The boron-mediated ketone-ketone aldol reaction

Article abstract of DOI:10.1016/j.tetlet.2005.01.011

The first examples of the directed, boron-mediated aldol reaction between different ketones are presented. Transformation of a variety of ketones to their corresponding boron enolates with Chx₂BCl/Et₃N, followed by reaction with acceptor ketones in diethyl ether, and oxidation of the resultant boron aldolate (H₂O₂, MeOH/pH 7 buffer), provided the aldol addition products. The reaction was most facile when cyclic ketones were used, with the highest yields obtained for the reaction of boron enolates with cyclohexanone as the acceptor.

Full text of DOI:10.1016/j.tetlet.2005.01.011

Tetrahedron
Letters
Tetrahedron Letters 46 (2005) 1505–1509
The boron-mediated ketone–ketone aldol reaction
Katie M. Cergol, Peter Turner and Mark J. Coster*
School of Chemistry, University of Sydney, NSW 2006, Australia
Received 9 November 2004; revised 19 December 2004; accepted 6 January 2005
Available online 21 January 2005
Abstract—The first examples of the directed, boron-mediated aldol reaction between different ketones are presented. Transforma-
tion of a variety of ketones to their corresponding boron enolates with Chx2BCl/Et3N, followed by reaction with acceptor ketones in
diethyl ether, and oxidation of the resultant boron aldolate (H2O2, MeOH/pH 7 buffer), provided the aldol addition products. The
reaction was most facile when cyclic ketones were used, with the highest yields obtained for the reaction of boron enolates with
cyclohexanone as the acceptor.
ꢀ 2005 Elsevier Ltd. All rights reserved.
The directed aldol reaction is one of the most valuable
carbon–carbon bond-forming processes available to
the organic chemist.1 Amid the plethora of available
procedures, the aldol reaction of boron enolates with
aldehydes has found particularly widespread application
in stereoselective organic synthesis.2 Attractive features
of this process include the highly predictable regio-
and stereoselective outcomes that can be achieved with
a variety of ketones and aldehydes, under mild condi-
tions. The many applications of this reaction to the
late-stage coupling of complex molecules in the area of
natural product synthesis,3 serve to highlight the re-
liability and functional group tolerance of this transfor-
mation. However, despite the exemplary nature of the
boron-mediated aldol reaction of ketone-derived eno-
lates with acceptor aldehydes, there have been no re-
ports of the reaction of boron enolates with acceptor
ketones to give aldol addition products in synthetically
useful yields. In contrast to the myriad of procedures
available for directed aldol reactions employing alde-
hydes as acceptors, there are relatively few general pro-
cedures for the directed aldol reaction of two different
ketones. Reactions of ketone-derived, Sn(II),4 Ce(III)5
and Ti(IV)6 enolates with acceptor ketones have been re-
ported, however, examples of the direct cross-coupling
of differing aliphatic ketones have been limited to the
last of these procedures.6a,c
As part of a program directed towards the synthesis of
sterically-congested 1,3-diols, we have investigated the
boron-mediated aldol reaction between two different
ketones for the construction of highly substituted b-
hydroxy ketones. We now report the reaction of dic-
yclohexylboron enolates, derived from an assortment
of aliphatic ketones, with a variety of acceptor ketones,
to provide the aldol addition products under mild
conditions.
Initial experiments involved the enolisation of cyclohex-
anone (Chx2BCl, Et3N, Et2O),7 to give the correspond-
ing boron enolate in situ, and subsequent reaction with
acetone or 3-pentanone (Et2O, 5 ꢁC, 16 h), followed by
treatment with H2O2 in MeOH/pH 7 buffer (Scheme
1). The reaction with acetone afforded the expected aldol
Keywords: Boron aldol; Ketones; Sterically congested.
*
Corresponding author. Tel.: +61 2 9351 2752; fax: +61 2 9351
3329; e-mail addresses: p.turner@chem.usyd.edu.au; m.coster@
chem.usyd.edu.au
To whom corrrespondence should be addressed regarding the crystal
structure.
Scheme 1.
0040-4039/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2005.01.011

1506
K. M. Cergol et al. / Tetrahedron Letters 46 (2005) 1505–1509
Table 1. Boron-mediated aldol reaction between cyclic ketones
product 1a in low yield (15%), along with a comparable
quantity of compound 2 (12%),8 presumably resulting
from the reaction of cyclohexanone with its correspond-
ing dicyclohexylboron enolate. In contrast, attempts to
extend this reaction to 3-pentanone failed to give any
of the desired aldol product 1b, providing 2 (11%) as
the sole isolated product. Changing the reaction solvent
to pentane led to only trace quantities of aldol products.
While the low yields obtained using the simple acyclic
ketones, acetone and 3-pentanone, suggested that
ketones are relatively unreactive towards boron eno-
lates, as implied by the lack of literature precedence in
this regard, we were intrigued by the apparent reactivity
of cyclohexanone as an acceptor under these conditions.
Entry
1
m
n
Products (yield)a
O
HO
1
2
—
—
The boron-mediated aldol reaction of simple cycloalkan-
ones was systematically studied for five- to seven-mem-
bered rings (Table 1).9 In all cases studied, the desired
cross-aldol product could be isolated in low to good
yields. The highest yields were obtained when cyclohex-
anone was used as the acceptor ketone (entries 1 and 6,
70% and 64%, respectively). The lower yields obtained
with cyclopentanone or cycloheptanone as the acceptor
under the standard reaction conditions, are attributed to
a slower rate of reaction. In the case of the cycloheptan-
one-derived enolate reacting with cyclopentanone (entry
5), increasing the reaction time from 16 to 40 h led to a
moderate increase in yield (41!61%).10 Furthermore,
reactions involving cyclohexanone as the donor ketone
(entries 3–4) resulted in the formation of small amounts
of 2 (5–8%). In contrast, when cyclopentanone or cyclo-
heptanone were used as the donor ketone, none of the
analogous Ôself-aldolÕ product was obtained (entries 1–
2, 5–6). These findings serve to demonstrate the high
reactivity of cyclohexanone as an acceptor ketone in
these reactions.
3a
(70%)
O
HO
2
1
2
2
3
3
1
3
1
3b
(39%)
O
O
HO
O
HO
3
3c
(47%)
2
(5%)
O
HO
HO
4b
3d
(8%)
2
(8%)
HO
O
Variation of the reaction conditions, e.g. rate/order of
addition, enolisation temperature (ꢀ78 ꢁC) and solvent
(pentane, CH2Cl2), failed to improve the overall yield
of aldol product or decrease the quantity of 2 produced
in these reactions. Furthermore, the use of the corre-
sponding di-n-butylboron enolates, gave lower yields
of the desired aldol products (<20%). In these cases,
the reactions were conducted in CH2Cl2, which we have
found to be a greatly inferior solvent for this aldol addi-
tion, and therefore the low yields likely illustrate the
effect of solvent on the rate of the reaction, rather than
the nature of the boron ligands.
5
—
3e
(41%)
c
O HO
6
3
2
—
3f
(64%)
a Isolated yields.
b Compounds 2 and 3d were not separable by flash chromatography.
c Aldol reaction at 5 ꢁC for 40 h gave a 61% yield of 3e.
The yields obtained using cyclic ketones as acceptors in
these aldol reactions correspond with their rates of
reduction with sodium borohydride,11 reflecting the rela-
tive reactivity of each of these ketones towards nucleo-
philic addition. Molecular mechanics calculations have
been used to rationalise the differing reactivity of cyclic
ketones according to ring size.12 Of the acceptor cycloal-
kanones studied here, only cyclohexanone has been cal-
culated to be more strained than the corresponding
hydrocarbon, cyclohexane. Moreover, the inability to
have an alkyl group eclipsing the carbonyl has been used
to rationalise the increased strain of cyclohexanone rela-
tive to 3-pentanone. Hence, cyclohexanone represents a
particularly reactive acceptor ketone for the boron-med-
iated aldol reaction, due to lack of steric hindrance
about the carbonyl group and release of ring strain upon
nucleophilic addition.13 Noting the superior results ob-
tained with cyclohexanone as the acceptor in these boron-
mediated ketone–ketone aldol reactions, the scope of
this transformation was further explored using a variety

K. M. Cergol et al. / Tetrahedron Letters 46 (2005) 1505–1509
1507
of ketone donors and cyclohexanone as the acceptor
(Table 2).
assignments of cis- and trans-4a (Table 2, entry 1), the
major diastereomer was subjected to NaBH(OAc)3 in
dichloromethane (Scheme 2).15 The corresponding, crys-
talline 1,3-diol 5 was produced as the major diastereo-
mer (96:4dr by GC and 1H NMR analysis).
The aldol reaction proceeded smoothly with 2-methyl-
cyclohexanone as the donor ketone (entry 1), via regio-
selective enolate formation, providing the aldol product
4a in good yield (75%), as a 2:1 mixture of cis and trans
diastereomers, respectively. The stereochemistry of each
of these were assigned on the basis of 1H NMR coupling
constants and NOESY spectra, and confirmed by X-ray
crystallographic analysis of a derivative, vide infra.
The structure and relative stereochemistry of 5 were
determined by single crystal X-ray diffraction,16–21
which confirmed the aldol reaction gave predominantly
the cis-diastereomer, while reduction occurred via equa-
torial delivery of hydride to cis-4a (Fig. 1). This is in
contrast to the reduction of syn- and anti-cyclohexan-
one–benzaldehyde aldol adducts with NaBH(OAc)3,
which provide as the major products, 1,3-diols resulting
from axial delivery of hydride.15
The aldol reaction of acyclic donor ketones, acetone and
3-pentanone, with cyclohexanone proceeded less rap-
idly, requiring 40 h to yield the corresponding aldol
products, 4b and 4c, respectively (entries 2 and 3).14
The aldol reaction with a heterocyclic donor ketone,
N-benzyl-4-piperidone, provided 4d (entry 4) in good
yield (75%).
Inspection of models, in relation to EvansÕ mechanistic
studies on triacetoxyborohydride mediated reduction
of hydroxyketones,22 suggests that the stereoselectivity
observed in the reduction of cis-4a to 5 may arise by
internal delivery of hydride via a boat-like transition
state TS-1, leading to the axial alcohol.23 The boat-like
transition state, TS-2, for internal delivery of hydride in
an axial sense suffers from severe 1,4- steric interactions
between the acetoxy group on boron and a CH2 of the
cyclohexanone ring. Examination of alternative, chair-
like transition states for internal delivery of hydride to
either face of the ketone, suggests these are unfavour-
able in each case due to 1,3-diaxial interactions between
The aldol products produced by the methods outlined
here could, in principle, be reduced to 1,3-diols by a
variety of methods. In order to verify the stereochemical
Table 2. Boron-mediated aldol reaction between ketones and cyclo-
hexanone
Entry
Donor ketone
Products (yield)a
O
HO
O
Me
Me
1
4a
(75%, 2:1 cis/trans-)
O
HO
O
O
2b
4b
(52%)
Scheme 2. Reduction of cis-4a.
O
HO
3b
4c
(63%)
O
N
HO
O
4
N
Bn
Bn
4d
(75%)
a Isolated yields.
b Aldol reaction at 5 ꢁC for 40 h.
Figure 1. ORTEP19,25 depiction of 5, with 50% displacement
ellipsoids.

1508
K. M. Cergol et al. / Tetrahedron Letters 46 (2005) 1505–1509
pH 7 buffer (1 mL) was added at 0 ꢁC, followed by a pre-
mixed solution of 2:1 MeOH (2 mL) and 30% H2O2
(1 mL), and the mixture was warmed to rt. After 2 h at rt
the mixture was diluted with Et2O (10 mL) and H2O
(10 mL), the layers were separated and the aqueous phase
was extracted with Et2O (3 · 10 mL). The combined
organic extracts were washed with satd NaHCO3
(15 mL), and brine (15 mL), dried (Na2SO4) and concen-
trated in vacuo. Purification by flash chromatography
(SiO2, 25:75 EtOAc/hexanes) provided 3a (128 mg, 70%)
as a colourless oil: IR (NaCl, thin film) 3493, 2934, 2858,
1724, 1448, 1402, 1356, 1321, 1271, 1250, 1209, 1155, 1065,
1034, 1001 cmꢀ1; 1H NMR (400 MHz, CDCl3) d 3.43 (1H,
s), 2.32–2.24(1H, m), 2.19–1.90 (4H, m), 1.72–1.52 (6H,
m), 1.44–1.30 (5H, m), 1.12–1.02 (1H, m); 13C NMR
(100.6 MHz, CDCl3) d 222.9, 72.6, 58.3, 39.7, 36.2, 32.6,
25.8, 25.5, 21.2, 21.1, 20.1; HRMS (+ESI) Calc. for
C11H18O2 Na [M+Na]+: 205.1199, found: 205.1200.
10. In cases where a low to moderate yield of the aldol
product was obtained, TLC analysis suggested that some
unreacted starting materials remained. However, these
starting materials were too volatile to recover from the
reaction mixtures. Conducting the reaction at tempera-
tures above 5 ꢁC led to lower yields and the production of
undesired side products.
an acetoxy group and an axially-oriented CH2 of one of
the six-membered rings.24
In summary, the boron-mediated ketone–ketone aldol
reaction has been shown to be a facile process with cyc-
lic acceptor ketones. In particular, cyclohexanone was
found to react with a variety of boron enolates to afford
the cross-aldol products in good yields. Further efforts
to explore the stereoselectivity of this reaction with
substituted ketones, and to functionalise the aldol prod-
ucts towards sterically-encumbered 1,3-diols and related
compounds, will be reported in due course.
Acknowledgements
We thank the University of Sydney for funding, Dr.
Kelvin Picker for assistance with HPLC and Dr. Ian
Luck for assistance with 2D NMR experiments.
References and notes
1. Heathcock, C. H. In Comprehensive Organic Synthesis;
Trost, B. M., Fleming, I., Eds.; Pergamon: Oxford, 1991;
Vol. 2, pp 181–238.
11. Brown, H. C.; Ichikawa, K. Tetrahedron 1957, 1, 221–230.
12. Allinger, N. L.; Tribble, M. T.; Miller, M. A. Tetrahedron
1972, 28, 1173–1190.
2. For reviews, see: (a) Cowden, C. J.; Paterson, I. Org.
React. 1997, 51, 1–200; (b) Paterson, I.; Cowden, C. J.;
Wallace, D. J. In Modern Carbonyl Chemistry; Otera, J.,
Ed.; Wiley–VCH: Weinheim, 2000; pp 249–297.
3. For a recent example, see: Paterson, I.; Chen, D. Y.-K.;
Coster, M. J.; Acen˜a, J. L.; Bach, J.; Gibson, K. R.;
Keown, L. E.; Oballa, R. M.; Trieselmann, T.; Wallace, D.
J.; Hodgson, A. P.; Norcross, R. D. Angew. Chem., Int.
Ed. 2001, 40, 4055–4060.
4. (a) Mukaiyama, T.; Stevens, R. W.; Iwasawa, N. Chem.
Lett. 1982, 353–356; (b) Stevens, R. W.; Iwasawa, N.;
Mukaiyama, T. Chem. Lett. 1982, 1459–1462; (c) Mukai-
yama, T.; Iwasawa, N.; Stevens, R. W.; Haga, T. Tetrahe-
dron 1984, 40, 1381–1390.
13. No aldol reaction is observed when 2-methylcyclohexa-
none is used as the acceptor ketone, demonstrating the
strong influence of a 2-alkyl substituent on the reactivity
of the acceptor. This may be attributed to the increased
steric hindrance about the carbonyl group, and a dimin-
ished contribution from strain relief.
14. The yield of 4c was also limited by difficulties in isolation,
due to the significant volatility of this compound.
15. Thompson, S. H. J.; Mahon, M. F.; Molloy, K. C.;
Hadley, M. S.; Gallagher, T. J. Chem. Soc., Perkin Trans.
1 1995, 379–383.
16. Structure Determination for 5: The crystals fragmented on
cutting, accordingly a larger than ideal crystal was used
for the data collection. A Bruker SMART 1000 CCD
diffractometer with graphite monochromated MoKa
radiation from a sealed tube was used to collect data at
150(2) Kelvin, using x scans. Data integration and
reduction were undertaken with SAINT and XPREP.17
and subsequent computations were carried out with the
WinGX18 and XTAL19 graphical user interfaces. The
structure was solved in the space group P21/n(#14) by
direct methods with SIR97,20 and extended and refined with
SHELXL-97.21 The non-hydrogen atoms were modelled with
anisotropic displacement parameters, and the hydrogen
atom sites were located and modelled with isotropic
displacement parameters. Crystal data: Formula
5. Imamoto, T.; Kusumoto, T.; Yokoyama, M. Tetrahedron
Lett. 1983, 24, 5233–5236.
6. (a) Yoshida, Y.; Hayashi, R.; Sumihara, H.; Tanabe, Y.
Tetrahedron Lett. 1997, 38, 8727–8730; (b) Yoshida, Y.;
Matsumoto, N.; Hamasaki, R.; Tanabe, Y. Tetrahedron
Lett. 1999, 40, 4227–4230; (c) Tanabe, Y.; Matsumoto, N.;
Higashi, T.; Misaki, T.; Itoh, T.; Yamamoto, M.; Mitarai,
K.; Nishii, Y. Tetrahedron 2002, 58, 8269–8280; (d)
Tanabe, Y.; Mitarai, K.; Higashi, T.; Misaki, T.; Nishii,
Y. Chem. Commun. 2002, 2542–2543.
7. (a) Brown, H. C.; Dhar, R. K.; Ganesan, K.; Singaram, B.
J. Org. Chem. 1992, 57, 499–504; (b) Brown, H. C.; Dhar,
R. K.; Ganesan, K.; Singaram, B. J. Org. Chem. 1992, 57,
2716–2721.
8. All new compounds gave spectroscopic data in agreement
with the assigned structures. The data for compound 2
match those reported previously: Yasuda, M.; Hayashi,
K.; Katoh, Y.; Shibata, I.; Baba, A. J. Am. Chem. Soc.
1998, 120, 715–721.
9. Representative procedure: A solution of Chx2BCl (241 lL,
234mg, 1.1 mmol, 1.1 equiv) in Et 2O (4mL) at 0 ꢁC was
treated with Et3N (167 lL, 121 mg, 1.2 mmol, 1.2 equiv)
dropwise, followed by cyclopentanone (88 lL, 84mg,
1.0 mmol) and the resulting mixture was stirred for
45 min. Cyclohexanone (114 lL, 108 mg, 1.1 mmol,
1.1 equiv) was added and the mixture was left at 5 ꢁC
for 16 h. A pre-mixed solution of 6:1 MeOH (6 mL) and
C13H24O2, M 212.32, space group P1/n(#14), a 11.783(2),
˚
3
˚
b 6.3284(11), c 15.914(3) A, b 97.095(3), V 1177.6(3) A , Z
4, crystal size 0.628 by 0.088 by 0.049 mm, colour
colourless, habit acicular, temperature 150(2) K,
˚
k(MoKa) 0.71073 A, l(MoKa) 0.078 mmꢀ1
,
2hmax
56.62, hkl range ꢀ1415, ꢀ8 8, ꢀ20 20, N 11150, Nind
2812(Rmerge 0.0616), Nobsd 1650(I > 2r(I)), Nvar 232,
residuals R1(F) 0.0397, wR2(F2) 0.0921, GoF(all) 1.085,
Dqmin, max ꢀ0.196, 0.286 eꢀ
A
ꢀ3. Crystallographic data
˚
(excluding structure factors) for 5 has been deposited with
the Cambridge Crystallographic Data Centre as supple-
mentary publication number CCDC 254676. Copies of the
data can be obtained free of charge, on application to
CCDC, 12 Union Road, Cambridge CB2 1EZ, UK [fax:
+44(0) 1223 336033 or email: deposit@ccdc.cam.ac.uk].

K. M. Cergol et al. / Tetrahedron Letters 46 (2005) 1505–1509
1509
17. Bruker. SMART, SAINT and XPREP. Area detector
control and data integration and reduction software.
Bruker Analytical X-ray Instruments Inc.: Madison, WI,
USA, 1995.
18. Farrugia, L. J. J. Appl. Cryst. 1999, 32, 837–838.
19. Xtal3.6 System; Hall, S. R., du Boulay, D. J., Olthof-
Hazekamp, R., Eds.; University of Western Australia:
Australia, 1999.
22. Evans, D. A.; Chapman, K. T.; Carreira, E. M. J. Am.
Chem. Soc. 1988, 110, 3560–3578.
23. Although it is surprising that this reaction occurs in the
absence of acetic acid as co-solvent, the one equivalent of
acetic acid released on exchange of the acetoxy for the
substrate alkoxy group, at boron, may be sufficient to
facilitate the reaction, given the facile nature of cyclohex-
anone reduction (Ref. 11).
20. Altomare, A.; Burla, M. C.; Camalli, M.; Cascarano, G.
L.; Giacovazzo, C.; Guagliardi, A.; Moliterni, A. G. G.;
Polidori, G.; Spagna, R. J. Appl. Cryst. 1998, 32, 115–119.
21. Sheldrick, G. M. SHELX97 Programs for Crystal Structure
Analysis; University of Go¨ttingen: Germany, 1998.
24. An alternative mechanism involving intermolecular
hydride delivery cannot be ruled out.
25. Johnson, C. K. ORTEP II. Report ORNL-5138. Oak
Ridge National Laboratory, Oak Ridge, TN, USA,
1976.