Organometallic reactions in aqueous media - Bismuth-mediated crossed aldol type reactions

Article abstract of DOI:10.1139/v03-142

Bismuth metal, upon activation by zinc fluoride, can effect the crossed aldol reaction between α-bromocarbonyl compounds and aldehydes in aqueous media. The reaction was found to be regiospecific and syndiastereoselective.

Full text of DOI:10.1139/v03-142

1406
Organometallic reactions in aqueous media —
Bismuth-mediated crossed aldol type reactions1
Yoon Joo Lee and Tak Hang Chan
Abstract: Bismuth metal, upon activation by zinc fluoride, can effect the crossed aldol reaction between α-
bromocarbonyl compounds and aldehydes in aqueous media. The reaction was found to be regiospecific and syn-
diastereoselective.
Key words: bismuth, zinc fluoride, aldol reaction, regioselectivity, aqueous media.
Résumé : En milieu acide, le bismuth métallique activé par le fluorure de zinc rend possible la réaction aldolique
croisée entre des composés α-bromocarbonylés et des aldéhydes. On a observé que la réaction est régiospécifique et
syn-diastéréosélective.
Mots clés : bismuth, fluorure de zinc, réaction aldolique, régiosélectivité, milieu aqueux.
[Traduit par la Rédaction] Lee and Chan 1412
Introduction
ium could mediate similar crossed aldol type reactions of α-
bromoacetophenone with a variety of aldehydes in a mixture
of THF and H2O. Because of our continuing interest in this
area, we have examined the bismuth-mediated reaction with
the following objectives in mind: (1) use bismuth metal di-
rectly in place of the reduction of bismuth chloride for the
crossed aldol reaction; (2) extend the reaction to aliphatic
acyclic ketones since most reported examples of α-
halocarbonyl compounds 1 have been limited to arylketones;
and (3) examine the regioselection of the reaction.
The aldol reaction has played an important role in organic
synthesis for the formation of carbon–carbon bonds, provid-
ing β-hydroxycarbonyl compounds. However, under the clas-
sical aldol reaction conditions involving basic aqueous media,
side products, such as regioisomers, dimers, polymers, self-
condensation products, and α,β-unsaturated carbonyl com-
pounds, are invariably formed as well. The formation of
these side products is often attributed to the fact that the
aldol reaction is an equilibrium process (1). To circumvent
this problem, useful modifications of the classical aldol reac-
tion, particularly using Lewis-acid-promoted reactions of
enol silyl or tin ethers with carbonyl compounds, have been
developed (2). In general, these methods employ organic sol-
vents as reaction media. Recently, there has been consider-
able interest in conducting organic reactions in aqueous
media (3) because of the need to reduce volatile organic
solvents as a burden to the environment. In 1990 and subse-
quently, we reported on the possibility of conducting metal-
mediated crossed aldol type reactions in aqueous media,
using zinc, tin, and indium (4). The yields of the crossed
aldol products were modest. In 1997, Shen et al. (5) reported
using bismuth to mediate the crossed aldol type reactions in
aqueous media. In the presence of bismuth(III) chloride and
metallic aluminum, α-bromocarbonyl compounds 1 were
found to react with various aldehydes 2 in water under a ni-
trogen atmosphere at 60 °C to afford the corresponding β-
hydroxycarbonyl compounds 3 in good yields together with
the dehalogenated products 4 (Scheme 1).
Results
Study of different fluoride salts as promoters
In general, mechanically atomized metals are less reactive
than metals chemically prepared by the reduction of metal
salts with alkali metals or other reducing agents (7). The low
reactivity of the bismuth metal powder may result from the
insufficient removal of metal oxide layers from the metal
surface and the low surface area of the metal particles. Thus,
it is obvious that activation of the bismuth metal is needed
for the reaction to proceed. Previous reports from our re-
search group have shown that fluoride salts are quite effec-
tive in activating metals, such as aluminum (8) and antimony
(9), to mediate organometallic reactions in aqueous media.
More recently, we found that ammonium hydrogen fluoride
(NH4HF2) could efficiently activate commercially available
bismuth metal to reduce α-halocarbonyl compounds 1 to the
corresponding dehalogenated products 4 in excellent yields
(10). The reaction is presumed to proceed through the
intermediacy of a bismuth enolate 5 (X = Br or F,
Scheme 2). We were able to demonstrate that starting with
2-bromoisobutyrophenone (1a, R1 = Ph, R2 = R3 = Me) and
benzaldehyde in the presence of bismuth and NH4HF2, the
crossed aldol product 3a (R1 = R4 = Ph, R2 = R3 = Me) was
obtained. The reaction was carried out in water as the reac-
tion medium. However, reduction was a serious side reaction
giving product 4a in 44% yield, substantially lowering the
yield of the aldol product 3a to 56%.
Very recently, J. Zhang and Y. Zhang (6) demonstrated
that a bimetallic system of bismuth(III) chloride and samar-
Received 3 June 2003. Published on the NRC Research Press
Web site at http://canjchem.nrc.ca on 10 November 2003.
Y.J. Lee and T.H. Chan.2 Department of Chemistry, McGill
University, Montreal, QC H3A 2K6, Canada.
1Corresponding author. (e-mail: tak-hang.chan@mcgill.ca).
Can. J. Chem. 81: 1406–1412 (2003)
doi: 10.1139/V03-142
© 2003 NRC Canada

Lee and Chan
1407
Scheme 1.
Scheme 2.
Table 1. Screening of promoters for the bismuth-mediated
crossed aldol type reaction of 2-bromoisobutyrophenone with
benzaldehyde in aqueous media.a
Table 2. Screening of solvents for the bismuth-mediated crossed
aldol type reaction of 2-bromoisobutyrophenone with benzalde-
hyde in the presence of ZnF2.a
Time Unreacted Yield
Entry Promoter (h)
Yield
Time Unreacted Yield of Yield of
(h)
1a (%)b
of 3a (%)b of 4a (%)b
Entry Solvent
1a (%)b
3a (%)b 4a (%)b
0
0
1
KF
24
24
24
24
12
24
18
16
18
24
24
24
24
12
12
24
100
100
100
40
0
78
0
1
2
Neat
H2O
24
12
21
Trace
2
CaF2
AIF3
NH4F
NH4HF2
Bu4NF
TIF4
0
0
85 (81c) 15
3
0
0
3
4
5
6
D2O
12
24
0
95 (90c)
5
0
4
53
82 (78c)
7
EtOH
100
0
0
77
0
5
18
H2O–THF (1:1) 24
23
6
100
0
0
0
THF 24
100
0
aRatio of 1a:benzaldehyde:Bi:ZnF2 = 1 : l.l : 1.2 : 1.2 mmol and 1 mL
7
75
25
of solvent
8
VF4
0
50
50
bDetermined by H NMR of the crude product.
1
cIsolated yield.
9
CrF3
0
75
25
10
11
12
13
14
15
16
MnF2
FeF2
99
84
89
87
0
Trace
15
Trace
Trace
Trace
Trace
42
hyde as a model reaction (Scheme 2), and the results are
shown in Table 1.
CoF2
NiF2
10
As can be seen from Table 1, the main group metal fluo-
rides (KF, CaF2, and AlF3) could not activate bismuth to
mediate the reaction in D2O (entries 1–3). Most of the tran-
sition metal fluorides tested were found to be inefficient in
activating bismuth leading to unsatisfactory yields of the
aldol product (entries 7–14). However, the reactions with
TiF4, VF4, CrF3, and CuF2, which are not commonly used
reagents in organic synthesis, afforded the corresponding
aldol adducts in moderate yields. In the case of zinc fluo-
ride, the aldol addition product was obtained predominantly
along with a minimal amount of the reduction product (entry
15). Fluoride anion seemed to have a special activating ef-
fect as ZnCl2 was ineffective (entry 16). Several different
solvents were also screened using the same model reaction
at room temperature with ZnF2. The results are summarized
in Table 2. Hardly any desired product was observed when
ethanol or THF was employed as the solvent (entries 4 and
12
CuF2
ZnF2
ZnCI2
58
95 (90c)
0
5
100
0
0
aRatio of 1a:benzaldehyde:Bi:promoter = 1 : 1.1 : 1.5 : 1.5 mmol and 1
mL of D2O.
bDetermined by H NMR of the crude product.
1
cIsolated yield.
In this reaction, the use of ammonium hydrogen fluoride
was found to be necessary since neither the crossed aldol
type reaction nor the reduction proceeded with bismuth alone.
In an attempt to search for a more effective promoter for bis-
muth metal activation, a wide range of fluoride salts was
screened using the reaction of compound 1a with benzalde-
© 2003 NRC Canada

1408
Can. J. Chem. Vol. 81, 2003
Table 3. Bismuth-mediated crossed aldol type reactions of 2-bromoisobutyrophenone with various aldehydes and ketones in the pres-
ence of ZnF2 in aqueous media.a
Entry
Electrophile
Time (h)
Unreacted 1a (%)b
Yield of 3 (%)b
Yield of 4 (%)b
95 (90c) (3a)
80 (66c) (3b)
Trace (3c)
69 (91c) (3d)
80 (98c) (3e)
53 (35c) (3f)
67 (56c) (3g)
0
0
10
98
26
19
0
5
1
2
3
4
5
6
7
8
PhCHO
12
24
24
16
16
12
12
12
p-CH3PhCHO
p-CH3OPhCHO
p-CIPhCHO
p-CNPhCHO
n-C5H11CHO
Ph(CH2)2CHO
10
Trace
5
Trace
47
20
7
13
93
9
20
10
0
90
aRatio of 1a:aldehyde:Bi:ZnF2 = 1 : 1.1 : 1.2 : 1.2 mmol and 1 mL of D2O.
1
bDetermined by H NMR of the crude product.
cIsolated yield based on reacted 1a.
6). In a mixture of water and THF (1:1), the reduction prod-
uct was formed to a significant extent as well (entry 5).
Aqueous media, water, or deuterium oxide, proved to be the
most suitable solvents for this bismuth-mediated crossed
aldol type reaction (entries 2 and 3).
media because of the easy formation of insoluble metal
oxide on the metal surface. Aluminum metal is a good ex-
ample. It has a first ionization potential (5.984 eV) which is
much lower than that of magnesium (7.646 eV) and is ex-
pected to be a reactive metal. On the other hand, aluminum
is resistant to water because it readily forms a layer of insol-
uble Al2O3 and this prevents the metal from further reaction.
It has been reported that trace amounts of fluoride anion in
water dramatically increased the corrosion of aluminum
metal (11). This effect is attributed to the reaction of fluo-
ride ion with Al2O3 to form various water soluble
fluoroaluminates (12). Bismuth metal may be activated by
fluoride salts in a similar manner in that fluorobismuthates
can be prepared from Bi2O3 and fluorides (13). The second
factor is that the zinc ion must also play a role. Recently,
Chan and co-workers (14) and others (15) have found that
zinc fluoride acts as an effective Lewis acid in aqueous me-
dia by the coordination of the zinc cation with the carbonyl
function. In the crossed aldol reaction, the bismuth enolate 5
may react with the Zn2+- coordinated aldehyde in either the
cyclic chair transition state (6) (16) or the extended
antiperiplanar transition state (7) (17).
Scope of the bismuth-mediated aqueous crossed aldol
type reaction
The bismuth-mediated aqueous crossed aldol type reac-
tions of compound 1a with a variety of aldehydes were
surveyed (Table 3). In general, both aromatic and aliphatic
aldehydes reacted smoothly with compound 1a to provide
the corresponding aldol products 3 in moderate to high
yields. The exception appeared to be p-methoxy-
benzaldehyde (entry 3). The presence of the methoxy group
seemed to have interfered with the activation process since
unreacted 1a could be recovered. On the other hand, for α-
branched aldehydes or ketones (entries 8 and 9), compound
1a was transformed efficiently to the presumed enolate and
was protonated faster than the aldol condensation.
The generality of the reaction was also explored with vari-
ous α-bromocarbonyl compounds using benzaldehyde under
similar reaction conditions. The results are summarized in
Table 4. Aromatic α-bromoketones generally reacted better
to provide the corresponding β-hydroxyketones in higher
yields (entries 1–3) in comparison with aliphatic α-
bromoketones (entries 4 and 5). The reactions of α-
bromocarboxylic acids with benzaldehyde resulted in the
formation of the corresponding reduction products as the
major product (entries 6 and 7).
Discussions
By comparing the products derived from 3-bromobutan-2-
one (Table 4, entry 4) and 1-bromobutan-2-one (Table 4, en-
try 5), it is possible to conclude that the reaction is
regiospecific. The two regioisomeric bismuth enolates (8 and
9) must have been generated separately. Reactions of the
enolates with benzaldehyde led to different isomeric aldol
The ability of zinc fluoride to promote the crossed aldol
reaction can be attributed to two factors. First, the fluoride
anion may activate the metal for chemical reactions in aque-
ous media. Many metals, though considered to be reactive
based on their ionization potentials, may be inert in aqueous
© 2003 NRC Canada

Lee and Chan
1409
Table 4. Bismuth-mediated crossed aldol type reactions of α-bromocarbonyl compounds with benzaldehyde in the presence of ZnF2 in
aqueous media.
1
aDetermined by H NMR of the crude product.
bRatio of 1:aldehyde:Bi:ZnF2 = 1 : 1.1 : 1.2 : 1.2 mmol / 12 h.
cRatio of 1:aldehyde:Bi:ZnF2 = 1 : 2 : 2 : 2 mmol / 40 h.
dRatio of 1:aldehyde:Bi:ZnF2 = 1 : 2 : 3 : 2 mmol / 40 h.
eRatio of 1:aldehyde:Bi:ZnF2 = 1 : 1.1 : 1.5 : 1.5 mmol / 6 h.
fRatio of 1:aldehyde:Bi:ZnF2 = 1 : 1.1 : 1 : 2 mmol / 24 h.
products. Furthermore, protonation of the enolates to give
butan-2-one must not be reversible, otherwise both enolates
would have been regenerated in the reaction mixture, which
would have led to a mixture of regioisomers in both reac-
tions (Scheme 3).
one explanation over the other in the present reaction, simi-
lar stereochemical preference was observed in the bismuth-
mediated coupling of crotyl bromide with aldehydes, and the
acyclic transition state explanation similar to 7 was invoked
(18).
In the three examples where diastereomeric aldol products
were formed, the syn-isomer was formed preferentially over
the anti-isomer (Table 4, entries 2, 4, and 6). Using the con-
densation of 3-bromobutan-2-one as the example (entry 4),
the syn/anti ratio was found to be 73:27. The preferred for-
mation of the syn-isomer is usually explained by one of the
following two (or more) possibilities: (1) in the formation of
the bismuth enolate, E-9 was formed preferentially, which
then reacted with benzaldehyde via the extended antiperi-
planar transition state 7 to give the syn- product; or (2) the
Z-9 bismuth enolate was formed preferentially, and it reacted
with benzaldehyde via the cyclic chair transition state 6.
While there is insufficient experimental evidence to prefer
Conclusion
The bismuth-mediated aqueous crossed aldol type reac-
tion was investigated. The study showed that bismuth
activated by several fluoride salts could mediate the aldol
addition reaction of α-bromocarbonyl compounds with alde-
hydes in aqueous media. Zinc fluoride was found to be the
promoter of choice. The scope of this bismuth-mediated
aqueous crossed aldol type reaction was examined and the
reaction was found to be regiospecific and diastereo-
selective.
© 2003 NRC Canada

1410
Can. J. Chem. Vol. 81, 2003
Scheme 3.
General procedure for bismuth-mediated aqueous
crossed aldol type reaction
Experimental
To a mixture of 2-bromoisobutyrophenone (1 mmol), an
aldehyde (1.1 mmol), bismuth powder (1.2 mmol), and ZnF2
(1.2 mmol) was added D2O (1 mL). The reaction mixture
was vigorously stirred at room temperature for the indicated
time in Table 3 (12–24 h). Additional distilled water
(10 mL) was added in the reaction mixture, and the product
was extracted with diethyl ether (3 × 30 mL). The combined
organic layer was dried over anhyd Na2SO4 and filtered. Af-
ter the evaporation of solvent under reduced pressure, the
crude product was purified by flash column chromatography
on silica gel using 5% ethyl acetate in hexane as an eluent to
give the corresponding pure aldol addition product.
General information
Chemicals were purchased from Aldrich or the Alfa Aesar
Chemical Company. All the aldehydes were purified by dis-
tillation or column chromatography prior to use. Thin layer
chromatography was performed on plastic plates precoated
with silica gel (60 F254), which were developed using a mix-
ture of hexane and ethyl acetate as an eluent. Analytes were
visualized by UV light or by dipping the plate into a devel-
oping agent (a solution of ammonium molybdate and ceric
sulfate in dilute sulfuric acid) and heating with a heat gun.
Column chromatography was performed on 230–400 mesh
silica gel.
Melting points were determined using a Gallenkamp ap-
paratus and were uncorrected. Infrared spectra were re-
corded on an Avatar 360 FT-IR spectrometer and reported
3-Hydroxy-2,2-dimethyl-1,3-diphenyl-1-propanone (3a)
(20)
White solid, mp 105–107 °C. IR (CHCl3) (cm–1): 3475
1
in cm–1. H NMR spectra and 13C NMR spectra were re-
1
(O-H), 1672 (C=O). H NMR (300 MHz, CDCl3) δ: 7.58–
corded on a Varian Mercury 300 or 400 MHz spectrometer
or a Varian Unity 500 MHz spectrometer at ambient temper-
ature. Chemical shifts (δ) were reported in parts per million
7.54 (m, 2H), 7.48–7.25 (m, 8H), 5.15 (d, J = 3.5 Hz, 1H),
2.92 (d, J = 4.0 Hz, 1H), 1.27 (s, 3H), 1.22 (s, 3H). 13C
NMR (75 MHz, CDCl3) δ: 211.59, 140.04, 139.51, 130.37,
127.86, 127.81, 127.64, 127.58, 126.98, 78.82, 52.52, 24.34,
19.70.
1
(ppm) and referenced to CDCl3 at δ 7.25 ppm for H and δ
77.00 ppm for 13C. The multiplicity of each signal was indi-
cated by s (singlet), d (doublet), t (triplet), q (quartet), quint
(quintet), sept (septet), and m (multiplet). Coupling con-
stants (J) were reported in Hz. The syn and anti configura-
tions of 3h, 3j, and 3m were assigned on the basis that the
coupling constants of the methyl-bearing methine proton and
the hydroxyl-bearing methine proton have greater values for
anti- than for syn-isomers (19). Mass spectra were recorded
on a Kratos MS25RFA mass spectrometer or on a HP 5980A
GC–MS.
3-Hydroxy-2,2-dimethyl-3-(4-methylphenyl)-1-phenyl-1-
propanone (3b) (20)
White solid, mp 83.5–85.0 °C. IR (CHCl3) (cm–1): 3490
1
(O-H), 1673 (C=O). H NMR (400 MHz, CDCl3) δ: 7.58–
7.55 (m, 2H), 7.47–7.42 (m, 1H), 7.40–7.36 (m, 2H), 7.23
(d, J = 8.4 Hz, 2H), 7.14 (d, J = 8.0 Hz, 2H), 5.10 (s, 1H),
2.94 (br s, 1H), 2.36 (s, 3H), 1.25 (s, 3H), 1.19 (s, 3H). 13C
NMR (100 MHz, CDCl3) δ: 211.62, 139.59, 137.19, 137.06,
© 2003 NRC Canada

Lee and Chan
1411
130.30, 128.30, 127.83, 127.68, 126.98, 78.67, 52.54, 24.34,
21.21, 19.62.
8.1 Hz, 1H), 3.84 (quint, J = 7.3 Hz, 1H), 3.09 (br s, 1H),
1.07 (d, J = 7.3 Hz, 3H). 13C NMR (75 MHz, CDCl3) δ:
204.63, 142.00, 136.54, 133.15, 128.50, 128.30, 127.77,
126.58, 76.68, 47.96, 15.76.
3-(4-Cyanophenyl)-3-hydroxy-2,2-dimethyl-1-phenyl-1-
propanone (3e)
Colorless oil. IR (CHCl3) (cm–1): 3477 (O-H), 2229
3-Hydroxy-1,3-diphenyl-1-propanone (3i) (23)
1
1
Colorless oil. H NMR (300 MHz, CDCl3) δ: 7.97–7.93
(CϵN), 1673 (C=O).; H NMR (400 MHz, CDCl3) δ: 7.55–
(m, 2H), 7.61–7.55 (m, 1H), 7.48–7.27 (m, 7H), 5.35 (dd,
7.49 (m, 4H), 7.43–7.31 (m, 5H), 5.13 (d, J = 3.3 Hz, 1H),
3.72 (d, J = 3.8 Hz, 1H), 1.17 (s, 3H), 1.12 (s, 3H). 13C
NMR (100 MHz, CDCl3) δ: 210.73, 145.52, 138.80, 131.30,
130.75, 128.54, 127.95, 127.03, 118.58, 111.18, 77.93,
52.38, 24.03, 19.63. MS (CI, NH3) m/z: 280 (M + H+, 5),
148 (51), 130 (48), 105 (100), 77 (13). HR-MS (FAB) for
C18H17NO2 + H+ calcd.: 280.1337; found: 280.1338.
J = 7.7, 4.4 Hz, 1H), 3.73 (br s, 1H), 3.39–3.36 (m, 2H). 13
C
NMR (75 MHz, CDCl3) δ: 199.74, 142.77, 136.29, 133.41,
128.46, 128.33, 127.95, 127.43, 125.56, 69.86, 47.34.
4-Hydroxy-3-methyl-4-phenyl-2-butanone (3j) (24)
syn-Isomer
Colorless oil. IR (CHCl3) (cm–1): 3443 (O-H), 1703
3-(4-Chlorophenyl)-3-hydroxy-2,2-dimethyl-1-phenyl-1-
propanone (3d)
1
(C=O). H NMR (400 MHz, CDCl3) δ: 7.33–7.21 (m, 5H),
5.05 (d, J = 4.1 Hz, 1H), 3.25 (br s, 1H), 2.82 (qd, J = 7.1,
4.1 Hz, 1H), 2.11 (s, 3H), 1.08 (d, J = 7.1 Hz, 3H). 13C
NMR (100 MHz, CDCl3) δ: 213.20, 141.54, 128.04, 127.14,
125.69, 72.96, 53.20, 29.43, 10.29.
White solid, mp 57–59 °C. IR (CHCl3) (cm–1): 3475 (O-
1
H), 1673 (C=O). H NMR (400 MHz, CDCl3) δ: 7.54–7.50
(m, 2H), 7.44–7.39 (m, 1H), 7.36–7.32 (m, 2H), 7.26–7.19
(m, 4H), 5.04 (s, 1H), 3.61 (br s, 1H), 1.18 (s, 3H), 1.12 (s,
3H). 13C NMR (100 MHz, CDCl3) δ: 211.25, 139.17,
138.53, 133.15, 130.46, 129.07, 127.83, 127.65, 126.96,
77.95, 52.36, 24.10, 19.53.
anti-Isomer
Colorless oil. IR (CHCl3) (cm–1): 3431 (O-H), 1705
1
(C=O). H NMR (400 MHz, CDCl3) δ: 7.36–7.25 (m, 5H),
4.72 (d, J = 8.8 Hz, 1H), 3.02 (br s, 1H), 2.92 (quint, J =
7.2 Hz, 1H), 2.22 (s, 3H), 0.92 (d, J = 7.6 Hz, 3H). 13C
NMR (100 MHz, CDCl3) δ: 213.19, 141.67, 128.32, 127.83,
126.46, 76.44, 53.66, 30.12, 14.21.
3-Hydroxy-2,2-dimethyl-1-phenyl-1-octanone (3f)
Colorless oil. IR (CHCl3) (cm–1): 3489 (O-H), 1671
1
(C=O). H NMR (400 MHz, CDCl3) δ: 7.61–7.58 (m, 2H),
7.44–7.40 (m, 1H), 7.37–7.33 (m, 2H), 3.85 (dd, J = 10.0,
2.1 Hz, 1H), 2.69 (br s, 1H), 1.61–1.14 (m, 14H), 0.87 (t,
J = 6.9 Hz, 3H). 13C NMR (100 MHz, CDCl3) δ: 210.75,
139.01, 130.52, 127.80, 127.19, 76.96, 52.11, 31.75, 31.59,
26.48, 22.89, 22.64, 21.20, 14.09. MS (FAB) m/z: 249 (M +
H+, 15), 231 (7), 149 (27), 131 (30), 105 (100), 77 (16). HR-
MS (FAB) for C16H24O2 + H+ calcd.: 249.1855; found:
249.1855.
Acknowledgment
We thank the Natural Sciences and Engineering Research
Council of Canada (NSERC) for financial support of this re-
search.
References
3-Hydroxy-2,2-dimethyl-1,5-diphenyl-1-pentanone (3g)
(21)
1. T. Mukaiyama. Isr. J. Chem. 24, 162 (1984).
2. (a) T. Mukaiyama. Org. React. (N.Y.), 28, 203 (1982);
(b) R.W. Stevens, N. Iwasawa, and T. Mukaiyama. Chem. Lett.
1459 (1982); (c) T.H. Chan. In Comprehensive organic synthe-
sis. Vol. 2. Edited by B.M. Trost and I. Fleming. Pergamon
Press, Oxford. 1991. pp. 595–628.
3. (a) C.J. Li and T.H. Chan. Organic reactions in aqueous media.
John Wiley and Sons, New York. 1997; (b) P.A. Grieco (Edi-
tor). Organic synthesis in water. Blackie Academic and Profes-
sional, London. 1998.
Colorless oil. IR (CHCl3) (cm–1): 3489 (O-H), 1669
1
(C=O). H NMR (400 MHz, CDCl3) δ: 7.63–7.60 (m, 2H),
7.48–7.44 (m, 1H), 7.41–7.36 (m, 2H), 7.31–7.27 (m, 2H),
7.22–7.18 (m, 3H), 3.91 (dd, J = 10.4, 2.4 Hz, 1H), 3.02–
2.95 (m, 1H), 2.77 (br d, J = 5.7 Hz, 1H), 2.71–2.63 (m,
1H), 1.85–1.71 (m, 2H), 1.34 (d, J = 2.4 Hz, 6H). 13C NMR
(100 MHz, CDCl3) δ: 210.72, 141.86, 138.73, 130.80,
128.32, 128.22, 127.95, 127.32, 125.68, 76.57, 51.93, 33.57,
33.06, 23.18, 21.24.
4. (a) T.H. Chan, C.J. Li, and Z.Y. Wei. J. Chem. Soc. Chem.
Commun. 505 (1990); (b) T.H. Chan, C.J. Li, M.C. Lee, and
Z.Y. Wei. Can. J. Chem. 72, 1181 (1994).
3-Hydroxy-2-methyl-1,3-diphenyl-1-propanone (3h) (22)
5. Z. Shen, J. Zhang, H. Zou, and M. Yang. Tetrahedron Lett. 38,
2733 (1997).
6. J.M. Zhang and Y.M. Zhang. Chin. J. Chem. 20, 111 (2002).
7. H. Kagoshima, Y. Hashimoto, D. Oguro, and K. Saigo. J. Org.
Chem. 63, 691 (1998).
8. L.H. Li and T.H. Chan. Org. Lett. 2, 1129 (2000).
9. L.H. Li and T.H. Chan. Tetrahedron Lett. 41, 5009 (2000).
10. (a) T.H. Chan, L.-H. Li, Y. Yang, and W. Lu. In Clean sol-
vents. ACS Symp. Ser. 819. Edited by M.A. Abraham and L.
Moens. American Chemical Society, Washington, D.C. 2002;
(b) J.Y.J. Lee and T.H. Chan. Can. J. Chem. Accepted for pub-
lication.
syn-Isomer
White solid, mp 71–73 °C. 1H NMR (300 MHz, CDCl3) δ:
7.96–7.92 (m, 2H), 7.62–7.56 (m, 1H), 7.50–7.23 (m, 7H),
5.25 (d, J = 2.6 Hz, 1H), 3.71 (qd, J = 7.3, 3.2 Hz, 1H), 3.69
(br s, 1H), 1.21 (d, J = 7.3 Hz, 3H). 13C NMR (75 MHz,
CDCl3) δ: 205.53, 141.63, 135.46, 133.47, 128.67, 128.36,
128.13, 127.19, 125.91, 73.03, 47.03, 11.24.
anti-Isomer
1
Colorless oil. H NMR (300 MHz, CDCl3) δ: 7.99–7.95
(m, 2H), 7.59–7.53 (m, 1H), 7.49–7.25 (m, 7H), 4.99 (d, J =
© 2003 NRC Canada

1412
Can. J. Chem. Vol. 81, 2003
11. K. Tennakone and S. Wickramanayake. Nature (London), 325,
202 (1987).
17. (a) S. Murata, M. Suzuki, and R. Noyori. J. Am. Chem. Soc.
102, 3248 (1980); (b) Y. Yamamoto, H. Yatagai, Y. Naruta,
and K. Maruyama. J. Am. Chem. Soc. 102, 7107 (1980); (c) Y.
Yamamoto. Acc. Chem. Res. 20, 243 (1987).
18. M. Wada, H. Ohki, and K. Akiba. Bull. Chem. Soc. Jpn. 63,
1738 (1990).
12. (a) T. Hurlen and K.H. Johansen. Acta Chem. Scand. Ser. A,
A39, 545 (1985); (b) K. Tennakone, S. Wickramanayake, and
C.A.N. Fernando. Environ. Pollut. 49, 133 (1988).
13. S.M. Godfrey, C.A. McAuliffe, A.G. Mackie, and R.G. Prit-
chard. In Chemistry of arsenic, antimony and bismuth. Edited
by N.C. Norman. Blackie Academic and Professional, London.
1998. pp. 168–179.
14. W. Miao, W. Lu, and T.H. Chan. J. Am. Chem. Soc. 125, 2414
(2003).
15. S. Kobayashi, T. Hamada and K. Manabe. J. Am. Chem. Soc.
19. F. Fringuelli, O. Piermatti, and F. Pizzo. J. Org. Chem. 60,
7006 (1995).
20. C.H. Harrison. Tetrahedron Lett. 28, 4135 (1987).
21. T. Hamada, K. Manabe, S. Ishikawa, S. Nagayama, M. Shiro,
and S. Kobayashi. J. Am. Chem. Soc. 125, 2989 (2003).
22. T. Harada and T. Mukaiyama. Chem. Lett. 467 (1982).
23. H.O. House, D.S. Crumrine, A.Y. Teranishi, and H.D.
Olmstead. J. Am. Chem. Soc. 95, 3310 (1973).
24. T. Hayashi, Y. Matsumoto, and Y. Ito. J. Am. Chem. Soc. 110,
5579 (1988).
124, 5640 (2002).
16. (a) H.E. Zimmerman and M.D. Traxler. J. Am. Chem. Soc. 79,
1920 (1957); (b) C.H. Heathcock, C.T. Buse, W.A. Kleschick,
M.C. Pirrung, J.E. Sohn, and J. Lampe. J. Org. Chem. 45,
1066 (1980); (c) R.W. Hoffman. Angew. Chem. Int. Ed. Engl.
21, 555 (1982).
© 2003 NRC Canada