AuBr3-catalyzed thiooxime-to-carbonyl conversion: From chiral aliphatic nitro compounds to ketones without racemization

Article abstract of DOI:10.1021/ol9017722

A new variant of the NO₂-to-CO transformation (the Nef reaction) that occurs at room temperature under neutral conditions is uncovered. After the conversion of secondary nitroalkanes to phenylsulfenylketlmines, these thlooximes are hydrolyzed quantitatively In situ, In THF-H₂O at pH 7, by addition of AuBr₃ (but not with other MX_n!). Adducts arising from asymmetric nitro-Mlchael and nltro-aldol reactions afford 1,4-dlketones and a-alkoxy ketones, respectively, with full retention of the configuration of the stereocenters a to the CHNO₂/C=N-SPh/C=o groups.

Full text of DOI:10.1021/ol9017722

ORGANIC
LETTERS
2009
Vol. 11, No. 19
4414-4417
AuBr3-Catalyzed Thiooxime-to-Carbonyl
Conversion: From Chiral Aliphatic Nitro
Compounds to Ketones without
Racemization
Jordi Bure´s, Carles Isart, and Jaume Vilarrasa*
Departament de Qu´ımica Orga`nica, Facultat de Qu´ımica, UniVersitat de Barcelona,
Diagonal 647, 08028 Barcelona, Catalonia, Spain
jVilarrasa@ub.edu
Received July 31, 2009
ABSTRACT
A new variant of the NO2-to-CO transformation (the Nef reaction) that occurs at room temperature under neutral conditions is uncovered. After the
conversion of secondary nitroalkanes to phenylsulfenylketimines, these thiooximes are hydrolyzed quantitatively in situ, in THF-H2O at pH 7, by
addition of AuBr3 (but not with other MXn!). Adducts arising from asymmetric nitro-Michael and nitro-aldol reactions afford 1,4-diketones and r-alkoxy
ketones, respectively, with full retention of the configuration of the stereocenters r to the CHNO2/CdN-SPh/CdO groups.
Many reactions of gold compounds have been described in
the past recent years.1 Here we report an unexpected
application of Au3+ complexes, which we have incidentally
discovered during a screening of inorganic salts in the search
for optimum conditions for the hydrolysis of sulfenylketimines
to ketones.
The conversion of secondary nitro alkanes to ketones is a
well-known useful reaction, which connects nitrogen and
carbonyl chemistry.2 Very recently, we developed a smooth
protocol that takes place at room temperature (rt) without
strong bases or acids; trimethylphosphine and an activator
(ArSSAr, PySeSePy, or PhthN-SePh) were the only reagents
required (Scheme 1).3
In spite of the mild conditions, the procedure has the
limitation that during the reduction of enantiopure nitro
compounds, substituted at positions R to the CHNO2 groups,
racemic ketones were obtained,3a as shown in the second
equation of Scheme 1. Such a racemization occurs at the
ketimine stage, via the enamines.3a,4 To solve this handicap,
we stopped the reduction cascade of nitro compounds (1) at
sulfenylimines 2, by means of our protocol that uses PMe3
and N-(phenylsulfenyl)phthalimide (PhthN-SPh) in THF at
rt.5 As indicated in Scheme 2, the key point was then to
(1) Among dozens of reviews on the uses of gold in organic chemistry
(the new “gold rush”, mainly involving interactions of Au(I) with multiple
C-C bonds in the key steps), see the following general summaries: (a)
Corma, A.; Garc´ıa, H. Chem. Soc. ReV. 2008, 37, 2096. (b) Hashmi, A. S. K.;
Rudolph, M. Chem. Soc. ReV. 2008, 37, 1766. (c) Arcadi, A. Chem. ReV.
2008, 108, 3266. (d) Li, Z.; Brouwer, C.; He, C. Chem. ReV. 2008, 108,
3239. For a mechanistic overview, see: (e) Soriano, E.; Marco-Contelles,
J. Acc. Chem. Res. 2009, 42, 1026.
(2) For recent reviews of Nef-like reactions, see: (a) Wolfe, J. P. In
Name Reactions for Functional Group Transformations; Li, J. J., Corey,
E. J., Eds.; Wiley: Hoboken, 2007; p 645. (b) Ballini, R.; Palmieri, A.;
Righi, P. Tetrahedron 2007, 63, 12099. (c) Ballini, R.; Petrini, M.
Tetrahedron 2004, 60, 1017. Classical review on the use of TiCl3: (d) Mc
Murry, J. E. Acc. Chem. Res. 1974, 7, 281. Many methods are too harsh
for our purposes (to be applied to NO2-containing polyfunctional fragments,
in advanced steps of total syntheses). We have sometimes used an oxidative
variant (2KHSO5·KHSO4·K2SO4, Oxone): (e) Ceccherelli, P.; Curini, M.;
Marcotullio, M. C.; Epifano, F.; Rosati, O. Synth. Commun. 1998, 28, 3057.
However, the necessity of having an alkaline aqueous medium to ensure
the presence of nitronate ions and to make Oxone partially soluble, as well
as the sensitivitity of several characteristic groups to peroxides, prevented
us from using it in other cases.
(3) (a) Bure´s, J.; Vilarrasa, J. Tetrahedron Lett. 2008, 49, 441. This
method was based on (what we call) the BMZ reaction. (b) Barton, D. H. R.;
Motherwell, W. B.; Zard, S. Z. Tetrahedron Lett. 1984, 25, 3707.
(4) This would not be the case if the imine/enamine equilibrium would
occur preferably via the CH2R′ moiety (e.g., for R′ ) Ar or EWG), but
speaking in general the handicap was significant and restricted too much
the application of our procedure.
10.1021/ol9017722 CCC: $40.75
Published on Web 09/04/2009
 2009 American Chemical Society

FeBr3 (entries 34 and 35), and InBr3 (entry 43) were the most
active. These metallic ions, being harder Lewis acids than
their respective M+ and/or M2+ ions, lower the pH of the
medium (by partial hydrolysis). Part of their performance
may be due to the inherently low pH of their aqueous
solutions, but not exclusively. Moreover, it is remarkable
that only AuBr3-derived species did not lose their activity
when the reaction medium was partially or fully neutralized
(AuBr3 was even better than AuCl3 in this regard, compare
entry 24 to entry 20). As known, insoluble hydroxides or
hydrated oxides of Cu(II), Fe(III), and In(III) were formed
when the solutions of their MXn salts were neutralized. On
the other hand, AuBr3 did not give a precipitate when
aqueous NaOH was added to neutralize the solution.8 Thus,
we attribute the activity of Au(III) to the formation of soluble
[AuBrx(OH)y(2a)]3-x-y species where 2a replaces one or two
ligands of the inner sphere of the central atom. Coordination
of Au(III) with the S atom and/or N-S group of thiooximes
is plausible.9
Scheme 1
.
PMe3-Mediated Conversion of Secondary Nitro
Groups to Imines and Ketones
find a way to hydrolyze 2 to 3 in situ, without first cleaving
the N-S bond.
Scheme 2. How to Hydrolyze Carefully N-(Phenylsulfenyl)-
At pH 7, when the amount of AuBr3 was reduced to 0.3
(entries 25-27) and to 0.1 equiv (entry 28) the hydrolysis
percentages fell. Given the price of gold and AuBr3 the
requirement of 0.5 equiv of catalyst would be a drawback
for large-scale applications.
ketimines (S-Phenylthiooximes, 2) to Ketones
Since the hydrolysis coproduct (PhSNH2) is expected to
be more basic and nucleophilic than 2a, it may coordinate
the central atom of the complex more strongly, “poisoning
the catalyst”. To eliminate PhSNH2 (and PhSNHSPh and
NH3)10 we added isopentyl nitrite to the reaction mixture.
At pH 7, isopentyl nitrite alone did not react with 2a, but it
did react with ArSNH2 (checked independently).11 The
hydrolysis of 2a was complete at pH 7 with 0.3 equiv of
AuBr3 and 0.4 equiv of RONO (overnight at rt). To our
delight, with 0.8 equiv of RONO we could reduce the amount
of AuBr3 to 10 mol %.12
With optimum protocols for the hydrolysis of 2a in hand,
we subjected another simple nitro compound, racemic 1b,
and the stereopure or scalemic nitro derivatives (1c-k)
shown in Table 2 to the cheapest protocol. Both stepss
reduction of 1 to 2 and hydrolysis of 2swere carried out in
one pot. The commercially available THF solution of PMe3
Simple N-phenylsulfenylimines such as that of cyclohex-
anone undergo hydrolysis on moist silica, but those that are
sterically crowded require strongly acidic media.6 Since
esters, acetals, and other protecting groups that could contain
polyfunctional nitro derivatives would not survive under
these conditions, and since chiral R-substituted thiooximes
and ketones may racemize at very low and high pH values,
we imposed ourselves the limitation of operating without
heating, in the absence of Bro¨nsted acids7 and as close as
possible to pH 7.
A screening of the potential catalysts (Lewis acids, with
thiophilic and relatively nontoxic transition-metal cations,
which do not decompose in water) was undertaken with a
model compound (2a),5 in THF-H2O, as shown in Table 1.
Polymeric salts (MX) and, in general, inorganic com-
pounds that are scarcely soluble in THF-H2O were inactive
(as it was CuO, not included in Table 1), whereas CuBr2
(entry 6), AuCl3 (entries 17 and 18), AuBr3 (entries 21-24),
(8) (a) Baes, C. F.; Mesmer, R. E. The Hydrolysis of Cations; Wiley:
New York, 1976; pp 279-285. (b) Usher, A.; McPhail, D. C.; Brugger, J.
Geochim. Cosmochim. Acta 2009, 73, 3359 (a spectrophotometric study of
aqueous Au(III) halide-hydroxide complexes). Also see ref 1a.
(9) On the other hand, oximes PhC(dN-OBn)Me and PhC(dN-OPh)Me
are not hydrolyzed under the conditions of entry 24 of Table 1.
(10) Simple sulfenamides (RSNH2) may disproportionate to RSNHSR
and NH3: (a) Bao, M.; Shimizu, M.; Shimada, S.; Tanaka, M. Tetrahedron
2003, 59, 303. (b) Davis, F. A.; Friedman, A. J.; Kluger, E. W.; Skibo,
E. B.; Fretz, E. R.; Milicia, A. P.; LeMasters, W. C.; Bentley, M. D.;
Lacadie, J. A.; Douglass, I. B. J. Org. Chem. 1977, 42, 967. For entries to
the chemistry of sulfenamides, see: (c) Koval, I. V. Russ. J. Org. Chem.
2005, 41, 386. (d) Davis, F. A.; Mancinelli, P. A. J. Org. Chem. 1978, 43,
1797. Recent review of N-S bond-containing compounds: (e) Davis, F. A.
J. Org. Chem. 2006, 71, 8993.
(5) Bure´s, J. Isart, C. Vilarrasa, J. Org. Lett. 2007, 9, 4635. In Table 1,
entry 4, the oxime should have been depicted as the Z isomer.
(6) (a) Most of our sulfenylimines underwent hydrolysis on warming
with 1 M HCl or with Amberlite IR-120 (pH 2.2), but racemization or
epimerization was then produced, as well as the cleavage of various
protecting groups. (b) To our knowledge, only the cleavage of tritylsulfe-
nylketimines, by an excess of AgNO3 (and a different mechanism) has been
described: Branchaud, B. P. J. Org. Chem. 1983, 48, 3531.
(11) (a) Any alkyl nitrite capable of nitrosating and hence decomposing
ArSNH2 and NH3 should work. No PhSH was detected; according to TLC
1
and H NMR, PhSSPh was formed predominantly. (b) C5H11ONO alone
reacted with 1a at low pH (1.5 equiv was required to fully decompose 1a,
overnight at rt, pH 4.2), but not at all at pH 7.
(12) On the other hand, the addition of NaNO2 (80 mol %) instead of
C5H11ONO to the mixture of 2a with AuBr3 (10 mol %) at pH 7 did not
improve the outcome of entry 28 of Table 1.
(7) By adding 1 M HCl or 1 M HBr to a THF solution of 2a, hydrolysis
to ketone 3a was complete after stirring overnight. However, as mentioned,
these conditions did not suit us, as we plan to apply the reaction on acid-
sensitive polyfunctional substrates.
Org. Lett., Vol. 11, No. 19, 2009
4415

Table 1. Potential Catalysts for the Hydrolysis of 2aa
Table 2. From Chiral Nitro Derivatives to Ketonesa
entry additive (MXn) equiv
pH conditions
% of 3a
1
2
3
4
5
6
7
8
Cu2Cl2
Cu2Cl2
CuI
1.0
1.0
1.0
0.5
1.0
1.0
0.5
0.1
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
0.5
0.5
0.5
1.0
0.5
0.5
0.5
0.3
0.3
0.3
0.1
1.0
1.0
1.0
1.0
1.0
1.0
0.5
0.1
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
0-20b
0-20b
0
11
43
100
71
23
5
0
16
0
0
0
buffered at pH 4.0
c
(CuOTf)2
CuCl2·2H2O
CuBr2
CuBr2
CuBr2
CuBr2
pH measured ) 3.0
pH measured ) 3.5
9
buffered at pH 7.0
basified up to pH 10.0
pH measured ) 5.5
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
CuBr2
Cu(OAc)2
Cu(acac)2
Cu(OTf)2
AgNO3
AgF
0
0
AuCld
AuCl3
AuCl3
AuCl3
100
100
63
10
100
100
100
100e
67
56
6
26
3
11
23
0
pH measured ) 0.9
adjusted at pH 6.0
adjusted at pH 7.0
pH measured ) 1.2
AuCl3
AuBr3
AuBr3
AuBr3
AuBr3
AuBr3
AuBr3
AuBr3
AuBr3
ZnCl2
ZnBr2
FeCl2
CoCl2·6H2O
NiCl2·6H2O
FeBr3
FeBr3
FeBr3
FeBr3
Sc(OTf)3
LaCl3·7H2O
CeCl3·7H2O
CANf
Yb(OTf)3
InBr3
InBr3
none
silicagel
dil. HBr
dil. HBr
adjusted at pH 4.0
adjusted at pH 7.0
pH 4.6
pH 6.7
pH 8.9
0
pH measured ) 0.2
100
100
24
8
62
0
0
0
6
100
0
adjusted at pH 4.0
a Unless otherwise indicated, both steps were carried out at rt. Workup:
after dilution of the final THF-H2O solution with more water and extraction
with CH2Cl2 several times, only 3, PhSSPh, and phthalimide derivatives
were extracted (as Me3PO is very soluble in water and remains in the
aqueous layer, together with the brownish gold complexes); PhSSPh was
easily removed by filtration through silica (elution with hexane). Ee values
were determined as explained in the Supporting Information. b t1, t2, and
experiments carried out at 0 °C (instead of at rt) are indicated. c Six-hundred
molar percent of PMe3 and 300 mol % of PhthN-SPh were added. d Nitro
compounds prepared by organocatalytic addition of nitroethane to the
corresponding enones, with trans-2,5-dimethylpiperazine as the base.13 e 1k
as a 1:1 syn/anti mixture.
adjusted at pH 7.0
adjusted at pH 7.0
pH measured ) 0.0
adjusted at pH 7.0
0
0
0
0
pH 1.5
pH 3.0
a To 2a (0.3 mmol) in 1 mL of THF (unless otherwise indicated) a
solution or suspension of the additive (possible catalyst) in 1 mL of H2O
(pH 6.9-7.0) was added, and the mixture was stirred vigorously overnight;
in some cases (indicated) the pH values were adjusted with 1 M NaOH
(buffering effects were noted) or by addition of standard phosphate buffers.
A Crison pHmeter was used. b We later confirmed that only commercial
samples contaminated with CuCl2 showed catalytic activity (no hydrolysis
occurred with pure Cu2Cl2). c Commercially available Cu2(OTf)2·C6H6;
identical result with commercial Cu2(OTf)2·C7H8. d AuCl is insoluble in
H2O. e Identical result in 9:1 THF-H2O; the reaction was slower (85% of
conversion) in 9:1 CH3CN-H2O. f Ce(NH4)2(NO3)6.
was used as the medium for the first step. After elimination
of the slight excess of PMe3 under vacuum, the aqueous
neutral (buffered) solution of Au(III) and the alkyl nitrite
were added and stirring was maintained at rt until the
complete disappearance of 2.
It is remarkable that the hydrolyses (second step) were
always practically quantitative; in fact, only the starting
4416
Org. Lett., Vol. 11, No. 19, 2009

materials and expected hydrolysis products were detected
by TLC and NMR.
Thus, the sulfenylimino groups can be hydrolyzed at
neutral buffered pH and at rt only (to date) with 50 mol %
of AuBr3 or with 2-10 mol % of AuBr3 and stoichiometric
or substoichiometric amounts of RONO.
On the other hand, the first step (when sulfenylketimines
were formed, isolated, and purified by chromatography) took
place in 80-90% yields. As an exception, in the 1j-2j-3j
sequence (entry 9) the yield was moderate, but it was due to
the unavoidable formation of Beckmann fragmentation
byproducts during the first step.5 In fact, these secondary
reactions are known to be inherent to all reactions involving
oximes, thiooximes, etc., if stable cationic intermediates can
be formed.
One advantage of our procedure is that it can be used to
prepare chiral 1,4-dicarbonyl compounds from the organo-
catalytic conjugate addition of nitroalkanes to enones (entries
6-8). In the case of 3g (entries 6 and 7), 2,5-dimethyl-3-
phenylpyrrole was not formed at all.14 Compounds arising
from asymmetric nitro-aldol (Henry) reactions, such as those
of entries 9 and 10, are also amenable to our protocol.
As final tests, we reduced the amount of AuBr3 to 5 mol %
and 2 mol %. We subjected 2a to hydrolysis at pH 7 and rt as
in Table 2, with 100 mol % of C5H11ONO. With 5 mol % of
AuBr3, stirring for 30 h was sufficient for the complete
disappearance of this sulfenylketimine and its full conversion
to ketone 3a. With 2 mol % of AuBr3, 5 days were required;15
the hydrolysis was slower, as expected, but still feasible.
In summary, a very smooth two-step one-pot procedure
for the conversion of secondary nitro groups to ketones has
been disclosed. As pursued by the senior author for a long-
time, it works at rt (or at 0 °C, if required) and under neutral
conditions. The first step, the conversion of secondary nitro
groups to sulfenylketimines,3a has been applied here suc-
cessfully, for the first time, to various chiral compounds
arising from organocatalytic reactions or stereoselective
variants of venerable reactions. The second step-the hy-
drolysis of these sulfenylimines-involves the use of AuBr3,
which among the large number of MXn salts examined is
the only one that catalyzes such hydrolyses at pH 7.
Therefore, for the first time to the best of our knowledge,
we have taken advantage of a practical feature of AuBr3:
the solubility and stability of [AuBrx(OH)y]3-x-y complexes,
which permits thiooximes to coordinate with them and
undergo the desired hydrolysis under the mildest possible
conditions (compatible with R-stereocenters and a plethora
of protecting groups). As the first step was originally inspired
in the BMZ reaction,3b the overall protocol might be called
the Vilarrasa-BMZ Nef-type procedure or something simi-
lar.
(13) (a) Hanessian, S.; Shao, Z.; Warrier, J. S. Org. Lett. 2006, 8, 4787,
and references therein. (b) Mitchell, C. E. T.; Brenner, S. E.; Garcia-Fortanet,
J.; Ley, S. V. Org. Biomol. Chem. 2006, 4, 2039. Instead of (S)-proline or
a bicyclo[3.1.0]-derivative of Pro and instead of (S)-5-(pyrrolidin-2-
yl)tetrazole, we used Seebach’s bicyclic oxazolidinone (the aminal of Pro
Acknowledgment. The Spanish Government (MEC, MI-
CINN) is acknowledged for grant CTQ2006-15393 (and
earlier grants, during which the objective was conceived) as
well as for a studentship to J.B. (Feb 2004-Jan 2008). The
Universitat of Barcelona is acknowledged for a studentship
to C.I. (since Oct 2006).
t
and BuCHO). See: (c) Isart, C.; Bure´s, J.; Vilarrasa, J. Tetrahedron Lett.
2008, 49, 5414.
(14) Pyrrole formation is unavoidable in most reductions of γ-nitro
ketones, since the intermediate oximes or imines react in situ with the CO
groups (formation of five-membered rings). Cf. refs 2d and 3b. Zard et al.
took advantage of this reaction to prepare various interesting pyrroles: (a)
Quiclet-Sire, B.; Thevenot, I.; Zard, S. Z. Tetrahedron Lett. 1995, 36, 9469.
(b) Barton, D. H. R.; Motherwell, W. B.; Simon, E. S.; Zard, S. Z. J. Chem.
Soc., Perkin Trans. 1 1986, 2243. (c) Barton, D. H. R.; Zard, S. Z. Chem.
Commun. 1985, 1098. Under the conditions of Scheme 1, γ-nitro acyclic
ketones give pyrrole derivatives almost quantitatively.
Supporting Information Available: Additional experi-
mental details. This material is available free of charge via
the Internet at http://pubs.acs.org.
(15) Experiments carried out in parallel, at rt and pH 7 as always, with
only C5H11ONO or NaNO2, without AuBr3, did not affect 2a.
OL9017722
Org. Lett., Vol. 11, No. 19, 2009
4417