"Design" of boron-based compounds as pro-nucleophiles and co-catalysts for indium(I)-catalyzed allyl transfer to various Csp 3-type electrophiles

Article abstract of DOI:10.1002/asia.201100096

We have recently uncovered a general indium(I)-catalyzed method for allylations and propargylation of acetals and ketals with a water- and air-stable allyl boronate. By using a more reactive allyl borane, we have successfully extended this methodology to the more challenging C-C coupling with ethers. Herein, we report an improved methodology for the indium(I)-catalyzed allylation of acetals and ethers, through combination of the allyl boronate with a commercially available "hard" Lewis acid, B-methoxy-9-BBN (BBN=borabicyclo[3.3.1]nonane), as an effective co-catalyst. Significantly, our work highlights for the first time the correlation between the Lewis acidity of "electrophilic" boron-based compounds and their " nucleophilic" reactivity in Csp³-Csp³ couplings, catalyzed by a "soft" low-oxidation main group metal. In addition, we also report several applications of these methodologies to the selective synthesis of various carbohydrate derivatives. Copyright

Full text of DOI:10.1002/asia.201100096

FULL PAPERS
DOI: 10.1002/asia.201100096
“Design” of Boron-Based Compounds as Pro-Nucleophiles and Co-Catalysts
for Indium(I)-Catalyzed Allyl Transfer to Various Csp3-Type Electrophiles
Hai Thanh Dao, Uwe Schneider, and Shu Kobayashi*[a]
¯
Abstract: We have recently uncovered
a general indium(I)-catalyzed method
for allylations and propargylation of
acetals and ketals with a water- and
air-stable allyl boronate. By using a
more reactive allyl borane, we have
successfully extended this methodology
ethers, through combination of the allyl
boronate with a commercially available
“hard” Lewis acid, B-methoxy-9-BBN
(BBN=borabicycloACTHNUTRGENUGN[3.3.1]nonane), as
an effective co-catalyst. Significantly,
our work highlights for the first time
the correlation between the Lewis acid-
ity of “electrophilic” boron-based com-
pounds and their “nucleophilic” reac-
tivity in Csp3–Csp3 couplings, catalyzed
by a “soft” low-oxidation main group
metal. In addition, we also report sev-
eral applications of these methodolo-
gies to the selective synthesis of various
carbohydrate derivatives.
ꢀ
to the more challenging C C coupling
Keywords: allylation
·
boron
·
·
with ethers. Herein, we report an im-
proved methodology for the indium(I)-
catalyzed allylation of acetals and
ꢀ
carbohydrates
indium
· C C coupling
Introduction
ing transformation proceeds by Lewis or Brønsted acid acti-
vation of the electrophile to form an oxocarbenium or car-
benium ion that can react with nucleophiles such as allyl si-
lanes (Hosomi–Sakurai allylation).[5]
Recently, we have disclosed a general catalytic method
for allylation of acetals and ketals with allyl boronates using
an indium(I) catalyst (Scheme 1).[6] To the best of our
knowledge, this work represents the first main group metal-
ꢀ
Innovative metal catalysis for C C bond formation plays a
vital role in organic chemistry.[1] In this context, boron-based
compounds[2] are among the most important and appealing
“standard” reagents in organic synthesis. Indeed, these mol-
ecules offer a unique combination of interesting properties.
Whereas trivalent boron compounds have an “electrophilic”
character owing to the vacant low-energy 2p orbital of the
boron atom, they may display “nucleophilic” behavior after
reaction with Lewis bases, thereby forming tetravalent
boron–ate complexes.[2c,d] Moreover, boron-based reagents
are attractive because of high functional group tolerance,
low toxicity, and convenient handling.[2]
ꢀ
catalyzed activation of allyl boronates for C C bond forma-
tion with noncarbonyls. This methodology features a broad
functional group tolerance, and is applicable to both allyla-
tions and propargylation.
The allylation of organic substrates is among the most im-
ꢀ
portant carbon carbon bond-forming reactions in organic
synthesis.[3] While allylation of Csp2-type electrophiles, such
as carbonyl compounds and imines, has been extensively
studied,[4] the coupling of allyl reagents with Csp3-type elec-
trophiles is relatively underexplored. Typically, this challeng-
Scheme 1. Indium(I)-catalyzed allylation of acetals and ketals with allyl
boronate 2a.
[a] H. T. Dao, Dr. U. Schneider, Prof. Dr. S. Kobayashi
Department of Chemistry, School of Science
The University of Tokyo
With detailed insight into the reaction mechanism, we
have successfully extended this dual activation concept to
the more challenging coupling with ethers, by switching
from allyl boronate 2a to allyl borane 2b (Scheme 2).[7] Our
present contribution provides a full account on issues relat-
The HFRE Division
ERATO (Japan) Science and Technology Agency (JST)
Hongo, Bunkyo-ku, Tokyo 113-0033 (Japan)
Fax : (+81)3-5684-0634
E-mail: shu_kobayashi@chem.s.u-tokyo.ac.jp
2522
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Asian J. 2011, 6, 2522 – 2529

Table 1. Screening of metal co-catalysts for the allylation of 1 with allyl
boronate 2a.
Entry
Ether 1
Co-Catalyst
Solvent
Conversion[a]
1
2
–
–
toluene
DCM
n.r.
n.r.
messy[b]
messy[b]
messy[b]
n.r.
Scheme 2. Concept of dual catalytic activation for “nucleophilic” substi-
tution reactions with sp3-type electrophiles employing “electrophilic”
allyl boron reagents.
3
4
5
6
7
8
9
10
InCl3
InACTHNGURETNNU(G OTf)3
GaCl3
–
InCl3
–
toluene
toluene
toluene
toluene
toluene
DCM
messy[b]
<20%
messy
messy
ed to allyl boron reagents for the alkyl–allyl coupling with
Csp3-type electrophiles, in order to 1) address limitations in
scope and 2) gain further mechanistic insight. We also
report applications in carbohydrate chemistry.
In
N
DCM
DCM
GaCl3
[a] The conversion of ethers 1 was determined by 1H NMR analysis of
aliquots of the reaction mixtures. [b] Friedel–Crafts-type by-products
were obtained. n.r.=no reaction.
Results and Discussion
2a, which is a prerequisite for the postulated transmetala-
tion mechanism. In addition, kinetic studies for the allyla-
tion of acetals with 2a revealed that the transmetalation was
the rate-determining step in the catalytic cycle. Therefore,
facilitation of this boron Lewis acid-involving step might be
the key for the acceleration of the coupling with less reac-
tive Csp3-type electrophiles, such as ethers. Based on this as-
sumption, we uncovered that, in contrast to allyl boronate
2a, the more reactive 9-BBN-derived (BBN=Borabicyclo-
While surveying the scope for the InOTf-catalyzed allylation
of acetals and ketals with allyl boronate 2a, we detected for
certain substrates the formation of diallylated products in
trace amounts. It was reasoned that these by-products might
be formed by further allylation of the initially generated ho-
moallylic ethers.[8] This observation prompted us to investi-
ꢀ
gate the C C coupling between 2a and ethers, another class
of Csp3-type electrophiles. However, owing to the relatively
ꢀ
strong C O bond, allylation of ethers is expected to be sig-
ACHTUNGTREN[NUGN 3.3.1]nonane), allyl borane 2b worked as an effective pro-
ꢀ
nificantly more challenging than that of acetals and
ketals.[5,9] Indeed, the InOTf-catalyzed reactions of several
ethers 1 with 2a, under optimized conditions for acetal ally-
nucleophile in the challenging C C bond formation with
ethers.
Various metal Lewis acids were tested for the allylation of
ether 1a with allyl borane 2b, and the results are summar-
ized in Table 2. With the exception of BiACTHNGUETRNNU(G OTf)3 (Table 2,
entry 2), all other metal triflates were found to be signifi-
cantly less effective than InOTf (Table 2, entry 1 vs en-
tries 3–11), or did not afford at all the desired product 3a
(Table 2, entries 12–14). Among these catalysts, stronger
ꢀ
lation, did not afford the desired C C bond-formed prod-
ucts 3 (Table 1, entries 1, 2, and 6), or only a very low yield
for 3 (Table 1, entry 8).
To overcome this higher energy barrier, our initial plan
was to facilitate the C O bond activation by employing
metal Lewis acids, such as InCl3, In
co-catalysts (Table 1, entries 3–
5, 7, 9, and 10). However, the
use of these stronger Lewis
acids in toluene resulted only
in messy mixtures, which con-
tained Friedel–Crafts-type side-
products. The formation of
ꢀ
ACHUTGTNERN(NGU OTf)3, and GaACHTUNGTERN(NUGN OTf)3, as
Table 2. Screening of metal catalysts for the allylation of ether 1a with allyl borane 2b.
these undesired compounds
suggests that the activation of
Entry
Catalyst
InOTf
Yield [%][a]
Entry
Catalyst
Cu(OTf)2
Al(OTf)3
Fe(OTf)2
Zn(OTf)2
Mg(OTf)2
Sm(OTf)3
Yield [%][a]
1
2
3
4
5
6
7
8
86
70
52
45
41
35
26
27
9
G
26
25
12
ꢀ
the C O bond indeed occurred
Bi
Sc
U
10
11
12
13
14
15
16
U
to afford both the correspond-
ing carbenium ions and the cor-
responding metal methoxide
species. However, in contrast
to indium(I) methoxide, these
intermediates may not be able
to deliver the required methox-
E
InU
E
n.r.[b]
n.r.[b]
n.r.[b]
43[c]
12[c]
N
ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
AgOTf
TfOSiMe3
TfOH
CuOTf[d]
1
[a] Yield of isolated product after purification on silica gel (PTLC). [b] Determined by H NMR analysis of ali-
quots of the reaction mixtures and by TLC monitoring of crude mixtures. [c] 10 mol% of the catalyst was
ide to activate allyl boronate added at ꢀ788C, then warming to RT. [d] Used as a toluene (0.5 equiv) complex.
Chem. Asian J. 2011, 6, 2522 – 2529
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemasianj.org
2523

S. Kobayashi et al.
FULL PAPERS
Lewis acids such as AlACHTNUGTRNEUNG(OTf)3 may suffer from undesired re-
actions, while other metal triflates, such as MgACTHUNRGTNEUNG(OTf)2, may
ꢀ
not successfully activate the C O bond of 1a, or the corre-
sponding metal methoxide intermediates may not be able to
deliver the required methoxide to activate allyl borane 2b.
Interestingly, metal-free acids, such as trimethylsilyl triflate
and triflic acid, also catalyzed the model reaction, albeit in
moderate and low yields (Table 2, entries 15 and 16).
Scope and limitation for the InOTf-catalyzed allylation of
ethers 1 with 2b are shown in Table 3. The present catalytic
alkyl–allyl cross-coupling showed good substrate generality
Table 3. Scope and limitation for the allylation of ethers 1 with allyl
borane 2b.
Figure 1. InOTf-catalyzed alkyl–allyl cross-coupling: boronate 2a versus
borane 2b.
thereby forming a carbenium ion species A (an oxocarbeni-
um ion species in the case of acetals) and indium(I) methox-
ide. The in situ-formed InOMe then triggers B-to-In trans-
metalation with borane 2b (or boronate 2a) to generate the
nucleophilic allyl indium(I) species (B), which subsequently
reacts with electrophile A to give the desired allylated prod-
uct 3. For the reaction using allyl boronate 2a, we believe
that this transformation exclusively follows the above trans-
metalative pathway (via boron–ate complex C).[6] In the
case of allyl borane 2b, however, we cannot rule out a non-
transmetalative pathway, which involves the direct reaction
of the more reactive boron–ate complex D with electrophile
A. Particularly, in the reactions catalyzed by metal-free
compounds (TfOSiMe3 and TfOH; Table 2, entries 15 and
16), the non-transmetalative mechanism should be the
major pathway for the formation of the allylated product
3a.[10]
The nature of the substituents on the boron atom of a tri-
valent boron compound determines its Lewis acidity, and
thus its electrophilicity.[2e,11] The relative strength in Lewis
acidity of boron-based compounds can be estimated by the
chemical shift d in 11B NMR spectroscopy, which is a
common tool for monitoring reactions involving boron-con-
taining reagents.[11] Although this method seems to be very
appealing for the prediction of the reactivity of boron-based
molecules, there have been only sporadic reports on the cor-
relation between 11B NMR values (boron Lewis acidity) and
the reactivity of boron-based compounds, such as 1,2-addi-
tions to aldehydes[12] or transition metal-triggered transmeta-
lation.[11] In the present indium(I) catalysis, we observed a
substantial improvement in reactivity by switching from allyl
boronate 2a to allyl borane 2b (Figure 1). Judging from our
11B NMR data, the Lewis acidity of the boron atom of
borane 2b (d2b =85 ppm) is expected to be significantly
higher than that of boronate 2a (d2a =32 ppm; Figure 2).
[a] Yield of isolated product after purification on silica gel (PTLC or
column chromatography). [b] Neat, 608C, 48 h.
including various primary, secondary, and tertiary benzylic,
allylic, and propargylic ethers, thereby providing the corre-
sponding allylated products in good to excellent yields in
most cases.[7]
We have proposed a transmetalative SN1 mechanism, in
which indium(I) plays a dual role, as shown in Figure 1.[7]
ꢀ
First, InOTf acts as a Lewis acid to activate the C O bond,
2524
www.chemasianj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Asian J. 2011, 6, 2522 – 2529

Boron-Based Compounds as Pro-Nucleophiles and Co-Catalysts
Table 4. Screening of metal-free co-catalysts for the allylation of 1a with
2a.
Entry
1
Co-Catalyst
–
Yield [%][a]
14
Figure 2. Correlation between Lewis acidity (“electrophilicity”) and “nu-
cleophilicity” of boron-based reagents.
2
55[b]
3
44
Therefore, allyl borane 2b would have a substantially in-
creased affinity toward the in situ-formed Lewis base, indiu-
m(I) methoxide, thus resulting in a faster generation of
boron–ate complex D (compared with the formation of
complex C from 2a). This notable change may push the
equilibrium in the first step of our mechanism to the right
side, thus leading to rate acceleration. In addition, boron–
ate complex D should also be more reactive than complex C
in the transmetalation step, or D as a nucleophile may un-
dergo a direct reaction with A. These combined effects ex-
plain why allyl borane 2b is a significantly more reactive
pro-nucleophile than allyl boronate 2a. Overall, 2b can
react with less reactive electrophiles, including primary
ethers (Table 3).
Being exceedingly reactive, thanks to its highly Lewis
acidic nature, allyl borane 2b exhibits a relatively low toler-
ance toward functionalities, such as hydroxy and ester
groups.[2c] Moreover, 2b is not very convenient in terms of
both preparation and storage, compared with the water- and
air-stable allyl boronate 2a. In addition, we experienced
4
5
6
7
BPh3 (4c)
41
14
19
5
BEt3 (4d)[c]
BF3·OEt2 (4e)[d]
BACHTNUTRGNENG(U C6F5)3 (4 f)
[a] Yield of isolated product after purification on silica gel (PTLC).
[b] Reaction time: 36 h; 61% yield. [c] A solution (1m in hexane) was
used. [d] The co-catalyst was added at 08C.
Next, we turned our attention to the mechanism of this
unprecedented indium(I)/boronACTHNUGRTNEUNG(III) co-catalysis. With the
aim to investigate the role of the most effective co-catalyst,
borinic ester 4a, we conducted H NMR studies. In deuteri-
um labeling experiments, our model substrate, ether 1a, was
combined with an equimolar amount of borinic ester
[D3]4a, in the absence and in the presence of indium(I) tri-
flate (Figure 3). These reactions were monitored with time
by H NMR analysis (15 min; 3 h; 16 h); the obtained spec-
tra were then compared with the data for reference 4a.
1
1
In the absence of InOTf, even after 16 h, there was no
1
some difficulties in cleanly synthesizing derivatives of 2b. In
change in the H NMR spectra of the mixture containing 1a
3
ꢀ
view of C C bond formations with challenging Csp -type
and [D3]4a (Figure 3a). Compound 4a was not detected;
thus, scrambling of the deuterium label did not occur. On
the other hand, in the presence of 20 mol% of InOTf, an in-
creasing amount of compound 4a was formed with time
(Figure 3b). Thus, scrambling of the deuterium label must
have occurred to convert 1a and [D3]4a into [D3]1a and 4a.
These results suggest that InOTf, rather than the boron
electrophiles (such as less reactive acetals and secondary
ethers), we sought an improved catalytic method to replace
allyl borane 2b with the more environmentally benign allyl
boronate 2a and derivatives thereof.
Based on the proposed mechanism in Figure 1, we envi-
sioned a “hard” Lewis acid co-catalyst to be appropriate to
“trap” in situ-formed indium(I) methoxide, which may force
the equilibrium in the first step to the formation of carbeni-
um ions A, thus accelerating the catalytic cycle with the less
reactive allyl boronate 2a. In this context, we screened vari-
ous boron Lewis acids (Table 4). To our delight, when the
commercially available “hard” boron Lewis acid, B-me-
thoxy-9-BBN (4a), was employed in a catalytic amount, the
yield was substantially improved (Table 4, entry 1 vs
entry 2). A similar rate enhancement was observed for the
catalytic use of boranes 4b and 4c (Table 4, entries 3 and 4),
while triethyl borane (4d) did not show any positive effect
(Table 4, entry 5). Strikingly, strong Lewis acids, such as
BF3·OEt2 (4e) and the perfluorinated borane 4 f, afforded
only low isolated yields of product 3a (Table 4, entries 6 and
7).
ꢀ
Lewis acid 4a, catalytically activates the C O bond of ether
1a. For the deuterium scrambling to occur, the in situ-
formed InOMe may be trapped by [D3]4a to form the corre-
sponding boron–ate complex [D3]5a (Figure 4a). This com-
plex may serve as a deuterio-methoxide source (ꢀOCD3)
that reacts with the generated carbenium ion A to form
[D3]1a and 4a. Therefore, in our catalytic reaction system,
borinic ester 4a may facilitate the first equilibrium by form-
ing the boron–ate complex 5a (Figure 4b). This complex
may work as an effective methoxide source (ꢀOCH3) to de-
liver the required Lewis base, in view of the activation of
allyl boronate 2a for B-to-In transmetalation. Stronger
boron Lewis acids, such as 4e and 4 f, may be able to “trap”
the methoxide species to generate the corresponding boron–
ate complexes, but those may not effectively transfer the
methoxide to 2a.[13] Finally, it is noted that we cannot rule
Chem. Asian J. 2011, 6, 2522 – 2529
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemasianj.org
2525

S. Kobayashi et al.
FULL PAPERS
Figure 4. Plausible mechanisms: a) scrambling of the deuterium label;
b) boron-to-indium transmetalation; c) boron-to-boron “transmetala-
tion”.
Table 5. Scope for the allylation of ethers and acetals 1 with allyl boro-
nate 2a.
Yield of isolated product after purification on silica gel (PTLC):
[a] 25 mol% of 4a was used; [b] 4a was not used. [c] A reaction was not
detected even at 508C. [d] The secondary allylic ether 1p’ (1-methoxy-1-
phenylpropene) was employed.
hancement was observed for the secondary ethers 1e, 1i,
and 1p. The reaction with the more reactive dibenzylic
ether 1j proceeded smoothly, even without the boron co-cat-
alyst; a rate acceleration was not observed. In the case of
the particularly challenging primary ether 1b, the desired
product was not obtained, even at an elevated temperature.
Finally, borinic ester 4a was also found to be an effective
co-catalyst for the allylation of the less reactive acetals
1v–x.
Figure 3. Labeling experiments with ether 1a and borinic ester [D3]4a
(a) without and (b) with InOTf.
out the possibility of a B-to-B “transmetalation” to form the
more reactive boron–ate complex D (Figure 4c). This com-
plex may serve as a precursor for allyl indium(I) (E+ =In+),
ꢀ
We then studied the scope for boronates 2 in catalytic C
ꢀ
or may be the active nucleophile in the C C bond formation
C bond formations with ether 1a (Table 6). In the absence
of co-catalyst 4a, the indium(I)-catalyzed reaction with a-
methylallyl boronate 2c afforded exclusively a-adduct 6 in
45% yield (Table 6, entry 2). This yield was improved to
70% by employing an additional 25 mol% of borinic ester
4a; the excellent constitutional selectivity was maintained
(E+ =carbenium ion R+).
Next, we investigated the scope for electrophiles by em-
ploying allyl boronate 2a with the new indium(I)/boronACHTUNTRGNEUNG(III)
dual catalyst system (Table 5). Under standard conditions
[InOTf (5 mol%) + 4a (25 mol%)], a substantial rate en-
2526
www.chemasianj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Asian J. 2011, 6, 2522 – 2529

Boron-Based Compounds as Pro-Nucleophiles and Co-Catalysts
ꢀ
Table 6. Scope for boronates 2 in C C bond formations with ether 1a.
Table 7. Allylation and propargylation of carbohydrate derivatives 8a–c.
Substrate
Method
Product
Entry
1
Pro-Nucleophile
Product
Yield[a] (Yield[b]
60% (20%)
)
2
70%[c] (45%)[c]
3
4
messy
messy
n.d.[d]
n.d.[d]
5
65%[e] (5%)
Yield of isolated product after purification on silica gel (PTLC):
[a] 25 mol% of 4a was used; [b] 4a was not used. [c] Reaction time: 14 h.
[d] A reaction was not detected even at 508C. [e] Propargylation/allenyla-
tion=>20:1. n.d.=not detected.
(Table 6, entry 2). This rare a-selectivity with 2c is a strong
ꢀ
indicator for catalytic B-to-In transmetalation prior to C C
bond formation. The reactions using crotyl boronates (E)-
2d and (Z)-2d gave only messy mixtures even in the pres-
ence of 4a (Table 6, entries 3 and 4). The low reactivity of
these boronates is consistent with our previous work, and
may be explained by a particularly slow transmetalation
owing to steric congestion at the g-position.[6] Finally, with-
out 4a, the reaction with allenyl boronate 2e hardly pro-
ceeded (Table 6, entry 5). On the other hand, in the pres-
ence of 4a (25 mol%), boronate 2e reacted with 1a regio-
specifically to afford the homopropargylic product 7 in 65%
yield (Table 6, entry 5). These improved results highlight the
[a] Conditions (“method 1”): 8a or 8b (0.4 mmol, 1.0 equiv), 2a or 2e
(1.5 equiv), InOTf (5 mol%), DCM (0.25m), 258C, 16 h. [b] Conditions
(“method 2”): 8b or 8c (0.2 mmol, 1.0 equiv), 2b (2.0 equiv), InOTf
(5 mol%), DCM (0.25m), 258C, 14 h. [c] Conditions (“method 3”): 8b or
8c (0.2 mmol, 1.0 equiv), 2a or 2e (2.0 equiv), InOTf (10 mol%), co-cata-
lyst 4a (50 mol%), DCM (0.25m), 508C, 24 h. [d] Diastereomeric ratios
appeal of the present indium(I)/boron
synthetic purposes.
ACHTUNGERTN(NUNG III) dual catalysis for
Having developed three general methodologies for Csp3–
Csp3 couplings of boron-based reagents with Csp3-type elec-
trophiles, we aimed at applying these methods to C-glycosy-
lation, in view of the selective synthesis of carbohydrates. C-
Glycosides are versatile building blocks for the preparation
of many biologically active compounds. This class of mole-
cules has the potential to serve as carbohydrate analogs re-
sistant to metabolic processes, and is a potential source for
therapeutic agents in view of various clinical applications.[14]
In this context, allylic and propargylic glycosides are attrac-
1
were determined by H NMR analysis.
The application of the above catalytic methodologies to
carbohydrate chemistry is summarized in Table 7. Treatment
of 3,4,5-tri-O-acetyl-d-glucal (8a) with allyl boronate 2a, in
the presence of indium(I) triflate (5 mol%), resulted in the
formation of 2,3-unsaturated allyl glycoside 9a in good yield
with high diastereoselectivity (“method 1”). Similarly, glucal
8a smoothly reacted with allenyl boronate 2e to provide the
corresponding propargylic C-pseudoglycal 10a in good yield
with moderate diastereoselectivity. However, when 3,4,5-tri-
O-benzyl-d-glucal (8b) was reacted under “method 1” con-
ditions, only the Ferrier rearrangement-type side-product
ꢀ
tive because the terminal, unsaturated C C bond can be
easily functionalized to generate other carbohydrate deriva-
tives.[15]
Chem. Asian J. 2011, 6, 2522 – 2529
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemasianj.org
2527

S. Kobayashi et al.
FULL PAPERS
8b’ was obtained in moderate yield. These results suggest
“electrophilic” boron-based compounds and their “nucleo-
philic” reactivity in Csp3–Csp3 couplings, catalyzed by a
“soft” low-oxidation main group metal (Figure 5). We be-
lieve that these findings will 1) provide more insight into the
understanding of boron-based pro-nucleophiles and poten-
tial co-catalysts and 2) significantly expand their utility in
various domains of organic synthesis. Further investigations
with the aim to apply this concept to asymmetric catalysis
are now ongoing in our laboratory.
ꢀ
that the C O bond activation in 8b occurred, but the B-to-
In transmetalation may be too slow compared with the un-
desired rearrangement. In addition, 8b’ may be less reactive,
and can therefore not be activated under the conditions of
“method 1”. However, when the more reactive borane 2b
was employed, we obtained the desired product 9b in good
yield with high diastereoselectivity (“method 2”). In analogy,
1-O-methyl-2,3,5-tri-O-benzylpentofuranose (8c) did not un-
dergo allylation with “method 1”, while providing high
yields and high diastereoselectivities with “method 2”. Im-
portantly, the challenging substrates 8b and 8c can indeed
be reacted with allyl boronate 2a when the newly developed
“method 3” was selected, although a higher loading of the
co-catalyst 4a (50 mol%) and an elevated temperature
(508C) were required. Interestingly, by employing “method
3”, carbohydrate 8c and allenyl boronate 2e could be selec-
tively converted into the desired propargylic product 10c,
whereas this compound would definitely not be accessible
with “method 1” or “method 2”.
Experimental Section
General
NMR spectra were recorded on a JEOL ECX-400, an ECA-500, or an
ECA-600 spectrometer, operating at 400, 500, or 600 MHz for 1H NMR,
and at 100, 125, or 150 MHz for 13C NMR, and at 128 MHz for
11B NMR. Chemical shifts are reported downfield from tetramethylsilane
(TMS). IR spectra were measured using a JASCO FTIR-610 spectrome-
ter. HRMS were recorded using a JEOL JMS-T100TD spectrometer
(DART). Preparative thin-layer chromatography (PTLC) was carried out
using Wakogel B-5F from WAKO. All organic solvents used were com-
mercially available dry solvents, which were distilled appropriately under
an argon atmosphere and stored over molecular sieves prior to use in an
argon box. Indium(I) triflate (InOTf) was prepared according to a report-
ed procedure, and stored in an argon box at ꢀ308C.[16] Ethers 1a–u were
prepared from the corresponding alcohols. Acetals 1v–x[17] and carbohy-
drates 8b–c[18] were synthesized following reported methods. Carbohy-
drate 8a (Aldrich) and boron Lewis acids 4c, 4d (1m in hexane), and 4e
are commercially available (all TCI), and were used without further pu-
rification. B-Methoxy-9-BBN (4a; 1m in hexane) is commercially avail-
able (Aldrich); the solvent was removed prior to its use. Allyl boronate
2a,[19] allyl borane 2b,[20] a-methylallyl boronate 2c,[21] crotyl boronates
(E)-2d and (Z)-2d,[19] allenyl boronate 2e,[22] and borane 4b[20] were pre-
pared by slightly modified procedures of reported methods. Allylations
and propargylation were performed according to the general procedure.
Products 3a–3r, 3t–u,[6] 3v,[7] 3w,[23] 8b’,[24] 9a,[18a] 9b,[25] 9c,[26] and 10a[27]
are literature-known compounds. Their analyses are in full agreement
with the reported data. The analytical data for unreported compounds
3x, 6, 7, and 10c are provided below.
Conclusions
We have uncovered an interesting rate enhancement for in-
dium(I)-catalyzed allylations and propargylation of Csp3-
type electrophiles with boronate reagents, through the use
of borinic ester 4a as a “hard” Lewis acid co-catalyst. Our
work not only represents a significant advance compared
with our earlier studies in terms of scope,[6,7] but also sheds
further light on the involved reaction mechanism. In addi-
tion, we have achieved the application of the present meth-
odologies to selective allylation and propargylation of carbo-
hydrate derivatives. Most importantly, we report here for
the first time the relationship between the Lewis acidity of
General Procedures
Indium(I) triflate (5–10 mol%), dry DCM or dry CDCl3 (0.25–0.50m),
and the corresponding substrate 1a–x or 8a–c were combined in a flame-
dried 5 mL-screw vial with magnetic stirring bar in an argon box. After
successive addition of the corresponding boron-based reagent 2a, 2b, 2c,
(E)-2d, (Z)-2d, or 2e (1.5–2.0 equiv) and co-catalyst 4a (25–50 mol%, if
applicable), the reaction mixture was stirred at the indicated temperature
for the indicated time. Quenching with aqueous K2CO3 (1m) yielded the
crude reaction mixture, which was then purified by preparative thin-layer
chromatography (PTLC; eluant: hexane!hexane/ethyl acetate=85:15;
in a small scale, the product may be purified without quenching).
1-(1-Methoxybut-3-en-1-yl)-2-nitrobenzene (3x)
Prepared from acetal 1x (0.4 mmol) and allyl boronate 2a according to
the general method (eluant for PTLC: hexane/ethyl acetate=95:5). Pale
yellow liquid; yield: 73%; 1H NMR (600 MHz, [D1]chloroform, 208C,
TMS): d=2.44–2.49 (m, 1H), 2.55–2.59 (m, 1H), 3.23 (s, 3H), 4.85–4.88
(m, 1H), 5.06–5.09 (m, 2H), 5.85–5.92 (m, 1H), 7.43 (t, 1H, J=7.6 Hz),
7.65 (t, 1H, , J=7.6 Hz), 7.72 (d, 1H, J=8.3 Hz), 7.94 ppm (d, 1H, J=
7.6 Hz); 13C NMR (150 MHz, [D1]chloroform, 208C, TMS, both diaste-
reoisomers): d=42.0, 57.4, 78.6, 117.6, 124.4, 128.2, 128.2, 133.5, 134.2,
Figure 5. Correlation between the Lewis acidity of “electrophilic” boron-
based compounds and their “nucleophilic” reactivity toward Csp3-type
electrophiles.
138.0, 148.9 ppm; IR (neat): n˜ =3077, 2932, 1526, 1345, 1102 cmꢀ1
;
HRMS (DART): calculated for C11H14N1O3+ =[M+H]+: m/z=208.09737,
found: m/z=208.09816.
2528
www.chemasianj.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Asian J. 2011, 6, 2522 – 2529

Boron-Based Compounds as Pro-Nucleophiles and Co-Catalysts
2-(1,2-Dimethyl-3-buten-1-yl)-naphthalene (6)
[3] Y. Yamamoto, N. Asao, Chem. Rev. 1993, 93, 2207.
[4] a) R. W. Hoffmann, W. Ladner, Tetrahedron Lett. 1979, 20, 4653;
b) R. W. Hoffmann, Angew. Chem. 1982, 94, 569; Angew. Chem. Int.
Ed. Engl. 1982, 21, 555; c) R. W. Hoffmann, Pure Appl. Chem. 1988,
60, 123; d) Y. Yamamoto, T. Komatsu, K. Maruyama, J. Am. Chem.
Soc. 1984, 106, 5031; e) S. E. Denmark, N. G. Almstead, in Modern
Carbonyl Chemistry, Vol. 10 (Ed.: J. Otera), Wiley-VCH, Weinheim,
2000, pp. 299; f) S. R. Chemler, W. R. Roush in Modern Carbonyl
Chemistry, Vol. 11 (Ed.: J. Otera), Wiley-VCH, Weinheim, 2000,
pp. 403.
[5] a) A. Hosomi, T. Imai, M. Endo, H. Sakurai, J. Organomet. Chem.
1985, 285, 95; b) G. Kaur, M. Kaushik, S. Trehan, Tetrahedron Lett.
1997, 38, 2521; c) M. Rubin, V. Gevorgyan, Org. Lett. 2001, 3, 2705;
d) S. H. Kim, C. Shin, A. N. Pae, H. Y. Koh, M. H. Chang, B. Y.
Chung, Y. S. Cho, Synthesis 2004, 1581; e) S. K. De, R. A. Gibbs, Tet-
rahedron Lett. 2005, 46, 8345; f) M. Yasuda, T. Saito, M. Ueba, A.
Baba, Angew. Chem. 2004, 116, 1438; Angew. Chem. Int. Ed. 2004,
43, 1414; g) T. Saito, Y. Nishimoto, M. Yasuda, A. Baba, J. Org.
Chem. 2006, 71, 8516.
[6] U. Schneider, H. T. Dao, S. Kobayashi, Org. Lett. 2010, 12, 2488.
[7] H. T. Dao, U. Schneider, S. Kobayashi, Chem. Commun. 2011, 47,
692.
[8] Double allylation of acetals has also been reported: H. Tanaka, S.
Yamashita, Y. Ikemoto, S. Torii, Tetrahedron Lett. 1988, 29, 1721.
[9] T. Saito, Y. Nishimoto, M. Yasuda, A. Baba, J. Org. Chem. 2007, 72,
8588.
[10] For an example of a stoichiometrically formed boron–ate complex
that reacts with carbenium ion centers, see: R. Hunter, G. D. Tom-
linson, Tetrahedron Lett. 1989, 30, 2013.
[11] K. Matos, J. A. Soderquist, J. Org. Chem. 1998, 63, 461.
[12] H. C. Brown, U. S. Racherla, P. J. Pellechia, J. Org. Chem. 1990, 55,
1868.
Prepared from ether 1a (0.4 mmol) and a-methylallyl boronate 2c ac-
cording to the general method (eluant for PTLC: hexane). Colorless
liquid; yield: 70%; 1H NMR (600 MHz, [D1]chloroform, 208C, TMS):
d=0.78 and 0.94 (both diastereoisomers; d, 3H, J=6.2 Hz and 6.9 Hz),
1.22 and 1.26 (both diastereoisomers; d, 3H, J=7.6 Hz and 7.6 Hz), 1.42
(s, 1H), 2.32–2.40 (both diastereoisomers; m, 1H), 2.58–2.78 (m, 1H),
4.79–4.98 (both diastereoisomers; m, 2H), 5.58–5.71 (both diastereoiso-
mers; m, 1H), 7.24–7.26 (m, 1H), 7.32–7.37 (m, 2H), 7.49–7.51 (m, 1H),
7.67–7.72 ppm (m, 3H); 13C NMR (150 MHz, [D1]chloroform, 208C,
TMS; both diastereoisomers): d=16.8, 18.1, 19.2, 19.9, 43.8, 44.9, 45.0,
45.6, 113.6, 114.0, 125.1, 125.7, 125.8, 126.0, 126.1, 126.2, 126.8, 127.4,
127.5, 127.5, 127.6, 127.7, 132.1, 132.2, 133.4, 133.5, 142.7, 143.1, 143.2,
143.9 ppm.
2-(1-Methyl-3-butyn-1-yl)-naphthalene (7)
Prepared from ether 1a (0.4 mmol) and allenyl boronate 2e according to
the general method (eluant for PTLC: hexane). Colorless liquid; yield:
65%; 1H NMR (600 MHz, [D1]chloroform, 208C, TMS): d=1.46 (d, 3H,
J=6.9 Hz), 1.51 (s, 1H), 1.97 (t, 1H, J=2.8 Hz), 2.47–2.52 (m, 1H), 2.56–
2.60 (m, 1H), 3.13–3.17 (m, 1H), 7.37–7.46 (m, 3H), 7.66 (s, 1H), 7.78–
7.80 ppm (m, 3H); 13C NMR (150 MHz, [D1]chloroform, 208C, TMS;
both diastereoisomers): d=20.8, 27.5, 39.0, 69.6, 83.0, 125.0, 125.4, 125.5,
125.9, 127.6, 127.7, 128.0, 132.4, 133.5, 143.0 ppm.
3-(2,3,5-Tri-O-benzyl-a-d-ribofuranosyl)-1-propyne (10c)
Prepared from carbohydrate 8b (0.2 mmol) and allenyl boronate 2e ac-
cording to the general method (eluant for PTLC: hexane/ethyl acetate=
85:15, three times). Colorless liquid; yield: 25%; 1H NMR (600 MHz,
[D1]chloroform, 208C, TMS): d=1.99 (s, 1H), 2.61–2.63 (m, 1H), 2.64–
2.71 (m, 1H), 3.49–3.51 (m, 1H), 3.60–3.62 (m, 1H), 4.07–4.09 (m, 1H),
4.12–4.13 (m, 1H), 4.21–4.24 (m, 2H), 4.47 (d, 2H, J=11.0 Hz), 4.58 (t,
2H, J=11.7 Hz), 4.67 (d, 1H, J=11.7 Hz), 4.79 (d, 1H, J=11.0 Hz),
7.26–7.39 ppm (m, 15H); 13C NMR (150 MHz, [D1]chloroform, 208C,
TMS; both diastereoisomers): d=20.0, 69.5, 69.9, 72.5, 73.4, 73.7, 79.0,
79.4, 80.0, 81.4, 127.6, 127.7, 127.7, 127.7, 127.8, 127.9, 128.3, 138.1,
138.2 ppm; IR (neat): n˜ =3030, 2916, 2862, 1453, 1122, 1087, 1048, 1026,
736, 698 cmꢀ1; HRMS (DART): calculated for C29H31O4+ =[M+H]+: m/
z=443.22010, found: m/z=443.22223.
[13] Boranes 4e and 4 f may work as hard Lewis acids for the catalytic
ꢀ
activation of C O bonds.
[14] The Chemistry of C-Glycosides (Eds.: D. E. Levy, C. Tang), Perga-
mon Press, Oxford, UK, 1995.
[15] G. J. McGarvey, C. A. LeClair, B. A. Schmidtmann, Org. Lett. 2008,
10, 4727.
[16] C. L. B. Macdonald, A. M. Corrente, C. G. Andrews, A. Taylor,
B. D. Ellis, Chem. Commun. 2004, 250.
[17] R. Kumar, D. Kumar, A. K. Chakraborti, Synthesis 2007, 299.
[18] a) R. Nakamura, K. Tanino, M. Miyashita, Org. Lett. 2003, 5, 3583;
b) G. Xu, K. D. Moeller, Org. Lett. 2010, 12, 2590.
[19] W. R. Roush, M. A. Adam, A. E. Walts, D. J. Harrish, J. Am. Chem.
Soc. 1986, 108, 3422.
Acknowledgements
This work was partially supported by a Grant-in-Aid for Scientific Re-
search from the Japan Society for the Promotion of Science (JSPS) and
the Global COE Program (Chemistry Innovation through Cooperation
of Science and Engineering), The University of Tokyo, MEXT, Japan.
[20] G. F. Fang, O. A. Wallner, N. Di Blasio, X. Ginesta, J. N. Harvey,
V. K. Aggarwal, J. Am. Chem. Soc. 2007, 129, 14632.
[21] R. W. Hoffmann, J. J. Wolff, Chem. Ber. 1991, 124, 563.
[22] K. Tonogaki, K. Itami, J.-i. Yoshida, J. Am. Chem. Soc. 2006, 128,
1464.
[23] M. Barbero, S. Bazzi, S. Cadamuro, S. Dughera, C. Piccinini, Synthe-
sis 2010, 315.
[24] P. Nagaraj, N. G. Ramesh, Eur. J. Org. Chem. 2008, 4607.
[25] J. S. Yadav, B. V. S. Reddy, K. Srinivasa Reddy, L. Chandraiah, V.
Sunitha, Synthesis 2004, 2523.
[26] T. Mukaiyama, S. Kobayashi, Carbohydr. Res. 1987, 171, 81.
[27] M. Isobe, W. Phoosaha, R. Saeeng, K. Kira, C. Yenjai, Org. Lett.
2003, 5, 4883.
[1] Comprehensive Organometallic Chemistry III (Eds.: R. Crabtree, M.
Mingos), Elsevier Ltd., Amsterdam, 2006.
[2] a) W. R. Roush in Comprehensive Organic Synthesis (Ed.: B. M.
Trost), Pergamon Press, Oxford, UK, 1991; b) D. S. Matteson in Ster-
eodirected Synthesis with Organoboranes, Springer, Berlin, 1995;
c) Science of Synthesis: Houben-Weyl Methods of Molecular Trans-
formation, Organometallics, Vol. 6 (Eds.: D. E. Kaufmann, D. S.
Matteson), Thieme, 2005; d) J. W. J. Kennedy, D. G. Hall, in Boronic
Acids, Vol. 11 (Ed.: D. G. Hall), Wiley-VCH, Weinheim, 2005,
pp. 241; e) N. Miyaura, A. Suzuki, Chem. Rev. 1995, 95, 2457.
Received: January 31, 2011
Published online: July 4, 2011
Chem. Asian J. 2011, 6, 2522 – 2529
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemasianj.org
2529