Reductive openings of benzylidene acetals. Kinetic studies of borane and alane activation by Lewis acids

Article abstract of DOI:10.1016/j.carres.2008.08.022

The reaction kinetics for a number of reductive openings of methyl 2,3-di-O-benzyl-4,6-O-benzylidene-α-d-glucopyranoside have been investigated. Openings to give free HO-6 (using BH₃·THF-AlCl₃-THF or LiAlH₄-AlCl₃

Full text of DOI:10.1016/j.carres.2008.08.022

Carbohydrate Research 343 (2008) 2997–3000
Contents lists available at ScienceDirect
Carbohydrate Research
journal homepage: www.elsevier.com/locate/carres
Note
Reductive openings of benzylidene acetals. Kinetic studies of borane
and alane activation by Lewis acids
Richard Johnsson, Risto Cukalevski, Fanny Dragén, Damir Ivanisevic, Ida Johansson, Linn Petersson,
*
Erika Elgstrand Wettergren, Ka Bo Yam, Beatrice Yang, Ulf Ellervik
Organic Chemistry, Lund University, PO Box 124, SE-221 00 Lund, Sweden
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 5 August 2008
Received in revised form 19 August 2008
Accepted 22 August 2008
Available online 29 August 2008
The reaction kinetics for a number of reductive openings of methyl 2,3-di-O-benzyl-4,6-O-benzylidene-a-
D
-glucopyranoside have been investigated. Openings to give free HO-6 (using BH
LiAlH –AlCl –Et O) follow first order kinetics, while reactions yielding free HO-4 (using BH
AlCl –THF or BH –BF ꢀOEt –THF) follow higher order kinetics. The addition of water to the
ꢀNMe
BH –AlCl –THF results in faster reactions. The BH –AlCl –THF system constitutes a border-
ꢀNMe
3
ꢀTHF–AlCl
3
–THF or
4
3
2
3
ꢀNMe
3
–
3
3
3
3
2
3
3
3
3
ꢀSMe
2
3
line case, yielding both free HO-6 (by a first order reaction) and free HO-4 (by a higher order reaction).
These results correlate well with the concept of regioselectivity by activation of borane complexes.
Ó 2008 Elsevier Ltd. All rights reserved.
Keywords:
Benzylidene acetals
Borane reduction
Regiochemistry
Lewis acids
Reductive openings of cyclic acetals (e.g., 4,6-O-benzylidene
acetals of hexopyranosides) are immensely important in modern
synthetic carbohydrate chemistry, and crucial for protective group
introduction and manipulation. In the original method, which was
HO
BnO
BnO
O
BnO
OMe
2
4 3
intensely explored by Lipták, the reagent combination LiAlH –AlCl
LiAlH , AlCl , CH Cl , Et O, 94% or
4
3
2
2
2
Ph
O
is used to open 4,6-O-benzylidene acetals to give 4-O-benzyl ethers
O
BH •NMe , AlCl , toluene, 50%.
3
3
3
and a free 6-hydroxyl group.1 However, these conditions are too
,2
O
BnO
harsh for many common protective groups (e.g., acetates) and
BnO
OMe
the reagent combination NaCNBH
3
–HCl–THF was introduced by
1
Garegg. Interestingly, these conditions gave the opposite regiose-
BnO
O
lectivity, compared to the original method.3 Furthermore, the
,4
HO
BnO
Garegg group found that the reagent combination BH
gave different regioselectivity by variation of the solvent, that is, 6-
O-benzyl ethers in THF and 4-O-benzyl ethers in toluene.
3
ꢀNMe
3 3
–AlCl
BnO
OMe
3
5
–7
How-
NaCNBH , HCl, THF, 81% or
3
ever, due to degradation, reactions in toluene usually gave low
yields. The methods are summarized in Scheme 1.
BH •NMe , AlCl , THF, 71%.
3
3
3
The underlying principles for the regioselectivity have not been
Scheme 1. Examples of regioselective openings of benzylidene acetals.
8
fully understood. The generally accepted mechanism, proposed by
9
Garegg, explains the regioselective outcome by the difference in
steric bulk between AlCl
problems associated with this mechanistic explanation, with the
major issue to explain being the regiochemical outcome of the
3
and a proton. However, there are several
AlCl
3
ꢀTHF would preferably associate with the less sterically hin-
dered O-6 to give 4-O-benzyl ethers.
Recently, we presented a mechanistic explanation for these
observations. From kinetic experiments, B NMR spectroscopy,
and Hammett plots, we found that BH is activated by AlCl ,
3
which renders the borane the most electrophilic species. Conse-
quently, the regioselectivity is directed by addition of the borane
1
0
11
reaction with BH
ethers, despite the obvious conclusion that the strongly solvated
3
ꢀNMe
3
–AlCl
3
in THF, which gives 6-O-benzyl
3
ꢀNMe
3
to the most basic acetal oxygen. In contrast, BH
vated by the Lewis acid and thus, regioselectivity is directed by
3
ꢀTHF is not acti-
*
Corresponding author. Tel.: +46 46 222 8220; fax: +46 46 222 8209.
E-mail address: ulf.ellervik@organic.lu.se (U. Ellervik).
0
008-6215/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carres.2008.08.022

2
998
R. Johnsson et al. / Carbohydrate Research 343 (2008) 2997–3000
the addition of water sped up otherwise slow reactions.21 The opti-
mized conditions turned out to be 2 equiv of water, 4 equiv of
BH , and 6 equiv of AlCl . To further investigate the impor-
ꢀNMe
tance of water in the reduction, the kinetics of the reduction of
Mechanism
directed by
Lewis Acid
Mechanism
directed by
borane
MeO
BnO
O
O
3
3
3
O
Ph
Path A
AlCl •THF
Path B
BH •NMe
3
AlCl •THF
compound 1 using the water–BH
gated (Fig. 1c).
3
ꢀNMe
3
–AlCl
3
system was investi-
OBn
3
1
3
3
To our surprise, we found that equimolar amounts of AlCl
3
and
water completely quenched the reaction, while a 3:1 ratio gave a
rate enhancement of approximately four times. However, the addi-
MeO
BnO
O
AlCl3
MeO
O
BH3
O
O
tion of even more AlCl
AlCl the rates were similar to reaction without water. These obser-
vations can be rationalized by the reactions of water with AlCl . In
a series of experiments, Gálová showed that the conductivity of a
solution of AlCl in THF reached a maximum at a molar ratio of
AlCl
the charged species AlCl
3
lowered the reaction rate, and at 8 equiv of
O
Ph
BnO
O
Ph
3
BnO
BnO
3
BH •THF
AlCl •THF
3
3
3
3
and water, of 1:2, which corresponds to the formation of
þ
ꢁ
22
MeO
BnO
O
MeO
BnO
O
ðH
2
OÞ and AlCl4
.
Further addition
2
4
OH
Ph
O
of water led to lowered conductivities due to the formation of un-
O
2
OH Ph
3 2 6
charged AlCl (H O) . It is reasonable to assume that the formed
BnO
BnO
3
aluminum complexes can act as proton sources that activate either
the borane or the acetal. Because equimolar amounts of water and
Lewis acid gave no reaction, the formed complex is not strong en-
ough for activation of both the borane and the acetal. Brown and
Scheme 2. The regiochemical outcome is directed by the relative nucleophilicity of
the acetal oxygens toward the Lewis acid (Path A) or the activated borane (Path B).
Murrey showed that the rate of hydrolysis of BH
3
ꢀNMe
3
in aqueous
diglyme is almost negligible, but that the complex can be activated
association of AlCl
product. To summarize, the regioselectivity is directed by associa-
tion of the most electrophilic species (i.e., BH or AlCl depending
3
to the same acetal oxygen, giving the opposite
by proton sources such as acetic acid or mineral acids.2
3,24
We
ꢀNMe
therefore propose that protons are strong activators of BH
which explains the enhanced reaction rates (Scheme 3c). At high
molar ratios of AlCl , other complexes are formed, and the effects
3
3
,
3
3
on activation) to the most electron-rich oxygen. The concept is de-
picted in Scheme 2.
3
of the added water are insignificant. It is reasonable that minute
amounts of water present in reagents and solvents generally en-
hance the reaction rates in reductive openings using borane
complexes.
As a consequence, the activation of BH
results in higher order reaction kinetics, with respect to AlCl
pared to first order kinetics, found for the reaction using BH
3
ꢀNMe
3
by the Lewis acid
, com-
ꢀTHF.
3
3
The original method for regioselective reductive openings of
cyclic acetals has been developed into an abundance of methods
The dissociation energy of dimethylsulfide borane (BH
DMSB) lies between those of the other investigated borane com-
plexes (BH ꢀTHF: 83 kJ/mol; BH ꢀSMe : 101 kJ/mol; BH ꢀNMe :
3 2
ꢀSMe ,
BOTf,11 Cu(OTf)
12
(
e.g., boranes activated by Bu
2
2
,
other metal tri-
or the use of DIBAL ), and we decided to
investigate the initial rate kinetics using quantitative thin layer
1
3
14,15
16
flates, BF
3
ꢀOEt
2
,
3
3
2
3
3
2
5
1
60 kJ/mol), and the possible activation by proton sources and
1
0,17,18
Lewis acids therefore represents a borderline case. BH ꢀSMe has
chromatography,
mechanism.
to shed further light on the reaction
3
2
26
earlier been used for acetal openings in conjunction with TMSOTf
2
7
or BF
3
ꢀOEt
2
,
and Saito and co-workers used the latter reagent
The rationale for the activation of BH
mation of the strong complex AlCl
ꢀNMe
While the AlCl complex is very strong (199 kJ mol ), the
ꢀNMe
analogous BF
ꢀNMe
would thus, according to theory, activate the borane to a less
3
ꢀNMe
3 3
by AlCl is the for-
combination for reductive opening of 1 to give a mixture of 2
and 3 in a 3:1 ratio. To compare the reactivity with other borane
3
3
according to Scheme 3a.
1
4
ꢁ1
3
3
ꢁ1
complexes, we subjected compound 1 to BH ꢀSMe –AlCl and
10
is weaker (Scheme 3b, 130 kJ mol ) and
3
2
3
3
3
investigated the initial rate kinetics (Fig. 2).
1
9
Reductive opening of 1 using BH ꢀSMe –AlCl in THF resulted in
extent.
We thus subjected compound 1 to BH
THF and followed the initial rate kinetics. The reaction yielded
the expected compound 3 (free 4-OH) and followed higher order
kinetics, that is, the borane was activated by the Lewis acid. How-
ever, the rates were considerably slower (approximately two times
3
3
3
the simultaneous formation of both compounds 2 and 3. However,
the formation of 2 and 3 showed different kinetics. Compound 2
was formed in a first order reaction, while compound 3 showed
3
ꢀNMe
3
and BF
3
ꢀOEt
2
in
higher order kinetics. At 8 equiv of AlCl
order reactions showed essentially the same rates. We conclude
that BH is a borderline case that, to a certain extent, can
3
, the first and the higher
3
ꢀSMe
3
slower using BF
3
ꢀOEt
2
at room temperature, compared to reaction
1
0
be activated by AlCl , but at low amounts of Lewis acid the first or-
with AlCl at 0 °C). The data are presented in Figure 1a and b.
3
3
der reaction predominates (no activation).
In a series of low-temperature experiments, Wei and co-work-
ers showed that the addition of acid scavengers (i.e., di-tert-bu-
tyl-4-methylpyridine) to ensure aprotic conditions severely
Finally, we decided to investigate the original LiAlH
4
proce-
and AlCl in
Et O and CH Cl , which yielded 2 in good yields. The initial rate
1
dure. Thus, we subjected compound 1 to LiAlH
4
3
20
decelerated the reduction rate. The importance of water in the
reaction was further showed by Nifantiev and co-workers, where
2
2
2
analysis using the weighed amounts of AlCl
first order kinetics (Fig. 3a). However, LiAlH
form AlH . Because 1 equiv of AlCl is used in this reaction, the real
concentration of Lewis acid is lowered (Scheme 4). Recalculation of
AlCl ] gave an excellent fit with first order kinetics (Fig. 3b). It is
thus reasonable to assume that alanes (probably AlH O) react
ꢀEt
in a similar way as BH
3
gave deviation from
4
reacts with AlCl to
3
3
3
[
3
(
(
(
a) 1 + BH •NMe + AlCl •THF
1•BH3 + THF + AlCl •NMe
3
3
3
3
3
3
2
3
ꢀTHF.
b) 1 + BH •NMe + BF •OEt2
1•BH3 + Et O + BF •NMe
3
3
3
3
2
3
In addition, we attempted several other reagent combinations,
but these reactions proved to be difficult to examine due to too fast
c) 1 + BH •NMe + H
1•BH3
+ HNMe3
3
3
(
BH
e.g.,
BH
3
ꢀNMe
3
–BF BH
3
ꢀOEt
2
–CH
2
Cl
2
,
3
ꢀNMe
3 3
–AlCl –toluene,
Scheme 3. Activation of BH
3
ꢀNMe
3
by Lewis acids and proton sources.
3
ꢀSMe
2
–Cu(OTf)
2
–THF, and NaCNBH –HCl–THF) or too slow
3

R. Johnsson et al. / Carbohydrate Research 343 (2008) 2997–3000
2999
Figure 1. Initial rate analysis for the formation of compound 3 using BH
3
ꢀNMe
3
. [1]
0
= 0.054 M, [BH
3
ꢀNMe
3
]0 = 0.217 M. (a) Variation of [AlCl
3
]
0
. Data from Ref. 10. Reactions
were run at 0 °C. (b) Variation of [BF . Reactions were run at room temperature. (c) Variation of [AlCl
3
ꢀOEt
2
]
0
3
]
0
with water added. [H O] = 0.108 M. Reactions were run at 0 °C.
2
0
All reactions were performed at least in duplicate.
Figure 2. Initial rate analysis for the formation of compound
2
(squares) and
3
(filled circles) using (a) BH
3
ꢀTHF or (b) BH
3
ꢀSMe
2 0
. [1] = 0.054 M,
[
BH
3
ꢀTHF]
0
= [BH ꢀSMe ]
3 2 0
= 0.217 M. (a) Variation of [AlCl
]
3 0
using BH
3
ꢀTHF. Data from Ref. 10. (b) Variation of [AlCl
3
]
0
using BH
3
ꢀSMe . All reactions were run at room
2
temperature and performed at least in duplicate.
reaction rates (e.g., BH
THF).
To summarize, we have shown the reaction kinetics for a series
of reductive openings of 4,6-O-benzylidene acetals. In all cases, the
3
ꢀTHF–BF
3
ꢀOEt
2
–THF, BH
3
ꢀNMe
3
–Cu(OTf)
2
–
openings to free HO-6 (i.e., 4-O-benzyl ethers) follow first order
kinetics, while reactions yielding free HO-4 (i.e., 6-O-benzyl ethers)
follow higher order kinetics. These results correlate well with the
concept of regioselectivity by activation of borane complexes.10

3
000
R. Johnsson et al. / Carbohydrate Research 343 (2008) 2997–3000
Figure 3. Initial rate analysis for the formation of compound 2 using LiAlH
4
3 0 4 0
and AlCl in THF. [1] = 0.072 M, [LiAlH ] = 0.288 M. (a) Initial rates plotted versus added amount
3
of AlCl . (b) Initial rates plotted versus real AlCl
3
concentration. All reactions were run at 0 °C and performed at least in duplicate.
5
0
8–230 mg/mL (2–8 equiv). To LiAlH
.53 mmol) was added a stock solution of 1 (0.480–0.615 mL,
O (0.480–0.615 mL), the mixture was
for 15 min. Then, AlCl (0.480–0.615 mL,
.220–1.00 mmol) was added.
4
(15–20 mg, 0.40–
3
LiAlH4 + AlCl3
4 AlH3 + 3 LiCl
Scheme 4. Formation of AlH3 by reaction of LiAlH
4
and AlCl
3
.
0.103–0.133 mmol) and Et
stirred at 0 °C under N
0
2
2
3
1
1
. Experimental
.1. General experimental details
Reactions were monitored by TLC using alumina plates coated
Acknowledgments
The Swedish Research Council, The Lennander Foundation, and
The Crafoord Foundation supported this work.
with silica gel and visualized by charring with p-anisaldehyde.
THF was distilled from sodium, and other reaction solvents were
dried on Al
2
O
3
. Known and commercially available compounds
Supplementary data
were in agreement with previously published data (e.g., NMR).
Samples (0.050 mL) were taken and quenched in micro vials with
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.carres.2008.08.022.
NaHCO
The samples were analyzed by quantitative TLC.
tions in THF: 1 100 mg/mL, 0.216 M; AlCl 145 mg/mL, 1.088 M;
BH 64 mg/mL, 0.877 M; H O 19 mg/mL, 1.077 M; BH ꢀSMe
ꢀNMe
was diluted from 2 M to 1 M and BF was not diluted.
ꢀOEt
3
(0.500 mL, satd aq) and extracted with ether (0.200 mL).
17,18
Stock solu-
References
3
3
3
2
3
2
1
2
.
.
Lipták, A.; Jodál, I.; Nánási, P. Carbohydr. Res. 1975, 44, 1–11.
Lipták, A.; Imre, J.; Harangi, J.; Nánási, P.; Neszmélyi, A. Tetrahedron 1982, 38,
3
2
3721–3727.
1
.2. Borane dimethylsulfide-complex—AlCl
3
3. Garegg, P. J.; Hultberg, H. Carbohydr. Res. 1981, 93, C10–C11.
4
5
.
.
Garegg, P. J.; Hultberg, H.; Wallin, S. Carbohydr. Res. 1982, 108, 97–101.
Ek, M.; Garegg, P. J.; Hultberg, H.; Oscarson, S. J. Carbohydr. Chem. 1983, 2, 305–
A mixture of stock solutions of 1 (0.250 mL, 0.054 mmol),
ꢀSMe (0.220 mL, 0.220 mmol), and THF (0.130–0.430 mL)
for 15 min. Then, a stock solution of AlCl
0.100–0.400 mL, 0.109–0.435 mmol) was added.
311.
BH
3
2
6. Fügedi, P.; Garegg, P. J.; Kvarnström, I.; Svansson, L. J. Carbohydr. Chem. 1988, 7,
89–397.
3
was stirred at rt under N
(
2
3
7
8
.
.
Fügedi, P.; Birberg, W.; Garegg, P. J.; Pilotti, Å Carbohydr. Res. 1987, 164, 297–312.
Stick, R. V. Carbohydrates: The Sweet Molecules of Life; Academic Press: London,
2001. p 53.
9
.
Garegg, P. J. In Preparative Carbohydrate Chemistry; Hanessian, S., Ed.; Marcel
Dekker: New York, 1996. pp 53–67.
1
.3. Borane trimethylamine—BF
3
ꢀOEt
2
10. Johnsson, R.; Olsson, D.; Ellervik, U. J. Org. Chem. 2008, 73, 5226–5232.
A mixture of stock solutions of 1 (0.250 mL, 0.054 mmol),
ꢀNMe (0.250 mL, 0.219 mmol), and THF (0.450–0.490 mL)
for 15 min. Then, BF (0.014–
ꢀOEt
.055 mL, 0.111–0.434 mmol) was added.
11. Jiang, L.; Chan, T.-H. Tetrahedron Lett. 1998, 39, 355–358.
12. Shie, C.-R.; Tzeng, Z.-H.; Kulkarni, S. S.; Uang, B.-J.; Hsu, C.-Y.; Hung, S.-C.
Angew. Chem., Int. Ed. 2005, 44, 1665–1668.
BH
3
3
was stirred at rt under N
0
2
3
2
13. Wang, C.-C.; Luo, S. Y.; Shie, C.-R.; Hung, S. C. Org. Lett. 2002, 4, 847–849.
14. Ishikawa, T.; Shimizu, Y.; Kudoh, T.; Saito, S. Org. Lett. 2003, 21, 3879–3882.
15. Oikawa, M.; Liu, W.-C.; Nakai, Y.; Koshida, S.; Fukase, K.; Kusumoto, S. Synlett
1
996, 1179–1180.
1
2 3
.4. Borane trimethylamine–H O–AlCl
1
1
1
6. Mikami, T.; Asano, H.; Mitsunobu, O. Chem. Lett. 1987, 2033–2036.
7. JustTLC, Version 2.4; Sweday: Lund, Sweden.
8. Johnsson, R.; Träff, G.; Sundén, M.; Ellervik, U. J. Chromatogr., A 2007, 1164,
A mixture of stock solutions of 1 (0.250 mL, 0.054 mmol),
ꢀNMe (0.250 mL, 0.219 mmol), H O (0.100 mL, 0.108 mmol)
for 15 min.
(0.100–0.400 mL, 0.109–0.435 mmol) was added.
2
98–305.
BH
3
3
2
1
9. Jonas, V.; Frenking, G.; Reetz, M. T. J. Am. Chem. Soc. 1994, 116, 8741–8753.
20. Hernández-Torres, J. M.; Achkar, J.; Wei, A. J. Org. Chem. 2004, 69, 7206–7211.
and THF (0–0.300 mL) was stirred at 0 °C under N
Then, AlCl
2
2
1. Sherman, A. A.; Mironov, Y. V.; Yudina, O. N.; Nifantiev, N. E. Carbohydr. Res.
003, 338, 697–703.
2. Gálová, M. Electrochim. Acta 1984, 29, 323–328.
3
2
2
1
.5. Lithium aluminum hydride—AlCl
3
23. Brown, H. C.; Murray, L. T. Inorg. Chem. 1984, 23, 2746–2753.
24. Ryschkewitsch, G. E. J. Am. Chem. Soc. 1960, 82, 3290–3294.
25. Rablen, P. R. J. Am. Chem. Soc. 1997, 119, 8350–8360.
26. Bartels, B.; Hunter, R. J. Org. Chem. 1993, 58, 6756–6765.
Compound 1 was dissolved in CH
00 mg/mL, AlCl was dissolved in Et
2
Cl
2
to give a stock solution of
1
3
2
O to give a stock solution of
27. Saito, S.; Kuroda, A.; Tanaka, K.; Kimura, R. Synlett 1996, 231–233.