Control of transient aluminum-aminals for masking and unmasking reactive carbonyl groups

Article abstract of DOI:10.1021/ol401265a

A new reagent, the dimethylaluminum N,O-dimethylhydroxylamine complex, is effective at masking reactive carbonyl groups in situ from nucleophilic addition. This reagent allows chemoselective addition of reducing reagents, Grignard reagents, organolithiums, Wittig reagents, and enolates into substrates with multiple carbonyl groups. Moreover, the trapped carbonyl group, a stable aminal, can be unmasked in situ for additional synthetic manipulations.

Full text of DOI:10.1021/ol401265a

ORGANIC
LETTERS
XXXX
Vol. XX, No. XX
000–000
Control of Transient AluminumÀAminals
for Masking and Unmasking Reactive
Carbonyl Groups
Francis J. Barrios,† Brannon C. Springer,† and David A. Colby*,†,‡
Department of Chemistry and Department of Medicinal Chemistry and Molecular
Pharmacology, Purdue University, West Lafayette, Indiana 47907, United States
dcolby@purdue.edu
Received May 6, 2013
ABSTRACT
A newreagent, thedimethylaluminumN,O-dimethylhydroxylaminecomplex, is effectiveatmasking reactivecarbonylgroupsinsitu fromnucleophilic
addition. This reagent allows chemoselective addition of reducing reagents, Grignard reagents, organolithiums, Wittig reagents, and enolates into
substrates with multiple carbonyl groups. Moreover, the trapped carbonyl group, a stable aminal, can be unmasked in situ for additional synthetic
manipulations.
Carbonyl groups are one of the most important func-
tional groups in organic chemistry and their reactivity
toward nucleophiles is well-known. One of the frequent
problems associated with compounds with multiple carbo-
nyl groups is the selective modification of a less reactive
carbonyl group in the presence of a more reactive group.
Even though using protecting groups or reduction/oxidation
sequences are reasonable strategies to control chemoselec-
tivity, both of the plans accumulate additional synthetic
steps.1,2 An alternative approach to control chemoselectivity
in the presence of multiple reactive carbonyl groups is to
mask the more reactive carbonyl group transiently.3À9
Although some in situ masking strategies for carbonyl
groups have been reported, most are limited in scope and
none have exploited the transient nature of the trapped
intermediate to unmask the captured carbonyl group in
situ for subsequent manipulation. Herein, we describe a
discrete reagent to mask carbonyl groups as stable aminals
from nucleophiles and demonstrate the in situ unveiling of
the trapped carbonyl group for immediate use.
In his pioneering work, Luche reported the first method
toreducea ketone selectively inthe presenceofanaldehyde
using CeCl3 and NaBH4 in aqueous ethanol.3 Although
the aldehyde is masked in situ from reduction, a disadvan-
tage is the requirement of aqueous solvent, which limits the
use of common nucleophiles. Reetz4 and Yamamoto5 ap-
plied titanium-dialkylamides and aluminum-dialkylamides
respectively, to transform an aldehyde into an unreactive
intermediate in situ. The drawback of these methods is the
requirement of low temperatures to prevent reversion to
the aldehyde. Yamamoto further refined his method to
create bulky aluminum-based Lewis acids (e.g., MAD)
† Department of Chemistry.
‡ Department of Medicinal Chemistry and Molecular Pharmacology.
(1) Wender, P. A.; Miller, B. L. Nature 2009, 460, 197–201.
(2) Young, I. S.; Baran, P. S. Nat. Chem 2009, 1, 193–205.
(3) Luche, J. L; Gemal, A. L. J. Am. Chem. Soc. 1979, 101, 5848–5849.
(4) Reetz, M. T.; Wenderoth, B.; Peter, R. J. Chem. Soc., Chem.
Commun. 1983, 406–408.
(5) Maruoka, K.; Araki, Y.; Yamamoto, H. Tetrahedron Lett. 1988,
25, 3101–3104.
(6) Maruoka, K.; Saito, S.; Concepcion, A. B.; Yamamoto, H. J. Am.
Chem. Soc. 1993, 115, 1183–1184.
ꢀ
(7) Bastug, G.; Dierick, S.; Lebreux, F.; Marko, I. E. Org. Lett. 2012,
14, 1306–1309.
(8) Fujioka, H.; Yahata, K.; Kubo, O.; Sawama, Y.; Hamada, T.;
Maegawa, T. Angew. Chem., Int. Ed. 2011, 50, 12232–12235.
(9) Barrios, F. J.; Zhang, X.; Colby, D. A. Org. Lett. 2010, 12, 5588–5591.
r
10.1021/ol401265a
XXXX American Chemical Society

that will selectively block carbonyl groups.6 More recently,
diethylaluminum benzenethiolate has been reported to en-
able the chemoselective reduction of carbonyl groups with
DIBALH in the presence of aldehydes.7 A more versatile
method has been described using phosphonium salts to mask
reactive carbonyl groups from reducing reagents, such as
Table 1. Chemoselective Additions to Carbonyl Groups
DIBALH and BH3 THF, and Grignard reagents.8 Our
3
laboratory has reported a strategy to mask reactive carbonyl
groups as an aminal using complexes of dialkylaluminum
and N,O-dimethylhydroxylamine.9 This approach was de-
signed from the stable intermediate formed following addi-
tion of a nucleophile to a Weinreb amide.10 The limitation of
this method was the necessity of 1 equiv of base to stabilize
the aminal intermediate. We speculated that if the require-
ment for additional base could be eliminated, a discrete
reagent for masking would be discovered and enable the
development of a subsequent unmasking strategy. These two
advances would be a substantial innovation over all other
reported methods and enable the broadest compatibility
with nucleophiles. Accordingly, we hypothesized that a
combination of N,O-dimethylhydroxylamine HCl with
3
n-BuLi and Me3Al would generate an aluminumÀamide
reagent to accomplish these objectives (Scheme 1).
Scheme 1. Reagent for Masking/Unmasking Carbonyl Groups
Cossy et al. have exploited the stable aminal formed after
nucleophilic addition to a Weinreb amide to prevent the
reduction of a carbonyl group under Birch conditions,11 and
Evans protected a ketone from a metalated hydrazone in a
similar fashion.12 Comins,13 Hoffmann,14 and Roschangar15
have used lithium amides to access stabilized aminals from
carbonyl groups, but the high basicity of these reagents
can lead to unavoidable deprotonation at other sites if
acidic protons are present.16 Our prior application of
aluminumÀamide complexes avoids the high basicity of
lithium amides.9 Accordingly, we investigated several con-
ditions for the preparation of the new aluminumÀamide
reagent, and the most efficient synthesis was accomplished
by first reacting HN(OMe)Me HCl with n-BuLi or
3
i-PrMgCl to consume the acidic proton of the hydrochloride
(10) Nahm, S.; Weinreb, S. M. Tetrahedron Lett. 1981, 22, 3815–3818.
(11) Taillier, C.; Bellosta, V.; Meyer, C.; Cossy, J. Org. Lett. 2004, 6,
2145–2147.
(12) Evans, D. A.; Bender, S. L.; Morris, J. J. Am. Chem. Soc. 1988,
110, 2506–2526.
(13) Comins, D. L. Synlett 1992, 615–625.
a Isolated yields. b i-PrMgCl was used instead of n-BuLi.
(14) Hoffmann,R.W.;Munster,I.Tetrahedron Lett. 1995,36, 1431–1434.
(15) Roschangar, F.; Brown, J. C.; Cooley, B. E.; Sharp, M. J.;
Matsuoka, R. T. Tetrahedron 2002, 58, 1657–1666.
(16) Kruger, J.; Hoffmann, R. W. J. Am. Chem. Soc. 1997, 119, 7499–
7504.
and then adding Me3Al to form the complex. Either an
organolithium or a Grignard reagent can serve as the
B
Org. Lett., Vol. XX, No. XX, XXXX

requisite base. Next, we investigated the utility of this
reagent for in situ masking with substrates 1À6, each
bearing two different carbonyl groups (Table 1). The
broad scope of this approach was striking, because high
levels of chemoselectivity were easily achieved during
nucleophilic addition with reducing reagents, Grignard
reagents, organolithium reagents, Wittig reagents, and
enolate anions. Following treatment with the complex on
4-acetylbenzaldehyde (1), MeLi addition occurred exclu-
sively at the ketone togive the tertiary alcohol 7 in excellent
86% yield. The Grignard reagents, EtMgBr and PhMgBr,
also provided high yields of the tertiary alcohols 8 and 9,
respectively. Indeed, the role of Grignard reagents in
synthesis continues to expand with recent advances in the
preparation of organomagnesium reagents.17,18 Wittig
olefination using Ph3PCH3Br and n-BuLi following in situ
trapping with the aluminumÀamide reagent gave the
alkene 10 in 80% yield. To our knowledge, prior in situ
masking strategies have not demonstrated compatibility
with Wittig olefinations.3À9 Also, a Claisen condensation
was performed with the lithium-derived enolate of EtOAc
after in situ masking of 1, and the product 11 was obtained
in 70% yield. This approach using dimethylaluminumÀ
dimethylhydroxylamine was then applied to the nonaromatic
substrate, 4-oxocyclohexanecarbaldehyde (2), and selective
Wittig olefination at the ketone provided the alkene 12 in
86% yield. Two additional selective Grignard additions using
MeMgBr and allyl MgCl were performed with 4-acetyl-
methylbenzoate (3), and double addition to the ester gave
alcohols 13 and 14, respectively. Ethyl 4-oxocyclohexanecar-
boxylate (4) participates in a similar fashion to provide the
double-addition product 15 in 79% yield. Also, methyl
4-formylbenzoate (5) was subjected to in situ trapping with
the aluminumÀamide followed by selective addition of the
organolithiums, EtLi and n-BuLi, at the ester to give 16 and
17, respectively. Grignard addition to the masked derivative
of 5 yielded 83% of the double-addition product 18. Com-
plete reduction of the ester in the presence of the aldehyde was
achieved using 5 and providing the primary alcohol 19.
Lastly, the reactive ethyl benzoylformate 6 was subjected to
in situ masking followed by Grignard addition, and despite
the proximity of the trapped R-ketoester, double addition
gave the tertiary alcohol 20 in 84% yield. Overall, this strategy
can efficiently use aromatic and alkyl substrates, regardless of
the presence of acidic R-protons, and also it is compatible
with many types of nucleophiles, including Wittig reagents
and enolate anions.
compound has been used as a tool to study chemical19 and
enzyme kinetics.20 Using the in situ masking followed by
reduction with DIBALH yielded the same target 21 in 88%
yield in a single step! This comparison clearly illustrates the
power of this process to eliminate protection/deprotection
sequences and enhance isolated yields.
Next, we turned our attention to developing a simple proto-
col to unmask the trapped carbonyl groups for immediate
synthetic manipulation. First, 19F NMR data were ac-
quired for 4-fluorobenzaldehyde (22), following treatment
with the dimethylaluminum N,O-dimethylhydroxylamine
complex (Figure 1A). A single signal was observed at À117
ppm, and this peak was distinct from the starting material
22 at À105 ppm (data not shown). After an extensive
screen of reagents and conditions, it was discovered that
the trapped intermediate could be quantitatively un-
masked to the precursor carbonyl group in the 19F NMR
spectrum using Dowex and sonication (Figure 1B). A new
peak at À105 ppm appeared, which indicates complete
regeneration of 4-fluorobenzaldehyde 22. The addition of
acid to collapse the intermediate aminal was anticipated,
because an acidic workup is required to hydrolyze this type
of aminal, which is usually formed following nucleophilic
addition to a Weinreb amide.
Scheme 2. Comparison of in Situ Masking versus Stepwise
A side-by-side comparison to demonstrate the in situ
masking approach with dimethylaluminumÀdimethyl-
hydroxylamine against a traditional protection/deprotection
sequence was executed (Scheme 2). A three-step synthesis
of 4-(hydroxymethyl)cyclohexanone 21 from 4-oxocyclo-
hexanecarboxylate (4) has been reported by acetal formation,
reduction of the ester, and acetal hydrolysis.19 The target 21
was isolated in 36% yield across these three steps, and this
Figure 1. Comparison of 19F NMR data of 4-fluorobenzalde-
hyde at 276 MHz. (A) 4-Fluorobenzaldehyde after treatment
with the dimethylaluminum N,O-dimethylhydroxylamine com-
plex. (B) After unmasking the mixture in 1A with Dowex and
sonication. C6H5CF3 was used as an internal standard.
(17) Krasovskiy, A.; Krasovskaya, V.; Knochel, P. Angew. Chem.,
Int. Ed. 2006, 45, 2958–2961.
(18) Fleury, L. M.; Ashfeld, B. L. Tetrahedron Lett. 2010, 51, 2427–2430.
(19) Kayser, M. M.; Clouthier, C. M. J. Org. Chem. 2006, 71, 8424–8430.
Org. Lett., Vol. XX, No. XX, XXXX
C

The compatibility of the unmasking strategy with nu-
cleophilic additions to the newly unveiled carbonyl group
was explored using substrates 1 and 5 (Table 2). A major
initial finding of these studies was that molecular sieves
must be included to remove small amounts of water that
are released from the Dowex resin upon sonication. The
overall process was to trap the more reactive carbonyl
group, add the first nucleophile, unmask the trapped
carbonyl group, and then add the second nucleophile.
The products 23À27 were obtained in good 57À76%
isolated yields, and these yields suggest that roughly 90%
conversion occurs during each individual step. The com-
patible nucleophiles listedin Table 2 span a reducing agent,
organolithiums, Grignard reagents, and a Wittig reagent.
These data provide substantial support for the flexibility of
the masking/unmasking protocol.
Scheme 3. Comparison of Masking/Unmasking versus Stepwise
would be to add the more reactive carbonyl group first and
the less reactive one second. Although this strategy is
logical for simple starting materials that bear two distinct
types of carbonyl groups, the real power of this method
becomes apparent when more complex substrates are
examined. Indeed, discriminating between two types of
ketones is a substantial synthetic challenge.21,22 For exam-
ple, aldol reaction between 4-acetylbenzaldehyde and acet-
ophenone gives the dione 28, and this product bears two
ketones that are nearly identical (Scheme 3). The chemo-
selective manipulation of one ketone on this substrate is
challenging, as Wittig olefination of 28 with Ph3PdCH2
(from MePPh3Br and n-BuLi) provided predominately
products from the β-elimination of the secondary
alcohol.23 Clearly, a protecting group strategy is required
for the secondary alcohol and to distinguish between the
ketones. The masking/unmasking approach offers a stark
contrast, because the reactivity of the two ketones can be
controlled and the desired product 29 was isolated in 60%
yield as single step!
Table 2. In situ Masking and Unmasking of Carbonyl Groups
In summary, we have discovered a novel strategy to mask
a reactive carbonyl group in situ using an aluminumÀamide
complex and regenerate the reactive carbonyl group after
manipulation of a less reactive carbonyl group. We demon-
strated the broad scope of substrates and nucleophiles that
are compatible in the process, and we have discovered that
this method is compatible with enolate anions and Wittig
reagents, which have not been used in any prior reports. We
have shown that this approach can provide superior yields
for chemoselective manipulations compared to protection/
deprotection sequences, and it can be easily incorporated
into the assembly of a complex structure that would require
multiple synthetic steps to build.
a Isolated yields. b i-PrMgCl was used instead of n-BuLi.
A cursory overview of the entries presented in Table 2
may lead to an obvious question about the synthetic utility
of the unmasking process, because the alternative method
Acknowledgment. These studies were funded by Purdue
University and the Midwest Crossroads Alliance for
Graduate Education and the Professoriate.
(20) Taschner, M. J.; Black, D. J.; Chen, Q.-Z. Tetrahedron: Asym-
metry 1993, 4, 1387–1390.
(21) Steward, K. M.; Corbett, M. T.; Goodman, C. G.; Johnson, J. S.
Supporting Information Available. Full experimental
details and spectroscopic data. This material is available
free of charge via the Internet at http://pubs.acs.org.
J. Am. Chem. Soc. 2012, 134, 20197–20206.
(22) Bhar, S.; Guha, S. Tetrahedron Lett. 2004, 45, 3775–3777.
(23) Using 1 equiv of Ph3PdCH2 provided mixtures of olefins, all
without the secondary alcohol, and using 2 equiv did not provide any
improvement.
The authors declare no competing financial interest.
D
Org. Lett., Vol. XX, No. XX, XXXX