Taking the F out of FLP: Simple lewis acid-base pairs for mild reductions with neutral boranes via borenium ion catalysis

Article abstract of DOI:10.1021/ja307374j

Discrete three-coordinate borenium salts 1c and 1d are accessed by cooperative Lewis acid-base pair-mediated heterolytic splitting of the B-H bond in pinacolborane by B(C₆F₅)₃DABCO and Ph ₃C⁺/DABCO, respectively. The resulting salts are competent catalysts in the reduction of a broad range of imines and can be generated in situ. Moreover, a mechanistic framework for borenium catalysis based on experimental evidence is proposed. The reaction is suggested to proceed by borenium activation of the imine substrate followed by counterintuitive hydride delivery from HBPin (with the assistance of DABCO) rather than from the HB(C₆F₅)₃^- anion, contrary to typical mechanisms of reduction in FLP systems.

Full text of DOI:10.1021/ja307374j

Communication
pubs.acs.org/JACS
Taking the F out of FLP: Simple Lewis Acid−Base Pairs for Mild
Reductions with Neutral Boranes via Borenium Ion Catalysis
Patrick Eisenberger, Adrian M. Bailey, and Cathleen M. Crudden*
Department of Chemistry, Queen’s University, 90 Bader Lane, Kingston, Ontario, Canada K7L 3N6
S
* Supporting Information
can be used catalytically in the presence of suitable receptors for
ABSTRACT: Discrete three-coordinate borenium salts 1c
the heterolytically cleaved E−H bonds as long as the
and 1d are accessed by cooperative Lewis acid−base pair-
mediated heterolytic splitting of the B−H bond in
pinacolborane by B(C6F5)3·DABCO and Ph3C+/
DABCO, respectively. The resulting salts are competent
catalysts in the reduction of a broad range of imines and
can be generated in situ. Moreover, a mechanistic
framework for borenium catalysis based on experimental
evidence is proposed. The reaction is suggested to proceed
by borenium activation of the imine substrate followed by
counterintuitive hydride delivery from HBPin (with the
regeneration of the Lewis acid catalyst (which is almost
exclusively B(C6F5)3) is suitably efficient.18,13,15
Recently, our group proposed the formation of borenium ion
1a in the reaction between B(C6F5)3 and HBPin in the presence
of a Lewis base by an FLP-type mechanism (eq 1).19 In the
−
assistance of DABCO) rather than from the HB(C6F5)3
anion, contrary to typical mechanisms of reduction in FLP
systems.
presence of an acceptor olefin and a Rh complex, this resulted in a
dramatically improved system for the hydroboration of olefins as
a result of accelerated cleavage of the B−H bond.19 Expanding on
this concept, we speculated that borenium ions such as 1a and 1b
could act as unique catalysts for the reduction of organic species
under mild, metal-free conditions because they combine a
strongly acidic borenium ion and a borohydride anion in the
same salt.
Borenium ion generation. As a starting point, we began by
investigating the ability of various Lewis acid−base pairs to effect
heterolytic splitting of the B−H bond of shelf-stable borane
reagents, such as catecholborane (HBCat) and HBPin. Because
of the propensity for tetrahydrofuran to polymerize when
exposed to strongly acidic catalyts,19,20 we focused on DABCO as
the Lewis base.
lectrophilic boron cations are interesting alternatives to
Eexpensive, potentially toxic, high-molecular-weight metal-
based Lewis acids. In particular, three-coordinate boron cations
(borenium ions) were rarities a decade ago but are now readily
accessible reagents.1−3 Despite this fact, until recently, virtually
the only application of these reactive species was as proposed
intermediates in oxazaborolidene reductions.4 However, in the
last two years, work has appeared that illustrates the significant
potential for discrete borenium intermediates to participate in
the synthesis of boron-containing organic compounds. For
example, electrophilic aromatic substitution reactions mediated
by electrophilic borenium ions have been reported by the
Ingleson group,5,6 and Vedejs has studied the stoichiometric,
intramolecular C(sp3)−H insertions of these species.7 Vedejs
and Curran recently reported the use of NHC·BH3 complexes in
the generation of borenium ion equivalents that are able to
catalyze hydroborations using the parent NHC·BH3 complex as
the stoichiometric reductant.8
As shown in eq 2, mixing equimolar amounts of B(C6F5)3 and
DABCO resulted in the formation of an insoluble Lewis acid−
Herein we report the generation and isolation of stable
borenium salts and the first example of their use as catalysts for
metal-free reductions with stable, neutral boranes.9 Isotopic
labeling studies and kinetic experiments provide evidence for
true borenium catalysis rather than frustrated Lewis pair (FLP)
chemistry, leading to the interesting conclusion that the 1,4-
diazabicyclo[2.2.2]octane (DABCO) complex of the neutral
borane pinacolborane (HBPin) is a better reducing agent than
base adduct since DABCO, unlike typical bulky FLP-type bases,
is sterically unhindered. Interestingly, exposure of this
DABCO·B(C6F5)3 complex to an equimolar amount of HBPin
resulted in the clean, high-yielding formation of borenium ion 1c
(eq 2), illustrating that even this classical Lewis base pair is
capable of cleaving nonpolar bonds such as the H−B bond in
HBPin.21,22
10
−
the anionic borohydride HB(C6F5)3 .
Significant advances have been made in the last five years in the
area of FLP chemistry.11 Typically, highly sterically hindered and
reactive Lewis acids and bases are used in concert to activate
small molecules such as silanes12−15 and H2.16,17 These systems
Received: July 26, 2012
Published: October 2, 2012
© 2012 American Chemical Society
17384
dx.doi.org/10.1021/ja307374j | J. Am. Chem. Soc. 2012, 134, 17384−17387

Journal of the American Chemical Society
Communication
Borenium ion 1c is characterized by two signals in the 11B
Table 1. B(C6F5)3·DABCO-Catalyzed Reduction of CN
NMR spectrum, one at 25.4 ppm due to the borenium ion and
Double Bonds
one at −25.0 ppm due to the borohydride counterion (1JB,H
=
89.3 Hz). This outcome parallels that of the FLP B(C6F5)3/
tBu3P, which is known to split the B−H bond of HBCat to form a
related phosphaborenium hydridoborate salt.23
However, all attempts to isolate 1c were unsuccessful. Since
trityl salts are isoelectronic with B(C6F5)3 and have similar
hydride affinities,24 we examined the formation of borenium ions
25
−
using Ph3C+B(C6F5)4 . When the reaction between this salt
and HBPin was carried out in the presence of DABCO, the
closely related borenium complex 1d was obtained (Chart 1).26
Chart 1. Borenium and Boronium Ions Prepared by Hydride
Abstraction with Trityl Salts
a
Yields relative to substrate; isolated yields are given in parentheses.
Isolated as the trifluoroacetamide. 8 h reaction time. 20 h reaction
b
c
d
time, the product was isolated along with hydrocinnamaldehyde in a
e
f
6:1 ratio. Only trace hydrocinnamaldehyde was observed. 100 °C.
g
The product was isolated as the pivalylamide.
Scheme 1. Potential Mechanism for Reduction by Borenium
Borohydride Salts in Analogy to FLP Hydrogenations
Using this procedure, we were consistently able to isolate 1d in
>90% yield by concentration, precipitation with pentanes, and
repeated extraction of the resulting solid with pentanes to
remove Ph3CH. The NMR data for the borenium fragment of 1d
matched those of 1c.
Other Lewis bases were also investigated (Chart 1). For
example, the reaction of 2,2,6,6-tetramethylpyridine and 1,8-
diazabicyclo[5.4.0]undec-7-ene with HBPin and Ph3C+B-
(C6F5)4− led to the clean generation of borenium ions 2 and 3,
respectively, as determined by NMR spectroscopy, although
neither of these was isolated in pure form.27 The chelating chiral
amine (−)-sparteine led to the corresponding four-coordinate
boronium ion 4.
Catalytic reductions. As previously noted, when generated by
FLP chemistry, salts 1a−c combine a highly electrophilic
borenium cation with a reactive borohydride (eqs 1 and 2) and
thus should be well-suited to catalyze mild, metal-free reductions
of unsaturated organic molecules. In the event, exposing N-
benzylidene-2-methylpropan-2-amine (5a) to 1 equiv of HBPin
and 5 mol % B(C6F5)3·DABCO (to generate 1c in situ) in PhCF3
resulted in clean reduction to the corresponding pinacolbor-
amide 6a (δB = 24.7 ppm) after 45 min in full conversion. As
shown in Table 1, this catalyst system effectively hydroborates a
range of N-alkyl and N-aryl aldimines.26
These results stand in contrast to related FLP-based
hydrogenation reactions of aldimines, where the nature of the
N-substituent is highly restricted because release of the Lewis
acid from the reduced product is the proposed turnover-limiting
step.16,28 In addition to aldimines, nitriles and N-heterocycles
such as acridine were also cleanly reduced to give the N-
protected primary amine and the 1,4 addition product,
respectively.
splits the H−B bond in HBPin, with DABCO acting as the Lewis
base. The reaction of the resulting borenium ion 1c with imine 5
gives activated iminium ion 12, which has a borohydride as its
counterion. By analogy with FLP chemistry, this borohydride
would then reduce the activated iminium 12, regenerating the
B(C6F5)3 catalyst.
Remarkably, however, we found that borenium ions 1d
prepared with unreactive (non-hydridic) counterions such as
B(C6F5)4− were also highly active catalysts for the reduction of
CN bonds in the presence of HBPin (Table 1, final column).
In all cases, except when elevated temperatures were employed,
borenium ion 1d with no apparent hydride source outperformed
1c.29 Thus, the mechanism of reduction under these conditions
must be different from that proposed for FLP reductions with H2.
In addition to its high reactivity in the reduction of regular
aldimines, 1d reduced α,β-unsaturated aldimines such as 5d
more cleanly than 1c did, and it also proved to be highly
competent in the reduction of a variety of ketimines (Table 2).
To provide a direct comparison of the catalytic efficiency of
B(C6F5)3 in an FLP-type mechanism with that of borenium ions
bearing unreactive counterions, we examined the kinetics of the
reduction of imine 5e with HBPin using catalytic quantities of
Mechanism. With some idea of the scope and limitations in
hand, we speculated on the mechanism of the reaction (Scheme
1). In a typical FLP mechanism, B(C6F5)3 acts as the catalyst and
17385
dx.doi.org/10.1021/ja307374j | J. Am. Chem. Soc. 2012, 134, 17384−17387

Journal of the American Chemical Society
Communication
of the B−H bond in HBPin takes place in the turnover-limiting
step. Consistent with this, the reaction was found to be first order
in HBPin over certain concentration ranges [see the Supporting
Information (SI)]. Since the direct reduction of a borenium-
activated iminium ion with HBPin would result in the unlikely
generation of the highly unstable borinium ion PinB+, we then
examined the role of DABCO in the reaction.
Table 2. Borenium-Catalyzed Reduction of Ketimines
Kinetic analysis of the role of DABCO was complicated since 1
equiv of DABCO was always present in the medium as a
stabilizing group for the borenium ion and the addition of excess
DABCO inhibited the reduction, likely by preventing borenium
transfer to the less Lewis basic imine.29a However, we were able
to prepare iminium ion 14 as a model for a borenium-activated
iminium ion (eq 4). Treatment of 14 with excess HBPin gave no
a
b
c
Average of two independent experiments. lsolated yield. Acid−base
d
treatment was required to cleave the N−B bond completely. The
solvent was 2:1 PhCF3:CH2CI2.
reaction, but in the presence of DABCO, 14 was fully consumed.
When DBPin was employed, deuterium incorporation in the
product was observed. When DABCO was replaced with imine
5e to probe the Lewis basicity of this species, we observed no
consumption of 14, indicating that the imine itself is not Lewis
basic enough to activate HBPin. In addition, low-temperature
NMR studies revealed an interaction between HBPin and
DABCO, while there was no observable interaction between
HBPin and imine 5e (see the SI). Taken together, these data
indicate that the strong Lewis base DABCO is required in the
hydride delivery step of the reduction.29b
−
borenium ions 1d [DABCO−BPin+]B(C6F5)4 and 1c
[DABCO−BPin+]HB(C6F5)3, the latter generated in situ from
B(C6F5)3·DABCO and HBPin. The reaction rates for the two
systems were on the same order of magnitude, with the non-
hydridic system 1d actually reacting somewhat faster than the
borohydride-containing Lewis acid−base pair 1c (Figure 1).
Thus, borenium ion 1d is clearly kinetically competent to
catalyze the hydroboration of imines.
As a final test of the mechanism of reduction, we employed
perdeuterated borenium 1d-d12 (prepared from HBPin-d12) at 10
mol % loading in the catalytic reduction of 5e by HBPin (eq 5).
The corresponding d12-labeled product was observed together
with the regular protio product by electron impact/time of flight
mass spectroscopy in a ratio of ca. 1:10 based on signal
intensities. Thus, the BPin group from the initial borenium ion is
incorporated into the product and then regenerated after transfer
of the hydride of HBPin to the active iminium ion.
Figure 1. Initial rates of reduction of 5e by HBPin catalyzed by 7.5 mol
% 1d (blue) and 7.5 mol % B(C6F5)3·DABCO (red).
A mechanism consistent with all of these observations is
shown in Scheme 2. The reaction begins with the transfer of a
borenium ion from catalyst 1c or 1d to imine 5. The resulting
boron-activated iminium ion 12 is then reduced by HBPin with
assistance from the Lewis base DABCO, likely by formation of a
precomplex as implied by the low-temperature NMR studies.
Reduction of 12 regenerates the borenium ion catalyst, which
then reenters the catalytic cycle. In the case where B(C6F5)3 is
employed to generate the borenium ion by FLP-type hydride
abstraction from HBPin, B(C6F5)3 acts only as an initiator and is
not itself involved in the catalytic cycle.
When 1c was employed stoichiometrically in the reduction of
imine 5e, only traces of product 6e were observed after 4 h at
room temperature (eq 3). This illustrates clearly that the
−
hydridoborate HB(C6F5)3 is considerably slower at reducing
imine 5e compared with the combination of 1d and HBPin under
the reaction conditions.
Competition experiments in which the rates of reduction with
HBPin and DBPin were compared gave a high primary kinetic
isotope effect of kH/kD = 6.7 0.1, clearly showing that cleavage
In conclusion, we have developed a novel metal-free catalytic
method for the reduction of imines using air-stable, non-
hazardous boranes. The catalysts for this process are discrete,
well-characterized borenium salts based on pinacolborane.
17386
dx.doi.org/10.1021/ja307374j | J. Am. Chem. Soc. 2012, 134, 17384−17387

Journal of the American Chemical Society
Communication
(8) Prokofjevs, A.; Boussonniere, A.; Li, L.; Bonin, H.; Lacote, E.;
Curran, D. P.; Vedejs, E. J. Am. Chem. Soc. 2012, 134, 12281.
(9) During review of this manuscript, the use of 9BBN·NHC adducts
as precursors for borenium-based catalysts for FLP-type hydrogenations
was reported. See: Farrell, J. M.; Hatnean, J. A.; Stephan, D. W. J. Am.
Chem. Soc. 2012, 134, 15728.
Scheme 2. Revised Mechanism of Borenium-Catalyzed Imine
Reduction
(10) In recent work, the Organ group has shown that the same is true
for HSnBu3-mediated reductions, namely, that the stannane itself, rather
−
than HB(C6F5)3 , is the true reducing agent in FLP-type hydro-
stannylations of alkynes. See: Oderinde, M. S.; Organ, M. G. Angew.
Chem., Int. Ed. 2012, 51, 9834.
(11) Stephan, D. W. Dalton Trans. 2012, 41, 9015 (special issue on
FLP chemistry).
(12) Piers, W. E.; Marwitz, A. J. V.; Mercier, L. G. Inorg. Chem. 2011,
50, 12252.
(13) Parks, D. J.; Blackwell, J. M.; Piers, W. E. J. Org. Chem. 2000, 65,
3090.
(14) Parks, D. J.; Piers, W. E. J. Am. Chem. Soc. 1996, 118, 9440.
(15) (a) Rubin, M.; Schwier, T.; Gevorgyan, V. J. Org. Chem. 2002, 67,
1936. (b) Gevorgyan, V.; Rubin, M.; Liu, J.-X.; Yamamoto, Y. J. Org.
Chem. 2001, 66, 1672.
(16) Chase, P. A.; Welch, G. C.; Jurca, T.; Stephan, D. W. Angew.
Chem., Int. Ed. 2007, 46, 8050.
(17) Welch, G. C.; Juan, R. R. S.; Masuda, J. D.; Stephan, D. W. Science
Mechanistic studies have revealed that this reaction constitutes
the first intermolecular process catalyzed by an isolated
borenium ion. This mode of reaction constitutes an attractive
alternative that avoids stoichiometric borohydride salts or
explosive hydrogen gas in the synthesis of bulky 2° amines.
Further work on the chemistry of these and related borenium
ions, including the development of an enantioselective process, is
underway in our laboratories.
2006, 314, 1124.
(18) Stephan, D. W.; Erker, G. Angew. Chem., Int. Ed. 2010, 49, 46.
(19) Lata, C. J.; Crudden, C. M. J. Am. Chem. Soc. 2010, 132, 131.
(20) Birkmann, B.; Voss, T.; Geier, S. J.; Ullrich, M.; Kehr, G.; Erker,
G.; Stephan, D. W. Organometallics 2010, 29, 5310.
(21) Non-FLP combinations have been shown to activate H2. See:
(a) Geier, S. J.; Stephan, D. W. J. Am. Chem. Soc. 2009, 131, 3476.
(b) Binding, S. C.; Zaher, H.; Chadwick, F. M.; O’Hare, D. Dalton Trans.
2012, 41, 9061. However, increasing the steric bulk in one of the
components (particularly the Lewis acid) resulted in remarkable,
increased functional group tolerance (e.g., see ref 22).
ASSOCIATED CONTENT
* Supporting Information
Detailed experimental procedures, NMR spectra, kinetic data,
and characterization data. This material is available free of charge
via the Internet at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
Corresponding Author
cruddenc@chem.queensu.ca
Notes
■
(22) (a) Ero
Tarkanyi, G.; Soos
Mehdi, H.; Papai, I.; Rokob, T. A.; Kiral
Angew. Chem., Int. Ed. 2010, 49, 6559.
̋
s, G.; Nagy, K.; Mehdi, H.; Pap
, T. Chem.Eur. J. 2012, 18, 574. (b) Ero
y, P.; Tarkanyi, G.; Soos
́
ai, I.; Nagy, P.; Kiral
s, G.;
, T.
́
y, P.;
́
́
́
̋
́
́
́
́
́
The authors declare no competing financial interest.
(23) Dureen, M. A.; Lough, A.; Gilbert, T. M.; Stephan, D. W. Chem.
Commun. 2008, 4303.
(24) Webb, J. D., Ph.D. Thesis, Queen’s University, Kingston, ON,
ACKNOWLEDGMENTS
■
Financial support from the National Sciences and Engineering
Research Council of Canada (NSERC) is gratefully acknowl-
edged in terms of operating, equipment, accelerator, and strategic
grants to C.M.C. The Canada Foundation for Innovation (CFI)
is thanked for support in terms of a Leader’s Opportunity Fund
(LOF) Grant to C.M.C. Queen’s University is thanked for
support in terms of QGA Awards to A.M.B. Dr. F. Sauriol is
thanked for assistance with NMR measurements. Prof. R. S.
Brown and Dr. A. A. Neverov are thanked for helpful
conversations regarding reaction kinetics.
2011.
(25) (a) De Vries, T. S.; Vedejs, E. Organometallics 2007, 26, 3079.
(b) Benjamin, E.; Carvalho, D. A.; Stafiej, S. F.; Takacs, E. A. Inorg.
Chem. 1970, 9, 1844. (c) Vedejs, E.; Nguyen, T.; Powell, D. R.;
Schrimpf, M. R. Chem. Commun. 1996, 2721.
(26) Replacing the noncoordinating perfluorophenylborate by
tetrafluoroborate under identical conditions resulted in the slow
formation of a four-coordinate F-bridged boronium salt, as indicated
by signals with chemical shifts close to 0 ppm in the 11B NMR spectrum.
The signals for the borenium cations in 1c and 1d exhibited the same
chemical shift in the 11B NMR spectrum, indicating absence of a
bridging μ-H interaction with the electrophilic boron center and
HB(C6F5)3− in 1c.
(27) The reaction generated only the borenium ions as drawn, but for 2
and 3, remaining starting material complicated the isolation of the pure
borenium salts, so we focused on 1d.
(28) Chase, P. A.; Jurca, T.; Stephan, D. W. Chem. Commun. 2008,
1701.
REFERENCES
■
(1) De Vries, T. S.; Prokofjevs, A.; Vedejs, E. Chem. Rev. 2012, 112,
4246.
(2) Piers, W. E.; Bourke, S. C.; Conroy, K. D. Angew. Chem., Int. Ed.
2005, 44, 5016.
(3) Koelle, P.; Noeth, H. Chem. Rev. 1985, 85, 399.
(4) Corey, E. J.; Bakshi, R. K.; Shibata, S. J. Am. Chem. Soc. 1987, 109,
5551.
(5) Grosso, A. D.; Helm, M. D.; Solomon, S. A.; Caras-Quintero, D.;
Ingleson, M. J. Chem. Commun. 2011, 47, 12459.
(6) Del Grosso, A.; Singleton, P. J.; Muryn, C. A.; Ingleson, M. J.
Angew. Chem., Int. Ed. 2011, 50, 2102.
(29) (a) HBPin was unable to reduce the imine in the absence of
B(C6F5)3·DABCO. Similarly, DABCO alone or HBpin and DABCO in
the absence of the borenium catalyst under otherwise identical
conditions showed no reactivity. (b) Query, I. P.; Squier, P. A.;
Larson, E. M.; Isley, N. A.; Clark, T. B. J. Org. Chem. 2011, 76, 6452.
(7) (a) Prokofjevs, A.; Vedejs, E. J. Am. Chem. Soc. 2011, 133, 20056.
(b) Prokofjevs, A.; Kampf, J. W.; Vedejs, E. Angew. Chem., Int. Ed. 2011,
50, 2098. (b) De Vries, T. S.; Prokofjevs, A.; Harvey, J. N.; Vedejs, E. J.
Am. Chem. Soc. 2009, 131, 14679.
17387
dx.doi.org/10.1021/ja307374j | J. Am. Chem. Soc. 2012, 134, 17384−17387