Mechanism of Mukaiyama-Michael reaction of ketene silyl acetal: Electron transfer or nucleophilic addition?

Article abstract of DOI:10.1021/jo9512968

Mechanism of Mukaiyama-Michael reaction of ketene silyl acetal has been discussed. The competition reaction employing various types of ketene silyl acetals reveals that those bearing more substituents at the β-position react preferentially over less substituted ones. However, when ketene silyl acetals involve bulky siloxy and/or alkoxy group(s), less substituted compounds react preferentially. The Lewis acids play an important role in these reactions. Enhanced preference for the more sterically demanding Michael adducts is obtained with Bu₂Sn(OTf)², SnCl₄, and Et₃-SiClO₄ in the former reaction while TiCl₄ gives the highest selectivity for the less sterically demanding products in the latter case. These results are interpreted in terms of alternative reaction mechanisms. The reaction of less bulky ketene silyl acetals are initiated by electron transfer from these compounds to a Lewis acid. On the other hand, bulkier ketene silyl acetals undergo a ubiquitous nucleophilic reaction. Such a mechanistic change is discussed based on a variety of experimental results as well as the semiempirical PM3 MO calculations.

Full text of DOI:10.1021/jo9512968

J . Org. Chem. 1996, 61, 2951-2962
2951
Mech a n ism of Mu k a iya m a -Mich a el Rea ction of Keten e Silyl
Aceta l: Electr on Tr a n sfer or Nu cleop h ilic Ad d ition ?
J unzo Otera,*,† Yukihiro Fujita,† Nobuyuki Sakuta,† Morifumi Fujita,‡ and
Shunichi Fukuzumi*,‡
Department of Applied Chemistry, Okayama University of Science, Ridai-cho, Okayama 700, J apan, and
Department of Applied Chemistry, Faculty of Engineering, Osaka University, Suita, Osaka 565, J apan
Received J uly 18, 1995X
Mechanism of Mukaiyama-Michael reaction of ketene silyl acetal has been discussed. The
competition reaction employing various types of ketene silyl acetals reveals that those bearing
more substituents at the â-position react preferentially over less substituted ones. However, when
ketene silyl acetals involve bulky siloxy and/or alkoxy group(s), less substituted compounds react
preferentially. The Lewis acids play an important role in these reactions. Enhanced preference
for the more sterically demanding Michael adducts is obtained with Bu2Sn(OTf)2, SnCl4, and Et3-
SiClO4 in the former reaction while TiCl4 gives the highest selectivity for the less sterically
demanding products in the latter case. These results are interpreted in terms of alternative reaction
mechanisms. The reaction of less bulky ketene silyl acetals are initiated by electron transfer from
these compounds to a Lewis acid. On the other hand, bulkier ketene silyl acetals undergo a
ubiquitous nucleophilic reaction. Such a mechanistic change is discussed based on a variety of
experimental results as well as the semiempirical PM3 MO calculations.
The Mukaiyama version of the Michael reaction of
ketene silyl acetals is recognized as an important reaction
in modern synthetic chemistry.1 Of particular signifi-
cance is the role of Lewis acids that can complement an
alternative alkali metal enolate method. The Lewis acids
are generally believed to activate carbonyl substrates
through coordination for nucleophilic attack of ketene
silyl acetals. In some cases, however, Lewis acids have
been reported to act as an electron acceptor from ketene
silyl acetals.2 In this context, we have recently reported
that â-methyl-substitution of ketene silyl acetals in-
creases the electron density rendering the donor more
susceptible to the electron transfer oxidation,3 although
it also generally increases the steric hindrance of the
reaction center, thereby reducing the reactivity of the
nucleophilic attack toward electrophiles.4 Operation of
these two opposing effects leads to the possibility of a
mechanistic change from ubiquitous nucleophilic pro-
cesses to novel electron transfer pathways. The electron
transfer vs nucleophilic process dichotomy is one of the
central propositions in the reaction mechanism,5 and thus
parallel assessment of both processes would be valuable.
Although discussion of the nucleophilic mechanism has
so far been advanced mostly on the basis of the product
stereochemistry, it is desirable to gain insight about the
transition state on the basis of the reactivities.
We report herein for the first time the extensive
comparison of the reactivities of various ketene silyl
acetals in the MukaiyamasMichael reaction.6 The rela-
tive reactivities of ketene silyl acetals have been found
to vary depending on the â-methyl-substitution and silyl
groups of ketene silyl acetals, substrates, and also Lewis
acids. Such change in the reactivity of ketene silyl
acetals particularly with â-methyl-substitution may pro-
vide valuable insight into the electron transfer vs nu-
cleophilic process dichotomy.
Resu lts
Com p etition Rea ction . The relative reactivities of
ketene silyl acetals toward R-enone in the presence of
Lewis acid were determined from competition experi-
ments. When two different ketene silyl acetals A and B
react with the same R-enone in the presence of a Lewis
acid, the ratio of reaction rate, {-d[A]/dt}/{-d[B]/dt} )
d[A]/d[B] may be given by eq 1,
d[A]/d[B] ) (kA/kB)[A]/[B]
(1)
† Okayama University of Science.
where kA and kB are the rate constants for ketene silyl
acetals A and B, and [A] and [B] are the concentrations,
respectively. From eq 1 is derived eq 2, where [A]0 and
[B]0 are the initial concentration of A and B, respectively.
‡ Osaka University.
X Abstract published in Advance ACS Abstracts, March 15, 1996.
(1) Oare, D. A.; Heathcock, C. H. In Topics in Stereochemistry; Eliel,
E. L., Wilen, S. H., Eds.; Wiley: New York, 1991; Vol. 20, p 87.
(2) (a) Reetz, M. T.; Schwellnus, K.; Hu¨bner, F.; Massa, W.; Schmidt,
R. E. Chem. Ber. 1983, 116, 3708. (b) Totten, G. E.; Wenke, G.; Rhodes,
Y. E. Synth. Commun. 1985, 15, 291. (c) Ali, S. M.; Rousseau, G.
Tetrahedron 1990, 46, 7011. (d) Quendo, A.; Ali, S. M. Rousseau, G. J .
Org. Chem. 1992, 57, 6890. (e) Odenkirk, W.; Whelan, J .; Bosnich, B.
Tetrahedron Lett. 1992, 33, 5729.
ln([A]/[A]0)/ln([B]/[B]0) ) kA/kB
(2)
Since the product yields YA (%) and YB (%) derived from
A and B are given by YA/100 ) 1 - [A]/[A]0 and YB/100 )
1 - [B]/[B]0, respectively, from eq 2 is derived the ratio
of the reactivity of two different ketene silyl acetals (r )
kA/kB) as given by eq 3.7
(3) Fukuzumi, S.; Fujita, M.; Otera, J .; Fujita, Y. J . Am. Chem. Soc.
1992, 114, 10271.
(4) Mayr, H.; Patz, M. Angew. Chem., Int. Ed. Engl. 1994, 33, 938.
(5) Pross, A.; Shaik, S. S. Acc. Chem. Res. 1983, 16, 363. Pross, A.
Acc. Chem. Res. 1985, 18, 212. Shaik, S. S. Prog. Phys. Org. Chem.
1985, 15, 197. Pradhan, S. K. Tetrahedron 1986, 42, 6351. Kochi, J .
K. Angew. Chem., Int. Ed. Engl. 1988, 27, 1227. Ashby, E. C. Acc.
Chem. Res. 1988, 21, 414. Save´ant, J .-M. Adv. Phys. Org. Chem. 1990,
26, 1. Shaik, S. S. Acta Chem. Scand. 1990, 44, 205. Fukuzumi, S. In
Advances in Electron Transfer Chemistry; Mariano, P. S., Ed.; J AI
Press: Greenwich, 1992; Vol. 2, p 65. Mattalia, J .-M.; Vacher, B.;
Samat, A.; Chanon, M. J . Am. Chem. Soc. 1992, 114, 4111 and
references cited therein.
r ) kA/kB ) ln(1 - YA/100)/ln(1 - YB/100) (3)
The reactivity ratios (r) of all combinations of â,â-
disubstituted ketene silyl acetals (2a -c), monosubsti-
S0022-3263(95)01296-5 CCC: $12.00 © 1996 American Chemical Society

2952 J . Org. Chem., Vol. 61, No. 9, 1996
Ta ble 1. Com p etition betw een â,â-Disu bstitu ted a n d Un su bstitu ted Keten e Silyl Aceta ls
Otera et al.
yield, %
entry
1
2
4
8
5
7
5:7
R1
R4
Et
Et
Et
Et
Et
Et
Et
r
(rpr)
1
2
3
4
5
6
7
8
9
10
11
12
13
1a (R2, R3 ) Me)
2a
4a
8a
8b
8c
8d
8b
8c
8d
8b
8b
8c
8d
8c
8d
5a , 85
95
7a , 0
100:0
100:0
100:0
97:3
100:0
93:7
(71:29)
100:0
9:91
7:93
5:95
6:94
0:100
Ph
Ph
Ph
Ph
100:0
100:0
100:0
99:1
100:0
95:5
77:23
100:0
7:93
6:94
3:97
6:94
0:100
(100:0)
(100:0)
(100:0)
(90:10)
(100:0)
(97:3)
(85:15)
(100:0)
(1:99)
(1:99)
(4:96)
(20:80)
(5:95)
0
0
3
93
97
1b (R2, R3 ) Me)
2a
4a
5b, 21
53
7b, 0
4
tBu
tBu
tBu
tBu
tBu
tBu
tBu
tBu
tBu
62
25
1c (R2 ) H, R3 ) Me)
2a
2b
4a
4b
5c, 45
5d , 3
3
7c, 0
7d , 34
38
Et
Mea
Mea
Mea
Bor
Bor
4
72
2c
4c
5e, 4
0
7e, 50
69
a
Determined as the methyl ester: see Experimental Section.
tuted (3a -c), and unsubstituted (4a -c) counterparts
toward R-enones (1a -c) were fully examined in the
presence of various Lewis acids: Bu2Sn(OTf)2 (8a ); SnCl4
(8b); Et3SiClO4 (TESClO4) (8c); TiCl4 (8d ). The results
of competition between 2a -c and 4a -c, 2a -c and 3a -
c, and 3a -c and 4a -c are shown in Tables 1-3. The
competition between â,â-disubstituted ketene silyl acetal
2a and the unsubstituted counterpart 4a toward R-enone
1a results in preferential formation of more hindered
adduct 5a over 7a (Table 1, entries 1-4). The preference
is remarkable as shown by the >99:1 reactivity ratio, r2/4
,
in the cases of Lewis acids 8a -c (the sum of ratio is taken
as 100).8 It should be noted, however, that 8d gave rise
to lower selectivity (97:3). Such remarkably enhanced
(6) For a preliminary communication: Sato, T.; Wakahara, Y.; Otera,
J .; Nozaki, H.; Fukuzumi, S. J . Am. Chem. Soc. 1991, 113, 4028.
(7) Although eq 3 can be reduced to a simple relation, kA/kB ) YA/
YB, under the experimental conditions such that the concentrations of
A and B are much greater than that of R-enone, the competition
experiments were performed under the normal conditions for the
synthetic purpose in order to avoid the use of extremely high
concentrations of A and B (.1 M).
reactivity by â-substitution with methyl group is also
observed for the competition between 2a and monosub-

Mechanism of Mukaiyama-Michael Reaction
J . Org. Chem., Vol. 61, No. 9, 1996 2953
Ta ble 2. Com p etition betw een â,â-Disu bstitu ted a n d â-Mon osu bstitu ted Keten e Silyl Aceta ls
yield, %
entry
1
2
3
8
5
6
5:6
R1
R4
r
1
2
1a (R2, R3 ) Me)
2a
3a
8a
8b
8c
8d
8b
8c
8d
8b
8b
8c
8d
8c
8d
5a , 84
89
6a , 0
100:0
92:8
Ph
Et
100:0
96:4
8
Ph
Et
3
88
12
28
6b, 2
27
32
6c, 6
6d , 86
55
86
6e, 36
74
88:12
70:30
85:15
67:33
63:27
91:9
14:86
7:93
1:99
Ph
Ph
Et
94:6
4
65
Et
76:24
92:8
5
1b (R2, R3 ) Me)
2a
3a
5b, 22
53
tBu
tBu
tBu
tBu
tBu
tBu
tBu
tBu
tBu
Et
6
Et
71:29
67:33
94:6
7
55
Et
8
1c (R2 ) H, R3 ) Me)
2a
2b
3a
3b
5c, 60
5d , 14
4
Et
9
Mea
Mea
Mea
Bor
Bor
3:93
10
11
12
13
5:95
1
1:99
2c
3c
5e, 25
29
41:59
28:72
39:61
20:80
a
Determined as the methyl ester; see Experimental Section.
Ta ble 3. Com p etition betw een â-Mon osu bstitu ted a n d Un su bstitu ted Keten e Silyl Aceta ls
yield, %
entry
1
3
4
8
6
7
6:7
R1
R4
r
1
2
1a (R2, R3 ) Me)
3a
4a
8a
8b
8c
8d
8b
8c
8d
8b
8b
8c
8d
8c
8d
6a , 67
69
7a , 0
100:0
100:0
86:14
69:31
100:0
90:10
69:31
97:3
Ph
Et
100:0
100:0
92:8
75:25
100:0
92:8
73:27
97:3
5:95
6:94
0
Ph
Et
3
78
13
26
7b, 0
5
23
7c, 1
7d , 43
36
47
7e, 48
59
Ph
Ph
Et
4
59
Et
5
1b (R2, R3 ) Me)
3a
4a
6b, 12
43
tBu
tBu
tBu
tBu
tBu
tBu
tBu
tRu
tBu
Et
6
Et
7
51
Et
8
1c (R2 ) H, R3 ) Me)
3a
3b
4a
4b
6c, 28
6d , 3
3
Et
9
6:94
Mea
Mea
Mea
Bor
Bor
10
11
12
13
7:93
38
6e, 22
14
45:55
31:69
19:81
43:57
28:72
16:84
3c
4c
a
Determined as the methyl ester: see Experimental Section.
stituted compound 3a and also between 3a and the
unsubstituted counterpart 4a as shown in Tables 2 and
3, respectively. The same trend also holds with R-enones
1b and 1c (entries 5-8, Table 1-3). Since the reactivity
ratio of 2a -c vs 4a -c in Table 1 can be derived from
that of 2a -c vs 3a -c together with that of 3a -c vs 4a -
c, the consistency of data is checked by comparing the
reaction with the aforementioned less sterically demand-
ing ketene silyl acetals (2a , 3a , 4a ). These normal
selectivities in terms of the conventional nucleophilic
mechanism can be attributed to incorporation of large
TBS and tert-butyl groups. Thus, ketene silyl acetals (2c,
3c, and 4c) which were derived from bornyl esters
exhibited similar tendencies (entries 12, 13 in Table 1-3).
The normal selectivity was attained even with less
bulky ketene silyl acetals under certain other conditions.
For example, the ether-LiClO4 system which had been
found to effect the Michael reaction of ketene silyl acetals
by Grieco et al.9,10 afforded predominantly the less
hindered products in the competition reactions (Scheme
1).11 Et2AlCl is another Lewis acid which gave rise to
the same outcome (Scheme 2). Further noteworthy is the
reaction of O,S-ketene silyl acetals: even the trivial
Lewis acids such as SnCl4, TESClO4, and TiCl4 exhibited
the normal selectivity (Table 4).
r2/4 values in Table 1 with the products (rpr ) r2/3 × r3/4
)
of the corresponding r values in Table 2 (r2/3) and 3 (r3/4),
respectively. The rpr values as listed in Table 1 agree
qualitatively with the experimental values determined
directly, demonstrating the validity of the reactivity
ratios.
Noticeably, a dramatic change was seen when a series
of tert-butyldimethylsilyl (TBS) enolates of tert-butyl
esters (2b, 3b, and 4b) were employed (entries 9-11 in
Table 1-3). These ketene silyl acetals sluggishly reacted
with â,â-disubstituted enones but with â-monosubsti-
tuted one 1c in reasonable yields. Among the Lewis
acids, TiCl4 induced the highest selectivities except in the
case of the entry 11 in Table 3 in sharp contrast to the
In flu en ce of Lew is Acid . It is apparent from the
above results that the Lewis acids have a profound
(9) Grieco, P. A.; Cooke, R. J .; Henry, K. J .; VanderRoest, J . M.
Tetrahedron Lett. 1991, 32, 4665.
(10) Reetz et al. has revealed the CH2Cl2-LiClO4 system to be more
powerful to trigger the Michael reaction: Reetz, M. T.; Fox, D. N. A.
Tetrahedron Lett. 1993, 34, 1119.
(8) The ratio of 100:0 includes the experimental error probably
within (2%. Thus, the ratio does not mean the infinite value. This
means that the rate constant of 2 is at least two orders of magnitude
greater than that of 4.

2954 J . Org. Chem., Vol. 61, No. 9, 1996
Otera et al.
Sch em e 1
Sch em e 2
Ta ble 4. Com p etition betw een â,â-Disu bstitu ted a n d
â-Mon osu bstitu ted O,S-Keten e Silyl Aceta ls
Ta ble 5. TiCl4-P r om oted Cou p lin g of Keten e Silyl Aceta l
3
Sil
R
yield, %
3a
3b
3c
TES
TBS
TBS
Et
41
tBu
Bor
63 (18)a
52 (48)a
a
yield, %
The recovery of the propionate ester is given in the parenthe-
ses.
8
14
15
14:15
8b
8c
8d
5
1
0
82
96
98
6:94
1:99
0:100
of ether to this solution caused no precipitation indicative
of absence of free SnCl4.13 The solution was warmed and
volatile materials were evaporated in vacuo. NMR
spectra of the residue in CD2Cl2 unambiguously con-
firmed R-stannyl esters to be produced quantitatively as
shown in Table 6.14
influence on the selectivity. Therefore, it is important
to disclose the difference between the Lewis acids. Ojima
et al.12 and Rhodes et al.2b reported that ketene silyl
acetals were converted to succinates upon treatment with
TiCl4 at room temperature. As Table 5 shows, the same
reaction occurs even at -78 °C, the temperature at which
our Michael reactions were carried out. In some cases,
parent esters were detected simultaneously. The reac-
tion with SnCl4, on the other hand, proceeded differently.
A CH2Cl2 solution of SnCl4 and ketene silyl acetals (1:1
molar ratio) was stirred at -78 °C for 30 min. Addition
(11) Use of acyclic R-enones resulted in poor yields under the same
reaction conditions.
(12) (a) Inaba, S.-I.; Ojima, I. Tetrahedron Lett. 1977, 2009. (b) Hirai,
K.; Ojima, I. Tetrahedron Lett. 1983, 24, 785.
(13) It was confirmed that, if SnCl4 was present, an exothermic
reaction immediately occurred to give precipitates of the SnCl4-ether
complex.
(14) Preparation of R-stannyl ketones from enol silyl ethers by an
analogous transmetalation was reported: Nakamura, E.; Kuwajima,
I. Tetrahedron Lett. 1983, 24, 3347.

Mechanism of Mukaiyama-Michael Reaction
Ta ble 6. F or m a tion of r-Sta n n yl Ester s
J . Org. Chem., Vol. 61, No. 9, 1996 2955
Ta ble 8. Isola tion of Silyl En ol Eth er In ter m ed ia tes
ketene silyl acetal
yield, %a
2a
2b
2d
3a
3b
3d
(R1, R2 ) Me, R3 ) Et, Sil ) TES)
(R1, R2 ) Me, R3 ) tBu, Sil ) TBS)
(R1, R2, R3 ) Me, Sil ) TMS)
81
89
100
96
96
94
(R1 ) Me, R2 ) H, R3 ) Et, Sil ) TES)
(R1 ) Me, R2 ) H, R3 ) tBu, Sil ) TBS)
(R1, R3 ) Me, R2 ) H, Sil ) TMS)
yield, %b
2 or 3
conditionsa
16
5 or 6
Determined by 1H NMR.
a
2a
A
B
A
B
A
B
16a , 56
69
16b, 77
64
16c, 58
55
5a , 38
<1
Ta ble 7. Effect of th e Am ou n t of Sn Cl4
2d
3a
5g, 6
0
6a , 42
20
a
b
A: Et3N/hexane quench; B: in CH2Cl2/THF. Isolated yield.
Ta ble 9. Isom er iza tion of â,â-Disu bstitu ted r-En on e
equiv amount
d u r in g Mich a el Ad d ition
ketene silyl acetal
of SnCl4
yield, %
3d (R1, R3 ) Me, R2 ) H, Sil ) TMS)
1.0
0.1
1.0
0.2
0.1
0.05
6f, 6
68
5f, 20
70
2a (R1, R2 ) Me, R3 ) Et, Sil ) TES)
67
80
ketene silyl
acetal
conversion,
%
E:Z of
diastereomer
1e
8
recovred 1e ratio of 6g
E:Z ) 99:1 2a (R1,
R2 ) Me)
8a
44
90:10
8b
8c
8a
8b
8c
74
84
49
61
88
79
92:8
88:12
22:78
22:78
23:77
E:Z ) 1:99
E:Z ) 99:1 3a (R1 ) Me; 8a
90:10
56:44
R2 ) H)
8b
68
64
68
52
88:12
12:88
33:67
13:87
55:45
45:55
42:58
44:55
E:Z ) 1:99
8a
8b
8c
low-field shift than SnCl4 and formation of a 1:2 TiCl4-
1c complex is evident.
F igu r e 1. Dependence of 13C chemical shifts of the C3 and C5
signals of 2,2-dimethyl-4-hexen-3-one (1c) upon the dosage of
Lewis acids. (1) C5 upon TiCl4, (2) C3 upon TiCl4, (3) C3 upon
SnCl4, and (4) C5 upon SnCl4.
Enol silyl ethers of the Michael adducts 16 were
obtained by quenching the Bu2Sn(OTf)2-catalyzed reac-
tion with Et3N (Table 8). Moreover, when the reaction
was conducted in CH2Cl2-THF (9:1), the silyl ethers are
feasible even after aqueous workup.
In connection with the above findings, a rather unex-
pected effect of the SnCl4 catalyst was found: the greater
the amount of SnCl4 catalyst used, the lower the yield of
the Michael adduct. Reaction of 3d with 1d in the
presence of a stoichiometric amount of SnCl4 afforded
only a 6% yield of the Michael adduct while a 68% yield
was obtained with 0.1 equiv of SnCl4 (Table 7). Similar
results are given for 2a .
Isom er iza tion of R-En on e. To gain further informa-
tion about the reaction mechanism we prepared an
unsymmetrically disubstituted R-enone 1e and investi-
gated the stereochemical isomerization during the Michael
reaction (Table 9). Exposure of Lewis acids (1.0 equiv)
such as 8a , 8b, or 8c to 1e in CH2Cl2 at -78 °C induced
no isomerization of the enone. TiCl4, however, resulted
in the isomerization probably due to the HCl concomitant
and, hence, was not used for the present purpose.
Reaction of 1e with ketene silyl acetals in the presence
of the above Lewis acids was quenched at lower conver-
sions, and the stereochemistry of the recovered enone was
analyzed by GLC. As seen in Table 9, 1e were isomerized
in all cases. These results give a clue to the reaction
mechanism as will be described later.
Such variations of the reactivity with SnCl4 and TiCl4
may be partly associated with the different coordinating
ability toward enone, which was confirmed by 13C NMR
spectroscopy. Figure 1 illustrates the change of the
chemical shifts of the C3 and C5 signals of 2,2-dimethyl-
4-hexen-3-one (1c) upon increasing the amount of added
Lewis acids. Obviously, TiCl4 causes more substantial

2956 J . Org. Chem., Vol. 61, No. 9, 1996
Otera et al.
F igu r e 2. Change of the enthalpy of reaction between 4a and
Ph(CO)CHdCH2-SnCl4 in terms of the dihedral angle (φ). The
sum of the ∆Hf values of the reactants is taken as zero.
MO Ca lcu la tion s for Nu cleop h ilic P r ocess. The
semiempirical MO calculations have been generally
considered the method of choice for studying the struc-
tures and energies of transition states of complicated
systems, otherwise impossible to tackle by ab initio
calculations.15,16 In the hope of obtaining better insight
into the nucleophilic mechanism, we invoked the theo-
retical calculations at the restricted Hartree-Fock (RHF)
level using the PM3 semiempirical SCF-MO method16-18
for the transitions states in the SnCl4-catalyzed nucleo-
philic reaction of ketene silyl acetals with R-enones.
First, the reaction 4a and Ph(CO)CHdCH2-SnCl4 com-
plex was analyzed. The activation enthalpy (∆Hq) is
evaluated as the difference in the ∆Hf (heat of formation)
values of the reaction pair from the sum of the ∆Hf values
of reactants. The ∆Hq is plotted against the dihedral
angle formed by the approaching two CdC bonds (φ) at
a fixed bond distance between the two reacting centers
(2.11 Å) as shown in Figure 2. There appear seven
minima, each of which corresponds to different transition
state geometries when the bond distance between the two
reacting centers is varied. Several transition state
geometries are feasible in this reaction for cases where
both reaction centers are not congested. By the same
method, the reactions between 2a and 1a -SnCl4 and
between 4a and 1a -SnCl4 were compared (the reaction
given as entry 2 in Table 1). In these cases, there also
appeared several minima by changing the dihedral angle
(Figure 3); however, one of them in each case (A an B) is
appreciably lower in energy than the others, indicative
of being the transition states with fixed geometries. The
F igu r e 3. Change of the enthalpy of reactions between 2a
and 1a -SnCl4 (upper) and between 4a and 1a -SnCl4 (lower)
in terms of the dihedral angle (φ).
reaction coordinate energy profiles of these transition
states in terms of the bond distance between the reacting
centers are shown in Figure 4.19 The structures of the
transition states A and B are illustrated in Figure 5. The
stationary points on the reaction coordinate energy
profiles were well characterized as the transition states
by calculating and diagonalizing the Hessian matrix
which had only one negative eigenvalue (see Experimen-
tal Section). The ∆Hq value of the reaction with 2a (A:
24.7 kcal mol-1) is significantly larger than that with 4a
(B: 10.1 kcal mol-1) by 14.6 kcal/mol (Figure 4). The
activation entropy (∆Sq) is also evaluated as the differ-
ence in the ∆S values of the reaction pair from the sum
of the ∆S values of the reactants. The activation free
energy (∆Gq ) ∆H - T∆Sq) of 2a is also significantly
(15) (a) Dewar, M. J . S.; Healy, E. F.; Stewart, J . J . P. J . Chem.
Soc. Faraday Trans. 2 1984, 80, 227. (b) Saa´, J . M.; Deya´, P. M.; Sun˜er,
G. A.; Frontera, A. J . Am. Chem. Soc. 1992, 114, 9093 and references
cited therein.
(16) The PM3 semiempirical calculations have been shown to be
highly useful for predicting the transition-state structures of reactions
of organic compounds containing heteroatoms. See: Alnajjar, M. S.;
Franz, J . A. J . Am. Chem. Soc. 1992, 114, 1052.
larger than that of 4a by 13.6 kcal mol-1 20,21
These
.
calculations suggest that nucleophilic attack by less
substituted ketene silyl acetals should be orders of
(17) The PM3 calculations were carried out using the MOPAC
Molecular Orbital Program Package, QCPE 455 (Ver. 6.0), Quantum
Chemistry Program Exchange, Department of Chemistry, Indiana
University, Bloomington, IN. (a) Stewart J . J . P. J . Comput. Chem.
1989, 10, 209 and 221. (b) Dewar, M. J . S. J . Comput. Chem. 1990,
11, 541. (c) Stewart, J . J . P. J . Comput. Chem. 1990, 11, 543.
(18) It was confirmed that fully optimized ab initio SCF-MO
calculation using the 3-21 G basis set gave essentially the same
structure of the R-enone-SnCl4 complex as the PM3 calculation (see
Experimental Section).
(19) Since the calculations for the reaction coordinate are performed
by optimizing the total molecular energy with respect to all structural
variables, the dihedral angles of the transition states (Figures 5a and
5b) are somewhat different from those of minima at a fixed bond
distance between the two reacting centers in Figure 3.
(20) The ∆S values were calculated at 298 K. The inclusion of
entropy term results in no significant change in the relative difference
in the transition state energies of 4a and 2a .

Mechanism of Mukaiyama-Michael Reaction
J . Org. Chem., Vol. 61, No. 9, 1996 2957
silyl acetals on their nucleophilicity is not apparent, the
comparison with trimethylsilyl allylic compounds with
methyl group(s) at the γ-positions is given: the nucleo-
philicity parameter (N) varies from 1.62 (allyltrimethyl-
silane) to 1.73-1.99 (crotyltrimethylsilane) and 0.84
(prenyltrimethylsilane). Evidently, monomethyl substi-
tution increases the nucleophilicity slightly, but dimethyl
substitution decreases half the reactivity. Accordingly,
the enhanced reactivity of ketene silyl acetals with more
â-substituents in the Mukaiyama-Michael reaction is
incompatible with the conventional nucleophilic mecha-
nism (Table 1-3). In fact, RajanBabu found that reaction
of an equimolar mixture of â,â-disubstituted and â-mono-
substituted ketene silyl acetals with cyclopentenone gave
the Michael adducts derived from both ketene silyl
acetals in ca. 1:1 ratio under thermal conditions where
the nucleophilic reaction is likely to proceed.28 Since the
facile connection of more hindered sites is characteristic
of radical reactions, we propose the mechanism involving
radical coupling that is shown representatively for the
SnCl4-mediated reaction in Scheme 3. The initial step
is electron transfer from ketene silyl acetal to SnCl4 to
generate a cation radical of the ketene silyl acetal and
an SnCl4 anion radical. The SnCl4 anion radical spon-
taneously decomposes to SnCl3 radical and Cl-. The
SnCl3 radical reacts with an R-enone in a 1,4-fashion to
give a stannyl enolate radical which subsequently couples
with the cation radical to give a stannyl enolate of the
Michael adduct. Transmetalation of this intermediate
with in situ-formed trialkylsilyl chloride produces the
corresponding silyl enol ether and regenerates SnCl4.
The results of the reaction of ketene silyl acetals with
TiCl4 (Table 5) and SnCl4 (Table 6) are consistently
accounted for in terms of electron transfer. As Reetz et
al.2a and Rhodes et al.2b suggested, an electron transfer
from ketene silyl acetal to TiCl4 to give an ester radical,
TiCl3 radical, and Cl-. TiCl3 radical may combine with
the ester radical but the titanium-carbon bond is so
thermally labile that titanium binds oxygen to generate
a titanium enolate. Alternatively, direct homo coupling
of the ester radical is also feasible without passing
through the titanium enolate intermediate. Analogous
electron transfer from SnCl4 produces an SnCl3 radical
that couples with the ester radical to afford the thermally
stable R-stannyl esters. Thus, both of the above reactions
can be explained by an initial electron transfer and the
discrepancy of the reaction course can be ascribed
straightforwardly to the difference in metal-carbon bond
character.
F igu r e 4. Reaction coordinate energy profiles of the transition
states A and B.
magnitude faster than those of more substituted. This
stands in sharp contrast to the experimental observa-
tions.
Con n ect ion of Con t igu ou s Qu a t er n a r y Ca r b on
Cen ter s. Another important synthetic aspect to be
emphasized in relation to the present process is its ability
to construct structures with contiguous quaternary car-
bon centers, a challenging task for synthetic chemists.22
Although Michael reactions hold great promise for this
purpose, conventional lithium enolate methodologies fail
to give satisfactory results.23 For example, Dauben et
al. utilized a high pressure technology to connect poten-
tial quaternary carbon centers, but succeeded only with
activated enolates.24 Casey revealed that a chromium
vinylcarbene complex worked as a Michael acceptor.25
Holton employed acceptors doubly substituted by electron-
withdrawing groups.26 Nonetheless, these methods suffer
from various problems such as low yields, difficulties in
preparing raw materials, and tedious procedures. Obvi-
ously, the easy preparation of Michael adducts disclosed
above proves the Mukaiyama-Michael reaction to be
highly promising. In fact, Mukaiyama has already shown
some simple examples along this line,27 but the versatility
of this method has not been fully recognized. It should
be noted that successful preparation of compounds 5i-k
broadens the scope of this method.
The suppression of the reaction with increasing amounts
of SnCl4 (Table 7) is also incompatible with the nucleo-
philic mechanism where the increase in the amount of
Lewis acid should afford higher yields or at least depres-
sion of the yield should not occur. The concentration of
free SnCl4 that is not coordinated with R-enone may
increase with an increase in the SnCl4 concentration. In
such a case, electron transfer from ketene silyl acetal to
free SnCl4 may compete with that to the SnCl4-R-enone
complex, resulting in the formation of R-stannyl ester via
the coupling of the ester radical and SnCl3 radical in
O
O
O
O
O
O
Ph
OMe
Ph
OMe Ph
OMe
5i (93%)
5j (92%)
5k (97%)
Discu ssion
Mayr has broadly assessed the reactivity of π-nucleo-
philes.4 Although the effect of â-substituents in ketene
(21) The ∆Gq value of the reaction between 4a and Ph(CO)-
CHdCH2-SnCl4 complex, where the reaction center is least congested,
is also evaluated as 19.6 kcal mol-1 (∆Hq ) 3.2 kcal mol-1), which is
the smallest among these reactions.
(22) For relevant studies, see literatures cited in ref 6.
(23) Oare, D. A.; Henderson, M. A.; Sanner, M. A.; Heathcock, C.
H. J . Org. Chem. 1990, 55, 132.
(24) Dauben, W. G.; Gerdes, J . M. Tetrahedron Lett. 1983, 24, 3841.
Dauben, W. G.; Bunce, R. A. J . Org. Chem. 1983, 48, 4642.
(25) Casey, C. P.; Brunsvold, W. R. Inorg. Chem. 1977, 16, 391.
(26) Holton, R. A.; Williams, A. D.; Kennedy, R. M. J . Org. Chem.
1986, 51, 5482.
(27) Saigo, K.; Osaki, M.; Mukaiyama, T. Chem. Lett. 1976, 163.
Kobayashi, S.; Mukaiyama, T. Chem. Lett. 1986, 1805. Kobayashi, S.;
Mukaiyama, T. Chem. Lett. 1987, 1183. Hashimoto, Y.; Sugumi, H.;
Okauchi, T.; Mukaiyama, T. Chem. Lett. 1987, 1691. Minowa, N.;
Mukaiyama, T. Chem. Lett. 1987, 1719. Kobayashi, S.; Tamura, M.;
Mukaiyama, T. Chem. Lett. 1988, 91.
(28) RajanBabu, T. V. J . Org. Chem. 1989, 49, 2083.

2958 J . Org. Chem., Vol. 61, No. 9, 1996
Otera et al.
F igu r e 5. (a) Transition state structure of A. (b) Transition state structure of B.
Sch em e 3
competition with the formation of the Michael adduct. It
was confirmed that no reaction occurred between the
R-stannyl ester and R-enone. No suppression of the
Michael reaction was observed by use of the stoichiomet-
ric TiCl4. The large formation constant for the TiCl4-
R-enone complex as compared with that for the SnCl4-
R-enone complex (Figure 1) accounts well for such
differences between SnCl4 and TiCl4.
The ketene silyl acetals with bulky group(s) exhibit the
normal selectivity that is deduced from the nucleophilic
process (Table 1-3). The nucleophilic reaction leading
to such selectivity is shown in the reaction employing
LiClO4 (Scheme 1) and Et2AlCl (Scheme 2), both of these
Lewis acids being incapable of working as an electron
acceptor. We have previously disclosed that the bulky
ketene silyl acetals are less readily oxidized than the

Mechanism of Mukaiyama-Michael Reaction
J . Org. Chem., Vol. 61, No. 9, 1996 2959
Sch em e 4
corresponding less bulky analogs.3 Thus, the competition
reaction employing O,S-ketene silyl acetals is noteworthy
since the oxidation potentials of these compounds are
higher than those of the O,O-analogs and even greater
than those of the bulky ketene silyl acetals involving TBS
by involvement of the electron transfer process. (The
relative stereochemistry was not determined.)
TiCl4 promotes a variety of reactions. However, reac-
tion between 1a and 2b afforded only a 10% yield of the
desired Michael adduct together with the recovered 1a
(80%) (Scheme 4). While one might be tempted to
attribute the poor yield to the homo coupling of 2b, this
is not the case; no succinate was detected. We reason
that electron transfer from 2b to TiCl4 is suppressed in
the presence of R-enone as described above. An alterna-
tive nucleophilic reaction between sterically crowded
latent quaternary centers would be difficult. When 1e
was exposed to bulky 2b in the presence of Et3SiClO4,
only isomerization of 1e occurred. Remarkably, when the
reaction was conducted in the presence of SnCl4, an
R-stannyl ester (60%) was produced in addition to the
isomerized R-enone. This contrasts with TiCl4 that failed
to provide the succinate in the analogous reaction be-
tween 2b and 1a . Obviously, the R-stannyl ester did not
result from the direct transmetalation between 2b and
SnCl4, but the R-enone 1e was involved in the enolate
exchange process. These results again lend support to
the facile electron transfer in SnCl4- and TESClO4-
mediated reactions. The failure of the Michael reaction
implies that sterically demanding reaction centers do not
readily undergo radical coupling.
t
and BuO groups.3 In line with this, O,S-acetals gave
rise to the normal selectivity (Table 4).
In the competition reactions between less bulky ketene
silyl acetals, the preference for the more hindered
Michael adducts was best achieved in the order SnCl4 >
Et3SiClO4 > TiCl4. On the other hand, the opposite order
is found for the less hindered adducts with bulky ketene
silyl acetals. These results are interpreted on the basis
of strong oxophilicity of TiCl4. The reduction potential
of TiCl4 may be decreased by the strong coordination with
R-enone, and thus the electron-transfer pathway of less
bulky ketene silyl acetals may be retarded to some extent,
resulting in the worst selectivity. By contrast, the
electron transfer is most efficiently suppressed in the
TiCl4-promoted reaction of less oxidizable bulky ketene
silyl acetals since TiCl4 is no longer a good electron
acceptor due to tight complexation.
In the electron-transfer initiated pathway in Scheme
3 which involves the equilibrium step, it is presumed that
the double bond of an R-enone might be isomerized
during the reaction. This is indeed the case (Table 9).
Such isomerization may also be envisioned to occur in
the nucleophilic process (eq 4). However, this possibility
has already been ruled out by Heathcock et al.; while k1
and k-1 are comparable with enol silyl ethers (R′ ) alkyl),
the reaction of strongly nucleophilic ketene silyl acetals
(R′ ) alkoxyl) has an early transition state, biasing the
equilibrium to the right (k1 . k-1).29 This is consistent
with our previous study on designing a diastereoselective
Michael reaction.30 If the equilibrium (eq 4) exists, the
reverse reaction should prevail for the bulkier ketene silyl
acetals more than for the less bulky ones, leading to lower
diastereoselectivity. The experimental outcome, how-
ever, is completely opposite; the high diastereocontrol (up
to >99:1 syn selectivity) is attained in reaction 4 when
R3 is a bulky alkoxy group and R43 ) tBuMe2. Such high
stereoselectivity is otherwise not observed. Hence, it is
reasonably concluded that the double bond isomerization
of R-enone is not caused by the equilibrium (eq 4) of the
nucleophilic process.
Con clu sion
The course of the Mukaiyama-Michael reaction has
proven to be highly dependent on ketene silyl acetal and
Lewis acid. We interpret these results in the following
way. The reaction is usually initiated by electron trans-
fer from ketene silyl acetal to Lewis acid. The reduced
Lewis acid radical species attacks R-enone to give an
enolate radical, which combines with a cation radical
derived from the ketene silyl acetal. However, a compet-
ing nucleophilic reaction pathway intervenes when the
following conditions are fulfilled: (1) A ketene silyl acetal
contains bulky substituents, which increase the ioniza-
tion potential of the compound in solution. (2) The Lewis
acid should be sufficiently oxygenophilic to form a tight
complex with R-enone. It should be noted again that the
competition reactions serve well for probing the reaction
mechanism.
Exp er im en ta l Section
All solvents were purified by standard methods before use.
Ketene silyl acetals were prepared according to literature
methods.31 Reagent-grade SnCl4 and TiCl4 were used as
32
33
received. Preparation of Bu2Sn(OTf)2 and TESClO4 was
described in the literature.
Reaction of â-monomethyl-substituted derivative 3a
also led to isomerization. Notably, the Michael adducts
showed low diastereoselectivity which might be caused
(31) Ireland, R. E.; Wipf, P.; Armstrong, J . D., III J . Org. Chem.
1991, 56, 650. Otera, J .; Fujita, Y.; Fukuzumi, S. Synlett 1994, 213.
Gennari, C.; Beretta, M. G.; Bernardi, A.; Moro, G.; Scolastico, C.;
Todeschini, R. Tetrahedron 1986, 42, 893.
(29) Heathcock, C. H.; Norman, M. H.; Uehling, D. E. J . Am. Chem.
Soc. 1985, 107, 2797.
(30) Otera, J .; Fujita, Y.; Sato, T.; Nozaki, H.; Fukuzumi, S.; Fujita,
M. J . Org. Chem. 1992, 57, 5054.
(32) Sato, T.; Otera, J .; Nozaki, H. J . Am. Chem. Soc. 1990, 112,
901.

2960 J . Org. Chem., Vol. 61, No. 9, 1996
Otera et al.
P r ep a r a tion of 1. To a THF solution (200 mL) of diiso-
propylamine (24.3 g, 0.24 mol) at 0 °C was added a 2.5 M
solution of butyllithium in hexane (96 mL, 0.24 mol). After
10 min, the solution was cooled to -78 °C and acetophenone
(24.0 g, 0.2 mol) was added over a 5 min period by syringe.
After 2.5 min, HMPA (40 mL) was added. After 5 min,
trimethylsilyl chloride (26.1 g, 0.24 mol) was added over a 30
s period. The solution was warmed to room temperature (30
min), diluted with cold pentane, and washed three times with
ice cold water. The organic layer was then washed with
saturated brine and dried over Na2SO4. The solvent was
removed with a rotary evaporator, and the resulting crude
product was purified by distillation with a Kugelrohr ap-
paratus (bath temperature, 90 °C/60 mmHg) to give 1-phenyl-
1-(trimethylsiloxy)ethene (31.1 g, 81%). To a CH2Cl2 solution
(200 mL) of 8d (45.5 g, 0.24 mol) was added acetone (11.0 g,
0.19 mol) at -78 °C. After 1 min, 1-phenyl-1-(trimethylsiloxy)-
ethene (31.1 g, 0.16 mol) was added. The solution was stirred
at -78 °C for 3 h. The reaction mixture was extracted with
EtOAc. The organic layer was washed with NaHCO3 solution
and brine and then dried (Na2SO4). After evaporation, the
resulting crude product was purified by distillation with a
Kugelrohr apparatus (bath temperature, 110 °C/60 mmHg) to
give C6H5COCH2C(OH)(CH3)2 (22.0 g, 77%). To a CH2Cl2
solution (100 mL) of this compound (21.98 g, 0.12 mol) was
added Et3N (36.4 g, 0.36 mol) at 0 °C. To this solution was
added (CF3CO)2O (37.8 g, 0.18 mol). The solution was stirred
at 0 °C for 10 min, and allowed to warm up to room
temperature. After 1 h, the reaction mixture was extracted
with CH2Cl2. The organic layer was washed with NaHCO3
solution and brine and dried (Na2SO4). The solvent was
removed to give crude C6H5COCH2C(OCOCF3)(CH3)2. To a
C6H6 solution (100 mL) of this compound (crude product, 0.12
mol) was added DBU (23.8 g, 0.16 mol) at room temperature.
The solution was stirred at room temperature for 1.5 h. The
reaction mixture was extracted with C6H6. The organic layer
was washed with cold 1 N HCl, cold NaHCO3 solution, and
brine and dried (Na2SO4). Evaporation and column chroma-
tography on silica gel (95:5 hexane-EtOAc) afforded 1a (13.5
g, 70%): 1H NMR δ 2.01 (s, 3H), 2.02 (s, 3H), 6.75 (s, 1H),
7.42-7.53 (m, 3H), 7.92 (d, 2H, J ) 7.69 Hz); 13C NMR δ 20.81,
27.60, 120.83, 127.84, 128.11, 131.93, 138.94, 156.27, 190.97;
MS (m/ z) 161 (M+ + 1); HRMS calcd for C11H12O (M+)
160.0880, found 160.0871. By an analogous procedure, 1e was
obtained. Column chromatography of the E/ Z mixture on
silica gel (98:2 hexane-EtOAc) provided (Z)-1e as the first
fraction and the (E)-isomer as the second one. (Z)-1e: 1H NMR
δ 1.13 (t, 3H, J ) 7.51 Hz), 2.00 (s, 3H), 2.63 (q, 2H, J ) 7.51
Hz), 6.70 (s, 1H), 7.42-7.53 (m, 3H), 7.93 (d, 2H, J ) 7.33
Hz). (E)-1e: 1H NMR δ 1.53 (t, 3H, J ) 7.33 Hz), 2.19 (s, 3H),
2.28 (q, 2H, J ) 7.33 Hz), 6.72 (s, 1H), 7.42-7.54 (m, 3H), 7.63
(d, 2H, J ) 7.33 Hz); MS (m/ z) 174 (M+); HRMS calcd for
C12H14O (M+) 174.1045, found 174.1037.
DBU (28.9 g, 0.19 mol) at room temperature. The solution
was stirred at room temperature for 1.5 h. The reaction
mixture was extracted with C6H6. The organic layer was
washed with cold 1 N HCl, cold NaHCO3 solution, and brine
and dried (Na2SO4). Evaporation and column chromatography
on silica gel (95:5 hexane-EtOAc) afforded 1c (14.9 g, 81%):
1H NMR δ 1.12 (s, 9H), 1.87 (d, 3H, J ) 6.96 Hz), 6.50 (dq,
1H, J ) 15.92, 1.63 Hz), 6.93 (dq, 1H, J ) 15.92, 6.81 Hz); MS
(m/ z) 126 (M+); HRMS calcd for C8H14O (M+) 126.1045, found
126.1044.
To a suspension of AlCl3 (40.0 g, 0.30 mol) in C6H6 (150 mL)
was added propionyl chloride (25.9 g, 0.28 mol) at 0 °C. The
mixture was stirred for 15 min at room temperature. The
reaction mixture was quenched with 200 mL of cold 2 N HCl.
The resulting solution was extracted with Et2O (140 mL × 3).
The organic layer was washed with a saturated aqueous
solution of NaHCO3 (50 mL × 2) and 50 mL of brine and then
dried over anhydrous Na2SO4. Evaporation and column chro-
matography of the residue on silica gel (95:5 hexane-EtOAc)
afforded 1d (31.2 g, 76%): 1H NMR δ 2.00 (d, 3H, J ) 8.24
Hz), 6.91 (dq, 1H, J ) 15.30, 1.44 Hz), 7.08 (dq, 1H, J ) 15.30,
7.08 Hz), 7.43-7.56 (m, 3H), 7.92 (d, 2H, J ) 8.06 Hz); MS
(m/ z) 147 (M+ + 1); HRMS calcd for C10H10O (M+) 146.1906,
found 146.0720. Analogous procedure provided 1b: 1H NMR
δ 1.14 (s, 9H), 1.91 (s, 3H), 2.11 (s, 3H), 6.31 (s, 1H); 13C NMR
δ 20.63, 26.45, 27.80, 43.42, 119.59, 154.96, 205.85; MS (m/ z)
141 (M+ + 1); HRMS calcd for C9H16O (M+) 140.1201, found
140.1193.
Mich a el Rea ction . A typical procedure is as follows. To
a suspension of 8a (26.6 mg, 0.05 mmol) in CH2Cl2 (5 mL) were
added 1a (160.2 mg, 1.0 mmol) and 2a (226.6 mg, 1.3 mmol)
at -78 °C. The resulting clear solution was stirred for 4 h at
this temperature. The reaction mixture was combined with
NaHCO3 solution and EtOAc. The organic layer was washed
with NaHCO3, 1 N HCl, and brine and then dried (Na2SO4).
Evaporation and column chromatography of the residue on
silica gel (98:2 hexane-EtOAc) afforded 5a (258 mg, 99%): 1H
NMR δ 1.06 (s, 6H), 1.22 (s, 6H), 1.29 (t, 3H, J ) 7.33 Hz),
3.07 (s, 2H), 4.16 (q, 2H, J ) 7.33 Hz), 7.43-7.56 (m, 3H), 7.93
(d, 2H, J ) 7.33 Hz); 13C NMR δ 14.08, 21.07, 22.44, 38.53,
43.31, 48.95, 60.22, 128.12, 128.34, 132.58, 138.76, 176.70,
200.72; MS (m/ z) 276 (M+); HRMS calcd for C15H19O2 (M+
OEt) 231.1385, found 231.1391. Anal. Calcd for C17H24O3: C,
73.88; H, 8.75. Found: C, 73.79; H, 9.00.
-
When ketene silyl acetals derived from tert-butyl esters were
employed, the crude reaction mixture, which consisted of the
tert-butyl ester and the corresponding carboxylic acid,29 was
stirred in 5 N HCl/THF at room temperature for 5 h. The
mixture was extracted with ether and the organic layer was
dried (Na2SO4). To this solution was added dropwise diaz-
omethane in ether until the color of this reagent began to
persist. Evaporation and column chromatography on silica gel
afforded the corresponding methyl esters of the Michael
adducts (5d , 6d , and 7d ).
Other reactions were carried out analogously. 5b: 1H NMR
δ 1.08 (s, 6H), 1.11 (s, 9H), 1.14 (s, 6H), 1.27 (t, 3H, J ) 7.13
Hz), 2.61 (s, 2H), 4.12 (q, 2H, J ) 7.13 Hz); MS (m/ z) 257 (M+
+ 1); HRMS calcd for C13H23O2 (M+ - OEt) 211.1698, found
211.1776. Anal. Calcd for C15H28O3: C, 70.27; H, 11.01.
Found: C, 70.59; H, 11.34.
5c: 1H NMR δ 0.79 (d, 3H, J ) 5.98 Hz), 1.11 (s, 9H), 1.11
(s, 6H), 1.23 (t, 3H, J ) 7.12 Hz), 2.31-2.46 (m, 3H), 4.11 (q,
2H, J ) 7.12 Hz); MS (m/ z) 243 (M+ + 1); HRMS calcd for
C12H21O2 (M+ - OEt) 197.1542, found 197.1556. Anal. Calcd
for C14H26O3: C, 69.38; H, 10.81. Found: C, 69.50; H, 11.09.
5d : 1H NMR δ 0.72 (d, 3H, J ) 6.22 Hz), 1.05 (s, 9H), 1.06
(s, 6H), 2.29-2.36 (m, 3H), 3.58 (s, 3H); 13C NMR δ 14.98,
21.61, 22,34, 26.08, 35.23, 38.97, 45.11, 51.33, 177.75, 214.23;
MS (m/ z) 229 (M+ + 1); HRMS calcd for C12H21O2 (M+ - OMe)
197.1542, found 197.1561. Anal. Calcd for C13H24O3: C, 68.38;
H, 10.59. Found: C, 68.76; H, 10.86.
To a THF solution (200 mL) of diisopropylamine (24.3 g,
0.24 mol) was added 2.5 M solution of butyllithium in hexane
(96 mL, 0.24 mol) at -78 °C. After 15 min, pinacolone (20.0
g, 0.2 mol) was added. After 30 min, acetaldehyde (10.6 g,
0.24 mol) was added. The reaction mixture was stirred at -78
°C for 3 h. The reaction was quenched by adding NH4Cl
solution. The reaction mixture was extracted with Et2O. The
organic layer was washed with NH4Cl solution and brine and
then dried (Na2SO4). The solvent was removed, and the
resulting crude product was purified by distillation with a
Kugelrohr apparatus (bath temperature, 100 °C/100 mmHg)
to give tBuCOCH2CH(OH)CH3 (21.1 g, 73%). To a CH2Cl2
solution (100 mL) of this compound (21.1 g, 0.15 mol) was
added Et3N (44.5 g, 0.44 mol) at 0 °C. To this solution was
added (CF3CO)2O (46.0 g, 0.22 mol). The solution was stirred
at 0 °C for 10 min and allowed to warm to room temperature.
After 1 h, the reaction mixture was extracted with CH2Cl2.
The organic layer was washed with NaHCO3 solution and
brine and dried (Na2SO4). The solvent was removed to give
t
crude BuCOCH2CH(OCOCF3)CH3. To a C6H6 solution (100
5e: 1H NMR δ 0.80-0.95 (m, 13H), 1.05-1.37 (m, 17H),
1.65-1.77 (m, 2H), 1.92-2.00 (m, 1H), 2.31-2.49 (m, 4H),
4.80-4.85 (m, 1H); MS (m/ z) 350 (M+); HRMS calcd for
C22H38O3 (M+) 350.2821, found 350.2819. Anal. Calcd for
C22H38O3: C, 75.38; H, 10.93. Found: C, 75.21; H, 11.04.
mL) of this compound (crude product, 0.15 mol) was added
(33) Lambert, J . B.; McConnell, J . A.; Schiff, W.; Schultz, W. J ., J r.
J . Chem. Soc., Chem. Commun. 1988, 455.

Mechanism of Mukaiyama-Michael Reaction
J . Org. Chem., Vol. 61, No. 9, 1996 2961
5f: 1H NMR δ 0.89 (d, 3H, J ) 6.52 Hz), 1.19 (s, 6H), 1.25
(t, 3H, J ) 7.13 Hz), 2.52-2.77 (m, 2H), 2.96-3.02 (m, 1H),
4.14 (q, 2H, J ) 7.12 Hz), 7.43-7.56 (m, 3H), 7.95 (d, 2H, J )
7.08 Hz); MS (m/ z) 262 (M+). Anal. Calcd for C16H22O3: C,
73.25; H, 8.45. Found: C, 73.60; H, 8.75.
6g: 1H NMR δ 0.83 (t, 3H, J ) 7.33 Hz), 1.07-1.20 (m, 9H),
1.49-1.76 (m, 2H), 2.83-3.26 (m, 3H), 4.03-4.11 (m, 2H),
7.42-7.56 (m, 3H), 7.95 (d, 2H, J ) 7.33 Hz); 13C NMR δ 8.14,
11.86, 14.07, 21.49, 29.50, 38.40, 42.67, 45.58, 59.87, 127.86,
128.41, 132.63, 138.45, 175.75, 199.72; MS (m/ z) 276 (M+);
HRMS calcd for C15H19O2 (M+ - OEt) 231.1385, found 231.1354.
Anal. Calcd for C17H24O3: C, 73.88; H, 8.75. Found: C, 74.08;
H, 8.91.
7a : 1H NMR δ 1.17 (s, 6H), 1.19 (t, 3H, J ) 7.33 Hz), 2.53
(s, 2H), 3.11 (s, 2H), 4.08 (q, 2H, J ) 7.33), 7.42-7.56 (m, 3H),
7.94 (d, 2H, J ) 7.33 Hz); 13C NMR δ 14.19, 28.25, 32.98, 45.05,
47.16, 59.93, 127.97, 128.47, 132.76, 138.24, 172.24, 199.63;
MS (m/ z) 248 (M+); HRMS calcd for C15H20O3 (M+) 248.1412,
found 248.1450. Anal. Calcd for C15H20O3: C, 72.55; H, 8.12.
Found: C, 72.87; H, 8.36.
7b: 1H NMR δ 1.08 (s, 6H), 1.12 (s, 9H), 1.24 (t, 3H, J )
7.14 Hz), 2.51 (s, 2H), 2.64 (s, 2H), 4.09 (q, 2H, J ) 7.14 Hz);
MS (m/ z) 229 (M+ + 1); HRMS calcd for C11H19O2 (M+ - OEt)
183.1385, found 183.1367. Anal. Calcd for C13H24O3: C, 68.38;
H, 10.59. Found: C, 68.66; H, 10.87.
7c: 1H NMR δ 0.94 (d, 3H, J ) 3.69 Hz), 1.11 (s, 9H), 1.24
(t, 3H, J ) 7.14 Hz), 2.13-2.57 (m, 5H), 4.10 (q, 2H, J ) 7.14
Hz); MS (m/ z) 215 (M+ + 1); HRMS calcd for C12H22O3 (M+)
214.1569, found 214.1557.
5g: 1H NMR δ 1.05 (s, 6H), 1.23 (s, 6H), 3.06 (s, 2H), 3.70
(s, 3H), 7.43-7.56 (m, 3H), 7.93 (d, 2H, J ) 7.33 Hz); 13C NMR
δ 21.13, 22.50, 38.63, 43.33, 49.18, 51.41, 128.18, 128.42,
132.66, 138.77, 177.25, 200.70; MS (m/ z) 262 (M+); HRMS
calcd for C15H19O2 (M+ - OMe) 231.1385, found 231.1355.
Anal. Calcd for C16H22O3 : C, 73.25; H, 8.45. Found: C,
73.15; H, 8.68.
5h : 1H NMR δ 0.80 (t, 3H, J ) 7.33 Hz), 1.05 (s, 3H), 1.17-
1.21 (m, 9H), 1.52-1.79 (m, 2H), 3.10 (ABq, 2H, J ) 17.75
Hz, ∆νAB )27.27 Hz), 4.06 (q, 2H, J ) 7.33 Hz), 7.40-7.52 (m,
3H), 7.94 (d, 2H, J ) 7.33 Hz); 13C NMR δ 9.75, 13.99, 20.86,
21.97, 27.97, 41.26, 41.85, 49.23, 60.24, 127.96, 128.38, 132.46,
138.75, 177.07, 200.50; MS (m/ z) 290 (M+); HRMS calcd for
C16H21O2 (M+ - OEt) 245.1542, found 245.1472. Anal. Calcd
for C18H26O3: C, 74.44; H, 9.02. Found: C, 74.81; H, 9.21.
5i: 1H NMR δ 1.02 (s, 6H), 1.22-2.21 (m, 10 H), 2.98 (s,
2H), 3.72 (s, 3H), 7.42-7.56 (m, 3H), 7.91 (d, 2H, J ) 7.33
Hz); 13C NMR δ 22.83, 24.07, 25.51, 28.70, 39.14, 43.27, 51.01,
54.72, 128.15, 128.38, 132.61, 138.83, 175.46, 200.82; MS (m/
z) 302 (M+); HRMS calcd for C15H18O3 (M+ - C4H8)246.1256,
found 246.1238. Anal. Calcd for C19H26O3: C, 75.46; H, 8.67.
Found: C, 75.15; H, 8.79.
7d : 1H NMR δ 0.89 (d, 3H, J ) 5.89 Hz), 1.06 (s, 9H), 2.11-
2.51 (m, 5H), 3.59 (s, 3H); 13C NMR δ 19.73, 25.80, 25.84, 26.03,
40.35, 42.45, 51.09, 172.75, 214.24; MS (m/ z) 200 (M+); HRMS
calcd for C11H20O3 (M+) 200.1413, found 200.1401.
7e: 1H NMR δ 0.80-0.95 (m, 13H), 1.04-1.32 (m, 11H),
1.63-1.78 (m, 2H), 1.86-1.96 (m, 1H), 2.15-2.56 (m, 6H),
4.84-4.87 (m, 1H); MS (m/ z) 322 (M+); HRMS calcd for
C20H34O3 (M+) 322.2508, found 322.2521. Anal. Calcd for
C20H34O3: C, 74.49; H, 10.63. Found: C, 74.54; H, 10.98.
5j: 1H NMR δ 0.82 (t, 3H, J ) 7.33 Hz), 1.05 (s, 3H), 1.11-
2.16 (m, 12H), 3.03 (ABq, 2H, J ) 16.48 Hz, ∆νAB ) 28.54 Hz),
3.63 (s, 3H), 7.42-7.56 (m, 3H), 7.93 (d, 2H, J ) 7.33 Hz);13C
NMR δ 9.81, 21.27, 24.16, 25,57, 27.84, 29.61, 41.08, 42.67,
51.03, 54.94, 127.93, 128.40, 132.43, 138.74, 175.68, 200.18;
MS (m/ z) 316 (M+); HRMS calcd for C18H25O (M+ - COOMe)
257.1906, found 257.1905. Anal. Calcd for C20H28O3: C, 75.91;
H, 8.91. Found: C, 76.12; H, 9.15.
Com p etition Rea ction . To a CH2Cl2 solution (5 mL) of
8a (26.6 mg, 0.05 mmol) was added 1a (160 mg, 1.0 mmol) at
-78 °C. To this solution was added a mixture of 2a (230 mg,
1.0 mmol) and 4a (202 mg, 1.0 mmol). The solution was stirred
at -78 °C for 4 h. The reaction was quenched by adding
NaHCO3 solution. The reaction mixture was extracted with
EtOAc. The organic layer was washed with NaHCO3 solution,
1 N HCl, and brine and then dried (Na2SO4). GLC analysis
of the crude mixture obtained by evaporation showed forma-
tion of 5a in 85% yield while 7a was not detected. Other
competition reactions were carried out analogously. When
ketene silyl acetals derived from tert-butyl esters were em-
ployed, the Michael adducts consisted of the tert-butyl esters
and the corresponding carboxylic acids. Accordingly, the
reaction mixture was treated with HCl and diazomethane (vide
supra), and the resulting methyl esters were subjected to GLC
analysis. The GLC peaks in the competition reaction were
fully confirmed by comparison with separately prepared
authentic Michael adducts.
Com p etition Rea ction w ith LiClO4 in Eth er . To an
ether solution (5 mL) of LiClO4 (26.6 mg, 0.25 mmol) was
added 2-cyclohexen-1-one (48.1 mg, 0.5 mmol) at room tem-
perature. To this solution was added a mixture of 2e (115.1
mg, 0.5 mmol) and 4d (101.7 mg, 0.5 mmol). After being
stirred at room temperature for 5 h, the solution was poured
into hexane (50 mL) containing Et3N (1 mL). The mixture was
filtered, and the filtrate was evaporated. GLC analysis of the
crude mixture showed 11 to be formed in 62% yield while 9
was not detected. Other reactions were conducted analogously.
5k : 1H NMR δ 0.84 (t, 3H, J ) 7.33 Hz), 0.98 (s, 3H), 1.09
(s, 3H), 1.16 (s, 3H), 1.35-1.40 (m, 1H), 2.05-2.10 (m, 1H),
3.05 (ABq, 2H, J ) 13.29 Hz, ∆νAB ) 24.11 Hz), 3.70 (s, 3H),
7.42-7.56 (m, 3H), 7.92 (d, 2H, J ) 7.33 Hz); 13C NMR δ 9.86,
16.55, 22.79, 25.99, 39.29, 43.31, 51.17, 53.71, 128.15, 128.39,
132.61, 138.79, 176.41, 200.70; MS (m/ z) 276 (M+); HRMS
calcd for C16H21O2 (M+ - OMe) 245.1542, found 245.1577.
Anal. Calcd for C17H24O3: C, 73.88; H, 8.75. Found: C, 73.62;
H, 8.92.
6a : 1H NMR δ 1.11 (s, 3H), 1.12 (s, 3H), 1.16 (d, 3H, J )
6.96 Hz), 1.18 (t, 3H, J ) 6.96 Hz), 2.76 (q, 1H, J ) 6.96 Hz),
2.84-3.22 (m, 2H), 4.09 (q, 2H, J ) 6.96 Hz), 7.42-7.56 (m,
3H), 7.94 (d, 2H, J ) 7.33 Hz); 13C NMR δ 12.33, 14.66, 25.41,
35.70, 45.89, 47.72, 59.96, 128.02, 128.48, 132.76, 138.43,
175.66, 199.79; MS (m/ z) 262 (M); HRMS calcd for C16H22O3
(M+) 262.1569, found 262.1558. Anal. Calcd for C16H22O3: C,
73.25; H, 8.45. Found: C, 73.27; H, 8.73.
6b: 1H NMR δ 1.05 (s, 3H), 1.06 (s, 3H), 1.08 (d, 3H, J )
7.17 Hz), 1.11 (s, 9H), 1.25 (t, 3H, J ) 7.17 Hz), 2.40-2.71 (m,
2H), 2.80 (q, 1H, J ) 7.17 Hz), 4.10 (q, 2H, J ) 7.15 Hz); MS
(m/ z) 242 (M+); HRMS calcd for C12H21O2 (M+ - OEt)
197.1542, found 197.1544. Anal. Calcd for C14H26O3: C, 69.38;
H, 10.81. Found: C, 69.02; H, 11.18.
6c: 1H NMR δ 0.75 (d, 3H, J ) 5.93 Hz), 1.05-1.10 (m,
3H), 1.07 (s, 9H), 1.19 (t, 3H, J ) 7.09 Hz), 2.27-2.42 (m, 4H),
4.07 (q, 2H, J ) 7.09 Hz); MS (m/ z) 228 [M+]. Anal. Calcd
for C13H24O3: C, 68.38; H, 10.59. Found: C, 68.76; H, 10.86.
6d : 1H NMR δ 0.85 (d, 3H × 0.96, J ) 6.54 Hz), 0.87 (d, 3H
× 0.04, J ) 5.80 Hz), 1.10-1.13 (m, 12H), 2.43-2.46 (m, 4H),
3.66 (s, 3H × 0.04) 3.67 (s, 3H × 0.96); MS (m/ z) 214 (M+);
HRMS calcd for C12H22O3 (M+) 214.1569, found 214.1571.
6e: 1H NMR δ 0.81-0.95 (m, 12H), 1.05-1.35 (m, 14H),
1.60-1.75 (m, 2H), 1.88-2.00 (m, 1H), 2.35-2.64 (m, 6H),
4.84-4.88 (m, 1H); MS (m/ z) 337 (M+ + 1); HRMS calcd for
C21H36O3 (M+) 336.2664, found 336.2655. Anal. Calcd for
C21H36O3: C, 74.95; H, 10.78. Found: C, 75.05; H, 10.61.
6f: 1H NMR δ 0.96 (d, 3H × 0.60, J ) 6.96 Hz), 0.98 (d, 3H
× 0.40, J ) 6.59 Hz), 1.16 (d, 3H × 0.60, J ) 6.96 Hz), 1.18 (d,
3H × 0.40, J ) 6.59 Hz), 2.51-2.61 (m, 2H), 2.75-2.83 (m, 1H),
3.01-3.19 (m, 1H), 3.67 (s, 3H × 0.40), 3.68 (s, 3H × 0.60),
7.44-7.58 (m, 3H), 7.95-7.97 (m, 2H); MS (m/ z) 234 (M+);
HRMS calcd for C14H18O3 (M+) 234.1256, found 234.1267.
Authentic samples of 9, 10, and 11 were prepared by TiCl4-
promoted reaction.
9: 1H NMR δ 0.11 (s, 3H), 0.13 (s, 3H), 0.91 (s, 9H), 1.08 (s,
3H), 1.10 (s, 3H), 1.25 (t, 3H, J ) 7.12 Hz), 1.45-1.63 (m, 3H),
1.78-1.89 (m, 1H), 1.91-2.07 (m, 2H), 2.52-2.60 (m, 1H), 4.12
(q, 2H, J ) 7.12 Hz), 4.67 (br s, 1H); MS (m/ z) 326 (M+). Anal.
Calcd for C18H34O3Si: C, 66.21; H, 10.49. Found: C, 66.60;
H, 10.69.
10: 1H NMR δ 0.11-0.13 (m, 6H), 0.90 (s, 9H × 0.61), 0.91
(s, 9H × 0.39), 1.09 (d, 3H × 0.61, J ) 6.96 Hz), 1.11 (d, 3H ×
0.39, J ) 7.11 Hz), 1.25 (t, 3H × 0.39, J ) 7.14 Hz), 1.26 (t,
3H × 0.61, J ) 7.11 Hz), 1.49-1.69 (m, 3H), 1.72-1.81 (m,
1H), 1.94-2.02 (m, 2H), 2,26-2.37 (m, 1H), 2.43-2.53 (m, 1H),
4.09-4.17 (m, 2H), 4.67 (br s, 1H × 0.61), 4.80-4.84 (m, 1H

2962 J . Org. Chem., Vol. 61, No. 9, 1996
Otera et al.
× 0.39); MS (m/ z) 312 (M+). Anal. Calcd for C17H32O3Si: C,
tert-Butyl 1-(trichlorostannyl)propionate: 1H NMR (CD2Cl2)
119
65.33; H, 10.32. Found: C, 65.11; H, 10.28.
δ 1.58 (s, 9H), 1.61 (d, 3H, J ) 7.63 Hz) with J
205.40;
Sn-H
117
11: 1H NMR δ 0.12 (s, 6H), 0.91 (s, 9H), 1.25 (t, 3H, J )
7.10 Hz), 1.52-1.65 (m, 2H), 1.68-1.78 (m, 2H), 1.94-2.04 (m,
2H), 2.24 (d, 2H, J ) 7.51 Hz), 2.60-2.70 (m, 1H), 4.13 (q,
2H, J ) 7.12 Hz), 4.76-4.78 (m, 1H); MS (m/ z) 298 (M+);
J
190.17 Hz, 3.70 (q, 1H, J ) 7.63 Hz).
Sn-H
Methyl 1-(trichlorostannyl)propionate: 1H NMR (CD2Cl2) δ
119
117
1.63 (d, 3H, J ) 7.50 Hz) with J
205.49 Hz; J
Sn-H
Sn-H
190.32 Hz, 3.67 (q, 1H, J ) 7.50 Hz), 3.76 (s, 3H).
HRMS calcd for C12H21O3Si (M+
241.1248. Anal. Calcd for C16H30O3Si: C,64.38; H, 10.13.
Found: C, 64.60; H, 10.29.
-
tBu) 241.1260, found
Isola tion of En ol Silyl Eth er In ter m ed ia te. Reaction
of 1a and 2a in the presence of 8a was carried out as described
above. The reaction mixture, after being stirred for 4 h, was
poured into hexane (50 mL) containing Et3N (1 mL). The
mixture was filtered and the filtrate was evaporated. The
residue was subjected to column chromatography on ammonia-
treated silica gel (98:2 hexane-EtOAc) to give 16a (215 mg,
56%) together with 5a (116 mg, 38%). 16a : 1H NMR δ 0.51
(q, 6H, J ) 7.96 Hz), 0.88 (t, 9H, J ) 7.69 Hz), 1.20 (s, 6H),
1.22 (t, 3H, J ) 6.96 Hz), 1.27 (s, 6H), 4.10 (q, 2H, J ) 6.96
Hz), 4.70 (s,1H), 7.23-7.27 (m, 3H), 7.29-7.34 (m, 2H); MS
(m/ z) 390 (M+); HRMS calcd for C21H33O2Si (M+ - OEt)
345.2250, found 345.2271. Upon acidic hydrolysis, 16a was
converted to 5a .
When the reaction was conducted in CH2Cl2 (4.5 mL)-THF
(0.5 mL), 16a was solely obtained even by aqueous workup.
The other reactions were carried out analogously. 16b: 1H
NMR δ 0.01 (s, 9H), 1.19 (s, 6H), 1.22 (s, 6H), 3.63 (s, 3H),
4.76 (s, 1H), 7.21-7.26 (m, 3H), 7.29-7.31 (m, 2H); MS (m/ z)
334 (M+); HRMS calcd for C16H21O3 (M+ - Me3Si) 261.1491,
found 261.1575.
Com p etition Rea ction of O-Silyl Keten e O,S-Aceta ls
w ith 1d . To a CH2Cl2 solution (5 mL) of 8b (26 mg, 0.1 mmol)
was added 1d (160 mg, 0.1 mmol) at -78 °C. To this solution
was added a mixture of 12 (232.5 mg, 1.0 mmol) and 13 (218.4
mg, 1.0 mmol). The solution was stirred at -78 °C for 4 h.
The reaction was quenched by adding NaHCO3 solution. The
reaction mixture was washed with EtOAc. The organic layer
was washed with NaHCO3 solution, 1 N HCl, and brine, and
then dried (Na2SO4). GLC analysis of the crude mixture
obtained after evaporation showed formation of 14 in 5% yield
and 15 in 82% yield; 14: 1H NMR δ 0.90 (d, 3H, J ) 6.35 Hz),
1.18 (s, 3H), 1.19 (s, 3H), 1.45 (s, 9H), 2.57-2.76 (m, 2H), 2.98-
3.10 (m, 1H), 7.43-7.55 (m, 3H), 7.96 (d, 2H, J ) 7.09 Hz);
MS (m/ z) 306 (M+). Anal. Calcd for C18H26O2S: C, 70.55; H,
8.55. Found: C, 70.67; H, 8.63; 15: 1H NMR δ 0.96 (d, 3H ×
0.34, J ) 6.35 Hz), 1.00 (d, 3H × 0.66, J ) 6.54 Hz), 1.14 (d,
3H × 0.34, J ) 6.53 Hz), 1.17 (d, 3H × 0.66, J ) 6.59 Hz),
1.45 (s, 9H × 0.66), 1.46 (s, 9H × 0.34), 2.52-2.62 (m, 2H),
2.67-2.80 (m, 1H), 3.10-2.23 (m, 1H), 7.42-7.49 (m, 2H),
7.53-7.58 (m, 1H), 7.93-7.97 (m, 2H); MS (m/ z) 292 (M+).
Anal. Calcd for C17H24O2S: C, 69.82; H, 8.27. Found: C,
70.01; H, 8.59.
Rea ction of 3 w ith TiCl4 (8d ). To a CH2Cl2 solution (5
mL) of 3b (244 mg, 1.0 mmol) was added a 1 M CH2Cl2 solution
of 8d (1.0 mL, 1.0 mmol) at -78 °C. The solution was stirred
for 30 min and combined with water. The mixture was
extracted with ether. The organic layer was washed with
water and dried (Na2SO4). The ether solution was filtered and
treated with diazomethane to afford the corresponding di-
methyl ester. The yield of dimethyl ester was determined on
the basis of GLC analysis by comparison with an authentic
sample. Other reactions were carried out analogously.
16c: 1H NMR δ 0.53 (q, 6H, J ) 8.06 Hz), 0.90 (t, 9H, J )
8.06 Hz), 1.15 (d, 3H, J ) 6.96 Hz), 1.25 (t, 3H, J ) 7.33 Hz),
1.27 (s, 6H), 2.80 (q, 1H, J ) 6.96 Hz), 4.10 (q, 2H, J ) 7.33
Hz), 4.68 (s, 1H), 7.23-7.34 (m, 5H); MS (m/ z) 376 (M+);
HRMS calcd for C22H36O3Si (M+) 376.2434, found 376.2346.
Isom er iza tion of 1e. To a suspension of 8a (26.6 mg, 0.05
mmol) in CH2Cl2 solution (5 mL) were added (E)-1e (174.2 mg,
1.0 mmol) and 2a (230.4 mg, 1.0 mmol) at -78 °C. The
resulting clear solution was stirred for 2 h at this temperature.
To the reaction mixture was added NaHCO3, 1 N HCl, and
brine. The organic layer was dried (Na2SO4). The mixture
was filtered, and the filtrate was evaporated. GLC analysis
of the crude mixture showed 5h to be formed in 44% yield.
HPLC analysis of the crude mixture showed that 1e consisted
of the E/ Z isomers in a 90:10 ratio. Other reactions were
carried out analogously.
Th eor etica l Ca lcu la tion s. The theoretical studies were
performed at the restricted Hartree-Fock (RHF) level using
the PM3 semiempirical SCF-MO method as implemented in
the MOPAC program (Ver. 6.0)17 with the MOL-GRAPH
program Ver. 2.8 by Daikin Industries, Ltd. or using the ab
initio method with Gaussian 92 program.35 Final geometries
and energetics were obtained by optimizing the total molecular
energy with respect to all structural variables with no sym-
metry constraints and further refined by using the key word
PRECISE. The transition state structures were refined by
using the key word TS and checked by diagonalizing the force
constant (Hessian) matrix and establishing that it had only
one negative eigenvalue.15
Rea ction of Keten e Silyl Aceta l w ith Sn Cl4 (8b). To a
CD2Cl2 solution (1 mL) of 8b (130 mg, 0.5 mmol) was added
2a (150 mg, 0.5 mmol) at -78 °C. The solution was stirred
for 30 min at this temperature. The solution was warmed up
to room temperature. A portion of this solution (ca. 0.4 mL)
was transferred to a nitrogen-filled NMR sample tube which
1
contained benzene (0.5 mmol) as an internal standard. A H
NMR spectrum unambiguously showed quantitative formation
of ethyl 1-(trichlorostannyl)isobutyrate: 1H NMR (CD2Cl2) δ
1.24 (t, 3H, J ) 7.33 Hz), 1.66 (s, 6H) with satellites induced
119
117
by Sn: J
201.13 Hz; J
192.33 Hz, quite reasonable
Sn-H
Sn-H
values for monoalkyltin trichlorides,34 4.20 (q, 2H, J ) 7.33
Hz). No other signals attributable to byproducts were de-
tected.
Su p p or tin g In for m a tion Ava ila ble: 1H NMR spectra of
1, 6d , 6f, 7d , and 16 (12 pages). This material is contained
in libraries on microfiche, immediately follows this article in
the microfilm version of the journal, and can be ordered from
the ACS; see any current masthead page for ordering
information.
Other reactions were conducted analogously. tert-Butyl
1-(trichlorostannyl)-isobutyrate: 1H NMR (CD2Cl2) δ 1.58 (s,
119
117
9H), 1.70 (s, 6H) with J
201.89 Hz; J
192.92 Hz.
Sn-H
Sn-H
Methyl 1-(trichlorostannyl)isobutyrate: 1H NMR (CD2Cl2)
119
117
δ 1.71 (s, 6H) with J
200.97 Hz; J
192.06 Hz, 3.79
Sn-H
Sn-H
(s, 3H).
Ethyl 1-(trichlorostannyl)propionate: 1H NMR (CD2Cl2) δ
J O9512968
119
1.28 (3H, J ) 7.14 Hz), 1.68 (d, 3H, J ) 7.51 Hz) with J
Sn-H
117
206.09 Hz; J
190.95 Hz, 2.54 (q, 1H, J ) 7.51 Hz), 4.23
Sn-H
(35) Gaussian 92, Revision C: Frisch, M J .; Trucks, G. W.; Head-
Gordon, M.; Gill, M. W.; Wong, M. W.; Foresman, J . B.; J ohnson, B.
G.; Schlegel H. B.; Robb, M. A.; Replogle, E. S.; Gomperts, R.; Andres,
J . L.; Raghavachari, K.; Binkley, J . S.; Gonzalez, C.; Martin, R. L.;
Fox, D. J .; Defrees, D. J .; Baker, J .; Stewart, J . J . P.; Pople, J . A.
Gaussian Inc., Pittsburgh, PA, 1992.
(q, 2H, J ) 7.14 Hz).
(34) Lorberth, J .; Vahrenkamp, H. J . Organomet. Chem. 1968, 11,
111.