Enzymatic polymerization of poly(ε-CL) containing an ethyl glucopyranoside head group: An NMR study

Article abstract of DOI:10.1366/0003702981942898

The usefulness of the 2D-NMR techniques in the structure determination of the oligo ε-caprolactone ethyl glucopyranoside (oligo(ε-CL) EGP) conjugate, synthesized by a lipase-catalyzed regioselective reaction, is described. Data from a ¹³C-¹³C COSY NMR spectrum of the ethyl glucopyranoside is used for complete and unambiguous assignments of the carbon resonances in the ¹³C NMR spectrum of the anomeric mixture of EGP. A comparison of the DEPT-135 spectrum of the product oligo(ε-CL) EGP conjugate with that of the starting material, ethyl glucopyranoside (EGP), led to the conclusion that the lipase-catalyzed ring-opening of ε-CL initiated by a multifunctional initiator, EGP, proceeded in a regioselective fashion. Also, a detailed analysis of the NMR data has allowed determination of the site for lipase-catalyzed ring-opening polymerization of ε-CL. Our study suggests that among the possible four initiation sites in EGP only the hydroxyl group on carbon C-6 took part in the initiation process.

Full text of DOI:10.1366/0003702981942898

e
Enzymatic Polymerization of Poly( -CL) Containing an
Ethyl Glucopyranoside Head Group: An NMR Study
KIRPAL S. BISHT,* RICHARD A. GROSS,² and ASHOK L. CHOLLI³
Department of Chemistry and Center for Advanced Materials, University of Massachusetts±Lowell, One University Ave., Lowell,
Massachusetts 01854
The usefulness of the 2D-NMR techniques in the structure deter-
ters for preparation of monoacylated sugars have been
reported in DMF and pyridine. However, yields reported
mination of the oligo e-caprolactone ethyl glucopyranoside (oilgo(e-
CL) EGP) conjugate, synthesized by a lipase-catalyzed regioselec-
tive reaction, is described. Data from a 13C±13C COSY NMR spec-
from these reactions are low.8±16 Moreover, using these
polar solvents prevents widespread use of such reaction
trum of the ethyl glucopyranoside is used for complete and unam-
schemes. Also, unfortunately, many enzymes are catalyt-
biguous assignments of the carbon resonances in the 13C NMR spec-
trum of the anomeric mixture of EGP. A comparison of the DEPT-
ically inactive in these solvent media.18 Direct enzymatic
esteri®cations of sugars with fatty acids in aqueous media
135 spectrum of the product oilgo(e-CL) EGP conjugate with that
have also been reported to give low yield and selectivi-
ty.19 Use of alkyl pyranosides offers advantages due to
their increased solubility in less polar solvents and higher
reactivity. Interestingly, a dramatic increase in the reac-
of the starting material, ethyl glucopyranoside (EGP), led to the
conclusion that the lipase-catalyzed ring-opening of e-CL initiated
by a multifunctional initiator, EGP, proceeded in a regioselective
fashion. Also, a detailed analysis of the NMR data has allowed de-
termination of the site for lipase-catalyzed ring-opening polymer-
ization of e-CL. Our study suggests that among the possible four
initiation sites in EGP only the hydroxyl group on carbon C-6 took
part in the initiation process.
tivity of ethyl -D-glucopyranoside as compared to either
a
glucose or methyl -D-glucopyranoside, in lipase-cata-
a
lyzed esteri®cation with dodecanoic acid, was reported
by Bjorkling and co-workers.20,21 The increased reactivity
of the ethyl glucopyranoside (EGP) was attributed to
higher solubilities of the reactants in one another. They
also reported regioselective acylation of ethyl glucopyr-
Index Headings: Lipase; Ethyl glucopyranoside; e-Caprolactone;
Ring-opening polymerization; Two-dimensional NMR; 13C NMR;
Regioselectivity; Multifunctional initiator; Site of reaction; Capro-
lactone; Glucopyranoside.
anoside, in bulk, for production of n-alkyl 6-O-acyl- -D-
a
glucopyranosides. Although high conversions were ob-
tained, contamination by 2±14% of the diester was ob-
served.20,21
INTRODUCTION
In addition to the dif®culties encountered in modi®-
cation of sugars, additional tasks exist in conclusive iden-
ti®cation of these compounds. Structure determination of
selectively protected sugars has been a challenge because
the proton and carbon-13 NMR spectra of these com-
pounds are often complicated due to (1) the overlapping
of several resonances, and (2) the presence of anomeric
mixtures in this class of compounds.
Amphiphilic compounds ®nd important applications in
industry and medicine as surface-active agents.1 Poly-
mers bearing sugar residues have been reported to be of
great importance as pharmacological and biomedical ma-
terials. The sugar groups in these materials are reported
to play an important role in these applications. Therefore
incorporation of sugar groups in polymer chains is an
important thrust area in synthesis of amphiphilic poly-
mers. Vinyl monomers with sugar groups such as glu-
We recently reported a one-pot highly regioselective
synthesis of oligomeric -caprolactone (oligo( -CL))
bearing an ethyl glucose head group by ring-opening po-
cosylethyl methacrylate, alkyl, or aryl 6-O-acryl- -D-glu-
a
Î
Î
copyranosides that can be subsequently polymerized have
been prepared.2,3 However, selective modi®cation of one
out of several hydroxyl groups in sugars has been a dif-
®cult challenge to synthetic chemists.4±6 The chemical
methods employed for this task are tedious and involve
several protection/deprotection steps that frequently re-
quire isolation/puri®cation at each step.4±7 Enzymatic
methods have increasingly been utilized for preparation
of selectively modi®ed sugars.8±17 Sugars are soluble only
in very polar solvents, viz., pyridine, dimethylformamide
(DMF), and dimethylsulfoxide (DMSO), due to the in-
creased number of hydroxyl functional groups in them.
Enzyme-catalyzed transesteri®cations using activated es-
lymerization of -CL initiated by ethyl glucopyranoside
Î
and catalyzed by lipase from procine pancreas (Scheme
I).22 We report here the application of one- and two-di-
mensional (1D and 2D) NMR techniques for structure
elucidation of the product. We also discuss how the one-
and two-dimensional NMR techniques are useful in iden-
tifying the key reaction sites in this reaction scheme.
EXPERIMENTAL
Materials and Methods. Analytical-quality D-glucose
was purchased from Sigma Chemical Company and was
dried in a vacuum oven at 40 C overnight prior to use.
D-Glucose-13C6 (99% 13C) was purchased from Aldrich
Chemical Company and was used as received. -Capro-
8
Received 10 April 1998; accepted 27 July 1998.
* Department of Chemistry, University of South Florida, 4202 East
Fowler Avenue, CHE 305, Tampa, FL 33620-1733.
² Present address: Department of Chemistry, Polytechnic University,
Brooklyn, NY 11201.
Î
lactone was also purchased from Aldrich and was dis-
tilled at 97±98 C over CaH at 10 mm of Hg prior to its
8
2
use. Porcine pancreatic lipase (PPL, type II) was pur-
chased from Sigma (25% protein, speci®ed activity 61
³ Author to whom correspondence should be sent.
0003-7028 / 98 / 5211-1472$2.00 / 0
q 1998 Society for Applied Spectroscopy
1472
Volume 52, Number 11, 1998
APPLIED SPECTROSCOPY

SCHEME I.
units/mg protein) and was stored at 4 C. One unit is
viscous oil. The crude product was found to be a 2.1:1
mixture of anomers, containing approxmately 1% un-
reacted glucose. The anomeric ratio of the product was
calculated from the peak areas of the doublets at 4.65 and
8
de®ned as the amount of protein that hydrolyzes 1 mi-
croequivalent of fatty acid from triacetin in 1 h at pH
7.4. Prior to use, lipase was dried over P2O5 overnight in
a vacuum desicator at 0.1 mm of Hg. Deuterated NMR
solvents were purchased from Aldrich and were used as
received.
/
a b
4.15 ppm corresponding to the proton on C-1 in the
a
and the anomer, respectively. Pure ethyl glucopyrano-
b
side was obtained after puri®cation by column chroma-
tography over silica gel (100±120 mesh) with the use of
chloroform/methanol (9:1) mixture as eluent. Pure ethyl
glucopyranoside, 10.30 g, was obtained as a colorless vis-
cous oil.
NM R Measurements. One- and two-dimensional
NMR spectral data were collected by using 250 and 500
1
MHz NMR instruments. H-NMR spectra were recorded
on a Bruker ARX-250 spectrometer at 250 MHz. The
NMR sample solutions were prepared with a concentra-
tion of 4% w/v in dimethylsulfoxide-d6. The instrumental
operating parameters were as follows: temperature 310
Preparation of Ethyl Glucopyranoside-13C6. The
procedure for the preparation of 13C-labeled ethyl glu-
copyranoside was similar to the one described above for
ethyl glucopyranoside except that 100 mg (0.0005 mol)
glucose-13C6 and proportional quantities of reagents were
used. The crude product was puri®ed by column chro-
matography to give 90 mg of pure ethyl glucopyranoside-
13C6. The 1H-NMR spectrum of 13C-labeled compound
was similar to that of ethyl glucopyranoside, but the pro-
ton-decoupled 13C-NMR spectrum showed resolved mul-
tiplets due to 13C±13C coupling.
K, pulse width 4.9 s (30 pulse), 32 KW data points,
m
8
3.17 s acquisition time, 1 s relaxation delay, and 16 tran-
sients. 13C-NMR spectra were recorded at 62.9 MHz on
a Bruker ARX-250 spectrometer. Samples were prepared
in dimethylsulfoxide-d6. Chemical shifts in parts per mil-
lion (ppm) were referenced relative to DMSO-d6 as an
internal reference at 39.7 ppm. Following are the exper-
imental parameters for 13C NMR data acquisition: 10%
w/v in DMSO-d6, 310 K, pulse width 4.9 s (30 ), 64
Enzyme-Catalyzed Ring Opening of e-Caprolactone
by Ethyl Glucopyranoside. The synthesis was carried
out as described elsewhere.22 In a typical run, 200 mg of
ethyl glucopyranoside and 500 mg of dry porcine pan-
m
8
KW data points, 1.638 s acquisition time, and 1 s relax-
ation delay. The number of transients varied from 15
000±18 000.
13C±13C COSY Experiment. The 2D 13C±13C COSY
spectral data were acquired on the DRX 500 MHz NMR
instrument by using the pulse sequence that was de-
scribed by Aue et al.23 The sample was prepared in the
deuterated-DMSO. The following parameters were used
for collecting 2D-NMR data: We accumulated 512 incre-
ments in t1 and 2048 data points for each free induction
decay (FID) in t2. Sixty-four scans were collected for
each FID with a relaxation delay of 1 s. The carbon-13
/2 pulse width was 11.8 s for 13C NMR observation
creatic lipase (0.1 mm Hg, 25 C, overnight) were trans-
8
ferred into an oven-dried 20 mL reaction vial under an
argon atmosphere. The vial was immediately stoppered
with a rubber septum and purged with argon. We added
0.8 mL of -caprolactone via syringe under argon atmo-
Î
sphere. The reaction vial was then placed in a constant
temperature oil bath maintained at 70 C for 96 h. In
8
addition, a control reaction was set up in the same bath.
The control sample was prepared in a similar way, except
that PPL was not added. The reaction was then quenched
by removing the enzyme with a vacuum ®ltration process
(glass-fritted ®lter, medium-pore porosity). The enzyme
was washed 3±4 times with 10 mL chloroform. All the
®ltrates were combined and the solvent was removed in
vacuo. The ®nal product was characterized by using its
1H- and 13C-NMR spectra.
p
m
at 125.768 MHz. Quadrature detection was used, FIDs
were multiplied by a sine function prior to Fourier trans-
formations, and t1 was zero-®lled once to 1024 w to yield
a processed data matrix of 1024 w
1024 w (reals only).
3
The spectral width was 9328 Hz in both dimensions. The
spectrum was symmetrized.
Preparation of Ethyl Glucopyranoside. The com-
pound was prepared according to the procedure described
elsewhere.20 In a typical experiment, in an oven-dried 250
mL round-bottom ¯ ask containing 25 g of absolute eth-
anol (0.55 mol), 10 g of vacuum-dried D-glucose (0.055
mol) and ion exchange resin Dowex-50 H1 (2.5 g) were
added. The reaction was then re¯ uxed (for 20 h) until
most of the glucose was dissolved. The reaction was then
allowed to cool to room temperature and it was ®ltered.
Ethanol was evaporated in vacuo to obtain ethyl gluco-
RESULTS AND DISCUSSION
Proton NMR Spectral Analysis of Starting Materi-
als: e-CL and EGP. First we discuss the common fea-
tures in one-dimensional proton NMR spectra of -CL
and EGP (Figs. 1 and 2a, respectively). EGP initiated
ring-opening polymerization of -CL in bulk catalyzed
Î
Î
by porcine pancreatic lipase was carried out (see Exper-
1
imental section). In the H NMR spectrum of -CL the
Î
pyranoside (12.65 g,
98% yield) as a light yellowish
methylene protons next to the oxygen (labeled as 6 in
.
APPLIED SPECTROSCOPY
1473

end group methylene resonances (±CH2±OH) appear as a
triplet at 3.75 ppm.
Ethyl glucopyranoside was prepared from D-glucose by
reaction with ethanol under strong acidic conditions (see
Experimental section). EGP has a complex resonance pat-
1
tern in its H NMR spectrum (Fig. 2a). Protons on the
carbon C-1 in the and anomers appear as doublets at
a
b
4.65 and 4.15 ppm, respectively. These resonances were
used to calculate the anomeric composition of the EGP.
NMR integrations suggest that the ratio of the relative
amount of and anomers in EGP was 2.1:1.0. Due to
a
the presence of and anomers in the sample, two trip-
b
a
b
lets appear for methyl protons at 1.15 ppm. The reso-
nances in the 2.85±3.80 ppm are due to the remaining
protons on the sugar ring.
1H-NMR Spectrum of the Product: EGP-Oligo (e-
1
CL). The 250 MHz H NMR spectrum of the product,
FIG. 1. The 250 MHz 1H Proton NMR spectrum of -caprolactone in
CDCl3.
Î
EGP-oligo( -CL), in Fig. 2b shows the resonances due to
both (1) the oligo( -CL), and (2) the ethyl glucose head
Î
Î
group. The multiplets at 1.35, 1.56, and 2.40 ppm and
the triplets at 3.35 and 4.05 ppm are representative for
oligo( -CL), and the resonances for ethyl glucose head
Scheme I) (±CH ±O±C O) and methylene protons next
5
to the carbonyl (±CH ±C O) (labeled 2) appear as trip-
2
Î
5
2
group appear as a triplet at 1.15, multiplets from 3.00±
3.70 and 4.05±4.44 ppm, and a doublet at 4.65 ppm. Thus
the presence of resonances due to ethyl glucopyranoside
lets at 4.25 and 2.65 ppm, respectively. The methylene
protons labeled as 3 and 5 appear as a multiplet at 1.90
ppm and a broad peak for methylene protons (labeled 4)
1
and poly( -CL) in the H NMR spectrum of the product
Î
at 1.75 ppm. The ring-opened product, poly( -CL), also
Î
1
suggested the formation of an oligomer containing an
EGP head group. On the basis of the NMR data, we will
also discuss later in this section the linkage between the
shows similar resonances in its H NMR spectrum. (The
proton NMR spectrum of this compound is not shown
here.) The resonances are, however, shifted up®eld for
head group and oligo( -CL).
poly( -CL) as expected due to its linear structure. As a
Î
Î
Reactions involving anomeric mixtures present the
possibility of substrate speci®city of the enzyme in pref-
erence of one or the other anomer. In fact, Mucor miehei
and Candida antarctica lipase catalyzed acylation of pure
result of ring-opening, the methylene resonances of -CL
Î
at 4.25 and 2.65 ppm are shifted to 4.05 and 2.35 ppm
in poly( -CL), respectively. The signal pattern in the
Î
spectral region of 1.5±1.9 ppm for methylene protons (la-
beled as 3 -5 in Scheme I) are also changed as compared
ethyl -glucopyranoside with fatty acids was reported to
proceed twice as fast as that of the anomer.21 However,
b
9
with the corresponding protons in -CL. Additionally, the
9
a
Î
FIG. 2. The 250 MHz Proton NMR spectra of (a) ethyl glucopyranoside, and (b) the product, oligo( -CL) EGP conjugate, in DMSO-d .
Î
6
1474
Volume 52, Number 11, 1998

FIG. 3. Region 170±180 ppm of the 62.9 MHz 13C-NMR spectra of the product, oligo( -CL) EGP conjugate, (a) before and (b) after derivatization
with diazomethane in CDCl3.
Î
in the present work, porcine pancreatic lipase catalyzed
ring-opening of -caprolactone with ethyl - and -glu-
copyranoside (anomeric mixture) did not indicate a pref-
erence towards either of the anomers. This observation
25.9, and 24.9 ppm due to the carbons of the poly( -CL)
Î
together with distinct low intensity peaks due to the car-
bons of ethyl glucopyranoside at 103.5, 99.3, 77.4, 74.5,
74.2, 74.1, 72.7, 71.4, 71.1, 70.58, 64.8, 64.5, 63.47,
16.0, and 15.8 ppm were present. The resonance ascrib-
a
b
Î
is based on our NMR determination of the ratio of and
b
spectrum (Fig. 2b) of the product, an EGP-oligo ( -CL),
a
anomers in the EGP-oligo( -CL). In the proton NMR
able to the -methylene carbon of the hydroxyl end group
a
Î
of the oligo chain was observed at 62.8 ppm. The as-
signments of resonances in Fig. 3a suggest the absence
of the carboxylic acid end group in the product. Water in
Î
the anomeric ratio of 2:1 was the same as in the starting
material, EGP (Fig. 2a). The area of the doublet at 4.65
ppm to that at 4.15 ppm was used to estimate the an-
omeric ratio.
lipase-catalyzed ring-opening polymerization of -CL is
Î
known to act as an initiator.24±26 Thus in the present re-
action, water may act as a competing initiator. If a frac-
tion of the oligo( -CL) chains were initiated by water,
Ethyl glucopyranoside as a multifunctional initiator
having four hydroxyl groups presents a challenging task
in the ring-opening of -CL. These hydroxyl groups have
Î
this procedure would result in carboxyl terminal chain
ends having a carbonyl signal in the region between 178
and 176 ppm.27 We do not observe these carboxyl carbon
resonances in Fig. 3a. The down®eld peaks (173±174
ppm region) in Fig. 3a are due to carbonyl carbons of
the oligo chain. The absence of carboxyl resonances in
Î
similar reactivities and they can equally participate in the
ring-opening reaction. It is dif®cult to distinguish various
reaction sites involving primary and secondary hydroxyl
groups. There is no general basis for positionally speci®c
participation among the three secondary hydroxyl groups.
In addition, distinction between the chemical reactions
involving primary and secondary hydroxyls is also not
easy. Therefore determination of the site(s) of reaction in
Fig. 3a suggests that the initiation of the oligo( -CL)
Î
chain was by EGP and not by the water.
The absence of carboxyl groups in the product was
also tested by derivatization of the product with diazo-
methane. As a result of diazomethane derivatization, con-
version of carboxyl terminal groups to their methyl ester
derivative would cause a 4 ppm up®eld shift in corre-
sponding carbonyl signals. In a control experiment, oli-
go(CL) was prepared by the PPL-catalyzed ring-opening
EGP initiated -CL ring-opening polymerization is im-
portant. In simple cases, it is possible to identify the site
of chemical linkage between EGP and oligo- -CL by
Î
Î
comparing the chemical shifts of ethyl glucopyranosides
before and after the reaction. However, as a result of
overlapped resonance peaks in the proton NMR spectrum
of the product, it is dif®cult to assign all the resonances
polymerization of -CL.24 Terminal carboxyl functional-
Î
to identify the possible links between the oligo( -CL)
ities of oligo(CL) were esteri®ed with diazomethane, and
its 13C-NMR spectrum before (Fig. 4a) and after (Fig. 4b)
derivatization were compared. The signals observed for
Î
chain and the EGP head group. Alternatively, 13C-NMR
can be used for determination of the position of the
link(s) due to the wide dispersion of chemical shifts in
oligo( -CL) at 177.8 and 178.1 ppm assigned to carbox-
Î
1
comparison with H NMR.
ylic acid C O groups were shifted up®eld by 4 ppm
5
13C NMR Spectrum of the Product, EGP-Oligo(e-
CL). Figure 3a displays the region from 170 to 180 ppm
of the 13C-NMR spectrum of the product. The 13C-NMR
spectrum of the product also supported attachment of
subsequent to esteri®cation. Importantly, upon derivati-
zation with diazomethane the 13C-NMR spectrum of the
product, oligo( -CL) EGP conjugate (Fig. 3b), did not
Î
show any changes in chemical shift of the peaks in the
173±174 ppm region. This observation con®rms our ear-
lier assignment of these peaks to carbonyl moieties. It
ethyl glucopyranoside to the poly( -CL) chain (full spec-
Î
trum not shown). Resonances at 173.8, 64.4, 34.4, 28.7,
APPLIED SPECTROSCOPY
1475

FIG. 4. Region 170±180 ppm of the 62.9 MHz 13C-NMR spectra of oligo( -CL) (a) before and (b) after derivatization with diazomethane in
CDCl3.
Î
means that the initiation in our reaction was caused by
EGP and not by water.
To establish the link(s) between poly( -CL) side
chain(s) and the EGP, we need conclusive assignments of
individual carbon resonances on the EGP ring. The in-
terpretation of the 13C-NMR spectrum is complicated
mixture of and anomers of EGP, a two-dimensional
a b
NMR experiment on 13C enriched EGP was performed.
Assignments of all carbons signals in EGP were made
from a 13C±13C COSY experiment of 99% 13C enriched
EGP (Fig. 6; for synthesis, see Experimental section).
Assignments for a and b Anomers. The expanded
region from 55 to 85 ppm of the 13C±13C COSY spectrum
of the 13C-labeled ethyl glucopyranoside showing meth-
ylene and methine carbon resonances is shown in Fig. 6.
The contour plot of the 13C±13C COSY experiment pro-
vides carbon±carbon connectivities. With knowledge of
the assignment for one carbon, it is easy to ®nd the con-
nectivities for other carbons in the molecule. Peaks at
103.4 and 99.2 ppm were assigned to the anomeric car-
bons in and anomers, respectively.21
Î
since the resonances from the
and methylene carbons from the poly( -CL) side chain
and anomers of EGP
b
a
Î
are crowded, especially in the 60±80 ppm region. The
13C-NMR DEPT (distortionless enhancement by polariza-
tion transfer) experiment is useful in distinguishing res-
onances from different types of carbons.28 With the as-
sistance of this experiment, it is possible to edit the 13C-
NMR spectra showing resonances due only to methyl,
methylene, methine, or quaternary carbons. The DEPT-
90 experiment provides a spectrum showing the reso-
nances only for the methine carbons. On the other hand,
the DEPT-135 experiment shows positive and negative
resonances to distinguish methyl, methine, and methylene
resonances. The positive peaks in DEPT-135 are due to
methyl and methine carbons, while negative peaks are
due to methylene carbons. We used the DEPT-135 ex-
periment to edit the complex 13C NMR spectrum of the
product. To assist in the assignments of resonances of
b
a
First we discuss the assignments of the carbons for the
anomer. The anomeric carbon ( C-1) at 103.4 ppm
b
b
b
shows a cross-peak to the resonances at 74.3 ppm. On
the basis of this connectivity, we assign the resonance
peak at 74.3 ppm to the carbon C-2 in the anomer (
b
b
C-2). Further, the resonance at 74.3 ppm ( C-2) shows
b
an additional cross-peak to the carbon resonance at 77.6
ppm. On the basis of connectivity, the resonance at 77.6
ppm is due to C-3 resonance shows the connectivity to
b
cross-peak to the resonance of the carbon
b
C-4 at 71.0 ppm. The carbon C-4 at 71.0 ppm has a
EGP-Oligo( -CL), we also recorded the DEPT-135 spec-
Î
trum of the starting material, EGP.
C-5 at 77.6
b
In the 13C DEPT-135 spectrum of EGP, all carbohy-
drate signals (i.e., 60±80 ppm) were resolved (Fig. 5a).
In Fig. 5, positive peaks are for resonances of ±CH,
±CH3 types of carbons, whereas negative peaks are for
±CH2± types of carbons in the molecule. In the DEPT-
135 spectrum, there are no resonances for the quaternary
carbons. Even though we were able to distinguish differ-
ent types of carbons in EGP, we were unable to make
complete assignments in spite of assistance from the pub-
lished literature on these molecules.20,21 In order to make
unequivocal assignments of all carbon resonances in the
ppm. The peak for
C-5 is also connected to the peak
b
of C-6 at 62.0 ppm. The ethyl group was not labeled
b
and therefore is not observed in the present setup of the
experiment. In a similar way, starting from the resonance
for the carbon C-1 at 99.2 ppm, we are able to assign
anomer. Assignments for the
a
a
all the peaks for the
a
anomer are also shown in Fig. 6.
Our data show that chemical shifts of different carbons
in the EGP are sensitive to the stereochemistry at the
anomeric carbon (C-1). The chemical shift difference be-
tween respective carbons in and anomers varied from
a
b
1476
Volume 52, Number 11, 1998

FIG. 5. Region 60±77 ppm of the 62.9 MHz DEPT-135 13C-NMR spectra of (a) ethyl glucopyranoside, and (b) the product, oligo( -CL) EGP
Î
conjugate, in DMSO-d6. The resonance at 69.2 ppm is due to residual unreacted -caprolactone.
Î
0.3 to 4.2 ppm (Table I). Carbon C-6 shows a chemical
shift difference of 0.1 ppm, while the chemical shift dif-
ference for carbon C-3 is 3.2 ppm. The largest difference
in the chemical shift was observed for the anomeric car-
bon itself (C-1).
13C-NMR Spectrum Analysis for the Oligo(e-CL)-
EGP. In Fig. 5, we show the 60±80 ppm region of the
13C-NMR (DEPT-135) spectrum for (a) ethyl glucopyr-
anoside (EGP) and (b) EGP oligo(CL). The assignments
and structures are also shown in Fig. 5. Comparison of
the 13C-NMR data for EGP and the oligomer suggests a
2.4 ppm down®eld shift for carbon C-6. This difference
of chemical shift is observed for both and
anomers
b
a
(Fig. 5). The peaks at 61.9 and 62.0 ppm, which are due
to carbons C-6 in and anomers, moved down®eld to
a
b
64.5 ppm and 64.8 ppm in the oligo(CL)-EGP (Fig. 5b),
respectively. The resonances at 64.3 and 61.2 ppm are
assigned to the caprolactone’s oxymethylenes and the end
group, respectively, in the oligo( -CL) chain.
Î
Furthermore, an up®eld shift of about 3.0 ppm is ob-
served for the carbon C-5 in both
the oligo( -CL) EGP conjugate when compared with that
and
anomers in
b
a
Î
of the EGP. The contributing factor for the up®eld shift
to these resonances for the oligomer is discussed below.
13C-NMR Identi®cation of the Site of Linkage. As
discussed here and elsewhere,21 the determination of re-
action site(s) in the EGP ring is particularly dif®cult due
to the presence of four possible reaction sites, i.e., hy-
droxyl groups. Three different possibilities exist: (1) ran-
dom participation of different hydroxyls, leading to a
TABLE I. Chemical shift differences in the a and b anomers of
ethyl glucopyranoside (DMSO-d6).
Difference in the
Chemical shifts (ppm)
chemical shifts
( ppm)
db 2 da 5 D
Carbons
Anomer (
)
da
Anomer (
)
db
a
b
C-1
C-2
C-3
C-4
C-5
C-6
99.2
72.8
74.2
71.3
73.5
61.9
63.4
14.6
103.4
74.3
77.6
71.0
77.6
62.0
64.6
14.8
4.2
1.5
3.2
0.3
4.1
0.1
1.2
0.2
1
1
1
2
1
1
1
1
FIG. 6. Region 55±85 ppm of the 125.8 MHz 13C±13C NMR COSY-
45 spectrum of the 13C enriched ethyl glucopyranoside (13C-EGP) in
DMSO-d6.
C-1
C-2
²
²
APPLIED SPECTROSCOPY
1477

complex mixture; (2) equal participation of different hy-
droxyls, resulting in a product in which all hydroxyls take
part in the ring-opening process; and (3) regioselective
reaction at a speci®c hydroxyl position. In any of these
cases, changes in chemical shifts of the participating car-
bons will provide important information regarding partic-
ipation of the individual hydroxyls. Now we discuss the
importance of the NMR data to determine the position of
the possible link(s) of EGP in the lipase-catalyzed ring-
CL. This observation allowed us to conclude that PPL-
catalyzed ring-opening polymerization of -CL initiated
by EGP is a highly regioselective reaction.
Î
1. C. J. Malm, J. W. Meneh, D. L. Kendall, and G. D. Hiatt, Ind. Eng.
Chem. 43, 684 (1951).
2. B. D. Martin, S. A. Ampofo, R. J. Linhardt, and J. S. Dordick,
Macromolecules 25, 7081 (1992).
3. I. Ikeda and A. M. Klibanov, Biotechnol. Bioeng. 42, 788 (1993).
4. J. M. Sugihara, Adv. Carbohydr. Chem. Biochem. 8, 1 (1953).
5. A. H. Haines, Adv. Carbohydr. Chem. Biochem. 33, 11 (1976).
6. M. L. Wolfrom and W. A. Szarek, in The Carbohydrates Chemistry
and Biochemistry, W. Pigman and D. Horton, Eds. (Academic
Press, New York, 1972), 2nd ed., Vol. IA, pp. 217±251.
7. A. H. Haines, Adv. Carbohydr. Chem. Biochem. 39, 13 (1981).
8. M. Therisod and A. M. Klibanov, J. Am. Chem. Soc. 108, 5638
(1986).
9. H. M. Sweers and C. H. Wong, J. Am. Chem. Soc. 108, 6421
(1986).
10. A. M. Klibanov, Biotechnol. Bioeng. 29, 648 (1987).
11. M. Therisod and A. M. Klibanov, J. Am. Chem. Soc. 109, 3977
(1987).
12. M. Kloosterman, E. W. J. Mosmuller, H. E. Schoemaker, and E. M.
Meijer, Tetrahedron Lett. 28, 2989 (1987).
13. J. Chopineau, F. D. McCafferty, M. Therisod, and A. M. Klibanov,
Biotechnol. Bioeng. 31, 208 (1988).
opening polymerization of -CL.
Î
The signi®cant changes in chemical shifts for the car-
bon C-6 and C-5 in Fig. 5 strongly suggest the in¯ uence
of the polymer link to the EGP at carbon C-6. The cou-
pling of EGP to the polymer chain is clearly evidenced
by the chemical shift change for the carbon C-6. There
is a 2.5 ppm down®eld shift as result of EGP linkage to
the polymer chain at the C-6. This chemical shift change
is possible only if the polymer is attached at the carbon
C-6. Additionally, an up®eld shift of 3.0 ppm for the
carbon C-5 is a strong indication of the effect to this
g
carbon by the ester carbonyl of the polymer chain.29 Im-
portantly, there is no signi®cant in¯ uence on carbons C-
2 to C-4. On the basis of these observations, our NMR
data strongly suggest that the C-6 hydroxyl of EGP
serves as a site for lipase-catalyzed ring-opening poly-
14. S. Riva, J. Chopineau, A. P. G. Kieboom, and A. M. Klibanov, J.
Am. Chem. Soc. 110, 584 (1988).
15. G. Carrea, S. Riva, F. Secundo, and B. Danieli, J. Chem. Soc.,
Chem. Commun. 1057 (1989).
merization of -CL. There is no signi®cant impact on the
Î
16. J. F. Shaw, V. Gotor, and R. J. Pulido, J. Chem. Soc., Perkin Trans.
1 491 (1991).
17. A. K. Prasad, M. D. Sorensen, V. S. Parmar, and J. Wengel, Tet-
rahedron Lett. 36, 6163 (1995).
18. A. M. Klibanov, CHEMTECH 16, 354 (1986).
19. H. Seino, T. Uchibori, T. Nishitani, and S. Inmasu, J. Am. Oil.
Chem. Soc. 61, 1761 (1984).
20. F. Bjorkling, S. E. Godfredsen, and O. Kirk, J. Chem. Soc., Chem.
Commun. 934 (1989).
chemical shifts of other carbons of EGP as a result of
lipase catalyzed ring-opening polymerization of -CL.
Î
The NMR data thus lead us to conclude that the ring-
opening polymerization of -CL was a regioselective one
Î
with exclusive participation of hydroxyl group on the car-
bon C-6 of EGP.
21. K. Adelhorst, F. Bjorkling, S. E. Godfredsen, and O. Kirk, Synthesis
112 (1990).
CONCLUSION
22. K. S. Bisht, F. Deng, R. A. Gross, D. L. Kaplan, and G. Swift, J.
Am. Chem. Soc. 120, 1363 (1998).
23. W. P. Aue, E. Bartholdi, and R. R. Ernst, J. Chem. Phys. 64, 2229
(1976).
24. R. T. MacDonald, S. K. Pulapura, Y. Y. Svirkin, R. A. Gross, D.
L. Kaplan, J. A. Akkara, G. Swift, and S. Wolk, Macromolecules
20, 73 (1995).
25. K. S. Bisht, Y. Y. Sivirkin, L. A. Henderson, R. A. Gross, D. L.
Kaplan, and G. Swift, Macromolecules 30, 7735 (1997).
26. K. S. Bisht, L. A. Henderson, R. A. Gross, D. L. Kaplan, and G.
Swift, Macromolecules 29, 7759 (1996).
27. H. Uyama and S. Kobayashi, Chem. Lett. 1149 (1993).
28. D. M. Doddrell, D. T. Pegg, and M. R. Bendall, J. Magn. Reson.
48, 323 (1982).
29. E. Breitmaier and G. Bauer, 13C NMR Spectroscopy: a Working
Manual with Exercises (Harwood Academic Publisher, New York,
1984), pp. 54±57.
We have demonstrated the applicability of 13C±13C
COSY and DEPT-135 NMR experiments in assignments
of carbon resonances in a mixture of
and
of ethyl glucopyranoside and the product obtained by
anomers
b
a
ring-opening polymerization of -CL initiated by EGP
Î
and catalyzed by porcine pancreatic lipase. The NMR
data allowed us to investigate the regioselective initiation
in the lipase-catalyzed ring-opening polymerization re-
action. Analysis of the data strongly suggested the incor-
poration of EGP to the polymer chain during ring-open-
ing polymerization. Among all the possibilities, NMR
data identi®ed the hydroxyl on carbon C-6 as the exclu-
sive participant in the ring-opening polymerization of -
Î
1478
Volume 52, Number 11, 1998