Dependence of the reactivities of titanium enolates on how they are generated: Diastereoselective coupling of phenylacetic acid esters using titanium tetrachloride

Article abstract of DOI:10.1021/jo952204h

Oxidative coupling of phenylacetic acid esters was easily achieved by treating the esters with TiCl₄ and then adding Et₃N to the resulting solution. The products consisted of dl- and meso-2,3-diphenylsuccinic acid esters with the Claisen condensation product, and the ratio of these products depended on the reaction conditions. Reaction conditions suitable for high dl selectivity were determined, and a dimer of titanium enolate was postulated as an intermediate responsible for the high dl selectivity. The selectivities were compared with those in known oxidative couplings in which titanium enolate intermediates are prepared through lithium enolates and silyl enol ethers. The results suggest that the reactivities of titanium enolates intermediates depend on how they are generated.

Full text of DOI:10.1021/jo952204h

J . Org. Chem. 1996, 61, 2809-2812
2809
Dep en d en ce of th e Rea ctivities of Tita n iu m En ola tes on How Th ey
Ar e Gen er a ted : Dia ster eoselective Cou p lin g of P h en yla cetic Acid
Ester s Usin g Tita n iu m Tetr a ch lor id e
Yoshihiro Matsumura,* Maiko Nishimura, and Hiroyuki Hiu
Faculty of Pharmaceutical Sciences, Nagasaki University, 1-14 Bunkyo-machi, Nagasaki 852, J apan
Mitsuaki Watanabe
Center for Instrumental Analysis, Nagasaki University, 1-14 Bunkyo-machi, Nagasaki 852, J apan
Naoki Kise
Department of Biotechnology, Faculty of Engineering, Tottori University, Koyama, Tottori 680, J apan
Received December 13, 1995X
Oxidative coupling of phenylacetic acid esters was easily achieved by treating the esters with TiCl4
and then adding Et3N to the resulting solution. The products consisted of dl- and meso-2,3-
diphenylsuccinic acid esters with the Claisen condensation product, and the ratio of these products
depended on the reaction conditions. Reaction conditions suitable for high dl selectivity were
determined, and a dimer of titanium enolate was postulated as an intermediate responsible for
the high dl selectivity. The selectivities were compared with those in known oxidative couplings
in which titanium enolate intermediates are prepared through lithium enolates and silyl enol ethers.
The results suggest that the reactivities of titanium enolates intermediates depend on how they
are generated.
Sch em e 1
In tr od u ction
Titanium enolates 1, which are useful intermediates
for carbon-carbon bond formation,1 can be generated
from carbonyl compounds 2 by three methods (Scheme
1): (i) transmetalation of lithium enolates 3, which can
be prepared from 2,2 (ii) transmetalation of silyl enol
ethers 4, which can be prepared from 3,3 and (iii) direct
generation from 2, when 2 are enolizable amides and
carbonyl compounds.4 These three methods give formally
identical titanium enolates 1, but it is not yet clear if
the reactivities of 1 depend on how they are generated.
Since titanium enolates play important roles in organic
synthesis, it is worthwhile to understand all aspects of
their reactivities.
Although Ojima et al. recently showed that there might
be a difference in regioselectivity between titanium
enolates generated by methods i and ii,2 there have been
no previous comparisons of the reactivities of titanium
enolates generated by these three methods. We identi-
fied an efficient homocoupling reaction of phenylacetic
acid esters 5 by method iii, the results of which could be
compared with those of methods i and ii.5 Although a
variety of methods for oxidative coupling of 5 have been
exploited,6 the stereoselectivity obtained by our new
method is the highest among those reported thus far,
and the procedure may be the simplest. This paper
describes our results and also discusses the dependence
of the reactivities of titanium enolates on how they are
generated.
Resu lts a n d Discu ssion
X Abstract published in Advance ACS Abstracts, March 15, 1996.
(1) Reetz, M. T. Organotitanium Reagents in Organic Synthesis;
Springer-Verlag: New York, 1986.
The reaction via method iii has been identified as an
oxidative homocoupling of methyl phenylacetate (5a ).
(2) Ojima, I.; Brandstadter, S. M.; Donovan, R. J . Chem. Lett. 1992,
1591.
(3) (a) Inaba, S.; Ojima, I. Tetrahedron Lett. 1977, 2009. (b) Hirai,
K.; Ojima, I. Tetrahedron Lett. 1983, 24, 785. (c) Reetz, M. T.;
Schwellnus, K.; Hu¨bner, F.; Massa, W.; Schmidt, R. E. Chem. Ber.
1983, 116, 3708. (d) Totten, G. E.; Wenke, G.; Rhodes, Y. E. Synth.
Commun. 1985, 12, 291.
(4) (a) Lehnert, W. Tetrahedron Lett. 1970, 4723. (b) Harrison, C.
R. Tetrahedron Lett. 1987, 28, 4135. (c) Brocchini, S. J .; Eberle, M.;
Lawton, R. G. J . Am. Chem. Soc. 1988, 110, 5211. (d) Evans, D. A.;
Clark, J . S.; Matternich, Nopvack, V. J .; Sheppard, G. S. J . Am. Chem.
Soc. 1990, 112, 866. (e) Evans, D. A.; Rieger, D. L.; Bilodeau, M. T.;
Urpi, F. J . Am. Chem. Soc. 1991, 113, 1047. (f) Evans, D. A.; Bilodeau,
M. T.; Somers, T. C.; Clardy, J .; Cherry, D.; Kato, Y. J . Org. Chem.
1991, 56, 5750. (g) Kise, N.; Tokioka, K.; Matsumura, Y. J . Org. Chem.
1995, 60, 1100. (h) Dikshit, K.; Bajpai, S. N. Tetrahedron Lett. 1995,
36, 3231.
(5) Although 4-isopropyl-3-(phenylacetyl)-2-oxazolidone was coupled
by method iii, it gave a complex mixture of products by method i; see
ref 4g.
(6) There are a variety of methods for the oxidative homocoupling
of phenylacetic acid derivatives. (i) Through enolate anions: (a) Kofron,
W. G.; Hauser, C. R. J . Org. Chem. 1970, 35, 2085. (b) Rathke, M. W.;
Lindert, A. J . Am. Chem. Soc. 1971, 93, 4605. (c) Tokuda, M.; Shigei,
T.; Itoh, M. Chem. Lett. 1975, 621. (ii) Through carboxylic dianions:
(d) Belletire, J . L.; Spletzer, E. G.; Pinhas, A. R. Tetrahedron Lett. 1984,
25, 5969. (e) Renaud, P. Fox, M. A. J . Org. Chem. 1988, 53, 3745. For
the oxidative homocoupling of other esters: (f) Chung, S. K.; Dunn, L.
B., J r. J . Org. Chem. 1983, 48, 1125. (g) Belletire, J . L.; Fremont, S.
L. Tetrahedron Lett. 1986, 27, 127. (h) Belletire, J . L.; Fry, D. F. J .
Org. Chem. 1987, 52, 2549. (i) Porter, N. A.; Su, Q.; Harp, J . J .;
Rosenstein, I. J .; McPhail, A. T. Tetrahedron Lett. 1993, 34, 4457. (j)
Langer, T.; Illich, M.; Helmchen, G. Tetrahedron Lett. 1995, 36, 4409.
0022-3263/96/1961-2809$12.00/0 © 1996 American Chemical Society

2810 J . Org. Chem., Vol. 61, No. 8, 1996
Matsumura et al.
Ta ble 1. Oxid a tive Cou p lin g of Meth yl P h en yla ceta te
(5a ) by Meth od iiia
product 6a
reaction yield
product 7
molar ratio
entry methodb 5a :TiCl4:Et3N temp (°C) (%) dl:meso
yield
(%)
1
2
3
4
5
6
7
8
1:1.1:0
1:0:1.1
0 or -45
0 or -45
-45
0
0
0c
0d
50
0
0
0
This reaction proceeds smoothly to give both stereo-
isomers of the dimer, dl-6a and meso-6a , with the Claisen
condensation product 7, and the ratios of these products
depend on the reaction conditions. The results are
summarized in Table 1.
A
A
A
A
B
B
1:0.5:0.5
1:1.1:1.1
1:2.2:1.1
1:2.2:1.1
1:1.1:1.1
1:2.2:1.1
18
86
83
73
46
20
75:25
95:5
99:1
85:15
75:25
80:20
-45
-45
0
-45
-45
24
50
The main features of this reaction are as follows.
Although the absence of either TiCl4 or Et3N did not
cause any oxidative coupling of 5a (Table 1, entries 1 and
2), the use of both TiCl4 and Et3N7 gave products
consisting of a mixture of stereoisomers of dimer 6a and
the Claisen condensation product 7. The ratio of these
products depended on the order of the addition, the
amounts of TiCl4 and Et3N, and the reaction tempera-
ture. The addition of more than 1 equiv of TiCl4 to a
solution of 5a in CH2Cl2 followed by the addition of more
than 1 equiv of Et3N to the resulting solution (method
A) at -45 °C exclusively gave dl-6a (Table 1, entries 4
and 5),8 while the use of less than 1 equiv of TiCl4 or
Et3N (Table 1, entry 3) resulted in the formation of 7 as
a main product with a mixture of dimer 6a in a dl/meso
ratio which was lower than those observed under the
conditions for entries 4 and 5 of Table 1. The dl/meso
ratio of 6a was also decreased by carrying out the
reaction at 0 °C (Table 1, entry 6). Furthermore, revers-
ing the order of addition of TiCl4 and Et3N (method B)
decreased the dl/meso ratio of 6a and increased the
formation of 7, even though more than 1 equiv of TiCl4
and Et3N was used (Table 1, entries 7 and 8).
a
b
The reaction was carried out for 1.5 h. Method A: addition
of TiCl4 to 5a followed by addition of Et3N. Method B: addition
of Et3N to 5a followed by addition of TiCl4. c Recovery of starting
d
compound was 93%. Recovery of starting compound was 91%.
reverse addition (method B) (Table 1, entries 7 and 8)
might be such cases.
Under reaction conditions similar to those in entry 5,
allyl phenylacetate (5b) gave the dimer dl-6b stereose-
lectively in 80% yield with no meso isomer, no condensa-
tion products, and no cyclized products (eq 2).
Using these data, we attempted to compare the selec-
tivities of methods i, ii, and iii. However, the homocou-
pling of 5a by method i has not been reported, and the
stereoselectivity of the homocoupling of 5a by method ii
has not been described.3a Thus, we attempted the
coupling of 5a by method i and reexamined the reaction
by method ii. The results with method iii are sum-
marized in Table 2, which shows the advantage of method
iii, with respect to product yields and diastereoselectivi-
ties, in comparison with methods i and ii.
The tentative reaction mechanism is shown in Scheme
2. The first step involves the formation of complex A
which is composed of 5a and TiCl4 in a 1:1 ratio. The
formation of A is supported by the fact that peaks of
methyl and methylene hydrogens of 5a in the NMR
spectrum were shifted to a lower field by the addition of
We also compared these three methods in the homo-
coupling reaction of methyl R-methylphenylacetate (5c)
(eq 3, Table 3). In this reaction, both stereo- and regio-
9
TiCl4 to a solution of 5a in CDCl3 and plateaus were
observed with the addition of 1 equiv of TiCl4 (Figure 1).
Upon reaction with an amine, complex A is converted to
titanium enolate B or dimer C, in which two phenyl
groups may be located opposite two titanium enolates in
separate planes. Oxidation of C may lead to dl-6a , while
homocoupling between monomer radical D generated
from B or C may favor the formation of meso-6a rather
than dl-6a , as recognized by the Newman projection E
which shows two D’s in close proximity. The exclusive
formation of dl-6a under the conditions in entries 4 and
5 (Table 1) suggests that such reaction conditions pro-
mote the formation of C. Increasing the reaction tem-
perature may enhance the rate of the transformation of
B or C to D and may result in increased meso-6a .
selectivities differed between the three methods. Thus,
both a mixture of the stereoisomers of the homocoupling
product 6c11 with a dl/meso ratio of ∼50/50 and the R,p-
coupling product 8 were formed using methods i and ii,
while method iii gave 6c with a dl/meso ratio of 30/70
and no 8. In the latter reaction, a dimeric intermediate
similar to C may also play an important role.
The absence of 5a in the reaction system at the
formation of B or C seems to be a key point in preventing
the formation of 7. If 5a exists in the reaction system
when B or C is formed, either of these as nucleophiles
may attack 5a to give 7. The reaction conditions which
use less than 1 equiv of TiCl4 (Table 1, entry 3) and the
Con clu sion
This paper presents a simple and convenient procedure
for the stereoselective homocoupling of phenylacetic acid
esters to give dl-2,3-diphenylsuccinic acid esters. The dl-
dimer was obtained with high stereoselectivity by treat-
ing phenylacetic acid esters with more than 1 equiv of
TiCl4 in CH2Cl2 and then adding Et3N at low tempera-
(7) Diisopropylethylamine and spartein also gave similar results.
(8) The coupling also took place at -78 °C, but a longer reaction
time was required (after 4.5 h, dl-6a /meso-6a ) 95/5, 85% yield).
(9) Oxidative coupling of 5a took place in CHCl3.
(10) We obtained a yield of 27%, while a yield of 76% was described
in ref 3a.
(11) de Luca, C.; Insei, A.; Rampazzo, L. J . Chem. Soc., Perkin Trans.
2, 1982, 1403.
(12) The numbers in parentheses are in ref 2.

Dependence of the Reactivities of Titanium Enolates
J . Org. Chem., Vol. 61, No. 8, 1996 2811
Sch em e 2
Ta ble 3. TiCl4-P r om oted Oxid a tive Cou p lin g of 5c
method
yield (%) of 6c
dl-6c/meso-6c
yield (%) of 8
ia
41 (63)d
50
87
51/49
50/50
30/70
22 (18)d
13
0
iib
iiic
a
Successive addition of LDA (1.0 equiv) and TiCl4 (1.0 equiv)
b
to a solution of 5c in THF at -78 °C. Addition of LDA to 5c,
addition of TMSCl, followed by addition of TiCl4 (1.1 equiv) at 0
°C. c Successive addition of TiCl4 (5.0 equiv) and Et3N (5.0 equiv)
d
to a solution of 5c in CH2Cl2 at -45 °C. See ref 12.
chloride with allyl alcohol in the presence of triethylamine in
CHCl3. Methyl R-methylphenylacetate (5c)2 was prepared by
esterification of commercially available phenylpropionic acid.
Selective F or m a tion of d l-2,3-Dip h en ylsu ccin ic Acid
Dim eth yl Ester (d l-6a ). To a solution of methyl phenylac-
etate (5a ) (3 mmol, 0.45 g) in CH2Cl2 (10 mL) was added TiCl4
(6.6 mmol, 0.72 mL) with a syringe at -45 °C, and the solution
was stirred for 30 min. Triethylamine (6.6 mmol, 0.92 mL)
was then added, and the solution was stirred at -45 °C for
1.5 h. The solution was quenched with saturated aqueous
NH4Cl solution and extracted with CH2Cl2. The extract was
dried on MgSO4, and the solvent was evaporated in vacuo to
give a white solid. The solid was subjected to column chro-
matography (Merck silica gel 60, 70-230 mesh) with hexane/
AcOEt (5/1) to give pure dl-2,3-diphenylsuccinic acid dimethyl
ester (dl-6a ) in 83% yield. Under these reaction conditions,
meso-6a was obtained in a trace amount, and the resulting
Claisen condensation product 7 was negligible. Although dl-
6a is a known compound, only its mp and elemental analysis
data are described in the literature.1 Spectroscopic data of
the resulting dl-6a were as follows.
F igu r e 1. NMR chemical shifts of methyl phenylacetate
(5a ): (9) methyl group, (O) methylene group.
Ta ble 2. TiCl4-P r om oted Oxid a tive Cou p lin g of 5a
method
yield (%) of 6a (dl + meso)
dl-6a /meso-6a
ia
40
50/50
63/37
99/1
iib
iiic
27 (76)d
83
a
Successive addition of LDA (1.25 equiv) and TiCl4 (1.25 equiv)
b
to a solution of 5a in THF at -78 °C. Addition of LDA to 5a ,
dl-6a : mp 163-164 °C (lit.14 mp 165-166 °C); IR (KBr)
addition of TMSCl, followed by addition of TiCl4 (1.1 equiv) at 0
°C. c Successive addition of TiCl4 (2.2 equiv) and Et3N (1.1 equiv)
3030, 2990, 2950, 1730, 1435, 1305, 1250, 1155, 740, 700 cm-1
;
d
1H NMR (CDCl3) δ 3.70 (s, 3H), 4.25 (s, 1H), 6.95-7.08 (m,
2H), 7.08-7.16 (m, 3H); MS 298, 266, 238, 206, 179, 165, 149,
139, 121 (P), 102, 91, 77, 60, 52.
to a solution of 5a in CH2Cl2 at -45 °C. See ref 10.
ture. Furthermore, the results in this paper clearly show
that the stereo- and regioselectivities in the homocoupling
reactions depend on the method used to generate tita-
nium enolates from phenylacetic acid esters. Although
further studies on the reaction mechanism must be
carried out, the high dl selectivity is tentatively explained
in terms of the formation of a dimer of titanium enolates.
Anal. Calcd for C18H18O4: C, 72.46; H, 6.08. Found: C,
72.38; H, 6.10.
P r ep a r a tion of d l-2,3-Dip h en ylsu ccin ic Acid Dim eth yl
Ester (d l-6a ), m eso-2,3-Dip h en ylsu ccin ic Acid Dim eth yl
Ester (m eso-6a ), a n d 2,4-Dip h en yl-3-oxobu tyr ic Acid
Meth yl Ester (7). To a solution of methyl phenylacetate (5a )
(3 mmol, 0.45 g) in CH2Cl2 (10 mL) was added TiCl4 (1.5 mmol,
0.16 mL) with a syringe at -45 °C, and the solution was stirred
for 30 min. Triethylamine (1.5 mmol, 0.21 mL) was then
added, and the solution was stirred at -45 °C for 1.5 h. The
Exp er im en ta l Section
Gen er a l. All solvents were dried and distilled by standard
techniques. Titanium tetrachloride, triethylamine, and methyl
phenylacetate (5a ) were obtained commercially. Allyl phen-
ylacetate (5b)13 was prepared by the reaction of phenylacetyl
(13) Cadogan, J . I. G.; Hey, D. H.; Ong, S. H. J . Chem. Soc. 1965,
1939.
(14) Wren, H.; Still, C. J . J . Chem. Soc. 1915, 107, 444.

2812 J . Org. Chem., Vol. 61, No. 8, 1996
Matsumura et al.
solution was quenched with saturated aqueous NH4Cl solution
and extracted with CH2Cl2. The extract was dried on MgSO4,
and the solvent was evaporated in vacuo to give a residue.
The residue was subjected to column chromatography (Merck
silica gel 60, 70-230 mesh) with hexane/AcOEt (5/1) to give
dl-6a , meso-6a , and 7.
d l-6b: mp 49-54 °C; IR (KBr) 3060, 3040, 2945, 1725, 1600,
1495, 1455, 1370, 1300, 1240, 1160, 985, 940, 735, 700, 570
cm-1; 1H NMR (CDCl3) δ 4.29 (s, 1H), 4.55-4.65 (m, 2H), 5.10-
5.28 (m, 2H), 5.73-5.95 (m, 1H), 6.95-7.20 (m, 5H); MS 352,
336, 292, 280, 264, 251, 223, 207, 193, 180, 165, 149, 147, 129,
119, 105, 91, 83, 69, 57.
d l-6a : 13.5%.
Anal. Calcd for C22H22O4: C, 75.41; H, 6.33. Found: C,
75.29; H, 6.41.
m eso-6a : 4.5%; sublimation temperature 174-179 °C (lit.14
mp 218-219 °C); IR (KBr) 2960, 1730, 1500, 1455, 1435, 1300,
1150, 995, 880, 733, 700 cm-1; 1H NMR (CDCl3) δ 3.40 (s, 3H),
4.40 (s, 1H), 7.25-7.42 (m, 3H), 7.43-7.53 (m, 2H); MS 298,
266, 238, 179, 165, 149 (P), 135, 121 102, 96, 91, 85, 77, 57.
Anal. Calcd for C18H18O4: C, 72.46; H, 6.08. Found: C,
72.05; H, 6.13.
P r ep a r a tion of d l- a n d m eso-2,3-Dim eth yl-2,3-d ip h en -
ylsu ccin ic Acid Dim eth yl Ester (6c). To a solution of
methyl R-methylphenylacetate (5c) (2 mmol, 0.33 g) in CH2-
Cl2 (10 mL) was added TiCl4 (10.0 mmol, 1.10 mL) with a
syringe at -45 °C, and the solution was stirred for 30 min.
Triethylamine (10.0 mmol, 1.40 mL) was then added, and the
solution was stirred at -45 °C for 1.5 h. The solution was
quenched with a saturated aqueous NH4Cl solution and
extracted with CH2Cl2. The extract was dried on MgSO4, and
the solvent was evaporated in vacuo to give a white solid. The
solid was subjected to column chromatography with CH2Cl2
to give a mixture of dl- and meso-2,3-dimethyl-2,3-diphenyl-
succinic acid dimethyl ester (6c) in 87% yield. dl- and meso-
6c are known compounds,11 and the dl/meso ratio was deter-
mined by NMR. R,p-Coupling product 8 was not observed in
this reaction. The authentic sample of 8 was prepared as
previously reported.2 The spectroscopic (IR and 1H NMR
spectra) data of 8 were as follows.
7: 50%; mp 58-60 °C; IR (KBr) 3480, 3140, 2950, 1745,
1720, 1500, 1460, 1435, 1220, 1165, 1060, 715, 700, 550 cm-1
;
1H NMR (CDCl3) δ 3.43 (s, 0.66H), 3.67 (s, 0.99H), 3.69 (d, J
) 16 Hz, 0.67H), 3.71 (s, 2.01H), 3.78 (d, J ) 16 Hz, 0.67H),
4.81 (s, 0.67H), 7.05-7.43 (m, 10H), 13.05 (s, 0.33H); MS 268,
236, 209, 191, 177, 165, 150, 145, 136, 116, 105, 91(P), 77, 65.
Anal. Calcd for C17H16O3: C, 76.10; H, 6.01. Found: C,
76.01; H, 6.08.
P r ep a r a tion of d l-2,3-Dip h en ylsu ccin ic Acid Dia llyl
Ester (d l-6b). To a solution of allyl phenylacetate (5b) (2
mmol, 0.35 g) in CH2Cl2 (10 mL) was added TiCl4 (4.4 mmol,
0.48 mL) with a syringe at -45 °C, and the solution was stirred
for 30 min. Triethylamine (4.4 mmol, 0.61 mL) was then
added, and the solution was stirred at -45 °C for 1.5 h. The
solution was quenched with a saturated aqueous NH4Cl
solution and extracted with CH2Cl2. The extract was dried
on MgSO4, and the solvent was evaporated in vacuo to give a
white solid. The solid was subjected to column chromatogra-
phy with hexane/AcOEt (10/1) to give pure dl-6b in 80% yield.
Under these reaction conditions, the meso-isomer and the
Claisen condensation product were not observed. The stereo-
chemistry of dl-6b was determined by its conversion to dl-6a :
i.e., heating of dl-6b in methanol containing NaOMe at reflux
temperature for 7 h followed by acidifying the resulting
solution to give the corresponding diacid, which was trans-
formed to dl-6a by heating in methanol containing hydrogen
chloride.
8: IR (KBr) 2988, 2951, 1728, 1599, 1495, 1433, 1240, 1167,
1080, 700 cm-1 1H NMR (CDCl3) δ 1.48 (d, J ) 7.1 Hz, 3H),
.
1.90 (s, 3H), 3.63 (s, 3H), 3.67 (m, 1H), 3.69 (s, 3H), 6.81-7.32
(m, 9H).
Oxid a tive Cou p lin g of 5a a n d 5c by Meth od s i a n d ii.
Oxidative coupling by methods i and ii was carried out by the
procedures described in the literature.2,3a
Ack n ow led gm en t. This work was supported in part
by a Grant-in-Aid for Scientific Research on Priority
Areas (No. 236) and for Scientific Research (No.
07672272) from the Ministry of Education, Science and
Culture, J apan.
J O952204H