Intramolecular Geminal and Vicinal Element Effects in Substitution of Simple Bromo(chloro)alkenes by Methoxide and Thiolate Ions. An Example of a Single Step Substitution?

Article abstract of DOI:10.1021/jo9709357

Intramolecular element effects k_Br/k_Cl for substitution of geminal bromochloroalkenes BrC(Cl)=C-(Br)Cl (1), BrC(Cl)=CCl₂ (2), Me₂C=C(Br)Cl (3), and XCH=C(Br)Cl (X = Cl, 4; X = Br, 5), with MeO^- and RS^- nucleophiles were investigated. 3 did not give substitution, and 4 and 5 gave substitution with MeO^- via an initial elimination (to acetylene)-addition route, followed by further reactions. In reactions of 4 with thiolates, geminal element effects of 2-10 were obtained. Formation of RSC(Cl)=C(Cl)Y, Y = SR, Br, is ascribed to an initial halophilic reaction, followed by addition of RSCl to the formed acetylene. Reaction of 2 with MeO^- gave a high vicinal element effect, and RS^- gave a high geminal element effect. Reaction of 1 with both MeO^- and RS^- ions gave high (2 orders of magnitude) geminal element effects, which were interpreted as indicating a rate-determining C-X bond cleavage. This is supported by the high k_Br/k_Cl intermolecular element effects (k(1)/k(Cl₂C=CCl₂) with MeO^- and PhCH₂S^- ions. Mechanistic alternatives based on these observations are discussed.

Full text of DOI:10.1021/jo9709357

J . Org. Chem. 1997, 62, 8049-8057
8049
In tr a m olecu la r Gem in a l a n d Vicin a l Elem en t Effects in
Su bstitu tion of Sim p le Br om o(ch lor o)a lk en es by Meth oxid e a n d
Th iola te Ion s. An Exa m p le of a Sin gle Step Su bstitu tion ?
Michal Beit-Yannai,† Zvi Rappoport,*,†,‡ Bagrat A. Shainyan,§ and Yuri S. Danilevich§
Department of Organic Chemistry and The Minerva Center of Computational Quantum Chemistry, the
Hebrew University, J erusalem 91904, Israel, and Irkutsk Institute of Organic Chemistry, Siberian
Division of the Russian Academy of Sciences, 1 Favorskii Str., 664033 Irkutsk, Russia
Received May 28, 1997X
Intramolecular element effects kBr/kCl for substitution of geminal bromochloroalkenes BrC(Cl)dC-
(Br)Cl (1), BrC(Cl)dCCl2 (2), Me2CdC(Br)Cl (3), and XCHdC(Br)Cl (X ) Cl, 4; X ) Br, 5), with
MeO- and RS- nucleophiles were investigated. 3 did not give substitution, and 4 and 5 gave
substitution with MeO- via an initial elimination (to acetylene)-addition route, followed by further
reactions. In reactions of 4 with thiolates, geminal element effects of 2-10 were obtained.
Formation of RSC(Cl)dC(Cl)Y, Y ) SR, Br, is ascribed to an initial halophilic reaction, followed by
addition of RSCl to the formed acetylene. Reaction of 2 with MeO- gave a high vicinal element
effect, and RS- gave a high geminal element effect. Reaction of 1 with both MeO- and RS- ions
gave high (2 orders of magnitude) geminal element effects, which were interpreted as indicating a
rate-determining C-X bond cleavage. This is supported by the high kBr/kCl intermolecular element
effects (k(1)/k(Cl2CdCCl2) with MeO- and PhCH2S- ions. Mechanistic alternatives based on these
observations are discussed.
In tr od u ction
reasons for this have been discussed.7 We know of only
two examples of an exclusive preferential substitution
of the fluorine atom occurring in the reactions of
PhC(F)dC(F)Cl and EtOC(F)dC(F)Cl with PhLi.9 How-
ever, these reactions apparently proceed through a dif-
ferent substitution route than eq 1 and include a four-
center transition state, so the result is determined by the
stronger Li-F bond as compared with the Li-Cl bond.
Despite the many available values of intermolecular
kBr/kCl EE,3,10 we know of only two cases of an intramo-
lecular geminal EE in a >CdC(Cl)Br system, i.e., in the
reactions of 1-bromo-1-chloro-2,2-bis(p-nitrophenyl)eth-
ylene or of 9-(bromochloromethylene)fluorene with thio
and oxygen nucleophiles, where the values of the EE were
ca. 2.0-3.211
The mechanistic significance of the element effects is
that they can give information on the involvement of the
C-X bond cleavage in the rate-determining step (RDS)
in nucleophilic vinylic substitution. For example, a high
intramolecular kBr/kCl EE will indicate that the RDS
involves a C-X bond cleavage step, either in the rare (or
unknown) single step process12 or in a second slow C-X
bond cleavage step (k2) from the intermediate carbanion
Element effects (EEs)1 studied in nucleophilic vinylic
substitution (SNV) proceeding via the addition-elimina-
tion route (eq 1) are mainly “intermolecular” EEs where
the reactivities of two systems >CdC-X differing only
in the nucleofuge X are compared. “Intramolecular” EEs
kX/kY, where the leaving abilities of the nucleofuges X and
Y are compared in systems possessing two different
leaving groups either in >CdCXY (geminal intramolecu-
lar EE) or in -CXdCY- (vicinal intramolecular EE),
were mostly measured for X ) F and Y ) Cl or Br.2 The
main difference between the inter- and intramolecular
EE is that, despite higher reactivities of vinyl fluorides
than those of the corresponding vinyl chlorides or
bromides in the intermolecular EE (i.e., kF/kCl . 1),3
in geminal vinyl halides >CdC(X)F the heavier
halogen atom is normally replaced predominantly or
exclusively,2,4-8 though kF/kCl values of ca. 1 were ob-
served for the intramolecular EE in systems CHXdCFX
and CX2dCFX (X ) Cl or Br)7 or R-C(F)dC(F)Cl.8 The
(5) (a) Koch, H. F.; Koch, J . G. In Molecular Structure and Energet-
ics: Fluorine Containing Molecules. Liebman, J . E., Greenberg, A.,
Dolbier, W. R., J r., Eds.; VCH Publ.: NY.; 1988; pp 99-121. (b) Koch,
H. F. In Comprehensive Carbanion Chemistry; Buncel, E., Durst, T.,
Eds.; Elsevier: Amsterdam, 1987; Part C, p 321;
(6) (a) Shainyan, B. A.; Vitkovskii, V. Yu.; Azarov, A. G. Zh. Org.
Khim. 1992, 28, 1711. (b) Shainyan, B. A.; Danilevich, Yu. S. Zh. Org.
Khim. 1996, 32, 209.
(7) Shainyan, B. A.; Vereshchagin, A. L. Zh. Org. Khim. 1993, 29,
2375.
† Department of Organic Chemistry, the Hebrew University.
‡ The Minerva Center of Computational Quantum Chemistry, the
Hebrew University.
(8) Norman, J .; Sauvetre, R.; Villieras, J . Tetrahedron 1975, 31, 891.
Sauvetre, R.; Norman, J .; Villieras, J . Tetrahedron 1975, 31, 897.
(9) Meier, R.; Buhler, F. Chem. Ber. 1957, 90, 2344.
(10) (a) Rappoport, Z. Adv. Phys. Org. Chem. 1969, 7, 1. (b) Modena,
G. Acc. Chem. Res. 1981, 14, 7. (c) Rappoport, Z. Acc. Chem. Res. 1992,
25, 474.
(11) Avramovitch, B.; Weyerstahl, P.; Rappoport, Z. J . Am. Chem.
Soc. 1987, 109, 6687.
(12) (a) Ochiai, M.; Oshima, K.; Masaki, Y. J . Am. Chem. Soc. 1991,
113, 7059. (b) Glukhovtsev, M. N.; Pross, A.; Radom, L. J . Am. Chem.
Soc. 1994, 116, 5961. (c) Lucchini, V; Modena, G.; Pasquato, L. J . Am.
Chem. Soc. 1995, 117, 2297.
§ Siberian Division of the Russian Academy of Sciences.
X Abstract published in Advance ACS Abstracts, October 1, 1997.
(1) Bunnett, J . F.; Garbisch, E. W., J r.; Pruitt, H. M. J . Am. Chem.
Soc. 1957, 79, 385.
(2) Shainyan, B. A.; Rappoport Z. J . Org. Chem. 1993, 58, 3421 and
references therein.
(3) (a) Rappoport, Z. Acc. Chem. Res. 1981, 14, 7. (b) Shainyan, B.
A. Russ. Chem. Rev. 1986, 55, 511.
(4) Burton, D. J .; Krutzch, H. C. J . Org. Chem. 1971, 36, 2351.
S0022-3263(97)00935-3 CCC: $14.00 © 1997 American Chemical Society

8050 J . Org. Chem., Vol. 62, No. 23, 1997
Beit-Yannai et al.
(I) formed in the first step (k1) of a multistep process (eq
1). An intramolecular EE close to unity will indicate
either a rate-determining k1 or that the transition state
for the C-X bond cleavage step k2 is early.
Compounds 1, 4, and 5 were obtained as mixtures of
the Z and E isomers. The isomeric ratio for 1 prepared
according to eq 2 was 3:4 based on the 13C NMR spectrum
that displayed two signals at 108.05 and 107.21 ppm. The
assignment was based on comparison of the 13C NMR
spectrum of 1 with that of its bromofluoro analogue for
which the E and Z isomers were assigned on the basis of
the F-F couplings on the 13C satellites.2 The 13C NMR
spectrum of BrC(F)dC(Br)F displayed two signals in a
The goal of the present work is to use this probe in
order to extend our previous mechanistic studies on
systems of relatively low reactivity, the polyhaloethyl-
enes, where a C-X bond cleavage may be involved in the
RDS. An experimental problem is that polysubstitution
can distort the product ratios. However, in the substitu-
tion of polychloroethylenes with sodium phenolate and
benzenethiolate in dipolar aprotic solvents,13 it was found
that substitution of one or more chlorine atoms can take
place depending on the reaction conditions. This and our
previous finding that monosubstitution can occur if low
nucleophile concentrations are used and the reaction is
monitored at early reaction times enabled us to overcome
this difficulty.11
1
ca. 25:75 ratio: the Z isomer at 128.51 ppm with J (13C-
F) 324.2 Hz and 2J (13C-F) 34.8 Hz and the main E isomer
at 123.58 ppm with 1J (13C-F) 252.8 Hz and 2J (13C-F)
103.0 Hz. Their ratio resembles that of 21:79 determined
1
from the H NMR spectrum of BrCHdC(Br)F.16 Hence,
we conclude that the downfield and upfield signals in the
13C spectra of BrCX)C(Br)X for both X ) F and X ) Cl
belong to the Z and E isomer, respectively. The chemical
shifts difference ∆δ(Z-E) between the isomers should
decrease on decreasing the difference in electronegativity
of the geminal substituents. Indeed, for BrC(F)dC(Br)F
and for 1 ∆δ(Z-E) ) 5 and 1 ppm, respectively. There-
fore, the E:Z ratio for 1 is 4:3. When compound 1 was
prepared by bromine addition to dichloroacetylene (eq 4)
the E:Z ratio was 11:1, as was expected from a predomi-
nant trans-addition to the acetylene.
Resu lts
Two types of compounds were investigated: (a) alkenes
lacking a vinylic hydrogen, such as BrC(Cl)dC(Br)Cl (1),
Cl2CdC(Br)Cl (2), and Me2CdC(Br)Cl (3), where compe-
tition from elimination-addition pathways is not a
problem and (b) trihaloethylenes bearing a vinylic hy-
drogen, i.e., ClCHdC(Br)Cl (4) and BrCHdC(Br)Cl (5),
where substitution via an elimination-addition route is
a mechanistic option. The nucleophiles used were sodium
methoxide in the dipolar aprotic solvents acetonitrile and
DMF, and sodium thiolates in acetonitrile.
For compound 2 the 13C NMR spectrum displayed two
signals at 106.14 and 121.83 ppm, corresponding to the
CClBr and CCl2 groups, respectively.17
The two vinylic carbons of compound 3 appear in the
13C NMR spectrum at 134.72 (CClBr) and 100.02 (CMe2)
ppm and the two methyl groups at 24.21 and 21.62 ppm.
1
Syn th esis. Chlorobromoethylenes 1, 2, and 4 were
prepared by bromination/dehydrobromination sequences
of dichloro- and trichloroethylenes according to eqs 2 and
3:
In the H NMR spectrum the two methyl groups appear
at 1.89 ppm due to accidental isochronicity.
In the 1H NMR spectrum of 4 two isomers are observed
in a 4:5 ratio. The more intense signal is at 6.70 ppm
and the less intense one is at 6.60 ppm. Their assign-
ment to the Z and E isomer, respectively, is based on an
additivity scheme18 [calculated to be 7.02 (Z) and 6.71
(E) ppm]. In the 13C NMR spectrum the signals were at
111.14 (CClBr) and 122.35 ppm (CHCl) for Z-4 and
108.18 (CClBr) and 120.54 ppm (CHCl) for E-4.
For compound 5, the two signals in the 1H NMR
spectrum at 7.08 and 6.76 ppm in a 1:2 ratio were
assigned by analogy to the Z and E isomers, respectively.
Rea ction s of th e Meth oxid e Ion . With 1. The
reaction of 0.02 M 1 with an equimolar amount of MeONa
in 50 mL of DMF at 100 °C for 5 h was slow, giving only
7% conversion. GC-MS analysis showed that the main
product is a Cl2BrC2OMe isomer showing a cluster of base
peaks (lower one at m/z 204; throughout this paper we
will give only the lowest m/z values for a cluster of peaks
for isotopomeric species) with the proper isotopic distri-
Compound 1 was also prepared by Pielichowski’s
method (eq 4)14 which was also used to obtain compound
5:
1
bution for the monobromo substitution product. The H
NMR spectrum of the reaction mixture after water-
chloroform workup showed two MeO signals at δ 3.64
and 3.76 ppm in a 2:1 ratio, indicating formation of the
Z/E mixture of BrC(Cl)dC(Cl)OMe (6). As with 4 and 5
(see below), no trace of the chlorine displacement product,
i.e., Br2ClC2OMe, nor any other substitution product was
Compound 3 was prepared from 1-chloro-2-methyl-
propene according to the method of Cunico and Han (eq
5):15
(13) Tanimoto, S.; Taniyasu, R.; Takahashi, T.; Miyake, T.; Okano,
M. Bull. Chem. Soc. J pn. 1976, 49, 1931.
(14) Pielichowski, J .; Popielaez, R. Synthesis 1984, 433.
(15) Cunico, R. F.; Han, Y.-K. J . Organomet. Chem. 1978, 162, 1.
(16) Demiel, A. J . Org. Chem. 1962, 27, 3500.
(17) Stothers, J . B. Carbon-13 NMR Spectroscopy; Academic Press:
New York, 1972; pp 183-195.
(18) Pascual, C.; Meier, J .; Simon, W. Helv. Chim. Acta 1966, 49,
164.

An Example of a Single Step Substitution?
J . Org. Chem., Vol. 62, No. 23, 1997 8051
Ta ble 1. Rea ction of Excess
1,2-d ibr om o-1,2-d ich lor oeth ylen e (1) w ith RSNa in
MeCNa
detected by GC-MS. When the reaction was carried out
in MeCN, the only products were again E-6 and Z-6,
displaying the same 1H NMR and GC-MS signals as
those formed in the reaction in DMF, although a sub-
stantial resinification also took place.
With 2. In the reaction of 0.02 M 2 with 0.02 M
MeONa in 50 mL of DMF at 100 °C for 5 h unreacted 2
was recovered in addition to four additional species which
were identified by GC-MS and NMR spectra:
b
c
product
R ) PhCH2
R ) PhCH2
R ) Me
ClC(Br)dC(Br)Cld
Cl2C(Br)CH(Br)Cle
RSC(Cl)dCHCl
RSC(Cl)dC(Br)Cl
RSSR
72.4
47.4
11.8
7.1
25.6
8.1
91.3
3.2
7.7
13.3
4.3
a
b
Numbers given are relative uncalibrated GC intensities.
1
(1) The ortho ester ClCHBrC(OMe)3 [MS: m/z 201 (21,
M - OMe), 105 (46, C(OMe)3). 1H NMR δ (CDCl3): 6.02
s (1H, CH), 3.45 s (9H, Me)], formed in ca. 1% yield [The
was prepared according to eq 2. c 1 was prepared according to eq
d
4. Also detected were 2.5 ( 0.3% of Br2C(Cl)CH(Br)Cl (the
precursor to 1) and 1.4 ( 0.2% of Br2CdC(Br)Cl (its HCl elimina-
tion product) which were not detected by NMR of precursor 1,
although they are presumably present in it in minor percentage.
e Presumably formed by addition of Br2 to trichloroethylene, which
is the first precursor to 1 in eq 4.
1
analogous Cl2CHC(OMe)3 is known. H NMR δ (CDCl3):
5.75 (CHCl2), 3.45 (OMe).7]; (2) trichloroethylene [1H
NMR δ (DMF-d6): 6.45], formed in traces; (3) E- and
1
Z-BrC(Cl)dC(Cl)OMe (6) (the same H NMR signals as
in the reaction of 1), formed in <1% yield; and (4)
ClCHdC(Cl)OMe (7) [1H NMR δ (CDCl3): 5.47 (1H,
dCH), 3.78 (3H, Me) (lit.7 5.45 and 3.78)], formed in <1%
yield.
Ta ble 2. Rea ction of Excess
1-Br om o-1,2,2-tr ich lor oeth ylen e (2) w ith RSNa in MeCNa
product
R ) PhCH2
R ) Me
Cl2CdC(Br)Cl
39.6
2.4
8.3
92
1.1
The reaction in MeCN was much faster even though
it took place at a lower temperature and for a shorter
period (80 °C, 1 h), and it gave 44% of unreacted 2 and
four additional species whose percentages are given below
on the basis of their uncalibrated integration values:
(1) trichloroethylene ClCHdCCl2 [m/z 130 (100, M), 95
(100, M - Cl), 60 (43, M - 2Cl)], formed in 42%; (2)
ClCHdC(Br)Cl (4) [m/z 174 (43, M), 139 (5, M - Cl), 95
(100, M - Br)], formed in 9% yield; (3) ClCHdC(Cl)OMe
(7) [m/z 126 (M), overlapped with trichloroethylene],
formed in 3% yield; and (4) dichloroacetylene ClCtCCl
(8) [m/z 94 (100, M), 59 (15, M - Cl), 47 (11, CCl)], formed
in 2%.
ClC(Br)dC(Br)Clb
RSC(Cl)dCHCl
RSC(Cl)dCCl2
9.9
0.8
6.1
RSC(Cl)dC(Br)Cl
31.0
a
Numbers given are relative uncalibrated GC intensities.
Presumably formed in a small quantity by elimination of HCl
rather than HBr from precursor BrCl2CCH(Cl)Br.
b
When the MeONa was dried for 2 days under high
vacuum and the mixture was protected from moisture,
the GC-MS of the crude reaction mixture showed (apart
from unreacted 4) a large peak of dichloroacetylene (8)
and only a small peak of 7. In the 13C NMR coupled
spectrum a singlet at 49.23 ppm, assigned to dichloro-
acetylene, was observed (calculated δ13C 53 ppm).
With 5. In the reaction of 1,2-dibromo-1-chloroethyl-
ene 5 with MeONa in DMF, no substitution products
were formed as judged from the lack of MeO signals in
the 3-4 ppm region in the 1H NMR spectrum of the crude
reaction mixture. In contrast, in the reaction in MeCN,
seven singlet signals were observed between 3.4 and 3.9
ppm, five of them at 3.43, 3.48, 3.76, 3.86, and 3.89 were
small (5, 6, 6, 8, and 12%, respectively) and the two most
intense ones were at 3.60 (30%) and 3.84 (33%) ppm. We
ascribe the signals at 3.84 and 5.99 ppm to the ester
ClBrCHCOOMe and the pair of signals at 3.60 and 5.2
ppm to the monobromo substitution product BrCHdC-
(Cl)OMe. When starting from a 1:2 Z-5:E-5 ratio, the
only isomer observed at the end of the reaction was E-5,
indicating that the reactivity of the Z isomer was much
higher than that of E-5.
Rea ction s w ith Th iola te Ion s. Chlorobromoethyl-
enes 1-5 in 8-fold molar excess were reacted with
methyl, tert-butyl, and benzyl thiolate ions (RS-, R ) Me,
t-Bu, PhCH2) under argon at room temperature in MeCN
for 24 h. These conditions ensure that only monosub-
stitution takes place at early reaction percentage, when
product composition was determined. Analysis was
mainly conducted by GC-MS, but 13C and 1H spectra
were also recorded when appropriate. The results for 1,
2, 4, and 5 are summarized in Tables 1-4. The results
for tri- and tetrachloroethylenes are given in Tables 5
and 6. The reaction of 3 at 78 °C with MeS- for 18 h or
with PhCH2S- for 23 h did not give any substitution
product.
With 4. The reaction of 4 with a suspension of an
equimolar amount of MeONa in MeCN proceeds exother-
mally at room temperature. The reaction mixture was
analyzed after 1 h by GC-MS and gave several com-
pounds. The main peak (75% by uncalibrated GC)
consisted of the overlapped peaks of unreacted 4 and the
monobromo substitution product, ClCHdC(Cl)OMe (7)
1
(for C3H435Cl2O m/z 126, M). The H NMR spectrum of
isolated 7 displayed only two signals at δ (CDCl3) 3.74
(MeO) and 5.43 (dCH) ppm (lit.7 3.78 and 5.45 ppm),
suggesting that only a single geometrical isomer was
formed. Likewise, the 13C NMR spectrum displayed only
three signals at 58.50 (OMe), 97.0 (dCHCl), and 144.65
(dCCl) ppm. These data are insufficient for distinguish-
ing between the Z and E isomers of 7, and the structure
1
was not assigned in the literature either.7 The H NMR
spectrum of the crude reaction mixture showed a Z-4:
E-4 ratio of 37:63, lower than the 45:55 ratio for the
precursor, indicating that the Z isomer reacted faster.
In addition to 7 and 8, five more compounds were
1
detected and identified by GC-MS and H NMR:
(1 and 2) methyl chloride (9) [m/z 50 (100, MeCl)] and
methyl bromide (10) [m/z 94 (50, MeBr), 79 (5.3, Br); the
two overlapping peaks consisted 10% of the observed
intensity]; (3) methyl chloroacetate (ClCH2COOMe, 11)
[1H NMR δ (CDCl3): 3.72 (Me), 4.11 (CH2)], formed in
3% yield; (4) methyl dichloroacetate (Cl2CHCOOMe) [m/z
127 (100, M - Me), 113 (78, M - CO - H), 95 (32, C2-
HCl2), 49 (73, CH2Cl). 1H NMR δ (CDCl3): 3.84 (Me),
5.94 (CH) (lit.7 5.97, 3.89).], formed in 2% yield; (5) BrC-
(Cl)dC(Cl)OMe (6) [m/z 204 (31, M), 161 (69, CCl2Br)],
as traces; and (6) compound 1 [m/z 252 (32, M), 173 (60,
M - Br), 138 (14, M - Br - Cl), 94 (46, M - 2Br), 79
(13, Br), 59 (26, M - 2Br - Cl)], formed in 4% yield.
1 vs Tetr a ch lor oeth ylen e Exter n a l Elem en t Ef-
fects. Because of expected complications due to the low

8052 J . Org. Chem., Vol. 62, No. 23, 1997
Beit-Yannai et al.
Ta ble 7. Rea ction of a 1:1 Mixtu r e of
1,2-Dibr om o-1,2-d ich lor oeth ylen e (1) a n d
Tetr a ch lor oeth ylen e w ith MeONa in MeCN (in p er cen ts
r ela tive to MeCN)
Ta ble 3. Rea ction of Excess
1-Br om o-1,2-d ich lor oeth ylen e (4) w ith RSNa in MeCNa ,b
c
product
R ) t-Bu
R ) Me
R ) PhCH2
ClCHdC(Br)Cl
RSCHdCHCl
RSC(Cl)dCHCl
RSC2H(Br)Cle
RSC(Cl)dC(Br)Cl
RSC(Cl)dC(Cl)SR
RSSR
58.9
0.4
4.8
2.3
16.3
31.5
d
5.8
45.5
12.9
2.4
8.1
3.2
17.5
4.0
13.1
3.0
29.3
time
(min)
no.
C2Cl4
1
6
EE
1
2
3
4
5
6
7
8
9
0
1
8.3
15.5
22.7
30
37
45
67
89
21.8
19.9
20.1
20.2
20.1
20.1
21.0
20.5
20.4
20.6
20.5
20.2
13.3
8.5
6.8
6.7
6.6
6.1
6.3
6.8
6.8
6.6
3.9
7.4
9.0
8.6
8.6
2.9
4.0
4.2
4.3
5.1
4.8
4.5
4.6
4.6
17.3
a
b
Numbers given are relative uncalibrated GC intensities. The
precursor contains traces of BrCHdC(Br)Cl. c 3.5% of an unidenti-
fied compound was also present. Its highest m/z was 147, it
8.5
11.5
10.3
12.1
11.3
+
contained no halogen, and it showed a C7H7 signal. It may be
PhCH2SCtCH (m/z 148) which by loss of H gives an S-benzylthi-
irenium ion (m/z 147), which is a known MS fragment. The
10
11
d
102
retention time is identical with that of 4, so that separate
integration is impossible. e Two compounds of this composition
with different retention times were observed in the GC-MS for R
) Me and PhCH2. For R ) Me, in the mass spectrum of the first
eluted compound (2.8%) there are fragments at m/z 125 and 127
in 1:1 ratio which are consistent with the fragment [CSBr]+. These
do not appear in the spectrum of the second compound (42.7%)
which showed a 1:1 ratio of m/z 126 and 128, which are consistent
with the fragment [MeSBr]+. Both spectra display 1:1 ratio of
signals at m/z 136 and 138, belonging to [C2HSBr]+ formed by loss
of Me and Cl from vicinal carbons. The first eluted compound was
tentatively assigned as RSC(Br)dCHCl, and the second one was
assigned as RSCHdC(Br)Cl.
Ta ble 8. Rea ction of 1:1 Mixtu r e of
1,2-Dibr om o-1,2-d ich lor oeth ylen e (1) a n d
Tetr a ch lor oeth ylen e w ith P h CH2SNa in MeCN (in
p er cen ts r ela tive to MeCN)
no.
time (min)
C2Cl4
1
EE
1
2
3
4
5
6
0
1.5
5.7
11.5
17.1
23.4
45.4
43.4
43.4
43.3
42.1
41.8
36.5
20.5
12.3
12.4
12.9
13.0
9.9
15
14.2
8.9
8.1
Ta ble 4. Rea ction of Excess
1,2-Dibr om o-1-ch lor oeth ylen e (5) w ith P h CH2SNa in
MeCNa
followed by GC. The results are presented in Tables 7
and 8, along with the calculated values of the intermo-
lecular EE for each point. In Table 7 the percentage of
a newly formed compound, which was shown by GC-
MS to be compound 6, is also given. It is evident from
these data that the content of the two substrates de-
creased with time at sharply different rates, the main
changes occurring in the first several minutes for both
nucleophiles. Values of the intermolecular kBr/kCl EE
calculated from eq 6,
productb
percent
productb
percent
BrCHdC(Br)Cl
22
18
6
PhCH2SC(Br)dCHCl
PhCH2SCH2Ph
PhCH2SSCH2Ph
4
7
37
PhCH2SCHdCHCl
PhCH2SCHdCHBr
PhCH2SC(Cl)dCHCl
3
a
b
Numbers given are relative uncalibrated GC intensities. 3%
of an unidentified compound was also present. It displayed the
+
highest m/z at 210, contained no halogen, and showed a C7H7
signal.
[1]o - [1]t
[1]o
Ta ble 5. Rea ction of Excess Tr ich lor oeth ylen e w ith
MeSNa in MeCNa
EE )
(6)
product
percent
product
percent
[C2Cl4]o - [C2Cl4]t
MeSC(Cl)dCHCl
MeSCHdCCl2
50.3
38
MeSCHdC(Cl)SMe
(MeS)2CdCHCl
3.5
8.0
[C2Cl4]o
a
Numbers given are relative uncalibrated GC intensities.
are rather sensitive to small variations of the percentage
of tetrachloroethylene and, hence, should be regarded as
a reasonable approximation. Nevertheless, we can con-
clude that, despite the large error obtained in averaging
the individual points for the EE, the intermolecular EEs
are the highest values obtained so far, being 9.1 ( 1.7
with MeO- and 11.2 ( 2.7 with PhCH2S-.
Ta ble 6. Rea ction of Excess Tetr a ch lor oeth ylen e w ith
RSNa in MeCNa ,b
product
Cl2CdCCl2
R ) t-Bu
R ) Me
25.1
R ) PhCH2
80
RSC(Cl)dCHCl
RSC(Cl)dCCl2
7.3
3.3
<0.1
67.6
c
d
(RS)2CdCCl2 or
1.3 or 5.6
RSC(Cl)dC(Cl)SRd
Discu ssion
RSSR
9.5
a
b
The reactions of the polyhaloethylenes studied here
with the oxygen and sulfur nucleophiles demonstrate
again “the rich mechanistic world of nucleophilic vinylic
substitution”.19 The outcome of a crowded array of
mechanistic routes is that small structural variations can
lead to elimination-addition (E-A), nucleophilic addi-
tion-elimination (AdN-E), and halophilic reactions, tak-
ing place either exclusively or competitively for the
different substrates. Moreover, the initial products
frequently react further under the reaction conditions,
leading to products which could be a posteriori rational-
ized but not necessarily predicted, as demonstrated
below.
Numbers given are relative uncalibrated GC intensities. The
mixture was evaporated and the remainder was analyzed. c Main
d
product. Isolated. Two products with the same m/z were formed
and were not identified.
solubility of the nucleophile and the possible formation
of multisubstitution products, the rates of the reactions
of 1 and tetrachloroethylene required for calculation of
“external” EE were not measured separately but in the
mixture of the two tetrahaloalkenes. The reactions of
mixtures of 1 and tetrachloroethylene were carried out
by dissolving together 5 mmol of each substrate in 1.5-2
mL of MeCN and adding 10 mmol of dry MeONa or
PhCH2SNa at room temperature and the reactions were

An Example of a Single Step Substitution?
J . Org. Chem., Vol. 62, No. 23, 1997 8053
Rea ction s of 1-Br om o-1,2-d ich lor oeth ylen e 4 a n d
1,2-Dibr om o-1-ch lor oeth ylen e 5. (a ) Rea ction s w ith
MeONa . The formation of both MeCl and MeBr (9 and
10) in reaction of 4 with MeONa is reminiscent of the
formation of dimethyl ether in the reaction of BrC(F)dC-
(Br)F with MeONa.2 In the absence of a nucleophilic
methyl anion source, 9 and 10 should be secondary
products derived from the initially formed species. The
most plausible route for their formation is a reaction
between the two products formed in the initial substitu-
tion reaction, i.e., an SN2 attack of Cl- or Br- on the
methoxydebromo substitution product 7 (eq 7).
Consequently, the reactions of 4 and 5 can not provide
data on the intramolecular EE in the AdN-E route.
(b) Rea ction s w ith Th iola tes. The reaction of
trichloroethylene with MeS- which gives both E- and
Z-ClCHdC(SMe)Cl and Cl2CdCHSMe is likely to proceed
via the AdN-E route rather than by an exclusive E-A
route, which should give only ClCHdC(SMe)Cl. We
reasonably assume that 4 and 5 will react similarly since
an E-A route will give only ClCHdC(SMe)Cl in the case
of 4. The AdN-E reaction of E/Z-4 with the three thiolate
nucleophiles RS- (Table 3) could give a priori three E/Z
pairs derived from substitution. For the reaction of 4
with t-BuS- and MeS-, both a bromine replacement
product and a single chlorine replacement product were
formed. With PhCH2S-, a bromine replacement and two
chlorine replacement products were observed, one of
which was <1% of the other. Whereas the bromine
replacement product should be RSC(Cl)dCHCl if no
rearrangement takes place, the single chlorine replace-
ment product (or major one with PhCH2S-) can be either
RSC(Br)dCHCl (13) or RSCHdC(Br)Cl, depending on
the regiochemistry of attack of the nucleophile. This
raises the question if the measured element effect is
geminal or vicinal.
Although the products were not isolated and our
information is derived from GC-MS, we believe that the
chlorine replacement product is 13 for the following
reason. The observed kBr/kCl element effects are 2, 2, and
10, for reaction with t-BuS-, MeS-, and PhCH2S-,
respectively, i.e., the element effect increases with the
increased softness of the nucleophile. We calculated by
AM1 the distribution of the LUMO of 4 and found that
only 6% of it is localized on the â-carbon (CHCl), while
35% is localized on the R-carbon (CClBr).20 As reactions
of soft nucleophiles should be governed by orbital rather
than by charge control, the reaction at the R-carbon
should predominate. In this case, the element effect is
geminal.
An intriguing process in the reaction of trihaloalkene
4 with all the three thiolate ions is the formation in
substantial amounts of the tetrasubstituted trihalovinyl
sulfides, RSC(Cl)dC(Br)Cl (14) (Table 3). Together with
their further substitution product RSC(Cl)dC(Cl)SR, R
) Me, PhCH2, they consist of ca. 16% of the total
substitution product. This can be rationalized in three
ways. (i) A partial disproportionation of 4 to dihalo- and
tetrahaloethylene, with a subsequent substitution of the
latter. This route requires the formation of comparable
amounts of products with two vicinal vinylic hydrogens
to those of the tetrasubstituted alkenes. Such products
were not formed, except for only 0.4-1.1% of chlorovinyl
sulfides RSCHdCHCl. However, since dichloroethylene
is volatile, some of it may be lost from the reaction
mixture before the reaction with the RS-. Consequently,
the theoretically expected 1:1 ratio of the disubstituted
to tetrasubstituted alkenes may be changed in favor of
the latter. (ii) An Ad-E process, initiated by addition of
disulfides RSSR (formed as side products in our reaction)
to 4 followed by elimination of a hydrogen halide mol-
ecule. This route have precedents in the addition of
disulfides21-24 or their R2S+SR‚2,4,6-(NO2)3C6H2SO3- salts23
to both simple and functionalized alkenes. Base-cata-
Although ethers are usually stable to reactions with
nucleophiles, the increased electrophilicity of the methyl
group in 7 due to electron withdrawal by the dihalovinylic
moiety is apparently responsible for the reaction with X-.
Furthermore, the other reaction product, i.e., the vinyloxy
anion 12, is protonated and by a further ketonization and
methanolysis affords methyl chloroacetate 11 which was
1
identified by H NMR spectroscopy.
Substitution of 4 with MeO- to form 7 apparently
proceeds by an elimination-addition pathway, as proved
by the almost exclusive formation of dichloroacetylene 8
under especially dry conditions. The Z isomers of triha-
loethylenes 4 and 5 are more reactive than the E isomers
toward MeONa, as judged by the increase in the percent-
age of the E isomers (from 1H NMR) in the recovered
unreacted precursor after the substitution. Since a base-
promoted elimination of HBr is faster than elimination
of HCl, and hydrogen and bromine are situated trans in
the Z isomer, this finding corroborates an elimination as
the first step in the reaction. Trans addition of MeOH
across the formed triple bond of dichloroacetylene formed
from 4 or bromochloroacetylene formed from 5, respec-
tively, affords the observed ether product 7 (eq 8). In
addition to the evidence based on the relative reactivity
of the E and Z isomers, the presence of a strong base,
and the exclusive formation of 8 under dry conditions,
corroborating evidence for the formation of the explosive
dichloroacetylene in the elimination step of the E-A
route is the fact that even when the MeONa was not
especially dried, the reaction mixture sometimes exploded
when heated in air, and it had the typical smell of
haloacetylene.
(19) Rappoport, Z. Recl. Trav. Chim. Pays-Bas. 1985, 104, 309.
(20) AM1 calculations were performed by using the MOPAC pro-
gram with full geometry optimization.
(21) Holmberg, B. Arkiv Kemi, Mineral., Geol. 1939, 13B, 6 (Chem.
Abstr. 1940, 34, 23419).

8054 J . Org. Chem., Vol. 62, No. 23, 1997
Beit-Yannai et al.
lyzed addition of disulfides to acetylenes25 and perfluo-
roacetylenes26 is also known as well as the formation of
trichlorovinyl sulfides RSC(Cl)dCCl2 in the reaction of
trichloroethylene with disulfides under free-radical con-
ditions.27 The iodine-catalyzed addition of disulfides to
styrenes21,22 implicates sulfenyl iodides RSI as the reac-
tive intermediates. This raises the most likely route (iii)
of sulfenyl bromide addition to 4. The sulfenyl bromide
may arise by a halophilic attack of RS- on the bromine
atom of 4. We therefore tentatively present the formation
of trihalovinyl sulfides RSC(Cl)dC(Br)Cl 14 in the reac-
tion of 4 with thiolate ions by the sequence of steps of eq
9. Note, however, that the reservation raised concerning
the product ratio in route (i) applies here too.
chlorophilic MeO- ion attack on the chlorine of the CCl2
group, followed by protonation of the resulting anion.
Consequently, the reaction of 2 with MeO- still cannot
give the desired geminal AdN-E intramolecular EE.
In the reaction of 2 with thiolate ions, we are again
confronted with the question of regioselectivity, namely,
whether the chlorine replacement product is RSC-
(Br)dCCl2 or RSC(Cl)dC(Br)Cl, and therefore, whether
the intramolecular element effect is geminal or vicinal.
The measured kBr/kCl EEs are 0.13 with MeS- and 0.32
with PhCH2S-, i.e., the value is higher for the softer
nucleophile. It seems highly unlikely that the intermedi-
ate carbanion [Cl2C-C(Cl)(Br)SR]- would expel Cl- much
easier than Br- when both halogens are geminal. More-
over, in contrast with 4, the AM1 calculations showed
that the LUMO of the substrate is almost equally
localized on both carbon atoms (44% on CClBr and 47%
on CCl2).20 Finally, the methylthio and benzylthio tri-
haloethylenes were identical with the monobromo re-
placement products in the reaction of 1 with MeS- and
PhCH2S-, indicating that they have structure 14. We
therefore interpret our reactions as reflecting a competi-
tive attack of the thio nucleophiles on both carbons of 2.
Attack at Câ generates [Cl2C(SR)C(Br)Cl]- (16), from
which only Cl- can be expelled, while attack at CR results
in Br- expulsion or Cl- expulsion. Consequently, two
element effects are obtained. The ratio of the chlorine
replacement to the bromine replacement product reflects
the ratio of attack on CR vs Câ and is therefore the vicinal
element effect. The fact that none of Cl2CdC(Br)SR is
formed indicates a high geminal element effect, as was
observed in the reactions of 1. Finally, 1-(benzylthio)-
1,2-dichloroethylene is also formed in the reaction with
PhCH2S-, most likely by a vinylic substitution of one
halogen and a halophilic reaction (followed by protona-
tion) on a different vicinal halogen.
R ea ct ion s of 1,2-Dib r om o-1,2-d ich lor oet h ylen e
(1). The only compound free from problems of both an
elimination-addition process and regioselectivity in the
attack of a nucleophile is 1,2-dibromo-1,2-dichloroethyl-
ene (1), which lacks a vinylic hydrogen and for which CR
) Câ. Consequently, it is suitable for evaluation of an
unequivocal geminal element effect. The reactions of 1
with either MeO- or PhCH2S- were found to be cleaner
than those of the other investigated substrates. With
MeO- the only product observed is the bromine replace-
ment product MeOC(Cl)dC(Br)Cl (6). None of the chlo-
rine replacement product was observed. In the reaction
with PhCH2S-, the only monosubstitution product ob-
served was again the benzylthio debromination product
NuC(Cl)dC(Br)Cl (17, Nu ) PhCH2), and none of the
benzylthio dechlorination product was observed. In
addition, a single monosubstituted , monohalogen reduc-
tion product PhCH2SC(Cl)dCHCl was obtained and no
chlorine was replaced during its formation. Reaction
with MeS- also gave only the methylthio debromination
product 17, Nu ) MeS, and none of NuC(Br)dC(Br)Cl
(18, Nu ) MeS). These are important observations since
they amount to a high kBr/kCl intramolecular geminal EE
with both nucleophiles. Unfortunately, this conclusion
Rea ction s of Br om otr ich lor oeth ylen e 2. In the
reactions of the tetrasubstituted 2 there is a regioselec-
tivity problem since attack can take place on either of
the two different sites, the dC(Cl)Br (CR) and dCCl2 (Câ).
If the attack is on CR, the measured element effect will
be an intramolecular geminal element effect, while when
it takes place at both CR and Câ, both intramolecular
geminal and vicinal effects will be obtained.
The reaction of 2 with MeO- in DMF, which gives
products that retained bromine, i.e., the Z and E isomers
of 6 as well as its further reaction product, the ortho ester
BrCHClC(OMe)3, indicates that the nucleophilic attack
takes place on the dCCl2 rather than on the dC(Cl)Br
group. This is consistent with the results of the AM1
calculations which showed that the electron density on
Câ (-4.079) is 0.13 e lower than on CR (-4.210).20
The other two products lack bromine, but they are not
the simple bromine replacement products since they
contain a vinylic hydrogen. They are most likely formed
by a bromophilic reaction of the MeO- on 2, forming
carbanion CCl2dCCl- (15), which is then protonated to
trichloroethylene, i.e., by a stepwise reduction. The main
product in the reaction in MeCN, which is trichloroeth-
ylene, seems also to be formed similarly. Its further
substitution gives 7. The trichlorovinyl carbanion 15 can
also expel a chloride ion, thus accounting for the forma-
tion of the observed dichloroacetylene (eq 10).
The second most abundant product is 1-bromo-1,2-
dichloroethylene 4, which is most likely formed by a
(25) Trofimov, B. A.; Shainyan, B. A. In: The Chemistry of Func-
tional Groups. Suppl. S. The Chemistry of Sulfur-Containing Func-
(22) (a) Schneider, H. J .; Bagnell, J . J . J . Org. Chem. 1961, 26, 1984.
(b) Schneider, H. J .; Bagnell, J . J .; Murdoch, G. C. J . Org. Chem. 1961,
26. 1987.
(23) Helmkamp, G. K.; Olsen, B. A.; Pettitt, D. J . J . Org. Chem.
1965, 30, 676.
(24) Bodrikov, I. V.; Kovaleva, L. I.; Zefirov, N. S. Zh. Org. Khim.
1976, 12, 2476.
tional Groups; Patai, S., Rappoport, Z., Eds.; J . Wiley
Chichester; 1993; pp 659-797.
& Sons:
(26) Krespan, C. G.; McKusick, B. C. J . Am. Chem. Soc. 1961, 83,
3434. Krespan, C. G. J . Am. Chem. Soc. 1961, 83, 3438.
(27) Voronkov, M. G.; Martynov, A. V.; Mirskova, A. N. Sulfur Rep.
1986, 6, Part 2, 77.

An Example of a Single Step Substitution?
J . Org. Chem., Vol. 62, No. 23, 1997 8055
could not be extended to t-BuS- ion since it did not give
any substitution product. If we assume that 0.1% of the
chlorine replacement product could have been detected,
as found for several other products obtained in this study,
and if we add the reduction/substitution product to the
monosubstitution product, then the geminal element
effects should be g107, g43, and g100 with PhCH2S-,
MeS-, and MeO-, respectively. Since (a) the structures
of the products in this system are unambiguous, (b) the
values are high, (c) the values are lower limits to the
geminal EE, and (d) highly unlikely errors of 100% will
still retain high EE values, we conclude that errors, e.g.,
due to the lack of calibration, are not going to affect our
major conclusion that the geminal EEs are high.
If the process investigated is one of the possible
variants of the “addition-elimination” type nucleophilic
vinylic substitution,3,10,19 in which the nucleophile attacks
the vinylic carbon and the leaving group is expelled from
the formed transition state (TS) or a carbanionic inter-
mediate, two cases can be envisioned.
(a) The reaction is a single step substitution occurring
via a single TS where bond formation to the nucleophile
and cleavage of the C-halogen bond occur concertedly,
i.e., eq 1 with k1 and k2 merging into a single step process,
i.e., I is a transition state rather than an intermediate.
In this case there are two different single step TSs,
chloride ion is expelled in one of them and bromide ion
is expelled in the other, and since bromide is a better
nucleofuge, the energy of the latter TS is much lower,
leading to the observed high geminal element effects. We
note that each one of these TSs can have two completely
different configurations: (i) the nucleophilic attack can
be on the π* orbital, perpendicular to the double bond,
and the process involves formation and cleavage of the
bonds as well as intramolecular rotation required to
achieve the proper stereoelectronic configuration (anti-
periplanar arrangement of the leaving group-carbon bond
and the carbanionic lone pair) in order to obtain the
substitution product; (ii) an in-plane SN2 nucleophilic
substitution where the nucleophile attacks from the back
of the leaving group.
tion (e.g., E- and Z-BrC(F)dC(Br)F are unstable to
mutual isomerization2), although both isomers could be
detected by 13C NMR. Consequently, the stereochemical
tool could not be applied and the remaining tools are the
intramolecular and intermolecular EEs. The ratio of the
bromine-replaced to the chlorine-replaced products in eq
11 (i.e., the geminal EE) is equal to the rate ratio for
expulsion of Br- vs Cl- from the same “ground state”
carbanion NuC(Br)(Cl)C(Br)Cl- (19).11 As expulsion of
different leaving groups occurs from different conforma-
tions of carbanion 19, characterized by antiperiplanar
arrangement of the carbanionic lone pair and the leaving
group expelled, there must be a bifurcation point on the
potential energy surface preceding to and resulting in two
different TSs. The EE in this case is a measure of how
early or late the corresponding TSs are. J udging from
intermolecular element effects in SN1 reactions, i.e., kion
-
(RBr)/kion(RCl), the ratios28 are mostly between 1 and 2
orders of magnitude in an endothermic carbocation
forming C-X bond cleavage, and a much lower value is
anticipated for an expected exothermic reaction in which
a â-halo carbanion is converted to a halide ion and a
neutral alkene. The TS in the latter reaction should be
much earlier than in an SN1 reaction and the element
effect is therefore expected to be lower. Note, however,
that an SN1, E1, or even E2 reaction differs somewhat
from what we have in our case, since in our case the
nonexpelled halogen stabilizes the TS, so we are compar-
ing expulsion of Cl- in a Br-stabilized TS to an expulsion
of Br- in a Cl-stabilized TS. Unfortunately, we have no
data for a proper comparison for the product ratio derived
from loss of bromide and chloride ion, i.e., on SN1
reactions of RR′C(Br)Cl solvolysis and on E1 or E2
reactions of RR′CH-CR′′(Br)Cl, respectively. Neverthe-
less, since the polar and resonance effects of Cl and Br
are similar, this effect could be neglected in the first
approximation. If so, an independent assessment of the
intramolecular EE values can be made based on the
known data for energy of Br- vs Cl- formation in
ionization processes, hydration, and transfer from water
to various solvents. The heterolytic bond dissociation
energies for RX f R+ + X- in the gas phase for R ) Me,
Et, n-Pr, i-Pr, t-Bu, Ph, vinyl, allyl, and PhCH2 are by
6-9 kcal mol-1, averaging 7.7 kcal mol-1, higher for X )
Cl than for X ) Br.29 For hydration, i.e., transfer from
(b) The reaction proceeds by the customary two-step
nucleophilic vinylic substitution (eq 11).3,10
the gas phase to water, ∆Go ) -81.3 (Cl-) and -75.3
hyd
(Br-) kcal mol-1 and ∆Hohyd ) -87.2 (Cl-) and -80.1 (Br-)
kcal mol-1. The maximum predicted EE in terms of ∆H
is therefore 0.7 kcal mol-1, i.e., less than 1 order of
magnitude. For transfer of X- from water to MeCN,
∆Gtrans ) 10.1 (Cl-), 7.4 (Br-) kcal mol-1 and ∆Htrans
)
4.5 (Cl-), 1.7 (Br-) kcal mol-1; for transfer from water to
DMSO, ∆Gtrans ) 9.8 (Cl-), 6.5 (Br-) kcal mol-1 and
∆Htrans ) 4.8 (Cl-), 1.2 (Br-) kcal mol-1 30
Hence, ∆∆Gtrans
.
(Cl- - Br-) ) 2.7 (MeCN), 3.3 (DMSO) kcal mol-1 and
∆∆Htrans (Cl- - Br-) ) 2.8 (MeCN), 3.6 (DMSO) kcal
(28) For few collections of several kBr/kCl ratios, see: Streitwieser,
A., J r. Solvolytic Displacement Reactions; McGraw-Hill: New York,
1962; p 30 and ref 40 therein: Rappoport, Z.; Gal, A. J . Chem. Soc.,
Perkin Trans. 2 1973, 301. For several other values, see: Ross, S. D.;
Labes, M. M. J . Am. Chem. Soc. 1957, 79, 4155. Gelles, E.; Hughes,
E. D.; Ingold, C. K. J . Chem. Soc. 1954, 2918; Pocker, Y. J . Chem.
Soc. 1960, 1972. Heinonen, K.; Tomilla, E. Suom. Kemistil. 1965, 38B,
9.
(29) Morrison, R. T.; Boyd, R. N. Organic Chemistry, 6th ed.;
Prentice Hall: New York, 1992; p 22.-
(30) Marcus, Y. In Liquid-Liquid Interfaces. Theory and Methodol-
ogy; Volkov, A. G., Deamer, D. W., Eds.; CRC Press: Boca Raton, 1996;
Chapter 3, p 321.
In this case, the expulsion of the leaving group(s) takes
place starting from a discrete intermediate carbanion.
Two mechanistic probes to distinguish some of these
alternatives are the stereochemistry of the reaction and
the “external” EE. The in-plane reaction should lead to
inversion, whereas the perpendicular attack either via a
single-step or a multistep route should lead to retention.
Unfortunately, we could not separate the E and Z isomers
of 1, and we do not know if they are stable to isomeriza-

8056 J . Org. Chem., Vol. 62, No. 23, 1997
Beit-Yannai et al.
Exp er im en ta l Section
mol-1. Consequently, the maximum EE (in kcal mol-1
)
is 3.5 (in ∆∆H) or 4.5 (in ∆∆G) in MeCN and 5.3 (∆∆G)
or 4.0 (∆∆H) in DMSO. At room temperature the
maximum kBr/kCl EE will be two to 3 orders of magnitude.
This is consistent with the observed values for SN1, E1
and E2 reactions which are mostly <100.28 Our intramo-
lecular EE can therefore be regarded as high for expul-
sion of halide ions from a carbanion and pathway a seems
a reasonable alternative.
Melting points were determined with a Thomas-Hoover
instrument and are uncorrected. GC analysis was performed
on a LKhM-80 chromatograph, using 3 × 2000 mm columns
of 5% silicon elastomer SE-30 on Chromaton N-AW HMDS,
a catharometer detector, and, helium as carrier gas, or on a
Packard-436 chromatograph, with FID and FPD detectors.
GC-MS analysis was performed on LKB 2091 and Finnigan
ITS-40 instruments at 70 eV on a 60 m capillary column with
SE-30 in a programming mode (injector temperature 250 °C,
initial temperature 100 °C, increase at a rate of 16-20°/min
for ca. 10 min, and a few minutes at the final temperature).
HPLC experiments were conducted with a Merck-Hitachi
L-6200 instrument with a UV detector. Mass spectra were
obtained on a MAT 311 mass spectrometer, UV spectra with
a UVIKON-930 Contron spectrophotometer, and IR spectra
with Perkin-Elmer 157G and 1600-FT-IR instruments. 1H
and13C NMR spectra were recorded on Bruker WP-200 (200
MHz for 1H, 50.3 MHz for 13C), Bruker AMX-400 (400 MHz
If we accept the high intramolecular EE as reflecting
a concerted transition state for the substitution, cf. 20,
this must be necessarily reflected also in a high inter-
molecular EE.
1
for H, 100.6 MHz for 13C), and Varian VXR 500 (500 MHz for
1H, 125.7 MHz for 13C) instruments, in CDCl3.
Solven ts a n d Ma ter ia ls. Ether (Frutarom) was dried over
sodium metal overnight before use. Methanol (99%) was dried
with magnesium metal. All other solvents and reagents were
commercially available and were used without further puri-
fication.
For this reason, the relative reaction rates of 1 and
tetrachloroethylene were compared in Tables 7 and 8.
Despite the relatively large error, the obtained intermo-
lecular EEs of 9.1 ( 1.7 with MeO- and 11.2 ( 2.7 with
PhCH2S- are significantly higher than any of the inter-
molecular kBr/kCl EEs obtained previously in vinylic
substitution.10 Hence, according to current thinking,
these values corroborate the deduction that a C-X bond
cleavage is part of the rate-determining step in the
substitution. Whereas this is an unusual conclusion, we
note that if a single step substitution will be found at
all, it is likely to be found in systems such as 1, where
the substituents are not capable of extensive stabilization
of the negative charge formed in the transition state.
Additional studies, preferably with structurally related
systems, using the EEs and other probes will be valuable
for strengthening our conclusion.
Syn th esis of th e P olyh a loeth ylen es. 1,2-Dibr om o-1,2-
d ich lor oeth ylen e (1). (1) Compound 1 was prepared from
trans-1,2-dichloroethylene by a sequence of two consecutive
brominations with Br2/dehydrobrominations (with liquid am-
monia) according to the method of Mkryan.34 This method was
applied after finding that van der Walle’s method,35 which
involves two consecutive brominations (with Br2)/dehydrobro-
minations (first with PhNH2, then with KOH/EtOH), failed
due to difficulties both in removing the excess aniline and in
the dehydrobromination with KOH. These difficulties disap-
peared when Et3N was used instead of aniline: To 1,1,2-
tribromo-1,2-dichloroethane (3 g) prepared by the above
sequence from dichloroethylene was added triethylamine (1.2
mL) in acetonitrile (10 mL), and the reaction mixture was kept
for 22 h at room temperature and then washed with water,
dried (MgSO4), and distilled under vacuo. The fraction boiling
at 88 °C/18 mmHg (0.8 g, 35%) was collected and identified
as an 4:3 E/ Z mixture of 1. For details and spectra see the
Results section. The yield was improved when piperidine was
used for the dehydrobromination instead of liquid ammonia,
as described below for preparation of compound 2.
J udging from the assignment made above and the
results of Table 1, the E isomer of 1 is more reactive than
the Z isomer. Analysis of the energies and localization
of MOs of E-1 and Z-1 revealed20 that the two lowest
unoccupied MOs of 1 are nearly similarly localized. The
LUMOs are >90% localized on the carbon atoms and they
are of virtually the same energy (-0.644 eV). The next
vacant orbital is 40% localized on the carbons and 60%
on the halogens. For the Z isomer it is 20% localized on
chlorines and 40% on bromines, while for the E isomer
it is >56% localized on bromines and only 4% on chlo-
rines. Its energy for the E isomer lies 3.4 kcal mol-1
lower than for the Z isomer. This can explain a higher
reactivity of the E isomer.
(2) Compound 1 was also prepared according to the method
of Pielichowski14 from dichloroacetylene which, in turn, was
obtained from trichloroethylene and aqueous NaOH solution
with a catalytic amount of benzyltriethylammonium chloride.
The formed ether complex of dichloroacetylene was brominated
by 1 equiv of Br2 in CCl4.
1-Br om o-1,1,2-tr ich lor oeth ylen e (2). Into trichloroeth-
ylene (100 mL, 1.1 mol) was added bromine (60 mL, 1.15 mol)
dropwise with stirring and irradiation with a sun lamp. The
mixture was washed after 24 h with 10% aqueous Na2S2O3
solution, dried (MgSO4), and distilled, affording 1,2-dibromo-
1,1,2-trichloroethane (257 g, 75%), bp 92-6 °C/18 mmHg. To
a solution of the dibromotrichloroethane (137 g, 0.46 mol) in
dry ether (100 mL) was added piperidine (45 mL, 0.46 mol)
with stirring. After 24 h the precipitated piperidine hydro-
bromide was filtered off and washed with dry ether and the
combined filtrate was distilled at 53-4 °C/18 mmHg, giving
1-bromo-1,2,2-trichloroethylene (81 g, 82%), which was pure
according to the 1H and 13C NMR spectra. Attempts to prepare
the compound according to Mkryan’s method34 using liquid
ammonia gave a mixture of the desired 2 and dibromotrichlo-
roethane which could not be separated efficiently by distilla-
tion.
The trisubstituted alkenes RSC(Cl)dCHCl formed in
the reaction with RS- seem to arise from protonation of
a trisubstituted vinyl anion, which is most likely formed
by a bromophilic reaction on either 1 or its monothio
substitution product. Such a process is well-known for
both simple and functionalized vinyl halides;2,6,7,31-33 as
also shown above. As shown by one of us, the Z-BrCHdC-
(F)Br is more reactive toward RS- and leads mainly to
the vinylic monosubstitution products RSC(F)dCHBr,
while the less reactive E isomer gives predominantly the
product RSC(F)dCH2 of halophilic reduction.33
(31) Chanet-Ray, J .; Vessiere, R. Bull. Soc. Chim. Fr. 1974, 1661.
(32) Verny, M. Bull. Soc. Chim. Fr. 1970, 1946.
(33) Shainyan, B. A.; Bel’skii, V. K. Zh. Org. Khim. 1991, 27, 2362.
(34) Mkryan, G. M. Bull. Armenian Branch Acad. Sci. USSR. 1944,
45 (Chem. Abstr. 1946, 40, 33925).
(35) van der Walle, H. Bull. Soc. Chim. Belg. 1920, 29, 307.

An Example of a Single Step Substitution?
J . Org. Chem., Vol. 62, No. 23, 1997 8057
1-Br om o-1-ch lor o-2-m eth ylp r op en e (3) was prepared
according to Cunico and Han’s procedure.15
based on the measured intensities of the signals without
calibration; the structural assignment was made by MS. In
addition, 1H and 13C NMR spectroscopies were used in ap-
propriate cases, and the single products from tetrachloro- and
trichloroethylenes that can give only the chlorine substitution
products were isolated.
1,2-Dich lor o-1-br om oeth ylen e (4) was obtained by addi-
tion of bromine to (E)-1,2-dichloroethylene followed by dehy-
drobromination of the mixture of diastereomers of 1,2-dichloro-
1,2-dibromoethane (two singlets in a 1:1 ratio at δ 6.00 and
6.05 in the 1H, and at δ 61.30 and 61.49 in the 13C NMR
spectra) with piperidine in ether. The fraction boiling at 45-
50 °C/30 mmHg was collected.
1,2-Dibr om o-1-ch lor oeth ylen e (5) was prepared accord-
ing to a procedure similar to that used for preparation of 1.14
Chloroacetylene was obtained as its ether complex by reaction
of an E/Z mixture of 1,2-dichloroethylene in ether with an
aqueous NaOH solution and a catalytic amount of benzyltri-
ethylammonium chloride. The ether complex which was
evolved as a gas from the solution was passed into the ethereal
solution of 1 equiv of bromine. Fractional distillation gave a
75% yield (fraction boiling at 50-55 °C/30 mmHg was col-
lected) of 5 as a 2:1 E/Z mixture.
Rea ction s w ith MeONa . The reactions of chlorobromo-
ethylenes 1, 2, 4, and 5 with an equimolar amount of dry
MeONa were conducted in MeCN at 80 °C and in DMF at 100
°C in the course of 30-60 min. Sodium methoxide was
prepared by dissolution of sodium metal in dry methanol,
evaporation of the solvent, and drying under high vacuo. Since
the conversion to products under these conditions was low,
individual products were not isolated. Instead, the reaction
mixtures in MeCN or DMF were poured into water and
extracted with chloroform, and the organic phase was analyzed
by GC, GC-MS, and 1H and 13C NMR spectroscopy.
Rea ction s w ith RSNa . Bromochloroethylenes 1, 2, 4, and
5 in ca. 8-fold molar excess were reacted with sodium thiolates
RSNa (R ) PhCH2, Me, t-Bu) in MeCN at room temperature
under nitrogen for 24 h in order to ensure nearly exclusive
formation of the monosubstitution products. Unfortunately,
separation of the latter by chromatography (silica columns,
preparative OV-101 column, analytical GC or HPLC PR and
cyanide columns) failed. Therefore, the analysis was mainly
conducted by GC-MS and the relative intensities given are
Rea ction s w ith 3. The reaction of an 8-fold excess of 3
with MeS-Na+ in acetonitrile was conducted under nitrogen
for 15-144 h at room temperature or for 18 h at 80 °C. No
reaction took place according to NMR or GC-MS analysis.
When the reaction between 3 and PhCH2S-Na+ was conducted
for 23 h at 80 °C, no substitution product was identified.
Dibenzyl disulfide (6.3%) and an unidentified product (5.4%)
whose higher m/z value is at 253 and which contains a benzyl
moiety and does not contain Br and Cl was formed. Appar-
ently, 3 is inert to substitution under the reaction conditions.
Mea su r em en t of Exter n a l EE. Compound 1 and tetra-
chloroethylene (5 mmol each) were dissolved in 1.5-2 mL of
MeCN, and 10 mmol of dry MeONa or PhCH2SNa was added
under an argon atmosphere. In the case of PhCH2SNa, the
reaction was self-heated. GC measurements were started
immediately (1-2 min after mixing), and the reactions were
followed until there was virtually no more changes in the ratio
of the substrates. The values of the external EE were
calculated from eq 6.
Ack n ow led gm en t. This work was supported in
Russia by the Russian Foundation for Basic Research
(Grant No. 96-03-33468a) and in Israel by the United
States-Israel Binational Science Foundation, J erusa-
lem (BSF). B.A.S. and Y.S.D. are thankful to Dr. L. V.
Klyba (IrIOCh) for the GC-MS analysis, and to Dr. S.
V. Zinchenko (Irkutsk State University) for the NMR
measurements. We are grateful to Prof. Y. Marcus for
discussion of solvation of halide ions and to Dr. D.
Danovitch for calculations.
J O9709357