Highly selective primary alkoxycarboxylation and esterification of unprotected pyranose derivatives mediated by scandium(III) triflate catalysis

Article abstract of DOI:10.1002/ejoc.201200261

A highly selective method for the alkoxycarboxylation and acylation of primary alcohols of pyranose derivatives is described. The reaction is high yielding and proceeds under mild conditions with 0.15-1 mol-% Sc(OTf) ₃ used in combination with anhydrides or pyrocarbonates at 40-50 °C. Selectivities observed for alkoxycarboxylation of unprotected pyranose derivatives are > 95 and this constitutes a significant advantage over existing methods. Mechanistic implications, including the role of steric demand and metal-heteroatom coordination are also discussed.

Full text of DOI:10.1002/ejoc.201200261

Job/Unit: O20261
/KAP1
Date: 08-05-12 10:03:07
Pages: 6
SHORT COMMUNICATION
DOI: 10.1002/ejoc.201200261
Highly Selective Primary Alkoxycarboxylation and Esterification of
Unprotected Pyranose Derivatives Mediated by Scandium(III) Triflate
Catalysis
Michael S. McClure,*[a] Malcolm B. Berry,[b] Darren Caine,[b] Claire Crawford,[b]
Brian C. Crump,[a] Bobby N. Glover,[a] Sandeep B. Kedia,[a] Alan Millar,[a]
Mark B. Mitchell,[a] Christopher J. Nichols,[b] Daniel E. Patterson,[a] and Jeremiah Powers[a]
Keywords: Scandium / Pyranose / Alkoxycarboxylation / Esterification / Chemoselectivity
A highly selective method for the alkoxycarboxylation and
acylation of primary alcohols of pyranose derivatives is de-
scribed. The reaction is high yielding and proceeds under
mild conditions with 0.15–1 mol-% Sc(OTf)3 used in combi-
nation with anhydrides or pyrocarbonates at 40–50 °C. Selec-
tivities observed for alkoxycarboxylation of unprotected pyr-
anose derivatives are Ͼ 95%, and this constitutes a signifi-
cant advantage over existing methods. Mechanistic implica-
tions, including the role of steric demand and metal–hetero-
atom coordination are also discussed.
Selective esterification or alkoxycarboxylation of the 6Ј-
hydroxy group of unprotected pyranose derivatives is a dif-
ficult synthetic challenge. There are a number of examples
in the literature of selective acylation of pyranose deriva-
tives, mostly employing lipase-catalyzed acetylations,[8] as
well as a few using bulky amine bases[9] or functionalized
DMAP catalysts.[10] In contrast, there are almost no reports
of selective alkoxycarboxylations of pyranose derivatives.[11]
The few reports of high selectivity were only achieved with
a large molar excess of the pyranose substrate.[12] For this
reason, a general, highly selective method for the alkoxy-
carboxylation of pyranose derivatives is desirable.
Introduction
Lanthanide-mediated catalysis has attracted much atten-
tion, and scandium(III) triflate has been one of the synthet-
ically most useful lanthanide catalysts.[1] Scandium(III)
triflate exhibits good stability in air and good solubility in
organic or aqueous media and can be recovered without
significant loss of activity.[2] Our interest in the area of scan-
dium catalysis stems from the need to effect a highly selec-
tive ethoxycarboxylation of an unprotected pyranose deriv-
ative. We report herein the successful outcome of our re-
search to develop a new catalytic reaction to achieve this
transformation; this communication includes our prelimi-
nary evaluation of the selectivity, scope, kinetics, and
mechanism.
Existing methods for alkoxycarboxylation include the use
of chloroformate[3] or imidazolecarboxylate derivatives, oxi-
dative carbonylation[4] and metal-catalyzed transcar-
bonylation with diethyl carbonate.[5] Bartoli and co-workers
have also reported the ethoxycarboxylation of phenols by
using diethyl pyrocarbonate with Lewis acid catalysis. Simi-
larly, the reactivity of dicarbonates and anhydrides with pri-
mary and secondary alcohols by using Lewis acid cataly-
sis[6] has been reported.[7]
Table 1. Initial method comparison.
[a] Product Development, GlaxoSmithKline,
Five Moore Drive, P. O. Box 13398, Research Triangle Park,
NC 27709, USA
Entry
X
Catalyst (mol-%)
Base
Ratio[a] 1/d[b] Conv.[a,c]
1
2
Cl[d]
OCO2Et[e]
pyridine (5)
Sc(OTf)3 (0.15)
2,6-lutidine
none
4.4:1
57:1
97%
100%
Fax: +1-919-315-8735
E-mail: michael.s.mcclure@gsk.com
[a] Ratios and conversions determined by HPLC at 228 nm. [b]
d = dicarbonate by-products, e.g. the 2Ј,6Ј-dicarbonate of 1. [c]
Conversion = 100ϫ[(prod + by-prod)/(prod + by-prod + starting
material)]. [d] Solvent: toluene. [e] Solvent: toluene/ethanol (4:1).
[b] Product Development, GlaxoSmithKline,
Gunnels Wood Road, Stevenage, Hertfordshire, SG1 2NY, UK
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201200261.
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1

Job/Unit: O20261
/KAP1
Date: 08-05-12 10:03:07
Pages: 6
M. S. McClure et al.
SHORT COMMUNICATION
In our initial attempt to prepare the primary ethyl car- relatively fast reaction. This reaction was successfully run
bonate derivative 2 from unprotected pyranose 1 by using on a multi-kilogram scale by using 0.15 mol-% Sc(OTf)3 to
ethyl chloroformate and amine bases gave a moderate selec- give a 93% yield (97% selectivity) of the desired product 2.
tivity with Ͼ 15% dicarbonates, see Table 1, vide infra. Op- The success of diethyl pyrocarbonate in this reaction was
timization of reaction parameters led to marginal gains in surprising, given that ethanol is a reaction byproduct. The
selectivity; therefore, canvassing a broader scope of reaction ethanol did not compete with the substrate towards ethoxy-
catalysts and reagents was essential.
carboxylation, and the reaction went to completion with
only 1 equiv. of diethyl pyrocarbonate. In fact, the reaction
could be run in ethanol as a solvent, without significant
competitive ethoxycarboxylation of ethanol. Furthermore,
use of ethanol as a solvent often provided a rate enhance-
ment.
When considering a plausible explanation for the ob-
served selectivity, we first postulated that pyrazole nitrogen
complexation with scandium may be involved, if not re-
quired. However, the selective reaction was not specific to
1 but proved to be general, giving high selectivities with a
variety of pyranoses, including substrates that did not bear
a pendant nitogen atom (Table 2). Whereas pyrazole 2 and
quinoline analogue 3 were produced with high selectivities,
non-heteroatom-containing analogues also gave selective re-
actions (4, 5, and 6). Other O-substitutions at the anomeric
Results and Discussion
Screening was conducted with a number of Lewis acid
catalysts in combination with ethyl chloroformate, diethyl
carbonate (DEC) or diethyl pyrocarbonate (DEPC). Di-
ethyl pyrocarbonate with Sc(OTf)3 gave complete conver-
sion with very high selectivity (Ͼ 95% C-6), whereas ethyl
chloroformate gave lower selectivity (15–20% dicarbon-
ates), and diethyl carbonate was unreactive. Generally, CuI
and CuII salts, including triflate and tetrafluoroborate, were
also effective and provided selectivities similar to those of
Sc(OTf)3 but required high catalyst loadings for a robust
reaction.
Because of the low natural abundance of various lanth-
anides,[13] further screening was conducted with Sc(OTf)3 to
see if low catalyst loading was achievable. In the case of 1,
only 0.15 mol-% of Sc(OTf)3 provided robust results with a
Table 2. Scope of the alkoxycarboxylation; selectivity (yield).[a]
[a] Selectivity = 100ϫ(desired monocarbonate/desired product + Figure 1. Competition experiments carried out in methyltetra-
starting material + byproducts) measured by HPLC at 228 nm; iso-
lated yields. [b] See the Supporting Information for experimental
details.
hydrofuran at 40 °C with 1 mol-% Sc(OTf)3. Plot of the formation
of 14 (squares), 17 (triangles), 13 (dashed line), 22 (solid line), 18
(dotted line) and 21 (circles).
2
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O20261
/KAP1
Date: 08-05-12 10:03:07
Pages: 6
Alkoxycarboxylation and Esterification of Pyranose Derivatives
position were well tolerated including simple alkyl (6), not likely to play a significant role as demonstrated by the
phenyl (4), formyl (10), and nitro (8). The C-glycoside 7 C-glycoside analogue 7 and by the insensitivity of the reac-
reacted also successfully, though notably slower than N- tion to the anomeric stereochemistry (4α, 4β). The role of
bearing substrates. Thus, nitrogen coordination may be oc- the endocyclic pyranose oxygen atom, however, was less evi-
curring, but is not required for the observed 6Ј-hydroxy dent.
selectivity. The reaction was insensitive to the anomeric
A number of competition experiments were devised to
stereochemistry as evidenced by the fact that 4α and 4β futher probe the effect of substrate structure on the reac-
could be synthesized with nearly identical results. The pyra- tion. Whereas substrates 11 and 12 are acyclic and arguably
nose stereochemistry at C-4 was equally unimportant as the simplistic (A, Figure 1), they approximate a glucose frame-
selectivities for the galactose analogue 5 and the glucose work without and with the pyranose oxygen atom, respec-
analogue 4β were effectively identical. Substrates containing tively. When equimolar amounts (0.5 equiv. each) of 4-
anilines were not amenable to the ethoxycarboxylation phenyl-1-butanol (11) and 2-[(phenylmethyl)oxy]ethanol
chemistry and were likely the result of coordinative effects (12) were treated with effectively 1 equiv. of diethyl pyrocar-
with the metal atom as well as by-product formation includ- bonate and a catalytic amount of Sc(OTf)3, the oxygen ana-
ing carbamation. Additionally, ribose sugars were problem- logue 12 proceeded at a remarkably faster rate compared to
atic in part, due to the higher propensity for acyl mi- the non-oxygenated analogue 11 (Figure 1). The effect of
gration.[14]
tether length was also examined in a similar manner (B,
To further establish the underlying basis for the observed Figure 1), and a longer tether led to a decrease in the reac-
selectivity, a purely steric argument was considered as well tion rate. To determine if oxygen–metal coordination could
as a potential templating effect with oxygen coordination be harnessed to overcome kinetic/steric differentiation in
through the pyranose ring. The anomeric oxygen atom is favor of secondary alcohol ethoxycarboxylation, the sec-
Figure 2. Catalytic cycle and graphical overlay plot for experiments 1 (diamonds), 2 (triangles), 3 (squares), 4 (open circles), 5 (crosses)
and
6
(circles). The regressed constants are K1
=
2.7ϫ10–3 Ϯ7.0ϫ10–4 mm–1, K2
=
2.6ϫ10–4 Ϯ6.6ϫ10–5 mm–1 and k3
=
1.3ϫ10–1 Ϯ3.2ϫ10–2 min–1.
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
3

Job/Unit: O20261
/KAP1
Date: 08-05-12 10:03:07
Pages: 6
M. S. McClure et al.
SHORT COMMUNICATION
ondary alcohol 19 with an oxygenated tether was compared though still provided excellent selectivities. Anhydrides in-
with the primary alcohol 20 containing a non-oxygenated cluding acetic and benzoic anhydride were also selective un-
tether (C, Figure 1). In this instance, the primary alcohol der these conditions (Entries 5, 6 and 10). Cyclic anhydrides
was still preferentially ethoxycarboxylated. These experi- (Entries 7 and 8), while reactive, never proceeded to com-
ments support the hypothesis that primary/secondary selec- pletion and resulted in lower isolated yields; this is likely
tivity is only driven in part by steric differentiation. Further, due to additional ligand exchange/coordination with the
the overall reaction rate is enhanced by heteroatom–metal scandium catalyst.
coordination, and this coordination enhances the rate suffi-
ciently that substrates can be ethoxycarboxylated with high
Table 4. Acylating agents.
selectivity over non-heteroatom-containing primary
alcohols even if in ethanol as the reaction solvent.
To gain insight into the catalytic cycle for the synthesis
of 2, reaction progress analysis was carried out by using
graphical overlay methodology.[15] Six experiments were
performed by utilizing a mixture of same and different ex-
cesses of diethyl pyrocarbonate over starting sugar
(Table 3). The experiments were performed in parallel,
tracking the disappearance of starting material by HPLC
to afford an excellent fit (r2 = 0.99) of the six reaction pro-
files to the catalytic rate equation for a two-intermediate
catalytic cycle (Figure 2). A key graphical overlay in the
analysis is shown where alignment of normalized rate vs. [1]
Entry Product
R
X
Selectivity[a] Yield
for all six experiments is apparent.
1
2
3
4
5
6
7
8
9
23
2
R1
R1
R1
R1
R1
R1
R1
R1
R2
R3
OCH3
OCH2CH3
O-tert-butyl
O-allyl
98%
97%
92%
95%
95%
83%
56%
74%
92%
82%
75%
93%
87%
79%
68%
75%
41%
46%
59%
50%
Table 3. Kinetic experiments.[a]
24
25
26
27
28
29
30
31
Entry 1 [mm] DEPC[b] [mm] DEPC excess[c] [mm] Sc(OTf)3 [mm]
CH3
phenyl
1
2
3
4
5
6
270
405
270
270
270
270
311
446
405
405
230
284
41
41
135
135
–40
14
41
41
41
81
41
41
(CH2)2CO2H[b]
(CH2)3CO2H[c]
OCH3
10
CH3
[a] Selectivity = 100ϫ(desired monocarbonate/desired product +
starting material + byproducts) measured by HPLC at 228 nm; iso-
lated yields. [b] Acylating agent: succinic anhydride. [c] Acylating
agent: glutaric anhydride.
[a] All kinetic experiments were performed at 58 °C in industrial
methylated spirit (IMS)/toluene (1:4). [b] Diethyl pyrocarbonate. [c]
Concentration of DEPC – concentration of 1.
This analysis revealed a number of facets of the catalytic
pathway. First, the same-excess experiments demonstrated
a robust catalyst system devoid of issues such as substrate
inhibition, product inhibition and catalyst degradation with
the expected first-order dependence upon catalyst concen-
tration. Second, the different-excess experiments allowed
the delineation of the following cycle features. Binding of 1
to free catalyst is the first step, followed by binding of di-
ethyl pyrocarbonate. The rate-limiting step in the process
is the binding of diethyl pyrocarbonate, and under typical
reaction conditions the catalyst is partitioned between un-
bound catalyst * and sugar-bound catalyst A*. Clearly, ki-
netics cannot reveal details of the exact nature of the bind-
ing between catalyst and substrate, but it does define the
sequence of events that are occurring.
Conclusions
A mild and highly chemoselective method for alkoxy-
carboxylation and esterification of pyranoses by using scan-
dium(III) triflate catalysis is presented. The scope of this
reaction was explored with variation of alcohol and acyl-
ation substrates. The reaction proceeds with low catalyst
loading and high selectivity for a range of pyranose sub-
strates making it a valuable synthetic method. In de-
veloping a preliminary mechanistic hypothesis, data were
also presented, which demonstrates the influence of steric
demand and heteroatom coordination.
The successful ethoxycarboxylation of a variety of sub-
strates by using diethyl pyrocarbonate and scandium(III) Experimental Section
triflate led to the question of whether this methodology
General Procedure for Carboxylation[16] (Glucopyranoside 4α): To
could be further extended to other bidentate acylating rea-
gents, including other pyrocarbonates and anhydrides. In-
deed other alkyl pyrocarbonates work well including
methyl, tert-butyl and allyl pyrocarbonate (Table 4, En-
a solution of phenyl α-d-glucopyranoside (150 mg, 0.57 mmol) in
toluene (1.1 mL) and ethanol (0.4 mL) was added scandium triflate
(3 mg, 0.006 mmol) and diethyl pyrocarbonate (104 mg,
0.64 mmol). The solution was heated to 50 °C for 30 min until the
tries 1–4, 9). tert-Butyl pyrocarbonate was markedly slower, reaction was complete according to HPLC [crude selectivity: 4α/
4
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O20261
/KAP1
Date: 08-05-12 10:03:07
Pages: 6
Alkoxycarboxylation and Esterification of Pyranose Derivatives
[4] Z. Zhang, X. Ma, P. Zhang, Y. Li, S. Wang, J. Mol. Catal. A:
Chem. 2007, 266, 202–206.
[5] M. Kantam, U. Pal, B. Sreedhar Choudary, Adv. Synth. Catal.
2007, 349, 1671–1675.
[6] G. Bartoli, M. Bosco, A. Carlone, M. Locatelli, E. Marcantoni,
P. Melchiorre, P. Palazzi, L. Sambri, Eur. J. Org. Chem. 2006,
4429–4434.
dicarbonates (95:5)] before concentrating to dryness. The residue
was chromatographed on silica [heptane/ethyl acetate (1:1)] to af-
ford the title compound as a white foam (168 mg, 87% yield).
FTIR (ATR): λmax = 3338, 3246, 2959, 2911, 1739, 1262, 1222,
1
1072, 1006, 752 cm–1. H NMR (500 MHz, [D6]DMSO): δ = 1.17
(t, J = 7.1 Hz, 3 H), 3.16–3.22 (m, 1 H), 3.38–3.43 (m, 1 H), 3.61–
3.67 (m, 1 H), 3.72 (ddd, J1 = 2.0, J2 = 6.2, J3 = 10.0 Hz, 1 H),
3.97–4.10 (m, 2 H), 4.14 (dd, J1 = 6.1, J2 = 11.6 Hz, 1 H), 4.26
(dd, J1 = 2.1, J2 = 11.7 Hz, 1 H), 5.05 (d, J = 5.2 Hz, 1 H), 5.13
(d, J = 6.3 Hz, 1 H), 5.29 (d, J = 5.9 Hz, 1 H), 5.40 (d, J = 3.7 Hz,
1 H), 7.01 (t, J = 7.3 Hz, 1 H), 7.06 (d, J = 8.2 Hz, 2 H), 7.30 (t,
J = 7.8 Hz, 2 H) ppm. 13C NMR (125 MHz, [D6]DMSO): δ = 14.0,
63.4, 66.6, 69.8, 70.4, 71.3, 72.8, 97.6, 116.8, 121.9, 129.3, 154.3,
156.9 ppm. HRMS (+ESI): calcd. for [M + H]+ 329.1231; found
329.1237.
[7]
[8]
a) G. Bartoli, M. Bosco, A. Carlone, R. Dalpozzo, M. Lo-
catelli, P. Melchiorre, L. Sambri, J. Org. Chem. 2006, 71, 9580–
9588; b) K. Ishihara, M. Kubota, H. Kurihara, H. Yamamoto,
J. Am. Chem. Soc. 1995, 117, 4413–4414; c) K. Ishihara, M.
Kubota, H. Kurihara, H. Yamamoto, J. Org. Chem. 1996, 61,
4560–4567.
a) M. Therisod, A. M. Kibanov, J. Am. Chem. Soc. 1986, 108,
5638–5640; b) A. Sharma, S. Chattopadhyay, Biotechnol. Lett.
1993, 15, 1145–1146; c) D. Bruno, M. Luisetti, G. Sampognaro,
G. Carrea, S. Riva, J. Mol. Catal. B 1997, 3, 193–201; d) S.
Park, R. J. Kazlauskas, J. Org. Chem. 2001, 66, 8395–8401.
K. Ishihara, H. Kurihara, H. Yamamoto, J. Org. Chem. 1999,
58, 3791–3793.
Supporting Information (see footnote on the first page of this arti-
cle): Experimental procedures, copies of the 1H and 13C NMR
spectra.
[9]
[10]
T. Kurahashi, T. Mizutani, J. Yoshida, Tetrahedron 2002, 58,
8669–8677.
[11] a) R. Pulido, V. Gotor, J. Chem. Soc. Perkin Trans. 1 1993,
589–592; b) L. Lay, L. Panza, S. Riva, M. Khitri, S. Tirendi,
Carbohydr. Res. 1996, 291, 197–204.
[12] a) M. Allainmat, D. Plusquellec, Tetrahedron Lett. 1991, 32,
2751–2754; b) M. Allainmat, P. L’Haridon, L. Toupet, D. Plus-
quellec, Synthesis 1990, 27–32.
Acknowledgments
We thank Hiromu Harada of Kissei Pharmaceutical Co., Ltd. for
thoughtful discussions.
[1] a) S. Mishra, Coord. Chem. Rev. 2008, 252, 1996–2025; b) A. [13] J. Royset, N. Ryrum, Int. Mater. Rev. 2005, 50, 19–44.
Abbasi, P. Lindqvist-Reis, L. Eriksson, D. Sandstroem, S. Li-
din, I. Persson, M. Sandstroem, Chem. Eur. J. 2005, 11, 4065–
4077; c) M. Distaso, E. Quaranta, J. Catal. 2008, 253, 278–
288; d) M. Skinner, B. Tyrrell, B. Ward, P. Mountford, J. Or-
ganomet. Chem. 2002, 647, 145–150; e) S. Cotton, Polyhedron
1999, 18, 1691–1715.
[14] O. Chevallier, M. Migaud Beil. J. Org. Chem. 2006, 2, No. 14.
[15] a) D. Blackmond, Angew. Chem. 2005, 117, 4374; Angew.
Chem. Int. Ed. 2005, 44, 4302–4320; b) D. Blackmond, J. Org.
Chem. 2006, 71, 4711–4722.
[16] See the Supporting Information for minor differences and com-
plete experimental details (e.g. the use of alternative solvents
due to compound-specific solubility differences).
Received: March 4, 2012
[2] a) S. Kobayashi, Eur. J. Org. Chem. 1999, 15–27; b) S. Kobaya-
shi, I. Hachiya, H. Ishitani, M. Araki, Synlett 1993, 472–474.
[3] W. Borsche, Ber. Dtsch. Chem. Ges. 1924, 57B, 1620–1622.
Published Online: ᭿
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5

Job/Unit: O20261
/KAP1
Date: 08-05-12 10:03:07
Pages: 6
M. S. McClure et al.
SHORT COMMUNICATION
Sc-Catalyzed Alkoxycarboxylations
A novel Sc-catalyzed alkoxycarboxylation
and esterification reaction of unprotected
pyranose substrates is described. This meth-
od gives nearly complete selectivity com-
pared with existing methods and avoids
known problems with polycarboxylation/
acylation. A preliminary mechanistic mo-
del is proposed.
M. S. McClure,* M. B. Berry, D. Caine,
C. Crawford, B. C. Crump, B. N. Glover,
S. B. Kedia, A. Millar, M. B. Mitchell,
C. J. Nichols, D. E. Patterson,
J. Powers ........................................... 1–6
Highly Selective Primary Alkoxycarboxyl-
ation and Esterification of Unprotected
Pyranose Derivatives Mediated by Scan-
dium(III) Triflate Catalysis
Keywords: Scandium / Pyranose / Alkoxy-
carboxylation / Esterification / Chemose-
lectivity
6
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0