Full chirality transfer in the conversion of secondary alcohols into tertiary boronic esters and alcohols using lithiation-borylation reactions

Article abstract of DOI:10.1002/anie.201001371

New conditions, full transfer: Using MgBr₂/MeOH as an additive now provides essentially 100 % retention of configuration in the lithiation-borylation reaction, thus leading to tertiary boronic esters (or tertiary alcohols) with exceptionally hi

Full text of DOI:10.1002/anie.201001371

Communications
DOI: 10.1002/anie.201001371
Enantioselective Synthesis
Full Chirality Transfer in the Conversion of Secondary Alcohols into
Tertiary Boronic Esters and Alcohols Using Lithiation–Borylation
Reactions**
Viktor Bagutski, Rosalind M. French, and Varinder K. Aggarwal*
Dedicated to Professor Saverio Florio on the occasion of his 70th birthday
A general method for the preparation of fully substituted
carbon atoms (e.g. quaternary stereogenic centers[1] or
tertiary alcohols[2]) that routinely gives high enantioselectiv-
ities (> 96% ee) with broad substrate scope is one of the most
challenging goals in organic synthesis.[3] Our research group
has recently described a conceptually new method that comes
close to meeting such a challenge (Scheme 1).[4] In this
process, readily available enantioenriched secondary alcohols
considerable erosion of enantioselectivity. Herein, we provide
a rationale for the observed decrease in enantioselectivity and
through understanding the intricacies of the process, we also
present a solution to the problem that now results in
> 98% ee for all the substrates tested, even the most
demanding.
The problem we encountered is illustrated in Scheme 2
(and expanded in the Supporting Information). Thus, reaction
of the para-chlorophenyl-substituted carbamate 1c with
Scheme 2. Example of the dependence of ee values on the stoichiom-
etry of the boronic ester.
Scheme 1. Lithiation–borylation of enantioenriched secondary carba-
mates.
were first converted into carbamate compounds. Lithiation
and subsequent reaction with boron reagents gave, after
oxidative workup, tertiary alcohols with high ee values.
Perhaps the most intriguing facet of this novel methodology
was that in all cases the reactions occurred with essentially
complete inversion of configuration when boranes were
employed but almost complete retention of configuration
when boronic esters were used. Although the reaction worked
well for simple substrates (> 90% ee), we have found that the
introduction of sterically more demanding groups in the
carbamate or boronic ester or the introduction of electron-
withdrawing aromatic groups in the carbamate resulted in a
EtBpin 2b (1.1 equiv) gave the tertiary boronic ester with
only 40% ee. Furthermore, we observed an unusual depen-
dence of the ee value on the stoichiometry: increasing the
stoichiometry of 2b from 1.1 equivalents to 3 equivalents led
to an increase in the ee value of the product 4cb from 40% to
96% ee.
Although a number of possible explanations can be
advanced for this observation, we focused on the possible
scenario shown in Scheme 3. We suspected that the critical
issue was not the selectivity in the formation of the ate-
complex, for which it would be difficult to rationalize the
dependence of the ee value on stoichiometry, but instead the
fate of the ate-complex upon warming. Even though the
desired stereospecific 1,2-migration occurred upon warming,
it was also possible that competing dissociation of the ate-
complex (kÀ1) back to the starting lithiated carbamate and
boronic ester species might also take place.[5] This dissociation
could be followed by racemization of the lithiated carbamate
and subsequent erosion of the ee value. The dependence of
the ee value on the stoichiometry can therefore be under-
stood. At low stoichiometry, reversion to starting materials
results in a low concentration of the boronic ester, and so the
subsequent recombination of the lithiated carbamate with
RBpin is inevitably slow. This pathway gives the lithiated
[*] Dr. V. Bagutski, R. M. French, Prof. V. K. Aggarwal
School of Chemistry, University of Bristol
Cantock’s Close, Bristol, BS81TS (UK)
Fax: (+44)117-925-1295
E-mail: v.aggarwal@bristol.ac.uk
[**] We thank the EPSRC for support of this work. V.K.A. thanks the Royal
Society for a Wolfson Research Merit Award and the EPSRC for a
Senior Research Fellowship. We thank Frontier Scientific for the
generous donation of boronic acids and boronic esters. We thank
Prof. Guy Lloyd-Jones for useful discussions.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201001371.
5142
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 5142 –5145

Angewandte
Chemie
Scheme 4. Suitable electrophiles for trapping experiments.
Scheme 3. Dissociation of ate-complex as a possible cause of erosion
in ee values.
ester to form the ate-complex. Next, D2O (electrophile 2;
1 equiv)—which is a more reactive electrophile than allyl
bromide towards the lithiated carbamate Li-1a—was then
added and subsequent warming would give a series of
products in ratios that reflected the efficiency of all three
individual reaction steps. Thus, any recovered carbamate 1
would show the efficiency of the deprotonation step (typi-
cally, 97%), the amount of allylated carbamate would show
the extent of the reaction of the lithiated carbamate with the
boronic ester, whereas the amount of deuterated carbamate
would reflect the fraction of the ate-complex that reverted
back into starting materials upon warming. Application of this
two-electrophile test shown in Scheme 5 to a series of
substrates is summarized in Table 1 and revealed the follow-
ing: 1) the reaction of the lithiated carbamate derived from 1a
with iPrBpin was incomplete after 30 minutes because 9% of
the allylated carbamate El1-1a was isolated, thus showing that
the formation of the ate-complex is slow enough to be
quantified (Table 1, entry 1). 2) A greater proportion of the
deuterated carbamate El2-1 was obtained with the more
electron-deficient aryl carbamate compared to the phenyl
carbamate (54% vs. 25%; compare Table 1, entries 2 and 1)
and with the more hindered pinacol ester compared to the
neopentyl glycol ester (54% vs. < 2%; compare Table 1,
entries 2 and 3). With boronic ester 3d bearing the less
hindered neopentyl boronic ester but more hindered boron
substituent together with the hindered aryl carbamate 1 f
considerable reversion of the ate-complex still occurred, and
generated 29% of the deuterated carbamate El2-1 f (Table 1,
entry 4). These results show that the stability of the ate-
complex depends on both the steric and electronic effects of
both the lithiated carbamate and the boronic ester species.
Although this strategy was clearly successful in enhancing
the ee value of the product, it did so at the expense of yield. To
address this issue we needed to enhance the relative rate of
the 1,2-migration (k2) and so we considered the use of Lewis
acids. After some preliminary experimentation[8] we found
carbamate time to racemize[6] prior to recombination, which
results in a low ee value. At high stoichiometry, the larger
concentration of the boronic ester results in a more rapid
recombination of the lithiated carbamate with RBpin and so
there is less time available for racemization to occur, thus a
higher ee value is observed. The fact that lower ee values were
observed with hindered boronic esters and electron-deficient
aromatic carbamates is consistent with this model because in
both cases the rate of the reverse reaction would be enhanced
and the rate of the forward reaction would be lower.
If this model is correct, the addition of a second, more
reactive electrophile than the boronic ester after ate-complex
formation should result in the trapping of any lithiated
carbamate formed from the reverse reaction. This trapping
would prevent recombination of the racemized lithiated
carbamate with the boronic ester and would therefore lead
to a higher ee value of the tertiary alcohol. We decided to test
this idea using three different electrophiles (D2O, TMSCl, and
allyl bromide) with the hindered boronic ester 2c (1.2 equiv;
Scheme 4). In all three cases the tertiary boronic ester 4ac was
formed with complete retention of configuration (99% ee),
which shows that the addition of the boronic ester to the
lithiated carbamate Li-1a is completely stereoselective.[7] In
fact, this simple test allows the determination of the maximum
ee value achievable in a lithiation–borylation reaction for any
substitution pattern.
To obtain a more complete picture of the fate of various
intermediates along the lithiation–borylation pathway, we
have developed a new reaction sequence (Scheme 5). Thus,
after treatment of the lithiated carbamate Li-1 with the
boronic ester over 30 minutes, allyl bromide (electrophile 1;
1 equiv) was added and the mixture was stirred for 15 minutes
at À788C. This allylation would result in the trapping of any
lithiated carbamate that had not reacted with the boronic
Scheme 5. Mechanistic representation of the “two-electrophile test” aimed to quantify the fate of reactive intermediates in the lithiation–borylation
reaction.
Angew. Chem. Int. Ed. 2010, 49, 5142 –5145
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5143

Communications
Table 1: Study of the borylation 1,2-migration reaction by the “two-
electrophile test”. Selected examples.
Table 2: Preparation of highly enantioenriched tertiary pinacolboronic
esters.
Entry
Substrate
R
R2Bpin
R2
Yield of 4 [%][a]
(ee [%][b])
1
R1
2
1
2
3
4
5
6
7
8
1a
1a
1b
1b
1b
1b
1b
1c
1c
1c
1d
1d
1e
1e
1 f
H
H
H
H
H
H
H
4-Cl
4-Cl
4-Cl
4-F
4-F
2-F
Me
Me
Et
Et
Et
2c
2d
2a
2c
2d
2e
2 f
2b
2c
2d
2c
2d
2c
2g
2g
2c
2d
2g
iPr
cHex
Me
iPr
cHex
cinnamyl
4-BrC6H4
Et
iPr
cHex
iPr
cHex
iPr
Ph
Ph
iPr
4ac
92 (99)
87 (99)
71 (99)
74 (99)
4ad
4ba
4bc
4bd
4be
4bf
4cb
4cc
4cd
4dc
4dd
4ec
4eg
4 fg
4gc
4gd
4gg
61 (99)
Et
Et
91 (99)[c]
82 (99)[c]
91 (99)
Entry
Substrate
Yield [%][a] (ee [%][b])
1
RAr
2 or 3 R2
El1-1 El2-1
4 or 5
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
9
89 (99)
75 (99)
88 (99)
69 (99)
1
2
3
4
1a
1c 4-Cl
1c 4-Cl
H
2c
2c
3c
iPr
iPr
iPr
cHex
9
17
<2
<2
25
54
<2
29
4ac 60 (99)
4cc 27 (99)
5cc 92 (99)
5 fd 48 (98)
10
11
12
13
14
15
16
17
18
1 f
2-Me 3d
64 (96)[d]
74 (99)[c]
64 (99)[c,e]
79 (99)[e]
65 (99)[c,e]
93 (99)[c,e]
[a] Yield of isolated product. [b] The ee values were determined after
oxidation of an aliquot of 4/5 (see the Supporting Information).
2-F
2-Me
2-MeO
2-MeO
2-MeO
1g
1g
1g
cHex
Ph
that use of MgBr2 in MeOH fulfilled our requirements, and
led to both high yields and high ee values (Table 2). Presum-
ably, the MgBr2 enhances the k2/kÀ1 ratio, while the MeOH
reprotonates any lithiated carbamate generated from the
dissociation of the ate-complex, thus preventing it from
eroding the ee value. These new reaction conditions were
found to be general over a broad range of substrates
(Table 2).
[a] Yield of isolated product. [b] The ee values were determined after
oxidation of an aliquot of 4 (see the Supporting Information). [c] Used
1.2 equivalents of boronic ester. [d] After addition of 2c the reaction
mixture was stirred at À788C for 3 hours. Note that the use of 3c instead
of 2c resulted in an increased yield and ee value for 5ec (Table 3, entry 8).
[e] The carbamate was deprotonated in the presence of TMEDA within
10 minutes. THF=tetrahydrofuran, TMEDA=N,N,N’,N’-tetramethyl-
ethylenediamine.
Carbamate 1a gave complete retention of configuration
with both iPrBpin 2c and cHexBpin 2d (Table 2, entries 1 and
2). The more hindered carbamate 1b also reacted with high
enantioselectivity with all the boronic esters explored from
the least hindered 2a to the most hindered alkyl 2d, as well as
allyl and aryl substrates (2e–f, respectively; Table 2, entries 3–
7). The dramatic enhancement of selectivity is illustrated in
the reaction outlined in entry 10 of Table 2: only 49% ee was
observed under the former conditions[4a] but under the new
conditions 99% ee was obtained. Aryl carbamates bearing
electron-withdrawing substituents in the para position also
gave high enantioselectivities even with the most sterically
demanding boronic esters (Table 2, entries 8–12). These
examples represent some of the most demanding reaction
partners as these substrates are most prone to reversibility in
ate-complex formation and thus erosion in ee values, but
nevertheless they gave essentially high enantioselectivities.
Additional steric hindrance in the ortho position was also
tolerated again, and led to high selectivities in the formation
of tertiary boronic esters (Table 2, entries 13–18).
oxidation. This change was especially needed the case of 4gc
and 4gd (Table 2, entries 16 and 17).
Because of the issues relating to slow oxidation, we
investigated the less hindered neopentyl boronic esters
instead of pinacol esters. Furthermore, as they are less
hindered than pinacol esters they were expected to be less
prone to undergo dissociation of the respective ate-complexes
(as revealed in the “two-electrophile test”; compare Table 1,
entries 2 and 3) and so more likely to give higher ee values. As
a slight drawback, the neopentyl boronic esters have to be
synthesized from the corresponding acids because unlike
many of the pinacol esters they are not commercially
available and are less stable to chromatography on silica
gel.[11]
Thus, a second set of reaction conditions was established
employing neopentyl boronic esters instead of pinacol esters
and the results are summarized in Table 3. This study revealed
that certain advantages accrued with these substrates: 1) high
ee values were achieved without the need for MgBr2/
MeOH[12] over a broad range of substrates, 2) the tertiary
boronic esters could be oxidized to the alcohols directly in one
pot without solvent exchange.
The tertiary pinacolboronic esters 4 provide a very
promising pool of chiral building blocks of exceptionally
high enantiomeric purity for many applications.[9,10] However,
they could not always be oxidized directly to the correspond-
ing alcohols by treatment with H2O2/NaOH (aq), but instead
required a solvent exchange from Et2O to THF prior to
Once again high yields and enantioselectivities were
obtained even with the most demanding substrates, for
5144
www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 5142 –5145

Angewandte
Chemie
Angew. Chem. Int. Ed. 2007, 46, 9291 – 9294; b) T. Tsubogo, S.
Table 3: Preparation of highly enantioenriched tertiary neopentylboronic
esters or tertiary alcohols.
Saito, K. Seki, Y. Yamashita, S. Kobayashi, J. Am. Chem. Soc.
2008, 130, 13321 – 13332; c) J. P. Das, H. Chechik, I. Marek, Nat.
Chem. 2009, 1, 128 – 132; d) Y. Lee, B. Li, A. H. Hoveyda, J. Am.
Chem. Soc. 2009, 131, 11625 – 11633; for a review, see: e) E. J.
Corey, A. Guzman-Perez, Angew. Chem. 1998, 110, 402 – 415;
Angew. Chem. Int. Ed. 1998, 37, 388 – 401; f) Quaternary Stereo-
centres: Challenges and Solutions for Organic Synthesis (Eds.: J.
Christoffers, A. Baro), Wiley-VCH, Weinheim, 2005; g) O.
Riant, J. Hannedouche, Org. Biomol. Chem. 2007, 5, 873 – 888.
[2] a) S.-J. Jeon, H. Li, P. J. Walsh, J. Am. Chem. Soc. 2005, 127,
16416 – 16425; b) U. Schneider, M. Ueno, S. Kobayashi, J. Am.
Chem. Soc. 2008, 130, 13824 – 13825; c) V. J. Forrat, D. J. Ramꢀn,
M. Yus, Tetrahedron: Asymmetry 2009, 20, 65 – 67; d) I. Chen, L.
Yin, W. Itano, M. Kanai, M. Shibasaki, J. Am. Chem. Soc. 2009,
131, 11664 – 11665; for a recent review, see: e) M. Hatano, K.
Ishihara, Synthesis 2008, 1647 – 1675.
[3] For general reviews, see: a) B. M. Trost, C. Jiang, Synthesis 2006,
369 – 396; b) M. Shibasaki, M. Kanai, Chem. Rev. 2008, 108,
2853 – 2873; c) M. M. Hussain, P. J. Walsh, Acc. Chem. Res. 2008,
41, 883 – 893; d) M. Bella, T. Gasperi, Synthesis 2009, 1583 –
1614.
[4] a) J. L. Stymiest, V. Bagutski, R. M. French, V. K. Aggarwal,
Nature 2008, 456, 778 – 782. These reactions are based on the
seminal studies by Hoppe on the lithiation and trapping of
secondary benzylic carbamates. See: b) D. Hoppe, A. Carstens,
T. Krꢁmer, Angew. Chem. 1990, 102, 1457 – 1459; Angew. Chem.
Int. Ed. Engl. 1990, 29, 1424 – 1425; c) A. Carstens, D. Hoppe,
Tetrahedron 1994, 50, 6097 – 6108.
Entry
Substrate
R
R2Bneop
3
R2
Yield of 5/6 [%][a]
(ee [%][b])
1
R1
Me 3d cHex 6ad
1
2
3
4
5
6
7
8
1a
1b
1b
H
H
H
–
–
–
–
–
94 (99)
98 (99)
94 (99)
95 (99)
95 (99)
79 (99)
87 (99)
74 (99)
Et
3c iPr
3d cHex 6bd
6cc
Me 3d cHex 6cd
Me 3c iPr
Me 3d cHex 6dd
Me 3c iPr 5/6ec 67
6bc
Et
1c 4-Cl
1c 4-Cl
1d 4-F
1d 4-F
1e 2-F
1 f 2-Me
Me 3c iPr
5/6dc 63
–
9
Me 3d cHex 5/6fd 65[c] 48 (98)[d,e]
10
11
1g 2-MeO Me 3c iPr
1g 2-MeO Me 3d cHex 5/6gd 70
5/6gc 89
92 (96)[d]
68 (96)[d,e,f]
[a] Yield of isolated boronic esters 5 (first column) or alcohols 6 (second
column; formed in one-pot operation). [b] The ee values were
a
[5] H. C. Brown, N. Vasumathi, N. N. Joshi, Organometallics 1993,
12, 1058 – 1067.
determined after oxidation of an aliquot of 5 (see the Supporting
Information). [c] 97% ee. [d] The carbamate was deprotonated in the
presence of TMEDA within 10 minutes. [e] Ethereal solution of D2O was
added at À788C after addition of R2Bneop. [f] Isolated boronic ester was
oxidized according to procedure GP3A in the Supporting Information.
[6] The lithiated secondary benzylic carbamates are generally
configurationally unstable above À608C.
À
[7] These reactions show that the pinacol ester B O bonds of the
ate-complex are essentially stable and do not undergo ring
opening under the reaction conditions employed. This ring
opening would have given an intermediate alkoxide which would
have been trapped by the electrophiles employed, but this was
not observed.
example, reactions between hindered or electron-withdraw-
ing carbamates with hindered boronic esters (Table 3).
In conclusion, we have demonstrated that the lithiation–
borylation of secondary carbamates to give tertiary boronic
esters (or tertiary alcohols) is accompanied by considerable
erosion in ee values as a result of the intermediate ate-
complex dissociating back to the starting lithiated carba-
mate—a species that is prone to racemization upon warming.
The dissociation–racemization pathway can be essentially
eliminated through the use of either MgBr2/MeOH or
neopentyl boronic esters in place of pinacol esters. These
two sets of reaction conditions now provide essentially
complete chirality transfer in the lithiation–borylation reac-
tion and lead to tertiary boronic esters (or tertiary alcohols)
with exceptionally high ee values in all cases, even the most
demanding.
[8] We first optimized the ate-complex formation using carbamate
1c and boronic ester 2c as these are a particularly challenging
pair of substrates. Using 1.5 equivalents of boronic ester 2c and
leaving it to react with the lithiated carbamate for 1 hour at
À788C led to complete ate-complex formation (no allylated
material was formed).
[9] Conversion of tertiary boronic esters into trifluoroborates: a) V.
Bagutski, A. Ros, V. K. Aggarwal, Tetrahedron 2009, 65, 9956 –
9960; and their reactions with aldehydes: b) A. Ros, V. K.
Aggarwal, Angew. Chem. 2009, 121, 6407 – 6410; Angew. Chem.
Int. Ed. 2009, 48, 6289 – 6292.
[10] Tertiary trifluoroborates can also be converted into the corre-
sponding tertiary amines: V. Bagutski, V. K. Aggarwal, unpub-
lished results.
[11] Tertiary neopentylboronic esters 5 could, in general, be isolated
by distillation or, in certain instances (Table 3, entries 6, 8–11),
by dry column vacuum chromatography (DCVC) techniques.
For details see the Supporting Information.
Received: March 8, 2010
Published online: June 11, 2010
[12] In fact, the use of MgBr2/MeOH with neopentylglycol esters
caused a significant decrease in yield of the desired products 5 as
À
result of B O bond cleavage (instead of 1,2-migration). This
Keywords: asymmetric synthesis · boronic esters · chirality ·
quaternary stereocenters · tertiary alcohols
outcome was determined by 11B NMR analysis of the reaction
mixture which showed a major signal for borinic esters at 50–
53 ppm.
.
[1] For recent examples, see: a) G. Kolodney, G. Sklute, S. Perrone,
P. Knochel, I. Marek, Angew. Chem. 2007, 119, 9451 – 9454;
Angew. Chem. Int. Ed. 2010, 49, 5142 –5145
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5145