Rearrangement during halogenation of 2-hydroxy-1,2-diphenylpropan-1-one (α-methylbenzoin)

Article abstract of DOI:10.1016/j.tet.2008.03.043

The reaction of 2-hydroxy-1,2-diphenylpropan-1-one 1 with SOCl₂ or PBr₃ gives, respectively, the 3-chloro- and 3-bromo-1,2-diphenylpropan-1-ones 4 and 6. The expected 2-chloro- and 2-bromo-1,2-diphenylpropan-1-ones 2 and 5 can, howev

Full text of DOI:10.1016/j.tet.2008.03.043

Tetrahedron 64 (2008) 5217–5220
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Rearrangement during halogenation of 2-hydroxy-1,2-
diphenylpropan-1-one (a-methylbenzoin)
*
Kati M. Aitken, R. Alan Aitken
EaStCHEM School of Chemistry, University of St Andrews, St Andrews, Fife KY16 9ST, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
The reaction of 2-hydroxy-1,2-diphenylpropan-1-one 1 with SOCl2 or PBr3 gives, respectively, the
3-chloro- and 3-bromo-1,2-diphenylpropan-1-ones 4 and 6. The expected 2-chloro- and 2-bromo-
1,2-diphenylpropan-1-ones 2 and 5 can, however, be formed by treatment of 1,2-diphenylpropan-1-one
8 with Cl2 or Br2/AlCl3. The four halogenated products are characterised by 1H and 13C NMR spectroscopy
for the first time.
Received 22 November 2007
Received in revised form 3 March 2008
Accepted 13 March 2008
Available online 15 March 2008
Ó 2008 Elsevier Ltd. All rights reserved.
1. Introduction
A similar confusion exists in the literature for the bromination of
1 using PBr3, which was first reported to give 5,5 but this result was
We recently required the monoimines of unsymmetrical 1,2-
diketones and were attracted by a procedure described by Boyer
and Straw in 1953,1 involving conversion of the a-hydroxyketone 1
into its chloride 2 followed by reaction with sodium azide and
thermal decomposition with rearrangement to give the product 3.
later found to be irreproducible,6 and the fact that the product is
actually 6 does not seem to have been clearly stated so far. The
synthesis of 5 was only reported in 1987.7 Most striking is the fact
that NMR spectroscopic data have apparently never been published
for any of the compounds 2, 4, 5 or 6. In an attempt to clarify this
area we describe in this paper a reinvestigation of these reactions
using modern analytical techniques, especially NMR spectroscopy.
O
O
O
Me
OH
Me NaN3
Cl
SOCl2
Me
N3
Ph
Ph
Ph
Ph
2. Results and discussion
Ph
Ph
Ph
2
1
heat
Authentic samples of the two non-rearranged halides 2 and 5
were first prepared by treatment of the a-methylketone 8, prepared
from deoxybenzoin 7 by methylation,8 with the appropriate
halogen (Scheme 1). Chlorination resulted in formation of some
ring-chlorinated products and, in common with the literature
procedure,4 we had to resort to chromatography to obtain a pure
sample of 2. In fact 1H NMR analysis of the reaction product
after chlorination of 8 for 14 h showed the presence of 2 (40%),
unreacted 8 (30%) and two ring-chlorinated products (each 15%).
While simple column chromatography was sufficient to separate 2
O
O
Me
Me
NPh
N:
Ph
Ph
3
In the event, treatment of 1 with thionyl chloride gave a product
that was clearly not 2 and was readily shown by 1H and 13C NMR to
have the isomeric structure 4. The fact that, contrary to an earlier
report,2 treatment of 1 with SOCl2 gives 4 and not 2 was already
demonstrated as early as 1955,3 and it was not until 1970 that
a genuine route to 2 appeared.4
O
O
Me
Me
Cl
X
Cl2
Ph
Cl
Ph
Ph
+
O
O
O
CHCl3
O
i. NaH, THF
ii. MeI
O
Me
Br
Ph
2
Me
Ph
Cl
Ph
Ph
Br
Ph
Ph
Ph
Ph
Ph
6
O
9 X = H
10 X = Cl
Ph
Ph
4
5
Me
Br
Br2, Et2O
cat. AlCl3
7
8
Ph
5
* Corresponding author.
E-mail address: raa@st-and.ac.uk (R. Alan Aitken).
Scheme 1. Formation of compounds 2 and 5.
0040-4020/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2008.03.043

5218
K.M. Aitken, R. Alan Aitken / Tetrahedron 64 (2008) 5217–5220
Table 1
and 8 from the ring-chlorinated products, the latter were very
similar to each other and were only separable using preparative
HPLC. This allowed isolation of the previously unknown o-chloro
compound 9 in pure form, and a second component, which was
contaminated with some 9 but could be identified as the corre-
sponding dichloro compound 10. Interestingly there was no trace
of the known p-chloro analogue of 9.9
Comparison of NMR data for 4 and 6 with that for 1216
X
2-H
3-H
Cl
Br
I
4.94 (dd, J 9.0, 5.7)
5.01 (dd, J 9.0, 5.4)
5.1 (dd, J 9, 6)
4.30 (dd, J 10.8, 9.0), 3.73 (dd, J 10.8, 5.7)
4.16 (dd, J 9.9, 9.0), 3.59 (dd, J 9.9, 5.4)
4.0 (t, J 9), 3.5 (dd, J 9, 6)
The bromination was much cleaner and proceeded readily using
Br2 with catalytic AlCl3 to afford 5 but since neither the original
work,7 nor later reports mention any analytical or spectroscopic
data for this compound, it was fully characterised.
X
C-1
C-2
C-3
Cl
Br
I
196.8
197.0
197.4
55.9
56.0
56.6
45.1
32.8
5.6
We next examined halogenation of 1, readily obtained by
addition of methylmagnesium iodide to benzil.10 Treatment with
thionyl chloride in the presence of pyridine using the reported
method,1,11 cleanly gave the rearranged product 4 (Scheme 2), in
contrast to benzoin itself, which does undergo chlorination without
rearrangement under these conditions.11 The reaction of 1 and
thionyl chloride to give 4 has already been noted,3,12 and the
product is identical to that obtained by addition of HCl to the enone
11,3,13,14 which was in fact present as an impurity in the sample of 4
formed from 1.15 Reaction of 1 with PBr3 using the method origi-
nally reported to give 5,5 gave a product that was obviously the
rearranged bromide 6, identical to that formed by addition of HBr to
11.14 With the four compounds 2, 5, 4 and 6 in hand, comparison of
their NMR spectra left no room for doubt as to which was which.
The iodide 12 has also been obtained from 11 by addition of HI,14
and also more recently by another method resulting in NMR data
being available for this iodide.16 This allowed a detailed assignment
of the spectra for 4, 6 and 12 (Table 1) showing a good degree of
agreement except for the well known shielding effect of the heavier
halogens on the C-3 signal.
X
(1-Ph)
C-10
C-20,60
C-30,50
C-40
Cl
Br
I
136.0
136.8
138.2
128.1
128.1
128.0
128.5
128.6
128.4
133.2
133.3
133.3
X
(2-Ph)
C-100
C-200,600
C-300,500
C-400
Cl
Br
I
136.0
136.0
128.7
129.1
129.2
129.3
128.6
128.7
128.6
128.0
128.1
127.9
alkene.18–20 Later studies showed that chlorination of atrolactinic
acid 16 with HCl could give either 17 or 18,21,22 while treatment
with PCl5 gave products derived from 19, as well as the expected
product 20.22 The fact that compounds 4 and 6 are known to be
formed by addition of HX to enone 9, which is actually formed from
1 under the conditions used to convert it to 4, and the ample
literature precedent of Scheme 3, leave little doubt that it is the
possibility of dehydration via the activated intermediates to give
the relatively stable enone 11 that makes compound 1 behave
anomalously towards attempted halogenation, as shown below.
O
O
O
Me
OH
SOCl2
pyridine
HCl
HBr
Ph
Cl
Ph
Ph
Ph
Ph
Ph
Ph
4
11
1
PBr3
Et2O
Me
BBr3
R
R
HI
O
OH
Br
Me
Me
Ph
Br
13 R = Me (10%)
O
R = Et (1–2%)
6
Me
Me
HCl
Ph
I
Ph
Ph
Ph
OH
Cl
CO2H
Cl
CO2H
14
not
Ph
12
CN
15
Scheme 2. Formation of compounds 4 and 6.
Me
Me
HCl
Ph
Ph
Ph
Ph
Ph
This type of halogenation with rearrangement is relatively
uncommon but some previously reported examples are shown in
Scheme 3. Hudson and co-workers observed a small amount of
isomerisation in the BBr3 mediated bromination of aliphatic
tertiary alcohols and ascribed this to elimination to form an alkene
followed by radical addition of HBr.17 Much earlier, the fact that
treatment of acetophenone cyanohydrin with concentrated HCl
gave the tropic acid derivative 14 rather than the expected atro-
lactinic acid derivative 15 was crucial in elucidating the structure of
atropine and was again postulated to involve an intermediate
OH
Cl
Cl
CO2H
17
or
CO2H
16
CO2H
18
PCl5
Me
Cl
Cl
and
COCl
19
COCl
20
Scheme 3. Mechanistically similar reactions.
To return to our original objective, once it became clear that the
unrearranged chloride 2 could only be obtained in low yield after
a difficult separation, we resorted to the direct reaction of 1-phe-
nylpropane-1,2-dione with appropriate aromatic amines,23 and in
this way were able to prepare 3 and its previously unknown
N-o-tolyl analogue.
X–
O
H
– HX
+ HX
O
11
4 or 6
Ph
Y
Ph
X = Cl, Y = S(O)Cl
X = Br, Y = –P<
3. Experimental
3.1. General
O
2'
3
1'
2
3'
4'
1
X
1"
6'
6"
2"
3"
5'
Infrared spectra were recorded for liquid films on a Perkin Elmer
Spectrum GX instrument. NMR spectra were obtained for 1H at
5"
4"

K.M. Aitken, R. Alan Aitken / Tetrahedron 64 (2008) 5217–5220
5219
300 MHz, and for 13C at 75 MHz using a Bruker Avance 500
3.4. 3-Chloro-1,2-diphenylpropan-1-one 4
instrument. All spectra were run on solutions in CDCl3 with internal
Me4Si as reference. Chemical shifts are reported in parts per million
to high frequency of the reference and coupling constants J are in
Hertz. Mass spectra were obtained on a Micromass GCT mass
spectrometer using chemical ionisation. Melting points were
determined in open capillary tubes on a Gallenkamp apparatus and
are uncorrected. Column chromatography was performed using
silica gel of 33–70 mm particle size and preparative HPLC was
carried out using a Waters system with a PhenomenexÒ Kingsorb
5 m C18 column.
A suspension of a-methylbenzoin 1 (14.0 g, 62 mmol) in dry
pyridine (7 cm3) was heated until the solid dissolved and then
cooled and stirred at 0 ꢀC while thionyl chloride (6.05 cm3, 9.87 g,
83 mmol) was added in small portions. After 1 h, water (25 cm3)
was cautiously added and the mixture was extracted with ethyl
acetate (3ꢂ25 cm3), which was dried and evaporated. The crude
product (13.89 g) consisted of a mixture of 4 (70%) and the enone 11
(20%). Recrystallisation of the residue from ethanol gave the
product (7.6 g, 50%) as colourless crystals, mp 57–59 ꢀC (lit.,3 60 ꢀC);
nmax/cmꢁ1 1683, 1597, 1493, 1448, 1339, 1235, 1176, 980, 944, 721
and 700; dH 3.73 (1H, dd, J 10.8, 5.7), 4.30 (1H, dd, J 10.8, 9.0), 4.94
(1H, dd, J 9.0, 5.7), 7.25–7.55 (8H, m) and 7.94–8.00 (2H, m); dC 45.1
(CH2), 55.9 (CH), 128.0 (CH), 128.1 (2CH), 128.5 (2CH), 128.6 (2CH),
129.1 (2CH), 133.2 (CH), 135.97 (C), 136.03 (C) and 196.8 (CO); m/z
247 (37ClꢁMþHþ, 10%), 245 (35ClꢁMþHþ, 40), 209 (MꢁClþ, 100)
and 105 (PhCO, 54).
3.2. 2-Chloro-1,2-diphenylpropan-1-one 2
A solution of a-methyldeoxybenzoin 8 (10.0 g, 47.6 mmol) in
chloroform (950 cm3) was stirred at rt while chlorine gas was
bubbled slowly through it for 14 h. Evaporation followed by
chromatography of the residual liquid (SiO2, hexane/EtOAc, 95:5)
gave the product 2 (3.2 g, 28%) as a colourless oil (lit.,4 mp
47–48 ꢀC); nmax/cmꢁ1 1688, 1597, 1490, 1447, 1375, 1253, 1183, 1055,
3.5. 3-Bromo-1,2-diphenylpropan-1-one 6
960, 841, 762 and 698; d 2.04 (3H, s), 7.24 (2H, t, J 6.0), 7.27–7.45
H
(4H, m), 7.49 (2H, d, J 6.0) and 7.75 (2H, d, J 6.0); d 33.2 (CH3), 73.6
(C), 125.2 (2CH), 127.8 (2CH), 128.1 (CH), 128.8 (2CH), 130.8 (2CH),
132.5 (CH), 133.5 (C), 141.5 (C) and 194.2 (CO); m/z 245
A solution of a-methylbenzoin 1 (15.0 g, 66 mmol) in diethyl
ether (110 cm3) was stirred at rt while PBr3 (9.25 cm3, 26.3 g,
97 mmol) was added dropwise and the mixture was then heated
under reflux for 4 h. The mixture was evaporated and the residue
dissolved in diethyl ether (100 cm3) and washed thoroughly with
water, followed by brine, dried and evaporated. The resulting
solid was recrystallised from toluene to give the product (12.0 g,
63%) as colourless crystals, mp 55–56 ꢀC (lit.,5 56 ꢀC); nmax/cmꢁ1
1679, 1596, 1580, 1491, 1447, 1356, 1262, 1225, 970, 759, 697 and
646; dH 3.59 (1H, dd, J 9.9, 5.4), 4.16 (1H, dd, J 9.9, 9.0), 5.01 (1H,
C
(
35ClꢁMþHþ, 2%), 209 (MꢁClþ, 100) and 105 (PhCO, 26). This was
followed by a fraction containing a mixture of two ring-chlorinated
products, and finally a fraction containing unreacted 8. Preparative
HPLC (5% H2O in MeCN) of the second fraction gave first
2(2-chlorophenyl)-1-phenylpropan-1-one 9.
3.2.1. 2(2-Chlorophenyl)-1-phenylpropan-1-one 9
35
Found: MþHþ, 245.0740. C15
H
ClO requires 245.0733; d 1.49
dd, J 9.0, 5.4), 7.25–7.55 (8H, m) and 7.9–8.0 (2H, m); d 32.8
14
H
C
(3H, d, J 7.0), 5.14 (1H, q, J 7.0), 7.13–7.17 (3H, m, Ar), 7.32–7.42
(3H, m, 3,5-H of Ph and Ar), 7.45–7.52 (1H, m, 4-H of Ph) and
7.92–7.95 (2H, m, 2,6-H of Ph); dC 17.8 (CH3), 44.2 (CH), 127.5
(CH), 128.2 (CH), 128.56 (3CH), 128.59 (2CH), 129.9 (CH), 133.0
(CH and C–Cl, confirmed by HMBC), 136.0 (C-1 of Ph), 139.2 (C-1
of Ar) and 200.1 (CO); m/z (CI) 247 (37ClꢁMþHþ, 7%), 245
(CH2), 56.0 (CH), 128.1 (3CH), 128.6 (2CH), 128.7 (2CH), 129.2
(2CH), 133.3 (CH), 136.0 (C), 136.8 (C) and 197.0 (CO); m/z 291
(
81BrꢁMþHþ, 43%), 289 (79BrꢁMþHþ, 52), 209 (MꢁBrþ, 100) and
105 (PhCO, 68).
3.6. 1-Phenyl-2-(phenylimino)propan-1-one 3
(
a
35ClꢁMþHþ, 28), 243 (12) and 209 (100). This was followed by
fraction consisting mainly of 2-chloro-2-(2-chlorophenyl)-
A solution of 1-phenylpropane-1,2-dione (0.80 g, 5.4 mmol)
and aniline (0.503 g, 5.4 mmol) in toluene (20 cm3) was heated
under reflux with a Dean–Stark trap for 6 h. Evaporation and
Kugelrohr distillation of the residue gave starting materials, bp
1-phenylpropan-1-one 10.
3.2.1.1. 2-Chloro-2-(2-chlorophenyl)-1-phenylpropan-1-one 10. Found:
35
MþHþ, 279.0349. C15
H
13
Cl2O requires 279.0343; d 2.17 (3H, s),
50–70 ꢀC/1 Torr, followed by the product (0.43 g, 36%) as
colourless liquid, bp 120 ꢀC/1 Torr (lit.,22 104–106 ꢀC/0.001 Torr);
2.18 (3H, s, Me), 6.86 (2H, dd, J 8.4, 1.2, 2,6-H of N–Ph), 7.16
a
H
7.15–7.45 (6H, m), 7.72–7.80 (2H, m) and 8.21–8.25 (1H, m); dC 29.2
(CH3), 73.2 (C), 127.7 (CH), 127.8 (2CH), 128.1 (CH), 129.6 (CH), 130.0
(2CH), 130.9 (CH), 131.1 (C), 132.5 (CH), 133.7 (C), 139.7 (C) and 192.0
(CO); m/z (CI) 281 (35Cl2ꢁMþHþ, 10%), 245 (36) and 243 (100).
d
H
(1H, tt, J 7.2, 1.2, 4-H of N–Ph), 7.39 (2H, dd, J 8.4, 7.2, 3,5-H of
N–Ph), 7.48 (2H, dd, J 8.4, 7.5, 3,5-H of C–Ph), 7.60 (1H, tt, J 7.5,
1.5, 4-H of C–Ph) and 8.15 (2H, dd, J 8.4, 1.5, 2,6-H of C–Ph); dC
16.5, 118.8 (C-2,6 of N–Ph), 124.6 (C-4 of N–Ph), 128.2 (C-2,6 of
C–Ph), 129.0 (C-3,5 of N–Ph), 130.7 (C-3,5 of C–Ph), 133.3 (C-4 of
C–Ph), 134.7 (C-1 of C–Ph), 149.1 (C-1 of N–Ph), 166.7 (C]N) and
193.2 (C]O).
3.3. 2-Bromo-1,2-diphenylpropan-1-one 5
A solution of a-methyldeoxybenzoin 8 (7.84 g, 37 mmol) in
diethyl ether (150 cm3) containing AlCl3 (0.25 g, 1.85 mmol) was
stirred at 0 ꢀC while bromine (2.02 cm3, 6.25 g, 39 mmol) was
added dropwise. The resulting HBr gas was trapped in aqueous
NaOH. After 3 h at rt, the colour of the Br2 had disappeared and the
solution was washed with water then brine, dried and evaporated.
The resulting solid was vacuum dried to give the product (10.6 þg,
3.7. 1-Phenyl-2-(o-tolylimino)propan-1-one
A preparation exactly as above but using o-toluidine in place of
aniline gave the product (0.74 g, 58%) as a colourless liquid, bp
150 ꢀC/1 Torr; dH 2.12 (3H, s, Me), 2.17 (3H, s, Ar–Me), 6.67 (1H, dd,
J 7.8, 1.5, 6-H of N–Ar), 7.07 (1H, td, J 7.5, 1.5, 4-H of N–Ar), 7.17–7.26
(2H, m, 3,5-H of N–Ar), 7.49 (2H, dd, J 8.4, 7.5, 3,5-H of Ph), 7.60 (1H,
tt, J 7.5, 1.5, 4-H of Ph) and 8.17 (2H, dd, J 8.4, 1.5, 2,6-H of Ph); dC
16.8, 18.1, 117.5 (C-6 of N–Ar), 124.6 (C-5 of N–Ar), 126.4 (C-4 of N–
Ar), 126.7 (C-2 of N–Ar), 128.3 (C-2,6 of Ph), 130.6 (C-3 of N–Ar),
130.7 (C-3,5 of Ph), 133.3 (C-4 of Ph), 134.8 (C-1 of Ph), 148.1 (C-1 of
N-Ar), 166.7 (C]N) and 193.2 (C]O).
98%) as faintly orange crystals, mp 59–61 ꢀC. Found: MþH ,
79
289.0226. C
H
BrO requires 289.0228; nmax/cmꢁ1 1677, 1595,
15 14
1447, 1369, 1252, 1231, 1047, 955, 844, 762 and 698; dH 2.22 (3H, s),
7.20–7.40 (6H, m), 7.47–7.52 (2H, m) and 7.75–7.80 (2H, m); dC 34.8
(CH3), 68.0 (C), 126.2 (2CH), 127.7 (2CH), 128.0 (CH), 128.8 (2CH),
130.9 (2CH), 132.5 (CH), 133.4 (C), 141.7 (C) and 193.8 (CO); m/z 291
(
81BrꢁMþHþ, 2%), 289 (79BrꢁMþHþ, 2), 210 (MꢁBrþHþ, 45), 209
(MꢁBrþ, 100), 193 (72) and 105 (PhCO, 22).

5220
K.M. Aitken, R. Alan Aitken / Tetrahedron 64 (2008) 5217–5220
13. Fiesselmann, H.; Ribka, J. Chem. Ber. 1956, 89, 27–39.
References and notes
14. Matti, J.; Laval-Verges, A. M.; Emod, I. Bull. Soc. Chim. Fr. 1963, 1176–1181.
15. Characteristic NMR signals for 11: dH 5.61 (1H, s, ]CH2) and 6.03 (1H, s, ]CH2);
dC 120.8 (]CH2), 148.1 (pC]) and 197.4 (C]O). See Hon, Y.-S.; Hsu, T.-R.; Chen,
C.-Y.; Lin, Y.-H.; Chang, F.-J.; Hsieh, C.-H.; Szu, P.-H. Tetrahedron 2003, 59,
1509–1520.
1. Boyer, J. H.; Straw, D. J. Am. Chem. Soc. 1953, 75, 1642–1644.
2. Garry, M. Ann. Chim. Appl. 1942, 17, 5–99; [Chem. Abstr. 1943, 37, 35729].
3. Matti, J.; Perrier, M. Bull. Soc. Chim. Fr. 1955, 525–530.
4. Luche-Ronteix, M.-J.; Bory, S.; Dvolaitzky, M.; Lett, R.; Marquet, A. Bull. Soc.
Chim. Fr. 1970, 2564–2572.
5. Buchmann, G.; Schmuck, R. J. Prakt. Chem. 1965, 28, 141–156.
6. Jochims, J. C.; Abu-Taha, A. Chem. Ber. 1976, 106, 139–153.
7. Dommro¨se, A.-M.; Gru¨tzmacher, H.-F. Org. Mass Spectrom. 1987, 22,
437–443.
8. Adam, W.; Bosio, S. G.; Turro, N. J. J. Am. Chem. Soc. 2002, 124, 8814–8815.
9. Grasa, G. A.; Colacot, T. J. Org. Lett. 2007, 9, 5489–5492.
10. Roger, R. J. Chem. Soc. 1925, 518–523.
11. Ward, A. M. Org. Synth. 1932, 12, 20–21.
12. Drehfahl, G.; Ho¨rhold, H.-H. Chem. Ber. 1961, 94, 1641–1656.
16. Ciganek, E.; Calabrese, J. C. J. Org. Chem. 1995, 60, 4439–4443.
17. Hudson, H. R.; Kinghorn, R. R. F.; Murphy, W. S. J. Chem. Soc.
3593–3596.
C 1971,
18. Spiegel, A. Ber. Dtsch. Chem. Ges. 1881, 14, 235–240.
19. Spiegel, A. Ber. Dtsch. Chem. Ges. 1881, 14, 1352–1355.
20. Ladenburg, A. Liebigs Ann. Chem. 1883, 217, 74–149.
21. Bougault, J. C. R. Seances Acad. Sci. 1908, 146, 766–769.
22. Staudinger, H.; Ruzicka, L. Liebigs Ann. Chem. 1911, 380, 278–303.
23. Alcaide, B.; Escobar, G.; Perez-Ossorio, R.; Plumet, J.; Rodriguez, I. M. An. Quim.,
Ser. C 1985, 81, 190–192.