Highly chemoselective direct crossed aliphatic-aromatic acyloin condensations with triazolium-derived carbene catalysts

Article abstract of DOI:10.1021/jo101791w

It has been shown for the first time that triazolium precatalysts promote (in the presence of base) highly chemoselective crossed acyloin condensation reactions between aliphatic and ortho-substituted aromatic aldehydes. An o-bromine atom can serve as a temporary directing group to ensure high chemoselectivity (regardless of the nature of the other substituents on the aromatic ring) which then can be conveniently removed. The process is of broad scope and is operationally simple as it does not require the preactivation of any of the coupling partners to ensure selectivtiy. Preliminary data indicate that highly enantioselective variants of the reaction are feasible using chiral precatalysts.

Full text of DOI:10.1021/jo101791w

pubs.acs.org/joc
Highly Chemoselective Direct Crossed Aliphatic-Aromatic Acyloin
Condensations with Triazolium-Derived Carbene Catalysts
Sarah E. O’Toole,‡ Christopher A. Rose,† Sivaji Gundala,‡ Kirsten Zeitler,*,† and
Stephen J. Connon*,‡
‡Centre for Synthesis and Chemical Biology, School of Chemistry, The University of Dublin, Trinity College,
Dublin 2, Ireland, and †Institut fu€r Organische Chemie, Universita€t Regensburg Universita€tsstrasse 31,
D-93053 Regensburg, Germany
kirsten.zeitler@chemie.uni-regensburg.de; connons@tcd.ie
Received September 12, 2010
It has been shown for the first time that triazolium precatalysts promote (in the presence of base)
highly chemoselective crossed acyloin condensation reactions between aliphatic and ortho-substi-
tuted aromatic aldehydes. An o-bromine atom can serve as a temporary directing group to ensure
high chemoselectivity (regardless of the nature of the other substituents on the aromatic ring) which
then can be conveniently removed. The process is of broad scope and is operationally simple as it does
not require the preactivation of any of the coupling partners to ensure selectivtiy. Preliminary data
indicate that highly enantioselective variants of the reaction are feasible using chiral precatalysts.
Introduction
The acyloin condensation (AC) is one of the oldest carbon-
carbon bond forming reactions in organic chemistry;with a
R-Hydroxy ketones are highly useful building blocks for
the synthesis of heterocycles, natural products, agrochemi-
cals, and (inter alia1) pharmaceuticals.2 In addition, the un-
symmetrical nature of the building block allows for access to
other important synthetic precursors, such as chiral 1,2-diols
and amino alcohols.3 As a consequence, the development of
routes to these compounds via metal-catalyzed heteroatom
transfer4 and organocatalytic R-oxidation chemistry5 has
been extensively investigated recently.6
€
rich history dating back to the pioneers Liebig and Wohler in
1832.7 For much of the intervening time, it has proven first
an interesting mechanistic challenge8,9 and later a process
which inspired the development of a suite of N-heterocyclic
(5) For selected examples on organocatalytic R-hydroxylations and
R-oxyaminations, see: (a) Zhong, G. Angew. Chem., Int. Ed. 2003, 42,
4247. (b) Brown, S. P.; Brochu, M. P.; Sinz, C. J.; MacMillan, D. W. C.
J. Am. Chem. Soc. 2003, 125, 10808. (c) Hayashi, Y.; Yamaguchi, J.; Hibino,
K.; Sumiya, T.; Urushima, T.; Shoji, M.; Hashizume, D.; Koshino, H.
Adv. Synth. Catal. 2004, 346, 1435. (d) Sibi, M. P.; Hasegawa, M. J. Am.
Chem. Soc. 2007, 129, 4124. (e) Gotoh, H.; Hayashi, Y. Chem. Commun.
(1) (a) For use as photocleavable protecting groups see: Pirrung, M. C.;
Shuey, S. W. J. Org. Chem. 1994, 59, 3890. (b) For use as an anchoring group
for materials see: Fuhrmann, G.; Nelles, G.; Zilai, B.; Obermaier, M. WO
2010/049042 A2.
ꢀ
2009, 3083. (f) Palomo, C.; Vera, S.; Velilla, I.; Mielgo, A.; Gomez-
Bengoa, E. Angew. Chem., Int. Ed. 2007, 46, 8054. (g) Koike, T.; Akita,
M. Chem. Lett. 2009, 38, 166. (h) Poe, S. L.; Bogdan, A. R.; Mason, B: P.;
Steinbacher, J. L.; Opalka, S. M.; McQuade, D. T. J. Org. Chem. 2009,
74, 1574.
(6) Recent reviews: (a) Merino, P.; Tejero, T. Angew. Chem., Int. Ed.
2004, 43, 2995. (b) Marigo, M.; Jørgensen, K. A. R-Heteroatom Function-
alization. Enantioselective Organocatalysis; Wiley: Weinheim, Germany,
2007.
ꢀ
(2) Hoyos, P.; Sinisterra, J. -V.; Molinari, F.; Alcantara, A. R.;
Domınguez de Marıa, P. Acc. Chem. Res. 2010, 43, 288.
(3) For instance, (-)-ephedrine is available from a thiamine-dependent
enzyme-catalyzed reaction which produces (R)-phenylacetylcarbinol:
Rosche, B.; Sandford, V.; Breuer, M.; Hauer, B.; Rogers, P. L. J. Mol. Catal.
B 2002, 19, 109.
€
(7) Wohler, F.; Liebig, J. Ann. Pharm. 1832, 3, 249.
(8) Breslow, R. J. Am. Chem. Soc. 1958, 80, 3719.
(4) For selected reviews, see: (a) Plietker, B. Tetrahedron: Asymmetry
2005, 16, 3453. (b) Davis, F. A.; Chen, B. C. Chem. Rev. 1992, 92, 919.
DOI: 10.1021/jo101791w
r
Published on Web 10/27/2010
J. Org. Chem. 2011, 76, 347–357 347
2010 American Chemical Society

O’Toole et al.
JOCFeatured Article
SCHEME 1. Challenges Associated with Direct Crossed Acyloin Condensations
carbene-based catalysts.10 While significant advances in the
catalysis of the (asymmetric) carbene-catalyzed homo-AC
reaction have been made recently,11-15 the absence of a
selective carbene-mediated methodology capable of promot-
ing the intermolecular reaction16 between two different alde-
hydes in a chemo- and enantioselective fashion curtails the
utility of the process. The challenges associated with the
development of an efficient and selective crossed acyloin
condensation protocol are considerable;the objective is to
exercise control (via the catalyst) over the process to the
extent that a single major adduct is formed from 8 possible
products (4 chiral ketones 3a-d ꢀ 2 enantiomers each,
Scheme 1) in good yield.17
Known Approaches. In 1930, Buck et al. published a report
concerning the crossed benzoin condensation of aromatic
aldehyde partners of contrasting electronic character in the
presence of high loadings of cyanide ion (Scheme 2B).18,19
Over 30 years ago, Stetter et al.,20 in an attempt to develop
efficient routes to 1,2-diketones, reported in a short, limited
study that an achiral thiazolium salt-derived carbene cata-
lyzed the crossed AC between aromatic and aliphatic alde-
hydes: good crossed product yields could be obtained if the
aliphatic aldehyde was utilized in quite high excess (3.0 equiv);
however, chemoselectivity was both highly variable and
substrate-dependent.21-23
Miller et al. carried out a single intermolecular AC reaction
reported to be selective involving o-tolualdehyde and hex-
anal in the presence of stoichiometric loadings of a triazo-
lium ion precatalyst. However, the yield of the only isolable
product was low (16%).24 Other approaches to the catalytic
synthesis of products (formally) derived from intermolecular
AC reactions have been developed, including the use of
enzyme catalysts25 and polymer-bound aldehydes,26 in addi-
tion to indirect methods where chemoselectivity is derived
from the preformation of an umpolung reagent, such as acyl
silanes,27 acyl phosphonates,28 and aldehyde thiazolium car-
bene adducts.29 Enders recently disclosed that aromatic
aldehydes could be coupled to R,R,R-trifluoroacetophe-
none in good to excellent yields under the influence of
triazolium carbene catalysis;30 however, to the best of our
knowledge a general carbene-catalyzed process capable of
promoting the direct, chemoselective31 (and enantioselective)
(22) For a later demonstration of the potential utility of formaldehyde as
a coupling partner in these reactions, see: Matsumoto, T.; Ohishi, M.; Inoue,
S. J. Org. Chem. 1985, 50, 603.
(23) In addition, most likely due to the lack of high-field NMR instru-
mentation and the focus of the study on the oxidized benzil products
(cf. their NMR characterization data), we have found that in some instances the
chemoselectivity reported in this study was over estimated due to misassignment of
products in the crude 1H NMR spectra. In our hands, Stetter’s conditions provided
compound 12 as major product (see Table 1 and ref 37.). Further details will be
reported in due course. For a more recent report using NMR spectroscopy for the
structural assignment in mixed aromatic benzoin systems, see: Simion, C.; Simion,
A. M. UPB Sci. Bull., Ser. B 2007, 69, 49.
(9) For a more recent kinetic study see: White, M. J.; Leeper, F. J. J. Org.
Chem. 2001, 66, 5124.
(10) Recent reviews: (a) Moore, J. L.; Rovis, T. Top. Curr. Chem. 2009,
291, 77. (b) Enders, D. J. Org. Chem. 2008, 73, 7857. (c) Rovis, T. Chem. Lett.
2008, 37, 1. (d) Zeitler, K. E. Schering Found. Symp. Proc. 2007, 2, 183.
(e) Enders, D.; Niemeier, O.; Henseler, A. Chem. Rev. 2007, 107, 5506.
(f) Marion, N.; Diez-Gonzalez, S.; Nolan, S. P. Angew. Chem., Int. Ed. 2007,
46, 2988. (g) Zeitler, K. Angew. Chem., Int. Ed. 2005, 44, 7506. (h) Christmann,
M. Angew. Chem., Int. Ed. 2005, 44, 2632. (i) Enders, D.; Balensiefer, T. Acc.
Chem. Res. 2004, 37, 534. (j) Johnson, J. S. Angew. Chem., Int. Ed. 2004, 43,
1326.
(24) Mennen, S. M.; Miller, S. J. J. Org. Chem. 2007, 72, 5260.
€
(25) (a) Lehwald, P.; Richter, M.; C. Rohr, C.; Liu, H.-W.; Muller, M.
Angew. Chem., Int. Ed. 2010, 49, 2389. (b) Muller, M.; Gocke, D.; Pohl, M.
€
€
€
€
FEBS J. 2009, 276, 2894. (c) Demir, A. S.; S-es-engolu, O.; Dunkelmann, P.;
Muller, M. Org. Lett. 2003, 5, 2047. (d) Dunkelmann, P.; Kolter-Jung, D.;
€
€
(11) Enders, D.; Breuer, K.; Teles, J. H. Helv. Chim. Acta 1996, 79, 1217.
(12) Knight, R. L.; Leeper, F. J. J. Chem. Soc., Perkin Trans. 1 1998,
1891.
Nitsche, A.; Demir, A. S.; Siegert, P.; Linden, B.; Baumann, M.; Pohl, M.;
€
Muller, M. J. Am. Chem. Soc. 2002, 124, 12084. (e) Demir, A. S.; S-es-engolu,
O.; Eren, E.; Hosrik, B.; Pohl, M.; Janzen, E.; Kolter, D.; Feldmann, R.;
€
€ €
Dunkelmann, P.; Muller, M. Adv. Synth. Catal. 2002, 344, 96. (f) Pohl, M.;
(13) Enders, D.; Kallfass, U. Angew. Chem., Int. Ed. 2002, 41, 1743.
(14) Enders, D.; Han, J. Tetrahedron: Asymmetry 2008, 19, 1367.
(15) Ma, Y.; Wei, S.; Wu, J.; Yang, F; Liu, B.; Lan, J.; Yang, S.; You, J.
Adv. Synth. Catal. 2008, 350, 2645.
(16) Highly enantioselective intermolecular AC reactions between alde-
hydes and ketones are known. For representative examples, see: (a) Enders,
D.; Niemeier, O.; Balensiefer, T. Angew. Chem., Int. Ed. 2006, 45, 1463.
(b) Takikawa, H.; Hachisu, Y.; Bode, J. W.; Suzuki, K. Angew. Chem., Int.
Ed. 2006, 45, 3492. (c) Enders, D.; Niemeier, O.; Raabe, G. Synlett 2006,
2431. (d) Li, Y.; Feng, Z.; You, S.-L. Chem. Commun. 2008, 2263. (e) For a
recent example on a sterically controlled intramolecular aldehyde ketone AC
reaction in the context of a multicatalytic cascade sequence, see: Lathrop,
S. P.; Rovis, T. J. Am. Chem. Soc. 2009, 131, 13628.
€
Lingen, B.; Muller, M. Chem.;Eur. J. 2002, 8, 5288. (g) Demir, A. S.; Pohl,
M.; Janzen, E.; Muller, M. J. Chem. Soc., Perkin Trans. 1 2001, 633.
€
(26) (a) Kuriakose, S.; Rajasekharan Pillai, V. N. Eur. Polym. J. 1994, 30,
881. (b) Kuriakose, S; Rajasekharan Pillai, V. N. Angew. Makromol. Chem.
1994, 221, 53. (c) Kuriakose, S; Rajasekharan Pillai, V. N. Polymer 1994, 35,
1339.
(27) (a) Linghu, X.; Johnson, J. S. Angew. Chem., Int. Ed. 2003, 42, 2534.
(b) Linghu, X.; Potnick, J. R.; Johnson, J. S. J. Am. Chem. Soc. 2004, 126,
3070. (c) Bausch, C. C.; Johnson, J. S. J. Org. Chem. 2004, 69, 4283.
(d) Linghu, X.; Bausch, C. C; Johnson, J. S. J. Am. Chem. Soc. 2005, 127,
1833. (e) Tarr, J. C.; Johnson, J. S. Org. Lett. 2009, 11, 3870.
€
€
(28) (a) Demir, A. S.; Esiringu, I.; Gollu, M.; Reis, O. J. Org. Chem. 2009,
74, 2197. (b) Demir, A. S.; Reis, B.; Reis, O.; Eymur, S.; Gollu, M.; Tural, S.;
€ €
€
€
€ €
(17) In addition, the use of enolizable aldehydes is potentially problematic
due to competitive aldol pathways under the basic reaction conditions.
(18) Ide, W. S.; Buck, J. S. Org. React. 1948, 4, 269.
€
€
Saglam, G. J. Org. Chem. 2007, 72, 7439. (c) Demir, A. S.; Reis, O.; Esiringu,
€
I.; Reis, B.; Baris, S. Tetrahedron 2007, 63, 160. (d) Demir, A. S.; Reis, O.;
ꢁ
€
(19) It is noteworthy that establishing the levels of chemoselectivity in this
study proved difficult due to the reactivity of the benzoins in the derivatiza-
tion process to establish the structures of the products. For a 1H NMR
spectroscopy based structural assignment, also see ref 23.
Cigdem, C.; Esiringu, I.; Eymur, S. J. Org. Chem. 2005, 70, 10584. (e) Bausch,
C. C.; Johnson, J. S. Adv. Synth. Catal. 2005, 347, 1207.
(29) (a) Mathies, A. K.; Mattson, A. E.; Scheidt, K. A. Synlett 2009, 377.
(b) Mattson, A. E.; Zuhl, A. M.; Reynolds, T. E.; Scheidt, K. A. J. Am. Chem.
Soc. 2006, 128, 4932. (c) Mattson, A. E.; Scheidt, K. A. J. Am. Chem. Soc.
2007, 129, 4508.
€
(20) (a) Stetter, H.; Dambkes, G. Synthesis 1977, 403. (b) Stetter, H.;
Dambkes, G. Synthesis 1980, 309.
(21) For a related report using aliphatic aldehydes see: Heck, R.;
€
(30) (a) Enders, D.; Henseler, A. Adv. Synth. Catal. 2009, 351, 1749.
(b) Enders, D.; Grossmann, A.; Fronert, J.; Raabe, G. Chem. Commun. 2010,
46, 6282.
€
Henderson, A. P.; Kohler, B.; Retey, J.; Golding, B. T. Eur. J. Org. Chem.
ꢀ
2001, 2623.
348 J. Org. Chem. Vol. 76, No. 2, 2011

O’Toole et al.
JOCFeatured Article
SCHEME 2. Previous Approaches and the Proposed Strategy Detailed in This Work
SCHEME 3. General Mechanism for NHC-Catalyzed Benzoin/
Acyloin Condensations
crossed AC reaction between two different aldehydes
remains elusive.32
Results and Discussion
In approaching this problem, we considered what we
regarded as the key question: the aldehyde is the electrophile
in both the Breslow intermediate (BI, Scheme 3: III) and
product (and stereocenter)-forming steps, so if, for instance,
aldehyde 2 is the superior electrophile in the BI formation
step (Scheme 3: IfIII), on what basis (using traditional
catalyst design strategies) can we expect 2 not to be the
superior electrophile in the subsequent stereocenter-forming
step (Scheme 3: IIIfIV), leading to the homodimer 3a
instead of cross-product 3d (Scheme 1B and Scheme 3: A
respectively D)?
(31) For contributions to the ongoing discussion on the importance of
chemoselectivity for efficient syntheses, see: (a) Trost, B. M. Science 1983,
219, 245. (b) Shenvi, R. A.; O’Malley, D. P.; Baran, P. S. Acc. Chem. Res.
2009, 42, 530. (c) Maurigo, M.; Melchiorre, P. ChemCatChem 2010, 2, 621.
(32) For an example of a preceding intermolecular aldehyde ketone cross-
acyloin reaction in the context of a cyclopentene-forming annulation via NHC-
catalyzed acyloin-oxy-Cope reaction, see: Chiang, P.-C.; Kaeobamrung, J.;
Bode, J. W. J. Am. Chem. Soc. 2007, 129, 3520.
(33) O’Toole, S. E.; Connon, S. J. Org. Biomol. Chem. 2009, 7, 3584.
(34) (a) Baragwanath, L.; Rose, C. A.; Zeitler, K.; Connon, S. J. J. Org.
Chem. 2009, 74, 9214. (b) The pKa of catalyst 8 (H2O) has been determined to
be 17.7. To the best of our knowledge, the corresponding pKa of 21 has yet to
be determined: Campbell, C. D.; Duguet, N.; Gallagher, K. A.; Thomson,
J. E.; Lindsay, A. G.; O’Donoghue, A. C.; Smith, A. D. Chem. Commun.
2008, 3528.
Recently, we reported that hydrogen bonding could be
utilized as a control element in enantioselective homo-
benzoin condensation reactions.33,34 The pentafluorophenyl-
substituted triazolium precatalyst 5 (which followed our
first generation system 4) could promote the formation of
J. Org. Chem. Vol. 76, No. 2, 2011 349

O’Toole et al.
JOCFeatured Article
benzoin (7) from benzaldehyde (6) with excellent efficiency
and stereocontrol (Scheme 2A). We postulated that the rigid,
hindered nature of 5, coupled with the presence of a catalyst
hydrogen bond donating group,35 would allow the catalyst to
potentially distinguish between the two aldehyde electrophiles
based on the recognition of two different substrate properties:
steric bulk and Brønsted basicity. However, as an important
further control element in addition to catalyst control,36 we
envisaged that a degree of synergistic substrate control could
be brought to bear on the process in the form of a removable
chemoselectivity-enhancing substituent in the ortho posi-
tion of the aromatic aldehyde. Stetter et al.20 had reported that
o-chloro-substituted benzaldehydes participated in more
selective crossed AC reactions than their unsubstituted counter-
parts. Thus, weproposed that a halogen atom could be used in
this capacity, which could later be either easily removed by
hydrogenolysis or utilized as a functional handle in further
structural elaboration of the product (Scheme 2C).
TABLE 1. Short study on the general applicability of known direct
cross-acyloin procedures using branched and unbranched aliphatic alde-
hydes with o-chloro benzaldehyde as aromatic aldehyde partner
Before testing this hypothesis, we wished to be certain that
highly chemoselective direct crossed acyloin chemistry was
not possible using existing technology. For instance, while
Buck et al.18 did not employ aliphatic aldehydes in their study
and Stetter20 utilized a large excess of one aldehyde component,
we felt it prudent to first examine these two protocols to ensure
that the dearth of a chemoselective protocol in the literature is
not due to either a simple oversight or omission by previous
researchers. Aiming for a widely applicable, practical procedure
that would also allow for the implementation of higher
advanced aldehyde building blocks, we focused on condi-
tions which would not employ a large excess of either
coupling partners (i.e., aliphatic aldehyde e1.7 equiv).37
Accordingly, we carried out experiments examining the
reaction between an aliphatic and an aromatic aldehyde
under the influence of either cyanide (conditions defined by
Buck et al.)18 or thiazolium-derived carbene (conditions
defined by Stetter et al.)20 catalysis (Table 1). In the pre-
sence of cyanide ion (65 mol %), the reaction between 9
and isobutyraldehyde (10) produced a poor yield of 12 as
the major product, while reaction with the unbranched
hydrocinnamaldehyde (11) failed to generate cross-product
at all (entries 1 and 2). Utilization of Stetter’s conditions
proved somewhat more successful, with the isolation of
36-38% yields of a “major” cross-product possible when
1-1.7 equiv of the aliphatic aldehyde component was em-
ployed (entries 3-6).
aFor details, see Supporting Information.
if unbranched aldehydes are employed. This encouraged us
to return our attention to the original proposal involving the
use of triazolium salt-derived systems.
Before attempting to examine the potential of hydrogen
bonding as a control element in these reactions,36 we first wished
to orient ourselves with respect to the natural bias (if any) a
triazolium-derived carbene devoid of protic substituents would
display toward one of the coupling partners in a crossed AC
reaction. As a model process, we chose the AC reaction between
a range of substituted benzaldeyhdes 9 and 15-20 (of variable
steric and electronic characteristics) and the relatively unhin-
dered hydrocinnamaldehyde (11) in the presence of the
achiral precatalysts 8 or 21 and base (i.e., conditions which
had proven conducive to the promotion of homobenzoin
condensation reactions in our previous studies33,34a). The
results of these experiments are outlined in Table 2.
As expected, the pentafluorophenyl-substituted catalyst 8
proved to be a superior system to 21 under these conditions
(entries 1 an 2).34 The coupling of benzaldehyde (6) and 11
proceeded with poor chemoselectivity; while a marked prefer-
ence for the formation of products derived from the aliphatic
Breslow intermediate (e.g., via initial attack of the catalyst on
11; i.e., 22a þ 22d vs 22b þ 22c) was observed, all four possible
products (homodimers 22a/22b and crossed products 22c/22d)
were formed without any one being present at synthetically
useful levels. The activation of the aromatic aldehyde compo-
nent with a chlorine atom in either the m- or p-position failed to
influence chemoselectivity to any appreciable extent (entries 3
and 4), and the preference for the cross-coupled products D was
slightly improved, at the expense of the formation of increased
From an analysis of the results of this study, a clear picture
of the limited potential of the current benchmark protocols
for the direct crossed AC reaction emerged: both do not
tolerate reduced amounts of the aliphatic aldehyde and also
fail to produce synthetically useful amounts of cross-product
(35) For a recent report on the synthesis of a bifunctional thiourea carbene
catalyst, see: Brand, J. P.; Siles, J. I.; Osuna; Waser, J. Synlett 2010, 881.
(36) Our investigations on catalyst control are ongoing and will be
reported in due course.
(37) Conducting the experiment with an excess of 3 equiv of the aliphatic
aldehyde (here, isobutyraldehyde) provides a main cross-product23 in 64%
yield in accordance with the results of Stetter et al.;20,23 however, application
of similar conditions to unbranched aldehydes such as 11 clearly revealed the
limitations of the process as this;even in combination with the potentially
“privileged” o-substituted aryl aldehydes;only yields a major product in
40% yield (together with significant amounts of the homodimer of
hydrocinnamaldehyde), providing some evidence for the requirement of
both large excess of aliphatic aldehydes and sterical demand of the aryl
aldehyde for successful transformations.
350 J. Org. Chem. Vol. 76, No. 2, 2011

O’Toole et al.
JOCFeatured Article
TABLE 2. Crossed AC Reactions with Unbranched Aldehydes: Preliminary Experiments
aYield determined by 1H NMR spectroscopy using styrene as an internal standard. Note: yields of 13 and 22-28a and 22-28b account for the 2:1
stoichiometry. To obtain the mol % of these materials, divide the yield by 2. bPhenyl-substituted triazolium precatalyst 21 was used instead. cWith 8 mol %
catalyst. dWith 10 mol % catalyst loading. eWith 1.3 equiv of 20 and 8 mol % catalyst. fWith 1.5 equiv of 20 and 8 mol % catalyst. gWith 1.3 equiv of
11 and 8 mol % catalyst. hWith 1.5 equiv of 11 and 8 mol % catalyst. iWith 1.7 equiv of 11 and 8 mol % catalyst.
amounts of the aryl homobenzoins B. However, the use of the
o-substituted analogue 9 generated 13d as the dominant pro-
duct in moderate yield (entry 5). Further investigation revealed
that the improved chemoselectivity associated with the use of
o-substituted aldehydes is primarily related to the steric require-
ment of the substituent, although its electronic characteristics
do also seem to play a minor role. For instance, the small but
highly electronegative fluorine atom does not confer high
chemoselectivity (entry 6); however, use of larger units such
as the electron-releasing methoxy and the electron-withdrawing
trifluoromethyl substituents (entries 7 and 8, respectively)
allows relatively selective crossed AC reactions to occur,
with the latter suppressing the pathways leading to 27a-c to
the extent that 27d was formed in 81% yield.
coupling reactions, etc.), while second, as mentioned earlier
(vide supra), we envisaged that it should be possible to cleanly
remove the halogen from the product, which allows one
to aspire toward the use of an o-bromo substituent as a re-
movable tool to control chemoselectivity in these processes,
thereby providing access to products (after debromination)
which would be otherwise difficult to prepare in good yield
via carbene-catalyzed crossed AC chemistry. It was found that
20 coupled to 11 with very good chemoselectivity and moderate
yield initially (entry 9), which could already be drastically
improved by employing 8 mol % of catalyst 8 (entry 10).
Subsequent optimization of the reaction conditions (entries
10-16) allowed the synthesis of 28d in 90% yield by employ-
ing a small excess of 11 (1.7 equiv) in the presence of 8 mol %
of 8.38
Particularly gratifying was the performance of the o-bromo
derivative 20. This coupling partner is of considerable potential
interest for two reasons: first, the bromine atom in the product
(i.e., 28d) can serve as a functional handle for further elaboration
(radical generation, participation in transition-metal-catalyzed
(38) The concurrently formed homodimer 28a can be easily separated by
column chromatography. In our subsequent studies employing unbranched
aliphatic aldehydes in 1.7 equiv (Table 3 and Table 4), we only report the
yields of the cross-product D for clarity reasons.
J. Org. Chem. Vol. 76, No. 2, 2011 351

O’Toole et al.
JOCFeatured Article
TABLE 3. Evaluation of Substrate Scope: Unbranched Aldehydes
Entry
‘Aliphatic’ Aldehyde
X
‘Aromatic’ Aldehyde
Product
Yield D38 (%)
1
2
3
4
5
29
29
30
30
30
31
31
11
11
32
32
1.7
10.0
1.7
1.7
2.5
1.7
1.7
1.7
1.7
1.7
1.7
19
19
19
20
20
19
20
19
19
19
20
33
33
34
35
35
36
37
27d
27d
38
50a
78
79
68a
73
6
7
84
77
8b
9c
10
11
60a
86
81
76
39
aYield determined by 1H NMR spectroscopy using styrene as an internal standard. bAt 5 °C, 17 and 16% yields of homo- and heterocoupling products
(B and C), respectively, derived from initial attack of the catalyst on 19 were obtained. cAt 18 °C, 10% yields of both homo- and heterocoupling products
(B and C) derived from initial attack of the catalyst on 19 were obtained.
TABLE 4. Exploitation of a Removable 2-Bromo Substituent
The scope of the process with respect to the “aliphatic”
or “umpolung” aldehyde component was next investigated.
o-Substituted electrophiles 19 and 20 were coupled to a range
of unbranched aldehydes 11 and 29-32 under our optimized
conditions in the presence of catalyst 8 at room temperature
(Table 3). Acetaldehyde (29) proved a challenging substrate
to utilize at ambient temperature due to its low boiling point
(entry 1); however, use of a 10-fold excess (feasible due to the
low cost of this reagent) resulted in good yield of its cross-
product with 19 (i.e., 33, entry 2). The less volatile un-
branched aldehydes n-propanal (30, entries 3-5), n-pentanal
(31, entries 6 and 7), hydrocinnamaldehyde (11, entries 8
and 9), and phenylacetaldehyde (32, entries 10 and 11) could
be efficiently coupled to either 19 or 20 with good product
yields without difficulty using a smaller excess of 1.7-2.5
equiv. It is perhaps interesting to note that coupling of 11 (1.7
equiv) to 19 at 5 °C is less chemoselective than an otherwise
identical reaction at 18 °C (entries 8 and 9). In the case of the
reaction at the lower temperature, 27d was still obtained as
the major product; however, significantly elevated levels of
products derived from initial attack of the catalyst on 19 (i.e.,
27b and 27c) were detected, indicating that these coupling
reactions may proceed under a significant degree of thermo-
dynamic control.39
To demonstrate the potential of the use of an ortho-bromo
substituent as a solution to circumvent the inherent lack of
chemoselectivity in crossed AC reactions involving aromatic
aldehydes and unbranched aliphatic aldehydes, we carried
out the coupling of a variety of o-bromobenzaldehydes (20
and 40-42) equipped with both electron-neutral (entry 1),
electron-donating (entries 2 and 3), and electron-withdrawing
(entry 4) substituents with 11 (Table 4). Good to excellent
yields of coupled products were obtained in each case under stan-
dard conditions. Adducts 28d and 43-45 were then smoothly
and conveniently debrominated under an atmosphere of
hydrogen in the presence of Pd/C to give hydroxyketones 22d
and 46-48, respectively, in uniformly excellent yields. Thus,
we would submit that the o-bromo substituent can be employed
as a temporary directing group which can first divert the course
of an otherwise relatively unselective (see Table 1, entry 2 vs
entries 9 and 16) coupling reaction toward the formation of a
single major product (irrespective of the overall electronic
nature of the aromatic aldehyde coupling partner) and then
either serve as a functional handle if required or be cleanly
(39) Carrying out the coupling reaction at temperatures higher than 18 °C
invariably led to lower overall yields due to decomposition of the
o-trifluoromethylbenzaldehyde. See also the results outlined in Scheme 4.
352 J. Org. Chem. Vol. 76, No. 2, 2011

O’Toole et al.
JOCFeatured Article
removed to give debrominated products not otherwise accessible
in high yield directly from a operationally simple carbene-
catalyzed AC process.
55 and 56, respectively, in good yield. o-Iodobenzaldehyde
(52) is also compatible with the methodology (the first
time this aldehyde has been evaluated as a AC substrate,
product 57). The use of the interesting substrate 50 led to
the formation of the densely functionalized cyclopropyl-
substituted ketone 58 in 80% yield. 2-Methylpropanal (10)
proved a challenging substrate due to its low boiling point
but could still be converted to 12 and 59 in good yields in the
presence of 9 and 20, respectively, while at this stage, it appears
that pivaldehyde is too bulky as substrate to form a nucleophilic
Breslow intermediate under these conditions. Importantly,
this cross-coupling procedure employing branched aldehydes
does not require an excess of one of the coupling partners, thus
providing a catalytic and relatively waste-free, selective access
to such valuable cross-acyloin products.
The question regarding the origin of the chemoselec-
tivity observed is both intriguing and difficult to defini-
tively answer at this juncture. It is reasonable to assume
that the presence of the o-substituent retards the rate of
attack of the carbene on the aromatic aldehyde, resulting
in increased concentrations of the Breslow intermediate
derived from initial attack on the aliphatic aldehyde. What
is unclear is why this intermediate (rather counterintuitively)
then prefers to react with the presumably more hindered
o-substituted benzaldehyde over another molecule of ali-
phatic aldehyde.
For this methodology to be genuinely synthetically useful,
its scope with respect to the “nucleophilic” component must
not be limited to unbranched aldehydes. Initial experiments
involving the coupling of R-substituted aldehydes at room
temperature resulted in poor conversion (<50%) even after
prolonged reaction times. At 60 °C, however, these sub-
strates will participate in efficient crossed AC reactions
(Scheme 4). The reaction between cyclohexane carbaldehyde
(49) and o-anisaldehyde (18) furnished 53 in good yield. We
found that o-trifluoromethyl benzaldehyde 19 performed
unsatisfactorily under these conditions (the stability of prod-
uct 54 under the reaction conditions appears to be problematic,
vide infra;39 product 54). o-Bromobenzaldehydes 20 and 40
both coupled with cyclohexane carbaldehyde (49) to afford
SCHEME 4. Evaluation of Substrate Scope: Branched Aldehydes
There are several possible explanations, such as a π-iminium
interaction in the developing TS as the enolamine attacks the
aromatic aldehyde,27,40 a stabilizing (and selectively formed)
hydrogen bond between the more basic aromatic aldehyde
carbonyl oxygen and the enolamine hydroxyl group, or per-
haps most importantly given the supporting evidence un-
covered as this study progressed;a degree of thermo-
dynamic product control. On the basis of the well-established
mechanistic picture of acyloin/benzoin condensations,10e,41
it is also reasonable to assume that the properties of the
Breslow intermediate should also be dependent to a sig-
nificant extent on the nature of the catalyst it is derived
from. What is certain is that these issues are now ripe for
investigation.
We were first interested in examining the influence of the
choice of catalyst on the outcome of these reactions from a
chemoselectivity perspective. We challenged our optimized
catalytic system (condition set A, Scheme 5) with the reaction
between benzaldehyde, lacking the selectivity-controlling
SCHEME 5. Influence of the Catalytic System on Chemoselectivitya
a Yields determined by 1H NMR spectroscopy using stilbene as internal standard. See Supporting Information for details.
J. Org. Chem. Vol. 76, No. 2, 2011 353

O’Toole et al.
JOCFeatured Article
SCHEME 6. Crossover Experiments under Optimized Conditions Using Triazolium Precatalyst 8
ortho-substituent, and isobutyraldehyde and then repeated
the experiment under Stetter’s conditions (condition set B,
Scheme 5), employing a thiazolium-derived carbene instead
of the triazolium precatalyst. In contrast to the results re-
ported by Stetter,20 in our hands, condition set B provides
both cross-products and the homoarylbenzoin in a ca. 1:1:1
ratio (combined yield of cross-acyloin products: 54%).42 Use
(40) Duddling, T.; Houk, K. N. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 5770.
(41) White, M. J.; Leeper, F. J. J. Org. Chem. 2001, 66, 5124.
354 J. Org. Chem. Vol. 76, No. 2, 2011

O’Toole et al.
SCHEME 7. Investigation of the Use of Cross-Products Derived from Activated Yet Unhindered Aldehydes
JOCFeatured Article
of condition set A on the other hand, involving the triazo-
lium catalyst 8, results in remarkably high selectivity for the
cross-coupled acyloin D, which could be isolated in 61%
yield. Neither the cross-coupled C nor homobenzoin B was
formed in significant amounts. It is therefore clear that the
catalyst exerts a significant degree of control over the process
from a chemoselectivity standpoint.
A number of conclusions can be drawn from these reac-
tions:
Aliphatic and o-substituted benzaldehydes dimerize, but
do so only slowly (Expts 1, 2, 4, and 5). This is central to
attaining high chemoselectivity in these processes. The
homodimers derived from aliphatic aldehydes do not parti-
cipate in retroacyloin chemistry under these conditions and
are essentially formed irreversibly (Expts 1 and 3). Unhin-
dered benzoins (i.e., homodimers of aromatic aldehydes) will
participate in retroacyloin chemistry under these conditions,
whereas o-substituted isomers will not (Expts 2-4). The
R-arylketone cross-product (the major product under our
conditions) from the reaction of an aliphatic and an aromatic
aldehyde undergoes retroacyloin either slowly (Expt 5) or
not at all (Expt 6). The crossed acyloin reactions involving
unhindered benzaldehydes are subject to a far greater degree
of thermodynamic control than those involving hindered
analogues (Expts 1-6). Given that the energy differences
between the acyloin products is likely to be small, this results
in relatively unselective reactions where benzaldehydes devoid
of o-substitution are employed.43
To support the theory that cross-products derived from
reactions involving activated, unhindered benzaldehydes are
more amenable to retroacyloin reactions (and hence are
formed in lower yields), we synthesized 48 and subjected it
to the reaction conditions in the presence of pentanal (31).
We were pleased to observe increased levels of products
derived from the retroacyloin reaction of 48 relative to those
observed using benzaldehyde as the reacting partner (see
Scheme 7 and Expt 6, Scheme 6). Thus, it is clear that the
reversibility of the process is also influenced by the electronic
nature of the benzaldehyde partner, with more activated
aldehydes participating in less chemoselective reactions.
Overall, it is clear that crossed-coupling is facilitated
by the slow dimerizability of the aliphatic aldehyde and
o-substituted benzaldehydes. Given that none of the pro-
ducts derived from the cross-coupling of these aldehydes
could demonstrably participate in retroacyloin reactions,
why cross-coupling is faster than dimerization and why the
R-arylketone cross-product is favored over the other still
require explanation. We would propose that it is reasonable
to assume that initial attack of the carbene on the aliphatic
aldehyde is preferred on electronic grounds; that is, the more
In an attempt to further shed light on the origins of the
observed chemoselectivity, a number of crossover experi-
ments were carried out. The results of these experiments
are outlined in Scheme 6. In the first instance, we wished to
establish the degree of reversibility of these processes. We
therefore treated the o-substituted aldehyde 20 with the
catalyst 8 under our standard conditions in the presence of
homodimer 61 (Expt 1). The slow dimerization of 20 was
observed, but no products (such as 28d) derived from the
retroacyloin of 61 could be detected. Next, we investigated
the opposite pairing of starting materials, that is, an aliphatic
aldehyde 11 and a homodimer derived from a (para-
substituted) aromatic aldehyde (i.e., 23b, Expt 2). Interest-
ingly, this experiment afforded significant amounts of the
cross-product 23d, along with free aldehyde 15 (which also
stems from a retroacyloin reaction) and the homodimer 61.
When the experiment was repeated, where the aliphatic
aldehyde 11 was replaced with its homodimer 61 (Expt 3),
again retroacyloin of 23b was observed, but in this case, no
coupling to form cross-product 23d occurred. These results
seemed to indicate that the benzoin 23b is able to revert to its
parent aldehyde under the reaction conditions, whereas the
homodimer 61 derived from hydrocinnamaldehyde is not.
To probe this further, the ortho-isomer of benzoin 23b (i.e.,
13b, Expt 4) was treated with an aliphatic aldehyde 31 in the
presence of the catalyst. Gratifyingly, no cross-product 63
was detected in this experiment (Expt 4).
Thus it would appear that the o-substituted 13b is more
stable toward the catalyst than its p-isomer 23b, which,
together with the rather slow rate of dimerization of
o-bromobenzaldehyde (20) and hydrocinnamaldehyde (11)
and the reluctance of the aliphatic homodimer 61 to undergo
a retroacyloin reaction, goes some way toward explaining the
chemoselectivity observed in these processes.
In a similar crossover experiment involving the cross-
product 22d and the aliphatic aldehyde 31, we observed trace
amounts of benzoin (22b, which could only arise from the
retroacyloin of 22d) and homodimer 62 (Expt 5). No cross-
product 64 was detected. Finally, exposure of the cross-
product 28d derived from reaction of o-bromobenzaldehyde
(20) and hydrocinnamaldehyde (11) to the catalyst under
standard conditions failed to produce any products.
(43) Calculation of the free energy of the two regioisomers of the cross-
acyloin products of o-chlorobenzaldehyde (9) and isobutyraldehyde (10)
revealed a preference for the R-aryl ketone 13d as compared to the aromatic
ketone 13c by 3.60 kcal/mol. Similar results were obtained for the calculation
using unbranched n-propanal as aliphatic aldehyde for the cross-acyloins
(lower energy for the R-aryl ketone vs aromatic ketone: 3.43 kcal/mol).
Calculations of the free energy were performed with SPARTAN-06 by
geometry optimization using Hartree-Fock methods (6-31G) starting from
the corresponding minimun conformer as obtained in a PM3-based Monte
Carlo conformer search.
(42) Stetter et al.20 report a combined yield of 56% for the cross-acyloins
C and D in a ratio of 1.9:1; no information is provided on the amount of
homobenzoin B.
J. Org. Chem. Vol. 76, No. 2, 2011 355

O’Toole et al.
JOCFeatured Article
SCHEME 8. Chemo- and Enantioselective Crossed Acyloin Condensation
electron-rich benzaldehyde carbonyl moieties make for
poorer electrophiles in the first step of the catalytic cycle.
This is supported by the observation that the use of more
activated, halogen-substituted benzaldehydes generates
greater levels of homobenzoin products derived from initial
attack on the benzaldehyde moiety (see entries 3-6 and 9,
Table 2). In the case of o-substituted benzaldehydes, this
preference for the aliphatic partner as the initial site of attack
would obviously be exaggerated for steric reasons; this
argument can be underlined by the decreasing amount of
these homobenzoins detected with increasing size of the
ortho-halogen substituent (o-F, o-Cl, o-Br with 45, 15, and
9% of homocoupled product; see entries 5, 6, and 9, Table 2).
It is difficult to establish why the BI then prefers to attack
the hindered aromatic aldehyde over another molecule of
aliphatic aldehyde. In a natural product synthesis study
involving an intramolecular AC step, Miller24 has suggested
that a stabilizing interaction between orthogonally aligned
carbonyl and aromatic moieties known to exist in R-phenyl
ketones (in cases where it is stereoelectronically permitted)
may influence the chemoselective outcome of AC reactions
between aliphatic and aromatic aldehyde components. It is
tempting to draw parallels in this study, that is, that the
observed preference for the R-arylketone cross-product over
the R-substituted aromatic ketone analogue is related to the
contribution of this interaction, which presumably results in
greater reversibility of the latter cross-product over the former.
However, it should be pointed out that the seemingly logical
extension of this argument to account for the preference for
cross-product formation over aliphatic aldehyde dimerization
is less sound at this juncture since we could not observe any
retroacyloin chemistry involving the aliphatic dimers.
What can be safely inferred is that chemoselectivity in
these processes is not governed by a single factor alone but
rather a confluence of factors depending on the catalyst
employed and the steric and electronic nature of the reac-
tants. That being said, it is clear that one can achieve high
selectivity in the diverse array of AC reactions examined in
this study by using catalyst 8 in the presence of an aromatic
aldehyde incorporating an (removable) o-bromo substitu-
ent, irrespective of other substrate characteristics.
While the methodologies outlined above allow one to
carry out highly chemoselective crossed AC reactions using
a combination of catalyst properties and the steric effects of
the substrates, the ability to control the stereochemical out-
come of these reactions is of course the ultimate goal. To this
end, 19 was coupled with 30 in the presence of the bifunctional
chiral triazolium salt 5 (10 mol %) to afford the expected
product 34 in good yield and enantiomeric excess (Scheme 8).
The novel precatalyst 67, which possesses a larger diaryl-
carbinol unit, is less active than 5 but promoted the same
reaction with improved enantioselectivity (81% ee). While
this aspect of the study is currently at an early stage of
development, it is clear from analysis of these preliminary
data that the process is amenable to the efficient transfer of
stereochemical information from catalyst to product.
Conclusions
In summary, we have developed the first chemoselective
intermolecular crossed AC reactions between two aldehyde
partners involving triazolium precatalysts. A key discovery is
the use of an o-bromo substituent as a temporary chemo-
selectivity-controlling group which can subsequently be re-
moved in high yield. The methodology is of very broad
scope: hindered, activated, and electron-rich aromatic alde-
hydes are compatible, as are both unbranched and more
hindered branched aliphatic aldehydes. Importantly, unlike
the previous benchmark study in the literature involving a
thiazolium catalyst, in these reactions, the expected product
from the cross-coupling of two aldehydes can be confidently
predicted beforehand, and the methodology is generally
complementary to existing methodologies based on enzy-
matic catalysis. The feasibility of highly enantioselective
crossed AC reactions has also been established;investigations
aimed at further refining the asymmetric catalysis and
elucidating the origins of the chemoselectivity are now
underway.
Experimental Section
General Methods. Unless otherwise noted, all commercially
available compounds were used as provided without further
purification.
NMR spectra were recorded on 300 MHz (300.13 MHz),
400 MHz (400.13 MHz), or 600 MHz (600.13 MHz) spectro-
meters using the solvent peak as internal reference (CDCl3 δ H
7.26, δ C 77.0; and DMSO-d6 δ H 2.51, δ C 39.5). Multiplicities
are indicated, s (singlet), d (doublet), t (triplet), q (quartet), quint
(quintet), sept (septet), m (multiplet); coupling constants (J) are
in Hertz (Hz). Mass spectra (MS ESI) were recorded using a
mass spectrometer equipped with a TOF analyzer. All reactions
were monitored by thin-layer chromatography using gel plates
60 F254; visualization was accomplished with UV light and/or
staining with appropriate stains (KMnO4, anisaldehyde, vani-
line, ninhydrin, or phosphomolybdic acid). Standard flash
chromatography procedures were followed (particle size 40-
63 μm). Infrared spectra were obtained using neat samples on
spectrometers equipped with a universal ATR (attenuated total
reflectance) sampling accessory. Optical rotation measurements
are quoted in units of 10-1 deg cm2 g-1. Analytical CSP-HPLC
was performed using an OJ-H (4.6 mm ꢀ 25 cm) column.
Tetrahydrofuran was distilled from sodium/benzophenone. All
reactions were carried out under a protective atmosphere of dry
nitrogen or argon using oven-dried glassware unless otherwise stated.
356 J. Org. Chem. Vol. 76, No. 2, 2011

O’Toole et al.
JOCFeatured Article
Catalyst 5 was prepared as per the procedure described by
Connon and Zeitler et al.34 and catalyst 8 as per the method
described by Rovis.44
0.36 mmol), 10% Pd/C (6.8 mg), and MeOH (6.0 cm3): 1H NMR
(600 MHz, CDCl3) δ = 7.18-7.06 (m, 3H), 6.97 (d, J = 7.0 Hz,
2H), 6.77 (d, J = 2.1 Hz, 2H), 6.59 (s, 1H), 4.91 (d, J = 4.1 Hz,
1H), 4.18 (d, J = 4.1 Hz, 1H), 3.81 (s, 3H), 3.75 (s, 3H), 2.87-2.69
(m, 2H), 2.60 (t, J = 7.5 Hz, 2H); 13C NMR (150 MHz, CDCl3)
δ = 209.0, 149.5, 149.3, 139.7, 130.1, 128.3, 128.0, 126.2, 120.5,
111.5, 109.8, 79.8, 55.8, 55.7, 39.4, 29.6; HRMS (C15H20O3 þ Na)
calcd 323.1259, found 323.1244; IR (cm-1) ν~ = 3448, 3064, 3028,
2927, 1715, 1603, 1580, 1496, 1467, 1454, 1404, 1361, 1264, 1229,
1147, 1100, 1079, 1055, 1025, 964, 875, 812, 745.
General Procedure for the Crossed Acyloin Condensation with
r-Branched Aldehydes. A flame-dried Schlenk flask equipped
with a magnetic stirring bar was charged with Rb2CO3 (12 mg,
0.05 mmol). The reaction vessel was evacuated, heated to 650 °C
for 1 min, and cooled to rt under N2. After this procedure had
been repeated once, triazolium precatalyst 8 (18 mg, 0.05 mmol)
was added. The precatalyst/base mixture was put under vacuum
and dried at rt for 45 min. Subsequently, absolute THF (0.45 mL)
was added, and the resulting mixture was allowed to stir at rt
for 15 min. After sequential addition of the appropriate
benzaldehyde (0.5 mmol) and R-branched aldehyde (0.5 mmol),
the reaction vessel was equipped with a reflux condenser and
stirred at 60 °C under an atmosphere of N2. After 20 h, the reaction
mixture was allowed to cool to rt, diluted with 5 mL of CHCl3, and
extracted with H2O. The aqueous layer was back-extracted twice
with 5 mL of CHCl3, and the combined organic layers were washed
with brine, dried over anhydrous Na2SO4, and filtered. Removal of
the solvent under reduced pressure affords the crude product,
which was purified by column chromatography.
General Procedure for the Crossed Acyloin Condensation with
Unbranched Aldehydes. To a 5 cm3 round-bottomed flask,
equipped with a magnetic stirring bar, was added Rb2CO3 (99.8%)
that had been finely ground using a mortar and pestle. The reaction
vessel was put under vacuum and heated with a heat gun for four 1
min intervals. When cooled to ambient temperature, the appropriate
catalyst (0.09 mmol) was added and the flask was fitted with a
septum seal. The flask was evacuated for 1 min and put under an
atmosphere of Ar. The required aldehydes were distilled under
vacuum and used directly. THF was charged to the reaction,
followed by consecutive addition of each aldehyde. The reaction was
stirred at room temperature for 40 h. CH2Cl2 (3.0cm3) and deionized
H2O (3.0 cm3) were added. The organic layer was removed, and the
aqueous layer was washed with CH2Cl2 (4 ꢀ 3.0 cm3). The organic
layers were combined, dried over MgSO4, filtered, and the solvent
was removed under reduced pressure. The product was purified using
column chromatography.
1-Hydroxy-4-phenyl-1-(2-trifluoromethyl-phenyl)-butan-2-one
(27d). Yield 275 mg; 81%, colorless to pale yellow oil; Rf
(CH2Cl2/hexanes 3/2) 0.25. Procedure used Rb2CO3 (20.3 mg,
0.09 mmol), 8 (31.9 mg, 0.09 mmol), R,R,R-trifluoromethyl
benzaldehyde (19) (145.1 μL, 1.10 mmol), hydrocinnamalde-
hyde (11) (246.2 μL, 1.87 mmol), and THF (610 μL): 1H NMR
(600 MHz, CDCl3) δ 7.74 (d, 1H, J = 7.5 Hz), 7.51-7.45 (m,
2H), 7.26-7.23 (app. t, 2H), 7.18 (t, 1H, J = 7.4 Hz), 7.12 (d,
1H, J = 7.5 Hz), 7.06 (d, 2H, J = 7.5 Hz), 5.48 (s, 1H), 4.39 (s
(broad), 1H), 2.95-2.86 (m, 2H), 2.75-2.70 (m, 1H), 2.58-2.53
(m, 1H); 13C NMR (150 MHz, CDCl3) δ 207.6, 139.8, 136.3,
132.5, 128.8, 128.8 (q, J = 30.2 Hz), 128.6, 128.4, 128.0, 126.2,
2-(2-Bromophenyl)-1-cyclohexyl-2-hydroxyethanone (55). Yield
117 mg; 79%, colorless to pale yellow oil; Rf (10%Et2O/n-pentane)
0.18; 1H NMR (300 MHz, CDCl3) δ 7.66-7.57 (m, 1H), 7.31 (td,
J = 7.5 and 1.3 Hz, 1H), 7.24-7.14 (m, 2H), 5.71 (d, J = 4.7 Hz,
1H), 4.45 (d, J = 4.7 Hz, 1H), 2.46 (tt, J = 11.3 and 3.4 Hz, 1H),
126.1 (q, J = 5.5 Hz), 124.0 (q, J = 274.1 Hz), 74.5, 39.4, 29.3; 19
F
NMR (376 MHz, CDCl3) δ -57.59; HRMS (C17H15F3O2þNa)
calcd 331.0922, found 331.0923; IR (cm-1) ν~ = 3456, 3031, 2930,
1717, 1605, 1497, 1454, 1311, 1159, 1108, 1034, 768, 747, 696.
1-(2-Bromo-phenyl)-1-hydroxy-4-phenyl-butan-2-one (28d). Yield
316 mg; 90%, colorless to pale yellow oil; Rf (CH2Cl2/hexanes 2/3)
0.17. Procedure used Rb2CO3 (20.3 mg, 0.09 mmol), 8 (31.9 mg,
0.09 mmol), ortho-bromobenzaldehyde (20) (128.4 μL, 1.10 mmol),
hydrocinnamaldehyde (11) (246.2 μL, 1.87 mmol), and THF
(630 μL): 1H NMR (600 MHz, CDCl3) δ 7.61 (d, 1H, J = 8.1 Hz),
7.28 (t, 1H, J= 7.3 Hz), 7.27-7.24 (app. t, 2H), 7.23-7.17 (m, 3H),
7.09 (d, 2H, J = 7.3 Hz), 5.59 (s, 1H), 4.39 (s (broad), 1H), 2.98-
2.93 (m, 1H), 2.88-2.82 (m, 2H), 2.70-2.64 (m, 1H); 13C NMR
(150 MHz, CDCl3) δ 207.8, 139.9, 137.1, 133.3, 130.0, 129.0, 128.4,
128.1, 128.0, 126.2, 123.7, 78.3, 39.4, 29.5; HRMS (C16H15BrO2 þ
Na) calcd 341.0153, found 341.0152; IR (cm-1) ν~ = 3456, 3063,
3028, 2925, 1713, 1536, 1496, 1454, 1024, 749, 698, 672.
General Procedure for the Hydro-debromination of Crossed
Acyloin Condensation Products. To a 25 cm3 round-bottomed
flask, equipped with a stirring bar, were added brominated
crossed acyloin (0.3 mmol), 10% Pd/C (6 wt %), Et3N (1.2 equiv),
and MeOH (0.36 M). The flask was evacuated and then purged
with N2. This cycle was performed twice. The reaction was stirred
vigorously at ambient temperature (ca. 20 °C) under an atmosphere
of hydrogen (hydrogen generator) overnight. The reaction mixture
was filtered and concentrated in vacuo. The residue was partitioned
between CH2Cl2 (10 cm3) and H2O (10 cm3), and the organic layer
waswashedwithbrine(10cm3), dried over MgSO4, filtered, and the
solvent removed under reduced pressure. The product was purified
using column chromatography.
2.04-1.92 (m, 1H), 1.84-1.55 (m, 3H), 1.51-0.96 (m, 6H); 13
C
NMR (75.5 MHz, CDCl3) δ = 211.9, 137.4, 133.4, 130.1, 129.2,
128.1, 124.1, 76.6 (overlapping with CDCl3 resonance), 46.2,
29.9, 27.7, 25.7, 25.5, 25.0; MS (EI) m/z (%) 217 (13) [M - Br]þ,
185 (19), 111 (52) [C6H11CO]þ, 83 (100) [C6H11]þ; HRMS
(C14H17BrO2 þ H) calcd 297.0490, found 297.0497; IR (cm-1
)
ν~ = 3452, 2930, 2854, 1707, 1470, 1448, 1369, 1313, 1192, 1127,
1058, 1023, 993, 754, 632, 537, 495.
2-(2-Bromophenyl)-1-cyclopropyl-2-hydroxyethanone (58). Yield
102 mg; 80%, colorless oil; Rf (10%Et2O/n-pentane) 0.13; 1HNMR
(300 MHz, CDCl3) δ 7.65-7.59 (m, 1H), 7.38-7.29 (m, 1H),
7.25-7.17 (m, 2H), 5.78 (d, J = 4.1 Hz, 1H), 4.45 (d, J = 4.1 Hz,
1H), 1.94 (tt, J = 7.5 and 4.9 Hz, 1H), 1.27-1.15 (m, 1H), 1.11-
0.96 (m, 2H), 0.90-0.79 (m, 1H); 13C NMR (75.5 MHz, CDCl3) δ
209.1, 137.6, 133.3, 130.1, 129.4, 128.2, 124.2, 78.7, 17.6, 12.8, 12.6;
MS (EI) m/z (%) 185 (56) [M - C4H5Oþ], 175 (31) [M - Br]þ, 77
(43) [C6H5]þ, 69 (100) [C4H5O]þ; HRMS (C11H11BrO2 þ H)
calcd 255.0021, found 255.0012; IR (cm-1) ν~ = 3450, 3009, 1695,
1568, 1471, 1438, 1373, 1271, 1193, 1134, 1051, 1017, 907, 757, 732,
672, 632, 531, 499.
Acknowledgment. This work emanates from a study funded
by Science Foundation Ireland (SFI). Financial support of
the DFG (Deutsche Forschungsgemeinschaft) and the Irish
Research Council for Science Engineering and Technology
(IRCSET) is gratefully acknowledged.
Supporting Information Available: Experimental details, gen-
eral procedures and spectroscopic and analytical data. This material
is available free of charge via the Internet at http://pubs.acs.org.
1-(3,4-Dimethoxyphenyl)-1-hydroxy-4-phenylbutan-2-one (46).
Yield 87 mg; 96%, yellow oil; Rf (hexanes/CH2Cl2 1/4) 0.9.
Procedure used 43 (113.7 mg, 0.30 mmol), NEt3 (50.1 μL,
Note Added after ASAP Publication. The Acknowledgment
was updated in the version reposted November 2, 2010.
(44) Kerr, M. S.; Read de Alaniz, J; Rovis, T. J. Org. Chem. 2005, 70,
5725.
J. Org. Chem. Vol. 76, No. 2, 2011 357