring under N2 for
90 min. The flask was wrapped in foil and allowed to stir under N2 for
36 h. The reaction mixture was filtered through a pad of Celite 545.
The solvent was removed under reduced pressure, and the resulting
solid was redissolved in CHCl3. This was filtered into a separatory
(C) Cellopentaose Heptadecaacetate (7). Cellopentaose (3) (40.0
mg, 0.0483 mmol, 1 equiv) and DMAP (1 mg, 0.008 mmol, 0.2 equiv)
were placed in a flask which was purged with N2, and pyridine (2.0
mL, 25 mmol, 510 equiv) and acetic anhydride (0.2 mL, 2.1 mmol, 49
equiv) were added via syringe. The solution was stirred for 3 days at
room temperature, then transferred to a flask containing 15 mL of ice
water slurry and stirred for an additional 90 min. The mixture was
extracted with CHCl3 (5 × 10 mL). The organic layers were combined
and washed with saturated CuSO4 (4 × 10 mL), saturated NaHCO3 (4
× 10 mL), and brine (10 mL). The organic layer was then dried
(MgSO4) and concentrated under reduced pressure. The residue was
dried to yield 56.0 mg (75%) of an off-white solid which was a mixture
1
of the R and â anomers of 7 by NMR. H NMR (500 MHz, CDCl3)
1.944 (s, 3H), 1.946 (s, 3H), 1.969 (s, 3H), 1.974 (s, 3H), 1.982 (s,
3H), 1.995 (s, 3H), 2.004 (s, 3H), 2.008 (s, 6H), 2.014 (s, 3H), 2.023
(s, 3H), 2.075 (s, 3H), 2.081 (s, 3H), 2.109 (s, 3H), 2.130 (s, 3H),
2.132 (s, 6H), 3.523-3.586 (m, 4H), 3.620 (ddd, J ) 9.9, 4.5, 2.3 Hz,
1H), 3.701-3.803 (m, 4H), 4.020 (dd, J ) 12.5, 2.1 Hz, 1H), 4.044-
4.119 (m, 4H), 4.343 (dd, J ) 12.6, 4.4 Hz, 1H), 4.367-4.481 (m,
4H), 4.419 (d, J ) 7.9 Hz, 1H), 4.429 (d, J ) 7.8 Hz, 1H), 4.460 (d,
J ) 7.9 Hz, 2H), 4.798 (dd, J ) 9.3, 7.7 Hz, 1H), 4.814 (dd, J ) 9.4,
7.7 Hz, 1H), 4.820 (dd, J ) 9.4, 7.9 Hz, 1H), 4.889 (dd, J ) 9.3 Hz,
8.0 Hz, 1H), 5.023 (t, J ) 8.9 Hz, 1H), 5.040 (t, J ) 9.8 Hz, 1H),
5.074 (t, J ) 8.9 Hz, 1H), 5.086 (t, J ) 9.2 Hz, 1H), 5.089 (t, J ) 9.2
Hz, 1H), 5.110 (t, J ) 9.3 Hz, 1H), 5.201 (dd, J ) 9.5, 8.9 Hz, 1H),
5.639 (d, J ) 8.3 Hz, 1H). 13C NMR (125 MHz, CDCl3) 20.394, 20.453,
20.489, 20.581, 20.717, 20.767, 20.845, 61.407, 61.506, 61.921, 61.980,
62.095, 67.634, 69.239, 69.285, 70.403, 71.484, 71.653, 71.702, 71.766,
71.807, 71.958, 72.176, 72.412, 72.482, 72.558, 72.692, 72.723, 72.807,
73.474, 75.846, 75.998, 76.033, 76.088, 100.286, 100.404, 100.501,
100.752, 168.818, 169.063, 169.169, 169.255, 169.310, 169.437,
169.595, 169.675, 169.702, 169.724, 170.157, 170.223, 170.461. Exact
mass calculated for C64H86O43K ) 1581.418. Found ) 1581.418
(HRFABMS, K° source).
(D) Cellohexaose Icosaacetate (8). Cellohexaose (4) (38.7 mg, 0.039
mmol, 1 equiv) and DMAP (1 mg, 0.008 mmol, 0.2 equiv) were placed
in a flask which was purged with N2, and pyridine (3.0 mL, 37 mmol,
950 equiv) and acetic anhydride (0.2 mL, 2.1 mmol, 54 equiv) were
added via syringe. The solution was stirred for 4 days at room
temperature, then transferred to a flask containing 15 mL of ice water
slurry and stirred for an additional 90 min. The mixture was extracted
with CHCl3 (5 × 10 mL). The organic layers were combined and
washed with saturated CuSO4 (4 × 10 mL), saturated NaHCO3 (4 ×
10 mL), and brine (10 mL). The organic layer was then dried (MgSO4)
and concentrated under reduced pressure. The residue was dried to yield
52.3 mg (73%) of an off-white solid which was a mixture of the R and
â anomers of 8 by NMR: mp 129-136 °C. 1H NMR (500 MHz,
CDCl3) 1.937 (s, 3H), 1.938 (s, 6H), 1.963 (s, 3H), 1.968 (s, 3H), 1.976
(s, 3H), 1.989 (s, 3H), 1.997 (s, 3H), 2.001 (s, 9H), 2.006 (s, 3H),
2.016 (s, 3H), 2.068 (s, 3H), 2.074 (s, 3H), 2.102 (s, 3H), 2.121 (s,
3H), 2.124 (s, 9H), 3.518-3.582 (m, 5H), 3.616 (ddd, J ) 10.0, 4.3,
2.3 Hz, 1H), 3.688-3.756 (m, 4H), 3.777 (t, J ) 9.4 Hz, 1H), 4.015
(dd, J ) 12.5, 2.3 Hz, 1H), 4.039-4.095 (m, 5H), 4.335 (dd, J ) 12.6,
4.4 Hz, 1H), 4.361-4.477 (m, 5H), 4.414 (d, J ) 7.8 Hz, 2H), 4.423
(d, J ) 7.8 Hz, 1H), 4.455 (d, J ) 7.8 Hz, 2H), 4.789 (dd, J ) 9.3, 7.8
Hz, 2H), 4.805 (dd, J ) 9.3, 7.8 Hz, 1H), 4.812 (dd, J ) 9.3, 7.9 Hz,

Cellulose Oligomers
J. Am. Chem. Soc., Vol. 121, No. 31, 1999 7237
funnel and washed with ice water (2 × 10 mL), ice/saturated NaHCO3
(2 × 10 mL), and again with ice water (10 mL). The combined aqueous
washes were back-extracted with CHCl3 (10 mL), the organic layers
were combined and dried (MgSO4), and the solvent was removed under
reduced pressure to yield 57.4 mg (86%) of 10 as an off-white solid:
mp 106-108 °C. 1H NMR (500 MHz, CDCl3) 1.994 (s, 3H), 2.011 (s,
3H), 2.017 (s, 3H), 2.030 (s, 3H), 2.043 (s, 3H), 2.054 (s, 3H), 2.062
(s, 3H), 2.072 (s, 3H), 2.074 (s, 3H), 2.124 (s, 3H), 2.164 (s, 3H),
2.176 (s, 3H), 2.182 (s, 3H), 3.506 (s, 3H), 3.580-3.635 (m, 3H), 3.668
(ddd, J ) 9.9. 4.3, 2.3 Hz, 1H), 3.782 (t, J ) 9.5 Hz, 1H), 3.790 (t, J
) 9.5 Hz, 2H), 4.065 (dd, J ) 12.5, 2.3 Hz, 1H), 4.118 (dd, J ) 12.0,
3.4 Hz, 1H), 4.126 (dd, J ) 12.1, 2.1 Hz, 1H), 4.135 (dd, J ) 12.1,
3.4 Hz, 1H), 4.394 (dd, J ) 13.0, 4.3 Hz, 1H), 4.404 (d, J ) 7.9 Hz,
1H), 4.423-4.450 (m, 2H), 4.475 (d, J ) 7.8 Hz, 1H), 4.504 (d, J )
7.9 Hz, 1H), 4.506 (d, J ) 7.9 Hz, 1H), 4.561 (dd, J ) 12.1, 2.2 Hz,
1H), 4.865 (dd, J ) 9.4, 7.8 Hz, 1H), 4.873 (dd, J ) 9.3, 7.9 Hz, 1H),
4.914 (dd, J ) 9.6, 7.9 Hz, 1H), 4.939 (dd, J ) 9.3, 8.0 Hz, 1H),
5.070 (t, J ) 9.6 Hz, 1H), 5.133 (t, J ) 9.2 Hz, 1H), 5.139 (t, J ) 9.1
Hz, 1H), 5.158 (t, J ) 9.3 Hz, 1H), 5.195 (t, J ) 9.3 Hz, 1H). 13C
NMR (125 MHz, CDCl3) 20.349, 20.372, 20.400, 20.424, 20.447,
20.461, 20.541, 20.664, 20.744, 56.878, 61.393, 61.687, 61.984, 62.002,
67.641, 71.469, 71.640, 71.781, 71.926, 72.333, 72.505, 72.543, 72.584,
72.624, 72.684, 72.780, 75.979, 76.076, 76.368, 100.432, 100.705,
101.320, 169.038, 169.197, 169.231, 169.277, 169.557, 169.694,
169.759, 170.121, 170.147, 170.285, 170.425. Exact mass calculated
for C51H70O34K ) 1265.339. Found ) 1265.339 (HRFABMS, K°
source).
61.925, 62.189, 62.229, 62.264, 67.861, 71.712, 71.885, 71.976, 72.010,
72.194, 72.586, 72.727, 72.798, 72.835, 72.882, 72.946, 73.047, 76.244,
76.294, 76.332, 76.657, 100.692, 100.701, 100.758, 100.997, 101.591,
169.314, 169.455, 169.507, 169.563, 169.850, 169.922, 169.952,
169.986, 170.009, 170.415, 170.554, 170.709. Exact mass calculated
for C63H86O42Na ) 1537.449. Found ) 1537.449 (HRFABMS, Na°
source).
(H) Methyl â-D-Cellohexaose Nonadecaacetate (12). Cellohexaose
icosaacetate (8) (52.3 mg, 0.0286 mmol, 1 equiv) was dried by
azeotropic distillation with anhydrous benzene distilled from sodium
(3 × 10 mL). Residual benzene was removed in vacuo. A solution of
30% HBr in HOAc (0.2 mL, 0.55 mmol, 19 equiv) was added to the
flask via syringe, and the mixture was stirred under N2 for 10 min.
CH2Cl2 (3 mL) was added to the flask, and the organic layer was washed
with ice water (2 × 10 mL), ice/saturated NaHCO3 (2 × 10 mL), and
ice water (10 mL). The organic layer was then dried (MgSO4) and
concentrated under reduced pressure. The product was dissolved in
CHCl3 and transferred to a flask containing CaSO4‚1/2 H2O (74.6 mg,
0.514 mmol, 18.0 equiv), HgO (81.6 mg, 0.377 mmol, 13.2 equiv),
HgBr2 (1 mg, 0.003 mmol, 0.10 equiv), CH2Cl2 (10 mL), and CH3OH
(1.5 mL, 37.03 mmol, 1295 equiv), which had been stirring under N2
for 90 min. The flask was wrapped in foil and allowed to stir under N2
for 36 h. The reaction mixture was filtered through a pad of Celite
545. The solvent was removed under reduced pressure, and the resulting
solid was redissolved in CHCl3. This was filtered into a separatory
funnel and washed with ice water (2 × 10 mL), ice/saturated NaHCO3
(2 × 10 mL), and again with ice water (10 mL). The combined aqueous
washes were back-extracted with CHCl3 (10 mL), the organic layers
were combined and dried (MgSO4), and the solvent was removed under
reduced pressure to yield 21.7 mg (42%) of 12 as a pale yellow solid:
mp 114-116 °C. 1H NMR (750 MHz, CDCl3) 1.944 (s, 6H), 1.945 (s,
3H), 1.963 (s, 3H), 1.968 (s, 3H), 1.983 (s, 3H), 1.995 (s, 3H), 2.004
(s, 3H), 2.006 (s, 6H), 2.012 (s, 3H), 2.021 (s, 3H), 2.023 (s, 3H),
2.074 (s, 3H), 2.112 (s, 3H), 2.120 (s, 3H), 2.126 (s, 3H), 2.128 (s,
3H), 2.129 (s, 3H), 3.457 (s, 3H), 3.531 (ddd, J ) 9.8, 4.9, 2.0 Hz,
1H), 3.551 (ddd, J ) 9.4, 5.1, 2.3 Hz, 2H), 3.564 (ddd, J ) 10.4, 5.2,
2.0 Hz, 1H), 3.571 (ddd, J ) 10.1, 5.1, 2.0 Hz, 1H), 3.622 (ddd, J )
9.9, 4.4, 2.3 Hz, 1H), 3.712 (t, J ) 9.4 Hz, 1H), 3.720 (t, J ) 9.5 Hz,
1H), 3.723 (t, J ) 9.3 Hz, 1H), 3.741 (t, J ) 9.5 Hz, 2H), 4.024 (dd,
J ) 12.5, 2.3 Hz, 1H), 4.065 (dd, J ) 12.4, 4.1 Hz, 2H), 4.076 (dd, J
) 12.2, 4.8 Hz, 1H), 4.082 (dd, J ) 12.2, 5.1 Hz, 1H), 4.085 (dd, J )
12.1, 4.6 Hz, 1H), 4.337 (dd, J ) 12.5, 4.4 Hz, 1H), 4.357 (d, J ) 7.9
Hz, 1H), 4.380 (dd, J ) 12.6, 2.7 Hz, 2H), 4.384 (dd, J ) 12.5, 3.0
Hz, 1H), 4.388 (dd, J ) 12.1, 2.5 Hz, 1H), 4.422 (d, J ) 7.8 Hz, 1H),
4.433 (d, J ) 7.8 Hz, 1H), 4.461 (d, J ) 7.8 Hz, 1H), 4.464 (d, J )
7.9 Hz, 2H), 4.510 (dd, J ) 12.0, 2.2 Hz, 1H), 4.793 (dd, J ) 9.3, 8.3
Hz, 1H), 4.794 (dd, J ) 9.3, 7.8 Hz, 1H), 4.810 (dd, J ) 9.3, 7.8 Hz,
1H), 4.819 (dd, J ) 9.4, 7.9 Hz, 1H), 4.863 (dd, J ) 9.6, 7.9 Hz, 1H),
4.887 (dd, J ) 9.3, 8.0 Hz, 1H), 5.038 (t, J ) 9.7 Hz, 1H), 5.071 (t,
J ) 9.2 Hz, 1H), 5.074 (t, J ) 9.2 Hz, 1H), 5.087 (t, J ) 9.1 Hz, 1H),
5.089 (t, J ) 8.6 Hz, 1H), 5.111 (t, J ) 9.3 Hz, 1H), 5.145 (t, J ) 9.4
Hz, 1H).13C NMR (188 MHz, CDCl3) 20.657, 20.731, 20.760, 20.863,
20.919, 20.988, 21.101, 21.076, 22.879, 57.210, 61.635, 61.915, 62.170,
62.200, 62.221, 62.273, 67.836, 71.707, 71.869, 71.970, 72.003, 72.181,
72.570, 72.675, 72.715, 72.792, 72.831, 72.873, 72.932, 73.042, 76.236,
76.267, 76.323, 76.658, 100.675, 100.699, 100.744, 100.993, 101.582,
169.282, 169.431, 169.461, 169.484, 169.533, 169.824, 169.897,
169.929, 169.952, 169.985, 170.386, 170.533, 170.679. Exact mass
calculated for C76H102O49Na ) 1825.534. Found ) 1825.535 (HR-
FABMS, Na° source).
(G) Methyl â-D-Cellopentaose Hexadecaacetate (11). Cellopen-
taose heptadecaacetate (7) (72.5 mg, 0.0470 mmol, 1 equiv) was dried
by azeotropic distillation with anhydrous benzene distilled from sodium
(3 × 10 mL). Residual benzene was removed in vacuo. A solution of
30% HBr in HOAc (0.2 mL, 0.545 mmol, 11.6 equiv) was added to
the flask via syringe, and the mixture was stirred under N2 for 20 min.
CH2Cl2 (3 mL) was added to the flask, and the organic layer was washed
with ice water (2 × 10 mL), ice/saturated NaHCO3 (2 × 10 mL), and
ice water (10 mL). The organic layer was then dried (MgSO4) and
concentrated under reduced pressure. The product was dissolved in
CHCl3 and transferred to a flask containing CaSO4‚1/2 H2O (50.4 mg,
0.347 mmol, 7.4 equiv), HgO (51.1 mg, 0.236 mmol, 5.0 equiv), HgBr2
(1 mg, 0.003 mmol, 0.06 equiv), CH2Cl2 (10 mL), and CH3OH (1.5
mL, 37 mmol, 790 equiv), which had been stirring under N2 for 90
min. The flask was wrapped in foil and allowed to stir under N2 for 36
h. The reaction mixture was filtered through a pad of Celite 545. The
solvent was removed under reduced pressure, and the resulting solid
was redissolved in CHCl3. This was filtered into a separatory funnel
and washed with ice water (2 × 10 mL), ice/saturated NaHCO3 (2 ×
10 mL), and again with ice water (10 mL). The combined aqueous
washes were back-extracted with CHCl3 (10 mL), the organic layers
were combined and dried (MgSO4), and the solvent was removed under
reduced pressure to yield 39.1 mg (55%) of 11 as an off-white solid:
mp 106-108 °C. 1H NMR (750 MHz, CDCl3) 1.937 (s, 3H), 1.938 (s,
3H), 1.955 (s, 3H), 1.963 (s, 3H), 1.976 (s, 3H), 1.990 (s, 3H), 2.000
(s, 3H), 2.001 (s, 3H), 2.008 (s, 3H), 2.017 (s, 3H), 2.019 (s, 3H),
2.069 (s, 3H), 2.109 (s, 3H), 2.117 (s, 3H), 2.126 (s, 3H), 2.127 (s,
3H), 3.452 (s, 3H), 3.546 (ddd, J ) 9.9, 4.9, 2.0 Hz, 1H), 3.553 (ddd,
J ) 10.0, 5.1, 2.0 Hz, 1H), 3.567 (ddd, J ) 10.1, 5.2, 2.0 Hz, 1H),
3.573 (ddd, J ) 10.1, 5.0, 2.0 Hz, 1H), 3.614 (ddd, J ) 9.9, 4.3, 2.3
Hz, 1H), 3.714 (t, J ) 9.9 Hz, 1H), 3.720 (t, J ) 9.4 Hz, 1H), 3.736
(t, J ) 9.5 Hz, 2H), 4.011 (dd, J ) 12.5, 2.3 Hz, 1H), 4.054 (dd, J )
11.9, 5.0 Hz, 1H), 4.064 (dd, J ) 11.9, 4.9 Hz, 1H), 4.070 (dd, J )
11.9, 3.2 Hz, 1H), 4.080 (dd, J ) 12.1, 3.2 Hz, 1H), 4.340 (dd, J )
12.5, 4.3 Hz, 1H), 4.350 (d, J ) 7.9 Hz, 1H), 4.385 (dd. J ) 11.9, 1.8
Hz, 1H), 4.392 (dd, J ) 12.2, 1.9 Hz, 2H), 4.411 (d, J ) 7.8 Hz, 1H),
4.421 (d, J ) 7.8 Hz, 1H), 4.451 (d, J ) 7.8 Hz, 1H), 4.453 (d, J )
7.9 Hz, 1H), 4.507 (dd, J ) 12.0, 2.1 Hz, 1H), 4.793 (dd, J ) 9.3, 7.5
Hz, 1H), 4.808 (dd, J ) 9.3, 7.8 Hz, 1H), 4.817 (dd, J ) 9.4, 7.9 Hz,
1H), 4.859 (dd, J ) 9.6, 7.9 Hz, 1H), 4.884 (dd, J ) 9.3, 7.9 Hz, 1H),
5.035 (t, J ) 9.7 Hz, 1H), 5.064 (t, J ) 9.3 Hz, 1H), 5.079 (t, J ) 9.2
Hz, 1H), 5.082 (t, J ) 9.2 Hz, 1H), 5.104 (t, J ) 9.0 Hz, 1H), 5.139
(t, J ) 9.1 Hz, 1H).13C NMR (125 MHz, CDCl3) 20.642, 20.733,
20.703, 20.752, 20.838, 20.887, 20.961, 20.982, 21.038, 57.189, 61.639,
(I) Methyl â-D-Cellotriose (13). Methyl-â-D-cellotriose decaacetate
(9) (26.8 mg, 0.0285 mmol) was placed in a 10 mL flask. A solution
of NaOMe in MeOH (1.5 mL) was made and added to the flask. The
reaction was allowed to stir for 16 h. Small portions of cation-exchange
resin (Dowex 50W-X8) were added until the solution was neutralized.
The reaction was then filtered to remove the resin, and the methanol
was removed under reduced pressure, resulting in 12.5 mg (84%) of
13 as a white solid. 1H NMR (750 MHz, D2O, benzene reference) 3.305
(dd, J ) 9.5, 7.6 Hz, 1H), 3.315 (dd, J ) 9.5, 7.9 Hz, 1H), 3.361 (dd,
J ) 9.3, 8.0 Hz, 1H), 3.418 (dd, J ) 9.9, 9.1 Hz, 1H), 3.487 (ddd, J
) 9.8, 5.9, 2.3 Hz, 1H), 3.507 (t, J ) 9.3 Hz, 1H), 3.575 (s, 3H),

7238 J. Am. Chem. Soc., Vol. 121, No. 31, 1999
Flugge et al.
3.597 (ddd, J ) 10.0, 5.1, 2.2 Hz, 1H), 3.618 (ddd, J ) 9.7, 5.2, 2.2
Hz, 1H), 3.634-3.663 (m, 3H), 3.677 (t, J ) 8.8 Hz, 1H), 3.736 (dd,
J ) 12.6, 6.0 Hz, 1H), 3.818 (dd, J ) 12.6, 5.4 Hz, 1H), 3.826 (dd, J
) 12.6, 5.4 Hz, 1H), 3.917 (dd, J ) 12.4, 2.3 Hz, 1H), 3.980 (dd, J )
12.3, 2.3 Hz, 1H), 3.993 (dd, J ) 12.3, 2.2 Hz, 1H), 4.403 (d, J ) 8.0
Hz, 1H), 4.509 (d, J ) 7.9 Hz, 1H), 4.533 (d, J ) 8.0 Hz, 1H). 13C
NMR (125 MHz, D2O, benzene reference) 57.109, 59.721, 59.806,
60.420, 69.303, 72.754, 72.809, 73.007, 73.908, 74.434, 74.623, 74.670,
75.328, 75.847, 78.202, 78.342, 102.221, 102.429, 102.947. Exact mass
calculated for C19H34O16Na ) 541.175. Found ) 541.174 (HRFABMS,
Na° source).
72.696, 72.749, 72.972, 73.795, 74.057, 74.592, 74.644, 75.266, 75.829,
78.041, 78.162, 78.283, 102.128, 102.348, 102.874. Exact mass
calculated for C31H54O26Na ) 865.280. Found ) 865.280 (HRFABMS,
Na° source).
(L) Methyl â-D-Cellohexaose (16). A small piece of Na° was
dissolved in EtOH (9 mL). CH2Cl2 (1 mL) was added to the solution,
then guanidine‚HCl (53.6 mg, 0.561 mmol). The resulting suspension
was subjected to a short anion-exchange column (Dowex 2-X8) and
dripped into a flask containing methyl cellohexaose nonadecaacetate
(12) (22.7 mg, 0.0126 mmol, 1 equiv) dissolved in a minimum amount
of CH2Cl2. After the guanidine solution was added, a precipitate formed.
The resulting suspension was stirred overnight, filtered, and dried. The
residue was triturated in 9:1 CH2Cl2/MeOH, filtered, dried, and weighed
to yield 9.3 mg (73%) of 16 as an off-white solid. 1H NMR (750 MHz,
D2O, benzene reference) 3.304 (dd, J ) 9.53, 8.29 Hz, 1H), 3.313 (dd,
J ) 9.4, 8.0 Hz, 1H), 3.358 (br t, J ) 8.5 Hz, 4H), 3.486 (ddd, J )
9.6, 5.7, 2.4 Hz, 1H), 3.506 (t, J ) 9.3 Hz, 1H), 3.574 (s, 3H), 3.59-
3.69 (m, 15H), 3.736 (dd, J ) 12.5, 5.9 Hz, 1H), 3.819 (dd, J ) 12.3,
5.1 Hz, 1H), 3.828 (dd, J ) 12.0, 5.0 Hz, 4H), 3.915 (dd, J ) 12.6,
2.1 Hz, 1H), 3.976 (br d, J ) 11.5 Hz, 4H), 3.991 (dd J ) 11.9, 2.0,
1H), 4.405 (d, J ) 8.1 Hz, 1H), 4.509 (d, J ) 7.8 Hz, 1H), 4.533 (d,
J ) 8.0 Hz, 4H). 13C NMR (188 MHz, D2O, benzene reference) 57.066,
59.652, 59.701, 60.376, 69.269, 72.710, 72.750, 72.972, 73.815, 74.101,
74.601, 74.650, 75.290, 75.810, 78.048, 78.181, 78.323, 102.179,
102.389, 102.907. Exact mass calculated for C37H64O31Na ) 1027.333.
Found ) 1027.333 (HRFABMS, Na° source).
(J) Methyl â-D-Cellotetraose (14). A small piece of Na° was
dissolved in EtOH (9 mL). CH2Cl2 (1 mL) was added to the solution,
then guanidine‚HCl (66.5 mg, 0.969 mmol). The resulting suspension
was subjected to a short anion-exchange column (Dowex 2-X8) and
dripped into a flask containing methyl cellotetraose tridecaacetate (10)
(53.0 mg, 0.0432 mmol, 1 equiv) dissolved in a minimum amount of
CH2Cl2. After the guanidine solution was added, a precipitate formed.
The resulting suspension was stirred overnight, filtered, and dried. The
residue was triturated in 9:1 CH2Cl2/MeOH, filtered, dried, and weighed
1
to yield 21.5 mg (73%) of 14 as an off-white solid. H NMR (750
MHz, D2O, benzene reference) 3.304 (t, J ) 8.0 Hz, 1H), 3.316 (t, J
) 8.7 Hz, 1H), 3.361 (t, J ) 8.6 Hz, 2H), 3.417 (t, J ) 9.5 Hz, 1H),
3.488 (ddd, J ) 9.3, 5.9, 2.2 Hz, 1H), 3.509 (t, J ) 10.0 Hz, 1H),
3.574 (s, 3H), 3.588-3.680 (m, 9H), 3.736 (dd, J ) 12.4, 5.9 Hz, 1H),
3.818 (dd, J ) 12.0, 6.4 Hz, 1H), 3.827 (dd, J ) 12.2, 5.5 Hz, 2H),
3.915 (dd, J ) 12.4, 2.2 Hz, 1H), 3.977 (dd, J ) 12.3, 1.9 Hz, 1H),
3.980 (dd, J ) 12.3, 1.8 Hz, 1H), 3.991 (dd, J ) 12.2, 2.0 Hz, 1H),
4.404 (d, J ) 8.0, 1H), 4.509 (d, J ) 7.8 Hz, 1H), 4.532 (d, J ) 8.1
Hz, 2H). 13C NMR (125 MHz, D2O, benzene reference) 57.066, 59.690,
59.723, 59.807, 60.402, 69.290, 72.722, 72.769, 72.792, 72.989, 73.847,
73.866, 73.885, 74.280, 74.113, 74.609, 74.658, 75.308, 75.823, 78.090,
78.208, 78.350, 102.180, 102.207, 102.407, 102.924. Exact mass
calculated for C25H44O21Na ) 703.227. Found ) 703.227 (HRFABMS,
Na° source).
Acknowledgment. This work has been supported by the
National Institute of Health, the Petroleum Research Fund, and
the American Heart Association. J.T.B. thanks the Colgate-
Palmolive Corporation for an undergraduate fellowship. Mark
Morris and Michael Williams are thanked for assistance in
isolation of the cellulose oligomers. NMR spectra were obtained
in the Varian Oxford Instrument Center for Excellence in NMR
Laboratory. Funding for this instrumentation was provided in
part from the W. M. Keck Foundation, the National Institutes
of Health (PHS 1 S10 RR10444-01), and the National Science
Foundation (NSF CHE 96-10502). The assistance of Vera
Mainz and Feng Lin of the VOICE NMR Spectroscopy Lab at
the University of Illinois is recognized. The members of the
Petillo group are acknowledged for numerous helpful discus-
sions.
(K) Methyl â-D-Cellopentaose (15). A small piece of Na° was
dissolved in EtOH (9 mL). CH2Cl2 (1 mL) was added to the solution,
then guanidine‚HCl (53.6 mg, 0.561 mmol). The resulting suspension
was subjected to a short anion-exchange column (Dowex 2-X8) and
dripped into a flask containing methyl cellopentaose hexadecaacetate
(11) (37.4 mg, 0.0247 mmol, 1 equiv) dissolved in a minimum amount
of CH2Cl2. After the guanidine solution was added, a precipitate formed.
The resulting suspension was stirred overnight, filtered, and dried. The
residue was triturated in 9:1 CH2Cl2/MeOH, filtered, dried, and weighed
1
to yield 16.2 mg (77%) of 15 as an off-white solid. H NMR (750
Supporting Information Available: Tables of the proton
and carbon assignments for 9-12, HMQC-COSY and HMQC-
TOCSY data for â-methyl cellotetraose, â-methyl cellopentaose,
and â-methyl cellohexaose, and plots of the 13C chemical shifts
vs ring size for the reducing end ring, middle rings, and
nonreducing end ring. This material is available free of charge
via the Internet at http://pubs.acs.org.
MHz, D2O, benzene reference) 3.304 (dd, J ) 9.5, 8.3 Hz, 1H), 3.315
(dd, J ) 9.5, 8.0 Hz, 1H), 3.360 (t, J ) 8.6 Hz, 3H), 3.416 (t, J ) 9.5
Hz, 1H), 3.486 (ddd, J ) 9.7, 5.7, 2.5 Hz, 1H), 3.507 (t, J ) 9.3 Hz,
1H), 3.574 (s, 3H), 3.59-3.69 (m, 12H), 3.737 (dd, J ) 12.4, 5.8 Hz,
1H), 3.820 (dd, J ) 12.2, 6.5 Hz, 1H), 3.829 (dd, J ) 12.1, 5.4 Hz,
3H), 3.916 (dd, J ) 12.4, 2.3 Hz, 1H), 3.977 (br d, J ) 12.0 Hz, 3H),
3.991 (dd, J ) 12.3, 2.2 Hz, 1H), 4.403 (d, J ) 8.0 Hz, 1H), 4.510 (d,
J ) 8.0 Hz, 1H), 4.533 (d, J ) 8.0 Hz, 3H). 13C NMR (188 MHz,
D2O, benzene reference) 57.066, 59.669, 59.749, 60.354, 69.227,
JA990561U