Kinetic resolution of chiral cyclohex-2-enones by rhodium(I)/binap- catalyzed 1,2- and 1,4-additions

Article abstract of DOI:10.1021/ol300387f

The feasibility of kinetic resolutions of racemic monosubstituted cyclohex-2-enones by Rh/binap-catalyzed reactions was investigated. 1,2-Addition of AlMe₃ to the 5-substituted derivatives furnished allylic alcohols in the matched case, while t

Full text of DOI:10.1021/ol300387f

ORGANIC
LETTERS
2012
Vol. 14, No. 8
1978–1981
Kinetic Resolution of Chiral Cyclohex-2-
enones by Rhodium(I)/binap-Catalyzed
1,2- and 1,4-Additions
Andreas Kolb, Sebastian Hirner, Klaus Harms, and Paultheo von Zezschwitz*
€
Fachbereich Chemie, Philipps-Universitat Marburg, 35032 Marburg, Germany
zezschwitz@chemie.uni-marburg.de
Received February 16, 2012
ABSTRACT
The feasibility of kinetic resolutions of racemic monosubstituted cyclohex-2-enones by Rh/binap-catalyzed reactions was investigated. 1,2-
Addition of AlMe3 to the 5-substituted derivatives furnished allylic alcohols in the matched case, while the less reactive enantiomers were either
left over or transformed into trans-3,5-disubstituted cyclohexanones in parallel or sequential 1,4-additions. Altogether, these represent
regiodivergent reactions on racemic mixtures. In contrast, 1,4-addition of aryl groups led to inferior results since either catalyst or substrate
control dominated.
Enantioselective additions of carbon nucleophiles to
cyclic enones, particularly cyclohex-2-enones, are highly
important transformations in organic synthesis. For 1,4-
additions, Cu-catalyzed processes are well established,1
yet the Rh-catalyzed 1,4-addition of aryl groups, the
HayashiÀMiyaura reaction, has also been developed
into a powerful tool.2 While boron organyls are the principal
reagents in this transformation, other metal organyls, e.g.
zinc reagents, can be used as well.2c,e Recently, we reported
on the first Rh/binap-catalyzed addition of aluminum
organyls,3 which surprisingly occurs in a highly selective
1,2-manner and thus yields valuable enantiopure tertiary
allylic alcohols.4
In the case of 1,4-addition, most studies were performed
on cyclohex-2-enone itself or on 3-substituted derivatives.5
In contrast, little is known about reactions with chiral
cyclohex-2-enones, even though such compounds are
the rule rather than the exception when it comes to the
preparation of synthetic targets. Additionally, the correla-
tion of substrate control and catalyst control is interesting
for mechanistic and synthetic reasons; it can open the door
to kinetic resolutions of racemic starting materials, if both
modes of control are comparably strong.6 To the best of
our knowledge, 4-substituted chiral cyclohex-2-enones
were never investigated under Rh-catalysis but show sig-
nificant substrate control in Cu-catalyzed 1,4-additions,7
€
(1) For reviews, see: (a) Alexakis, A.; Backvall, J. E.; Krause, N.;
ꢀ
ꢁ
Pamies, O.; Dieguez, M. Chem. Rev. 2008, 108, 2796–2823. (b) Jerphagnon,
T.;Pizzuti, M. G.;Minnaard, A. J.;Feringa, B. L.Chem. Soc. Rev. 2009, 38,
1039–1075. (c) Ji, J.-X.; Chan, A. S. C. Catalytic Asymmetric Conjugate
Addition. In Catalytic Asymmetric Synthesis, 3rd ed.; Ojima, I., Ed.; John
Wiley & Sons, Inc.: Hoboken, NJ, 2010; pp 439À495.
(4) Siewert, J.; Sandmann, R.; von Zezschwitz, P. Angew. Chem., Int.
Ed. 2007, 46, 7122–7124.
(5) For a review about 1,4-additions to β,β-disubstituted enones, see:
(a) Hawner, C.; Alexakis, A. Chem. Commun. 2010, 46, 7295–7306. For
a seminal study of organocatalytic 1,4-additions of thiols to substituted
cyclic enones, see: (b) Hiemstra, H.; Wynberg, H. J. Am. Chem. Soc.
1981, 103, 417–430.
(6) Vedejs, E.; Jure, M. Angew. Chem., Int. Ed. 2005, 44, 3974–4001.
(7) (a) Corey, E. J.; Boaz, N. W. Tetrahedron Lett. 1985, 26, 6015–
6018. (b) Lopez-Pelegrin, J. A.; Janda, K. D. Chem.;Eur. J. 2000, 6,
1917–1922. (c) William, A. D.; Kobayashi, Y. Org. Lett. 2001, 3, 2017–
2020. (d) Naasz, R.; Arnold, L. A.; Minnaard, A. J.; Feringa, B. L.
Angew. Chem., Int. Ed. 2001, 40, 927–930.
(2) (a) Fagnou, K.; Lautens, M. Chem. Rev. 2003, 103, 169–196. (b)
Hayashi, T.; Yamasaki, K. Chem. Rev. 2003, 103, 2829–2844. (c)
Shintani, R.; Tokunaga, N.; Doi, H.; Hayashi, T. J. Am. Chem. Soc.
2004, 126, 6240–6241. (d) Yoshida, K.; Hayashi, T. Rhodium(I)-Cata-
lyzed Asymmetric Addition of Organometallic Reagents to Electron-
Deficient Olefins. In Modern Rhodium-Catalyzed Organic Reactions;
Evans, P. A., Ed.; WILEY-VCH Verlag GmbH & Co. KGaA: Weinheim,
2005; pp 55À77. (e) Hargrave, J. D.; Allen, J. C.; Frost, C. G. Chem.;
Asian J. 2010, 5, 386–396. (f) Edwards, H. J.; Hargrave, J. D.; Penrose,
S. D.; Frost, C. G. Chem. Soc. Rev. 2010, 39, 2093–2105.
(3) von Zezschwitz, P. Synthesis 2008, 1809–1831.
r
10.1021/ol300387f
Published on Web 04/11/2012
2012 American Chemical Society

thus allowing for a moderately effective kinetic resolu-
tion.7d Rh-catalyzed 1,4-additions to racemic 5-TMS- or
5-phenylcyclohex-2-enone proceeded with high enantio-
selectivity but low diastereoselectivity, thus revealing the
predominance of catalyst control.8 Substrate control,
however, can prevail under Cu-catalysis resulting in only
one diastereomer, but with low ee.8c With a different Cu-
catalyst, however, Feringa et al. achieved very impressive
kinetic resolutions of various 5-substituted derivatives.7d
Finally, 6-substituted cyclohex-2-enones exerted very low
substrate control in both Cu- and Rh-catalyzed 1,4-additions,
and mixtures of diastereomers were obtained which, via
their enolates, could be epimerized into enantiopure trans-
2,5-disubstituted cyclohexanones.9,10 Herein, we report on
kinetic resolutions of 4- and 5-substituted cyclohex-2-
enones by Rh/binap-catalyzed 1,2-addition of AlMe3 or
1,4-addition of boron and zinc organyls.
First, addition reactions to geminal dimethylated cyclo-
hex-2-enones were studied to determine whether substitu-
ents in the respective positions would interfere with the
catalyst. As already reported,4 1,2-addition of AlMe3 to
the 4,4-dimethylated 1b occurred in the same high yield as
in the case of the unsubstituted 1a (Table 1, entries 1, 2).
The 5,5-dimethylated 1c furnished only a moderate yield,
while the 6,6-disubstituted 1d showed again a very high
reactivity (entries 3, 4). In contrast, 1b did not undergo a
1,4-addition of PhB(OH)2 but slowly decomposed, while
1c smoothly reacted to cyclohexanone 3cm (entries 5, 6).
These results led to the hypothesis that the 1,2-addition
is most promising for kinetic resolutions of 5-substituted
cyclohex-2-enones whereas the 1,4-addition might be sui-
table for those with substituents in the 4-position. There-
fore, racemic 5-methylcyclohex-2-enone (1e) was reacted
with AlMe3 (1.2 equiv) and GC analysis revealed the
formation of the 1,2-adduct cis-(R,R)-2e in 50% yield with
79% ee, together with the 1,4-adduct (S,S)-4e (15% yield,
96% ee) and traces of trans-2e (Table 2, entry 1). Enantio-
pure (R)-1ewas then treated withthe Rh-catalysts contain-
ing (S)- and (R)-binap, respectively, and the former turned
out to be the matched pair furnishing only cis-2e in an
excellent yield (entry 2). In the case of the mismatched pair,
however, both diastereomers of the 1,2-adduct 2e were
obtained in 9% combined yield and, additionally, the 1,4-
adduct (R,R)-4e wasformedasasinglediastereomer in 16%
yield, most likely the consequence of a trans-selectivity
Table 1. Addition to Geminal Dimethylated Cyclohex-2-enones
a A: [Rh(cod)OMe]2 (2.5 mol %), (S)-binap (6.0 mol %), THF, rt,
15 min, then enone, AlMe3 (1.0 equiv), temp, time. B: [Rh(cod)OH]2
(1.5 mol %), (R)-binap (3.6 mol %), dioxane/H2O (10:1), rt, 1 h, then
enone, PhB(OH)2 (2.5 equiv), temp, time. b Yield of isolated product,
GC yield in parentheses. c Determined by chiral GC. d See ref 4. e [Rh-
(cod)Cl]2 and 1.2 equiv of AlMe3 were used. f Slow decomposition.
under substrate control (entry 3).11 Obviously, steric inter-
actions hampered the regular coordination of (R)-1e to the
(R)-catalyst, but some sort of unspecific activation must
have occurred. These results elucidate that, under the
conditions of entry 1, (S)-1e is transformed unspecifically
leading to the moderate 79% ee in the formation of cis-2e.
Finally, rac-1e was treated with the racemic catalyst, which
furnished almost exclusively cis-2e (entry 4). Thus, the
inherent 1,2-selectivity of the Rh-catalyzed AlMe3 addition
is maintained and combined with strong substrate control,
leading to a highly selective approach of the methyl group
from the face opposite to the substituent in the 5-position.
Under the conditions for a kinetic resolution with 0.5 equiv
of AlMe3, rac-1e reacted extraordinarily well furnishing
almost exclusively the expected cis-(R,R)-2e in 46% yield
with 95% ee together with unreacted (S)-1e in 32% yield with
97% ee (entry 5). Similar results were obtained at a lower
temperature, with the air-stable Lewis acidÀbase pair
DABCO 2AlMe3 (DABCO = 1,4-diazabicyclo[2.2.2]-
3
octane)12 or with the bulkier 5-isopropylcyclohexenone
1f(entries 6À8). Comparedtotheknown kineticresolution
of 5-substituted cyclohex-2-enones by Cu-catalyzed 1,4-
additions,7d this procedure furnishes synthetically interest-
ing allylic alcohols in high yield and optical purity while the
amount of recovered starting material is lower due to
unspecific side reactions.
(8) Rh-catalysis: (a) Chen, Q.; Kuriyama, M.; Soeta, T.; Hao, X.;
Yamada, K.-i.; Tomioka, K. Org. Lett. 2005, 7, 4439–4441. (b) Tomioka,
K. Pure Appl. Chem. 2006, 78, 2029–2034. Cu-catalysis: (c) Soeta, T.; Selim,
K.; Kuriyama, M.; Tomioka, K. Tetrahedron 2007, 63, 6573–6576.
(9) Rh-catalysis: (a) Mediavilla Urbaneja, L.; Krause, N. Tetrahe-
dron: Asymmetry 2006, 17, 494–496. (b) Chen, Q.; Soeta, T.; Kuriyama,
M.; Yamada, K.-i.; Tomioka, K. Adv. Synth. Catal. 2006, 348, 2604–
2608. Cu-catalysis: (c) Mediavilla Urbaneja, L.; Alexakis, A.; Krause,
N. Tetrahedron Lett. 2002, 43, 7887–7890. (d) Selim, K.; Soeta, T.;
Yamada, K.-i.; Tomioka, K. Chem.;Asian J. 2008, 3, 342–350.
(10) For Cu-catalyzed asymmetric 1,4-additions to other chiral
cyclohex-2-enones, see: (a) Imbos, R.; Minnaard, A. J.; Feringa, B. L.
Tetrahedron 2001, 57, 2485–2489. (b) Jagt, R. B. C.; Imbos, R.; Naasz,
R.; Minnaard, A. J.; Feringa, B. L. Isr. J. Chem. 2001, 41, 221–230. For
1,4-additions to chiral cyclic substrates catalyzed by achiral Rh-com-
plexes, see ref 2f and (c) Ramnauth, J.; Poulin, O.; Bratovanov, S. S.;
Rakhit, S.; Maddaford, S. P. Org. Lett. 2001, 3, 2571–2573.
Besides the possibility of a kinetic resolution, the Rh-
catalyzed addition of AlMe3 to 1e is an unprecedented
example of a regiodivergent reaction on a racemic mixture
(regiodivergent RRM)6 due to the formation of the
(11) For a discussion on the origin of trans-selectivity in cuprate
additions, see: Corey, E. J.; Hannon, F. J. Tetrahedron Lett. 1990, 31,
1393–1396.
(12) Biswas, K.; Prieto, O.; Goldsmith, P. J.; Woodward, S. Angew.
Chem., Int. Ed. 2005, 44, 2232–2234.
(13) For asymmetric 1,2-additions of mixed aluminum-zinc organyls
to aldehydes, see: Shannon, J.; Bernier, D.; Rawson, D.; Woodward, S.
Chem. Commun. 2007, 3945–3947.
(14) See Supporting Information for details.
Org. Lett., Vol. 14, No. 8, 2012
1979

Table 2. Rhodium-Catalyzed Addition of AlMe3 to 5-Substituted Cyclohex-2-enones
a GC yield, isolated yield in parentheses. b Determined by chiral GC. c 96% ee. d 0.5 equiv of DABCO 2AlMe3 was used.
3
1,2-adduct cis-2e from (R)-1e and the 1,4-adduct 4e from
(S)-1e (entry 1). Because the attempted optimization of the
yield of (S,S)-4e met with little success we envisaged a trans-
formation which may be called “sequential regiodivergent
RRM”: The kinetic resolution of 1e by a 1,2-addition to the
(R)-enantiomer followed by a 1,4-addition to the less
reactive (S)-enantiomer in the same reaction vessel.
As the achiral complex [Rh(cod)Cl]2 (cod = cyclo-
octadiene) catalyzes the racemic 1,4-addition of AlMe3 to
cyclohex-2-enone,4 experiments were performed in which
in situ formed [Rh(binap)Cl]2 for the 1,2-addition and an
excess of AlMe3 were present from the beginning; addi-
tional [Rh(cod)Cl]2 was added after a time delay (time 1,
Table 3). A delay of 15 min was obviously too long as (S)-
1e had already partially reacted in an unspecific manner
(concluded from the low ee of cis-2e and the low yield of
(S,S)-4e, entry 1). Addition of [Rh(cod)Cl]2 after only
5 min furnished 4e in higher yield with a low ee, indicating
incomplete conversion of (R)-1e at this time (entry 2). The
best results were obtained with a time delay of 10 min at rt
leading to a 75% combined yield of 1,2- and 1,4-adduct,
both with >90% ee (entry 3).
directs the 1,2-addition of AlMe3 to the si face, the 1,4-
addition of aryl groups occurs on the opposite face.2,4
Thus, we envisaged another special type of divergent RRM,
a parallel kinetic resolution (PKR)6 with the same chiral
precatalyst but two different organometallic reagents. Cata-
lyzed by Rh/(S)-binap, the 1,2-addition of AlMe3 to (R)-1e
is the matched pair (cf. Table 2, entry 5). We expected that
the 1,4-addition of PhZnCl to(S)-1etoform trans-(3S,5S)-
3-methyl-5-phenylcyclohexanone 3em should be the matched
pair in the conjugate addition because it would fit with the
usual trans-selectivity of substrate control. Although both
addition reactions must proceed via entirely different me-
chanisms, and thus different catalytic species, it was indeed
possible to perform them in parallel (Scheme 1).13 Both,
product (R,R)-2e from the 1,2-addition of AlMe3 and pro-
duct 3em from the 1,4-addition of PhZnCl were formed in
33% yield with ee’s g95%; however, the latter was formed as
atrans/cis mixture with a 4:1 dr. Using the bulkier substrate 1f
the diastereoselectivity was only slightly increased.
Scheme 1. Parallel Rh-Catalyzed 1,2-Addition of AlMe3 and
1,4-Addition of PhZnCla
Table 3. Kinetic Resolution of 1e Followed by 1,4-Addition of
Remaining (S)-1e
a Yields determined by GC. bAdditionally 19% of (S)-1e with 96% ee.
These results are highly interesting from a conceptual
and mechanistic point of view, but not completely
a GC yield. b Determined by chiral GC. c Reaction carried out at 0 °C.
(15) cis-Selective 1,4-additions of cuprates to enantiopure 5-alkoxy-
cyclohex-2-enones have been described: Hareau, G. P. J.; Koiwa, M.;
Hikichi, S.; Sato, F. J. Am. Chem. Soc. 1999, 121, 3640–3650 and
references therein.
The facial selectivities of the Rh-catalyzed 1,2- and 1,4-
addition to cylohex-2-enone are reversed: while (S)-binap
1980
Org. Lett., Vol. 14, No. 8, 2012

satisfying due to the moderate dr in the formation of the
1,4-adducts. To analyze the substrate control, PhZnCl or
PhB(OH)2 were added to enone 1e furnishing 1:1 mixtures
of cis- and trans-3em in excellent optical purities (Table 4,
entries 1, 2). From GC analyses of the reaction progress,
the ratio of rate constants kfast/kslow (the s-factor)6 was
calculated as 5.8À6.1 with PhZnCl and 1.3À1.5 with
PhB(OH)2.14 These low selectivities could not be increased
using the bulkier enone 1f or the bulkier (o-Tol)ZnCl
(entries 3, 4). In the absence of any catalyst control, i.e.
with rac-binap, formation of the trans-diastereomers was
favored with selectivities from 4:1 to 12:1 (entries 5À8).
Thus, the strong catalyst control of the HayashiÀMiyaura
reaction outpaces the significant substrate control of such
enones.8a,b This contrasts sharply with not only the corre-
sponding Cu-catalyzed 1,4-additions7d,8c but also the [Rh-
(cod)Cl]2-catalyzed 1,4-addition of AlMe3 which led to
pure trans-product 4 (cf. Table 3). Although this is a
problem for the PKR described above (Scheme 1), it
enables the diastereoselective formation of cis-3,5-disub-
stituted cyclohexanones from enantiopure 5-substituted
cyclohex-2-enones: The transformation of (R)-1e cata-
lyzed by the (S)-catalyst furnished almost pure cis-3em
(entry 9) and is, to the best of our knowledge, the first cis-
selective 1,4-addition to a 5-substituted cyclohex-2-enone
which is not based on precoordination of the reagent to a
heteroatomic substituent.15
(cf. Table 1, entries 5, 6). Yet, when 1g was treated with
PhB(OH)2 and [Rh(binap)OH]2, no conversion occurred
and a change in color from red to yellow indicated de-
composition of the catalyst. Transformation with the more
reactivePhZnCl, however, was possibleand furnished only
the trans-diastereomer 3gm in a good yield but with very
low ee (Scheme 2). Inferior results were achieved when
varying the aryl group, and time-resolved GC analysis of
the reaction with 0.5 equiv of PhZnCl revealed a maximum
eeof 74% for 3gm at13% conversion and a maximum eeof
15% for 1g at 21% conversion.14 The substrate control of
1g thus almost completely overrides catalyst control,
making kinetic resolutions hardly possible.
Scheme 2. 1,4-Addition of Various Aryl Groups to Enone 1g
In summary, it can be stated that the ratio of rate con-
stants kfast/kslow is very low in Rh/binap-catalyzed 1,4-
additions of aryl groups to chiral cyclohex-2-enones be-
cause catalyst control is too strong in the case of a 5- or
6-substitution8,9 and substrate control is too strong in the
case of a 4-substitution. The 1,2-addition of AlMe3, how-
ever, allows for a kinetic resolution of 5-substituted deri-
vatives, and interesting regiodivergent reactions can be
accomplished either as sequential reactions of the same
metal organyl with two different catalysts or as parallel
reactions of two different metal organyls with the same
catalyst. This special behavior of 5-substituted cyclohex-
enones in 1,2- and 1,4-additions and the opposite facial
selectivity strongly suggest two entirely disparate reaction
mechanisms. Contrary to the conjugate addition, the
methyl group might be delivered directly from the alane
and not via transmetalation to rhodium.
Table 4. 1,4-Addition of ArZnCl and PhB(OH)2 to Enones 1e,f
Acknowledgment. We thank the Fonds der Chemischen
Industrie and the Bengt Lundqvist foundation for scholar-
ships for A.K. and S.H., respectively.
a A: (1) [Rh(cod)Cl]2 (1.5 mol %), binap (3.6 mol %), THF, rt, 15 min;
(2) 1, PhZnCl (1.5 equiv), 0 °C, 30 min. B: (1) [Rh(cod)OH]2 (1.5 mol %),
binap (3.1 mol %), dioxane/H2O(10:1),rt, 1h;(2) 1e, PhB(OH)2 (2.5 equiv),
35 °C, 4 h. C: (1) [Rh(cod)Cl]2 (1.5 mol %), binap (3.6 mol %), THF, rt,
15 min; (2) 1e, (o-Tol)ZnCl (1.5 equiv), 0 °C, 1.5 h. b Isolated yield of
Supporting Information Available. Experimental proce-
dures and data from the characterization of compounds,
determination of s-factors for 1,4-additions to 1e, attempted
kinetic resolution of 1g by 1,4-addition of ArZnCl, and
ORTEP drawing of cis-3em. This material is available free
of charge via the Internet at http://pubs.acs.org.
1
inseparable mixtures of trans- and cis-3. c Determined by H NMR of
the crude product. d Determined by chiral HPLC. e Reaction time 1.5 h.
f Diastereomers separable by chromatography.
Drawn from the 1,4-additions to the geminal disubsti-
tuted cyclohexenones a much stronger substrate control
was expected in the case of 4-substituted cyclohexenones
The authors declare no competing financial interest.
Org. Lett., Vol. 14, No. 8, 2012
1981