Triflate/mesylate ratios and competing C-O and S-O bond cleavages in nucleophilic vinylic substitution

Article abstract of DOI:10.1021/jo961108t

In an attempt to develop the k_OTf/k_OMs ratio as a mechanistic tool for the "addition-elimination" route in nucleophilic vinylic substitution, several pairs of vinyl mesylates and triflates were prepared. Whereas reactions of ArC(LG)=C(CO₂Et)₂ (LG = OTf, OMs) with piperidine and morpholine in MeCN or THF gave the normal substitution product with k_OTf/k_OMs ratios of 4.3-10.6, the reaction of the mesylates, Ar = p-O₂NC₆H₄, and of PhC(OMs)=C(Me)CN with MeS^- gave a ketone via an S-O bond cleavage. A related mesityl-substituted tosylate also reacted with p-MeC₆H₄X^- (X = O, S) via an S-O bond cleavage. Hence, k_OTf/k_OMs ratios cannot be used as a general mechanistic tool. Several reactivity ratios in vinylic substitution are briefly discussed.

Full text of DOI:10.1021/jo961108t

8536
J . Org. Chem. 1996, 61, 8536-8543
Tr ifla te/Mesyla te Ra tios a n d Com p etin g C-O a n d S-O Bon d
Clea va ges in Nu cleop h ilic Vin ylic Su bstitu tion
Ettie Z. Schottland and Zvi Rappoport*
Department of Organic Chemistry, The Hebrew University, J erusalem 91904, Israel
Received J une 12, 1996X
In an attempt to develop the kOTf/kOMs ratio as a mechanistic tool for the “addition-elimination”
route in nucleophilic vinylic substitution, several pairs of vinyl mesylates and triflates were prepared.
Whereas reactions of ArC(LG)dC(CO2Et)2 (LG ) OTf, OMs) with piperidine and morpholine in
MeCN or THF gave the normal substitution product with kOTf/kOMs ratios of 4.3-10.6, the reaction
of the mesylates, Ar ) p-O2NC6H4, and of PhC(OMs)dC(Me)CN with MeS- gave a ketone via an
S-O bond cleavage. A related mesityl-substituted tosylate also reacted with p-MeC6H4X- (X ) O,
S) via an S-O bond cleavage. Hence, kOTf/kOMs ratios cannot be used as a general mechanistic
tool. Several reactivity ratios in vinylic substitution are briefly discussed.
In tr od u ction
The leaving group (LG) effect is a valuable tool in the
study of nucleophilic vinylic substitution.1 The only
extensively studied tool is the kBr/kCl ratio, which com-
pares the rates of a pair of otherwise similar vinyl
bromide and vinyl chloride. In reaction routes involving
rate-determining C-halogen bond cleavage such as
elimination-addition, SN1, or halophilic reaction, the kBr/
The relative reaction rates of a vinyl trifluoromethane-
sulfonate (triflate) ROTf, compared with the correspond-
ing vinyl methanesulfonate (mesylate) ROMs, with a
nucleophile seem appropriate for this purpose. Whereas
kBr/kCl ratios for aliphatic SN2 or aliphatic or vinylic SN1
k
Cl ratio should be much greater than unity. In contrast,
reactions are 10-102 (up to 103 in a few cases),4 the kOTf
/
in the two-step addition-elimination route with rate-
determining nucleophilic attack (eq 1, k1 is the slow step)
the similar electronic effects of the two halogens lead to
kBr/kCl ∼ 1 or slightly > 1, and this is one of the strongest
arguments for the nonconcerted nature of this route.1
kOMs ratios are much higher, 104-(5 × 105).5 At the
reaction site the bulk of the two groups is very similar,
closer than those of Cl and Br, thus excluding the steric
effect on the ratios. The two anions are structurally very
similar, except for the higher stability of TfO-, arising
from the stronger electron withdrawal and therefore
negative charge delocalization of the CF3 group. Al-
though the effect of mesylate and triflate on the stereo-
chemistry of vinylic substitution was compared in a single
system,6 there are no kinetic data on the ratios in vinylic
substitution of the addition-elimination route.
For a proper analysis of the use of the kOTf/kOMs tool
the ratio should be first evaluated for a case of a bona
fide two-step process (eq 1). Due to the higher electrone-
gativity of triflate, a kOTf/kOMs ratio higher than the kBr/
kCl ratio in the same system is expected. Second, it
should be investigated if the ratio depends (and to what
extent) on the reactivity of the system, as determined by
the activation by its electron-attracting groups Y and Y′.
Note that no discernible dependency on Y,Y′ is known
for the kBr/kCl ratio in eq 1.1 Finally, the substitution of
a system with very low activation by Y and Y′, where
both a perpendicular and an in-plane single step route
may have a chance to occur,3 should be investigated.
Since the electronegativity of fluorine is higher than
that of chlorine, whereas the C-Cl bond cleavage is more
facile than the C-F bond cleavage, for rate-determining
nucleophilic addition kF/kCl . 1 and for rate-determining
C-LG bond cleavage kF/kCl , 1. This sharper tool than
the kBr/kCl ratio was used to corroborate the multistep
route.1 However, due to the different bulk of F and Cl,
the ratio may contain a steric contribution. Moreover,
since fluorine is not a very good nucleofuge, its expulsion
can compete with the protonation of the intermediate
carbanion,2 thus reducing the utility of the kF/kCl probe.
We are interested in the possibility of a single step
“addition-elimination” type vinylic substitution. This
can involve perpendicular nucleophilic attack, resulting
in retention of configuration (eq 2),1b or an in-plane
nucleophilic attack that will lead to inversion of config-
uration.3 We expect a kBr/kCl > 1, but if the transition
state is early, the ratio is not expected to be very high.
We therefore look for a mechanistic tool that will ac-
centuate the differences between the two routes.
(3) (a) Ochiai, M.; Oshima, K.; Masaki, Y. J . Am. Chem. Soc. 1991,
113, 7059. (b) Shainyan, B.; Rappoport, Z. J . Org. Chem. 1993, 58,
3421. (c) Glukhovtsev, M. N.; Pross, A.; Radom, L.; J . Am. Chem. Soc.
1994, 116, 5961. (d) Lucchini, V.; Modena, G.; Pasquato, L. Ibid. 1995,
117, 2297.
(4) (a) Stang, P. J .; Rappoport, Z.; Hanack, M.; Subramanian, L. R.
Vinyl Cations; Academic Press: New York, 1979. (b) For values in
several reactions see: Avramovitch, B.; Weyerstahl, P.; Rappoport, Z.
J . Am. Chem. Soc. 1987, 109, 6687.
(5) For a recent reference see: Bentley, T. W. In The Chemistry of
Sulphonic Acids, Esters and their Derivatives; Patai, S., Rappoport,
Z., Eds.; Wiley: Chichester, 1991; Chapter 16, pp 671-696.
(6) Avramovitch, B.; Rappoport, Z. J . Am. Chem. Soc. 1988, 110,
911.
X Abstract published in Advance ACS Abstracts, November 1, 1996.
(1) (a) Rappoport, Z. Adv. Phys. Org. Chem, 1969, 7, 1. (b) Rappoport,
Z. Acc. Chem. Res. 1981, 14, 7.
(2) Marchese, G.; Naso, F. Chim. Ind. (Milano) 1971, 53, 760.
S0022-3263(96)01108-5 CCC: $12.00 © 1996 American Chemical Society

Nucleophilic Vinylic Substitution
J . Org. Chem., Vol. 61, No. 24, 1996 8537
Ta ble 1. Selected Bon d Len gth s a n d An gles for 4a
In the present paper we measured a few kOTf/kOMs ratios
for substitution proceeding via the multistep addition-
elimination route, but we found that nonaddition-
elimination routes can also take place, invalidating the
mechanistic generality of the kOTf/kOMs tool.
distance, Å
deg
bond
A
B
angle
A
B
C(1)-C(2) 1.359(6) 1.329(5) C(1)C(2)C(3)
C(1)-O(1) 1.412(5) 1.410(4) C(1)C(2)C(4)
C(1)-C(6) 1.478(6) 1.459(5) C(2)C(1)C(6)
C(2)-C(3) 1.353(6) 1.446(7) O(1)C(1)C(2)
C(2)-C(4) 1.501(6) 1.502(6) C(3)C(2)C(4)
O(1)-S(1) 1.608(3) 1.618(3) O(1)C(1)C(6)
O(2)-S(1) 1.415(3) 1.411(3) N(1)C(3)C(2)
C-C (ring) 1.358(6)- 1.372(7)- O(1)C(1)C(2)C(4)
1.390(6)a 1.398(5)
119.5(4) 118.3(4)
125.7(4) 126.8(4)
127.9(4) 128.4(4)
117.2(4) 116.1(4)
114.7(4) 114.9(4)
114.5(4) 115.1(3)
178.3(5) 176.6(5)
4.4(6) -3.1(6)
Resu lts
Syn th esis. Compounds reacting by the multistep
addition-elimination route should have at least one and
preferably two electron-withdrawing groups. Hence,
compounds 1-3 having no stereochemistry and the
isomeric pairs 4 and 5, carrying triflate and mesylate
leaving groups in each case, were prepared.
C(3)C(2)C(1)C(6) -2.9 (7)
2.1 (6)
a
Shorter bonds C(9)-C(10) 1.358(6) Å; C(8)-C(9) 1.364(7) Å;
C(9′)-C(10′) 1.375(7) Å; C(8)-C(5) 1.372(7)Å.
4a established the stereochemistry in this case and by
analogy for 5a . Selected crystallographic data are given
in Table 1.7
Su bstitu tion . Compounds 1a , 1b, 2a and 2b reacted
with piperidine and morpholine in both THF and MeCN
to give the substitution products 9a and 9b (eq 6). The
reactions of 1a were studied earlier.8 The reactions were
The vinyl sulfonates were prepared by reaction of the
enol/pyridine or triethylamine, i.e., via the corresponding
enolate salt with the appropriate sulfonic anhydride (eq
3). The precursors, written as the enol 6 in eq 1, were
1
ca. 50:50 and 85:15 ketone:enol (according to H NMR)
when Ar ) p-O2NC6H4 and Ph, respectively.
followed spectrophotometrically at the λmax of the prod-
ucts 9 at 303 and 313 K. The relative concentrations of
the amines to the substrates were sufficiently high to
ensure pseudo-first-order reactions. Five different nu-
cleophile concentrations in the two solvents were used
in each system. Clean second-order reactions, first each
in the amine and the substrate, were obtained with no
evidence for amine catalysis. Isosbestic points were
retained during the measurements, and the infinity
readings were stable for several half-lives. The second-
order rate constants and the approximate activation
parameters (due to the narrow temperature range, we
estimate ∆Hq as (1.5 kcal mol-1 and ∆Sq as (4.5 cal K-1
mol-1) are given in Table 2. The calculated kOTf/kOMs
ratios are given in Table 3, and kpiperidine/kmorpholine, kMeCN
/
k
THF, and kp-O NC H /kPh ratios are given in Table 4. The
2
6
4
agreement with previous data8 is very good.
When the reactions were conducted in EtOH the
behavior of the triflates 1b and 2b with piperidine at
30 °C was identical to that described above, giving
appreciably lower rate constants than in MeCN. How-
ever, the reactions of 1a and 2a did not show a steady
isosbestic point during the reaction, and the absorp-
tion vs time plot did not reach a plateau even after
long reaction times. In a quantitative determination of
the products of piperidine in EtOH, 2b gave completely
the substitution product 9a , whereas 2a gave 95% of
9a together with 5% of the enol/ketone 6 (Ar ) Ph)
(according to the 1H NMR data). Reaction of 2a with
ethanol in the absence of the amine had shown that a
spectral change takes places with time, and λmax shifts
to a longer wavelength with increase in the absorption
3a was prepared from the enol 7 (eq 4). Attempted
preparation of 3b in an analogous way gave a yellow oil
with an appropriate H NMR spectrum. However, the
1
compound was unstable and turned black rapidly, so that
further reaction with this system was not continued. The
reaction of R-benzoylpropionitrile 8 with the two sulfonic
anhydrides in the presence of base gave 4:5 mixtures rich
in the E isomer (85%:15%) (eq 5). X-ray diffraction of
(7) The author has deposited atomic coordinates for this structure
with the Cambridge Crystallographic Data Centre. The coordinates
can be obtained, on request, from the Director, Cambridge Crystal-
lographic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK.
(8) Rappoport, Z.; Topol, A. J . Chem. Soc., Perkin Trans. 2 1975,
863.

8538 J . Org. Chem., Vol. 61, No. 24, 1996
Schottland and Rappoport
Ta ble 2. Secon d Or d er Ra te Con sta n ts (in M-1 s-1) for th e Rea ction of System s 1 a n d 2 w ith Am in es
102k2 at
∆Hq
∆Sq
substratea
solvent
MeCN
nucleophile
[Nu]
303 K
313 K
(kcal mol-1
)
(cal mol-1 K-1
)
1a
piperidine
morpholine
piperidine
morpholine
piperidine
morpholine
piperidine
morpholine
piperidine
morpholine
piperidine
morpholine
piperidine
morpholine
piperidine
morpholine
piperidine
piperidine
0.0014-0.035
0.016-0.27
0.018-0.1
0.065-0.22
0.00057-0.00528
0.00311-0.038
0.0029-0.018
0.065-0.2
19.7
1.25
4.73
0.60
149
35.1
2.00
7.11
0.76
234
13.6
44.3
4.19
6.62
0.75
2.52
0.46
60.3
3.87
14.8
2.0
10.3
8.2
7.1
3.9
7.9
8.4
2.0
2.0
4.3
5.3
3.7
3.6
1.4
6.9
2.6
6.5
-28
-40
-41
-47
-32
-36
-54
-58
-50
-51
-54
-58
-55
-43
-54
-46
THF
1b
2a
2b
MeCN
THF
8.43
38.6
3.64
5.11
0.55
2.01
0.37
54.2
2.60
12.5
1.37
9.31
2.99
MeCN
THF
0.0025-0.071
0.0039-0.32
0.018-0.2
0.2-1.0
MeCN
THF
0.00057-0.0049
0.0075-0.081
0.0029-0.037
0.044-1.0
1b
2b
EtOH
EtOH
0.00266-0.0133
0.00426-0.0213
a
[Substrate] ) (4.7-6.3) × 10-5 M-1
.
Ta ble 3. Mesyla te/Tr ifla te Ra tios in th e Rea ction of 1
a n d 2 w ith Am in es
in MeOH gives ketone 8 (eq 11).
THF
MeCN
piperidine morpholine piperidine morpholine
303 313 303
313
K
303 313 303
313
K
kOTf/kOMs
K
K
K
K
K
K
k(1b)/k(1a ) 8.2
k(2b)/k(2a ) 6.2
6.2
5.9
6.1
3.7
5.5
4.3
7.6 6.7
10.6 9.1
6.5
4.8
6.8
5.2
without an isosbestic point. In contrast, neither the
spectrum of 2b nor that of 4a change in EtOH after a
long time.
The reaction of 5a with piperidine in MeCN is slow at
moderate [amine]/[5a ] ratios. When ratios > 100 were
used the reaction is faster but the amine absorption
overlaps that of 5a , so that spectrophotometic kinetic
study of the reaction becomes impossible.
Rea ction s w ith Oth er Nu cleop h iles. The reactions
of 1 and 2 with MeS-Na+ are too fast to follow by
conventional spectrophotometry.9 The reaction of triflate
4b with MeS-Na+ led to replacement of the leaving
group. However, product analyses show a different
reaction course for the triflates and the mesylates.
Whereas the triflates give a normal nucleophilic substi-
tution product (eq 7), the vinyl mesylates 4a and 5a
undergo a cleavage reaction to their precursor enol (eq
8).
In order to see if a vinylic tosylate will behave similarly
to a vinylic mesylate (as expected) or like a triflate, the
R-mesityl [R-(tosyloxy)vinylidene]malonate (13) was pre-
pared from the keto diester 12 (eq 12) and substituted
by p-toluenethiolate and by p-cresolate anions (14). No
substitution product was formed, and the only product
obtained was the ketone 12, presumably formed by S-O
bond cleavage (eq 13).
Solid Sta te Str u ctu r e of 4a . Two independent
molecules of 4a , A and B, are present in the unit cell.
The ORTEP drawing of A is given in Figure 1. A and B
differ appreciably (by 0.03 Å) in the double bond C(1)-
C(2) length and by 0.113 Å in the dCsCN bond length
in a complementary way. All of the other differences are
minor. Angles between geminal double bond substitu-
ents are ca. 115°, whereas bond angles CdCsC(4) and
CdCsC(6) are 126-128°. The double bond torsional
angle is 7.7° (A) and 6.4° (B), and the torsional angles
between the Ph and Câ are 41.8° (A) or 43.0 (B).
The reaction of 1b with sodium p-toluenethiolate gives
the substitution product and 4b gives the retained
substitution product 10 with MeS- (eq 9). In contrast,
reactions of 1a and 4a with MeSNa in EtOH give the
ketones 11 and 8 (eq 10). Reaction of 4a with NaOMe
Discu ssion
(9) The reaction of p-MeC6H4S- with 1a in EtOH was somewhat
irreproducible,8 and this may be due to a competition between attack
on dC and on S, which is sensitive to minor changes in the reaction
conditions.
Com p etition betw een C-O a n d S-O Bon d Clea v-
a ges. The desired application of kOTf/kOMs ratios as a

Nucleophilic Vinylic Substitution
J . Org. Chem., Vol. 61, No. 24, 1996 8539
Ta ble 4. P ip er id in e/Mor p h olin e, MeCN/THF , a n d p-O2NC6H4/P h Ra te Ra tios in th e Rea ction s of 1 a n d 2
kpip/kmorp kMeCN/kTHF
k
/kPh
p-O2NC6H4
compd
solvent
303 K
313 K
amine
303 K
313K
amine/solvent
nucleofuge
303 K
313 K
1a
MeCN
THF
MeCN
THF
MeCN
THF
MeCN
THF
15.6
7.9
17.7
10.6
9.3
5.4
20.8
9.1
17.5
9.4
17.2
10.6
8.8
5.5
15.8
7.4
piperidine
morpholine
piperidine
morpholine
piperidine
morpholine
piperidine
morpholine
4.2
2.1
3.9
2.3
2.6
1.5
4.3
1.9
4.9
2.6
5.3
3.3
2.6
1.5
4.1
1.9
piperidine/MeCN
piperidine/THF
Morpholine/MeCN
morpholine/THF
piperidine/MeCN
piperidine/THF
morpholine/MeCN
morpholine/THF
OMs
OMs
OMs
OMs
OTf
OTf
OTf
OTf
2.3
1.6
2.3
1.6
2.7
3.1
3.2
2.7
2.7
1.7
2.7
1.7
3.9
3.0
3.6
2.1
1b
2a
2b
ing triflate increases its sensitivity to reaction with
nucleophiles, such as adventitious water in the solvent
or in a humid atmosphere. Hence, the kOTf/kOMs probe
will be difficult to use with highly activated systems.
This is a limitation only if there is a trend in the kOTf
/
kOMs ratios as a function of the electrophilicity of the
(sulfonyloxy)alkene. If this is not the case, as found for
the kBr/kCl ratios that are relatively insensitive to the
electrophilicity of the haloalkene in the addition-elimina-
tion regime,1 the ratios found for 1 and 2 can be regarded
as “normal” values for the addition-elimination route.
The second obstacle is more severe. Comparison with
systems of differing electrophilicity should be with the
same nucleophile. Whereas the reactions of 1 and 2 with
piperidine and morpholine in the aprotic solvents proceed
by the desired addition-elimination route, the observed
rates are not very high and the rates for less activated
systems, e.g., only by a singly activating group, such as
4 and 5 will be low. Compensation by higher nucleophile
concentrations makes it unpractical to follow the reac-
tions spectrophometrically, due to the end absorption of
the nucleophile at the λmax of the (sulfonyloxy)alkene.
Moreover, the intermediate in the reaction with amino
nucleophiles is a zwitterion rather than a carbanion, and
the mechanistic behavior may be different. Since the
transition to a single-step route should be at systems with
a much lower reactivity, the amino nucleophiles are
expected not to be sufficiently reactive with the systems
of low reactivity.
F igu r e 1. ORTEP drawing and numbering for 4a .
probe to distinguish between rate-determining nucleo-
philic attack on the vinylic carbon and a rate-determining
C-LG bond cleavage in nucleophilic vinylic substitution
requires the study of a variety of substrates ranging from
the most reactive to the least reactive ones. The present
work had shown two major obstacles to such a study; i.e.,
(i) the “instability” of highly reactive triflates and (ii) the
competition or predominance of nucleophilic attack on
the sulfonate group rather than on the vinylic carbon.
Whereas vinylic triflates of (mostly) nucleophilic olefins
are usually stable species that are used extensively in
synthesis,10 especially in organometallic coupling reac-
tions, the situation apparently differs with triflate-substi-
tuted highly electrophilic olefins. Not many of those are
known, and when they are not highly activated like 1b,
2b, 4b, or 5b they seem sufficiently stable. However,
replacing an ester by a cyano group, which moderately
increases the double bond electrophilicity, as judged by
previous studies,8,11 apparently decreases the stability of
the vinylic triflate. We have previously found that p-O2-
NC6H4C(OTf)dC(CN)CO2Me is very reactive, although
it was sufficiently long-lived for a stereochemical study,6
and we observed now that the dicyano-substituted triflate
3b blackens, i.e., deteriorates a short time after its
preparation. The reason for this was not investigated,
but by analogy with the low stability of highly reactive
aliphatic tosylates in solvolysis12 we assume that the
higher reactivity due to the strongly electron-withdraw-
Since the lower reactivity range is the one of the
highest interest, anionic, and especially the highly reac-
tive nonconjugated thio nucleophiles, like MeS-, are the
nucleophiles of choice. At the higher reactivity range,
their rate could be followed by a fast reaction technique.
The reactions of the vinyl mesylates 1a and 4a with
MeS- and of 4a and 5a with MeO- ions take a different
reaction course than those of the triflates 1b and 4b with
these nucleophiles or of both the vinyl mesylates and
triflates with the cyclic amines. The product obtained,
either completely with 1a or 4a with RS- or partially in
addition to the substitution product in the reaction of 2a
with piperidine in EtOH, are the enol (ketone) precursors
of the mesylates. Their formation is ascribed to nucleo-
philic attack of the nucleophile at the mesylate sulfur
by an O-S bond cleavage, displacing the enolate ion as
a nucleofuge (eq 14). The latter is protonated in the
protic medium or on workup.
(10) (a) Stang, P. J .; Hanack, M.; Subramanian, L. R. Synthesis
1982, 85. (b) Scott, W. J .; McMurry, J . E. Acc. Chem. Res. 1988, 21,
47.
(11) Patai, S.; Rappoport, Z. J . Chem. Soc. 1962, 392.
(12) H. M. R. Hoffmann, J . Chem. Soc. 1965, 6748.

8540 J . Org. Chem., Vol. 61, No. 24, 1996
Schottland and Rappoport
This reaction is not an isolated case. The structurally
closely related tosylate 13 also reacts with the softer
p-toluenethiolate and p-cresolate nucleophiles to give a
similar S-O bond fission to the enolate. This cleavage
competes, although to a lower extent, even in the reaction
with piperidine in EtOH, where the C-O:S-O cleavage
ratio, judged by the products, is 95:5.
to the intervention of this route for the triflates is
unlikely with our moderately activated system based on
what is known for nucleophilic vinylic substitutions.1,3,6
Electron-withdrawing CF3 and Y,Y′ groups affect the
electrophilicity of both the sulfonyloxy-substituted carbon
and the sulfonyl sulfur electrophilic reaction centers. The
closer the substituent is to one of these centers the higher
is its effect on its electrophilicity. Hence, it is reasonable
that as the electron-withdrawing effect of Y and Y′
decrease, the relative C-/S-electrophilicity also decreases
and the effect of the constant and strongly electron-
withdrawing CF3 group on the sulfur electrophilicity will
relatively increase with a consequently higher impor-
tance of the S-O bond cleavage. This prediction is in
contrast with our observation that the mesylates rather
than the triflates undergo the preferential S-O bond
cleavage. Moreover, this effect cannot be dominated by
hard-hard, soft-soft interactions since it takes place
with the mesylates and tosylate 13 with both the harder
oxygen and the softer thiolate anions. The solvent also
affects the site of attack since slight S-O bond cleavage
takes place in EtOH compared with MeCN or THF. We
conclude that with the present limited and unsystematic
data we cannot predict the dC- vs S-regioselectivity of
the nucleophilic attack and why a certain combination
of electrophilic sulfonyloxyalkene/nucleophile/solvent gives
an AdN-E substitution, while other combinations lead
to reaction at sulfur.
Mesyla te/Tr ifla te Rea ctivity Ra tios. There is evi-
dence, e.g., from stereochemistry,5,18 that compounds such
as 1 and 2 are likely to react via the stepwise nucleophilic
addition-elimination route. Additional evidence are the
activation parameters. Whereas the temperature span
of 10 °C is insufficient to give accurate activation
parameters it is clear that the ∆Hq values are very low
and the ∆Sq values are highly negative. This is typical
for many similar multistep reactions with amines that
proceed via a polar zwitterionic intermediate.8,19 We
therefore ascribe the kOTf/kOMs reactivity ratios of Table
3 to a different effect of the two electron-withdrawing
sulfonyloxy groups on the electrophilicity of CR.
The clean second-order kinetics is usually ascribed to
a rate-determining nucleophilic attack k1. Indeed, no
amine catalysis was observed, although this strong
criterion for a multistep route was previously observed
for p-Me2NC6H4C(LG)dC(CN)2 (LG ) Cl, F, OR) with
amines.19,20 However, amine catalysis was never found
for dicarboalkoxy-activated vinyl chlorides and bromides
and is therefore not expected for the better sulfonyloxy
nucleofuges.
In a previous analysis of the kOMs/kCl ratios for 1a and
its corresponding chloride with piperidine and morpho-
line in THF and MeCN, it was suggested that the
relatively high ratios should be interpreted in terms of a
scheme similar to eq 1, where the rate of both an amine-
catalyzed route (k3 [amine]) and the expulsion rate of the
nucleofuge (k2) are smaller than k1. Under these condi-
tions (k-1 . k2 + k3[amine]) the observed rate constant
is given by kobs ) k1k2/k-1. This conclusion was derived
from the observation that in the substitution of system
1 with ArS- kOMs/kCl ) 1.5 ( 0.18 and since in the reaction
Consequently, we conclude that great care should be
exercised in trying to use kOTf/kOMs ratios as a mechanistic
probe. In addition to the kinetics of the disappearance
of the precursor alkene, the products should be carefully
and completely identified as the substitution products.
S-O bond cleavage in nucleophilic reactions of vinyl
sulfonates were previously observed for nucleophilic vinyl
sulfonates.13 Whereas most cycloalkenyl triflates solvo-
lyze via SN1,4b cyclopentenyl triflate gives [16O]cyclopen-
tanone in 50% ethanol-H218O, excluding an SN1 route
and suggesting reaction via an S-O bond cleavage.13b
Other cyclic triflates with geometrical constraints
react similarly.13c The k(cyclohexenyl-OTf)/k((E)-MeC-
(OTf)dCHMe) ratios in 50% dioxane are 106 and ca. 4
with and without added OH-, respectively. The high
sensitivity in the absence of OH- was ascribed to an SN1
reaction, whereas the low value with OH- was ascribed
to reaction on sulfur, which is relatively insensitive to
the effect of remote substituents.14
An interesting use of a sulfonate/sulfonate leaving
group effect was to distinguish between rate-determining
electrophilic attack on the CdC bond (AdE-E substitution
route) and C-X bond cleavage (vinylic SN1 route). p-
Bromobenzenesulfonate (OBs) is more electron-withdraw-
ing than p-methylbenzenesulfonate (tosylate, OTs), but
-OBs is a better nucleofuge than -OTs, a situation
resembling the OTf/OMs pair. Consequently, a kOBs/kOTs
>1 is expected for the SN1 route and kOBs/kOTs < 1 for the
AdE-E route. Indeed, for the reaction of (E)-MeCHdC-
(Me)LG, kOBs/kOTs ) 3.8 in 50% MeOH (SN1) and 0.28 in
HCOOH/HCOO- (AdE-E).15a For cyclohexenyl-LG (kOBs
/
kOTs ) 0.30 (AdE-E) whereas cyclohexenyl-OTf in AcOH/
AcO- reacts via SN1.15b
All of these examples are for competition of reactions
of nucleophilic olefins. We know of only two cases where
attack on a multiatom ester nucleofuge predominated
completely over the competing reaction at the vinylic
carbon.16 The nucleophile was OH-/H2O. In the analo-
gous reaction with p-MeC6H4S- the products indicated a
ca. 1:1 competition between reaction on the vinylic carbon
and the carbonyl.16
Consequently, the reactions demonstrated by eqs 8, 10,
11, and 13 add one more variant to the multitude of
reactions routes leading to or competing with vinylic
substitution.17
Whereas triflates should be more prone than mesylates
to react via a concerted route (eq 2) the possibility that
the different behavior of triflates and mesylates is due
(13) (a) Frydman, N.; Bixon, R.; Sprecher, M.; Mazur, Y. Chem.
Commun. 1969, 1044. (b) Eymann, E.; Hanack, M. Tetrahedron Lett.
1976, 3507. Hanack, M.; Benz, H.; Markl, R.; Subramanian, L. R.
Liebigs Ann. Chem. 1978, 1894. (c) Subramanian, L. R.; Hanack, M.;
Chang, L. W. K.; Imhoff, M. A.; Schleyer, P.v.R.; Effenberger, F.; Kurtz,
W.; Stang, P. J .; Dueber, T. E. J . Org. Chem. 1976, 41, 4099.
(14) Summerville, R. H.; Schleyer, P. v. R. J . Am. Chem. Soc. 1972,
96, 1110.
(15) (a) Peterson, P. E.; Indelicato, J . M. J . Am. Chem. Soc. 1969,
91, 6194. (b) Pfeifer, W. D.; Bahn, C. A.; Schleyer, P. v. R.; Bocher, S.;
Harding, C.E.; Hummel, K.; Hanack, M.; Stang, P. J . J . Am. Chem.
Soc. 1971, 93, 1513.
(16) Strauss, M. J .; Rappoport, Z. J . Org. Chem. 1982, 47, 1009.
(17) Rappoport, Z. Recl. Trav. Chim. Pays-Bas 1985, 104, 309.
(18) Rappoport, Z.; Gazit, A. J . Am. Chem. Soc. 1987, 109, 6698.
(19) E.g. (a) Rappoport, Z.; Greenzaid, P.; Horowitz, A. J . Chem. Soc.
1964, 1334. (b) Rappoport, Z.; Peled, P. J . Chem. Soc., Perkin Trans.
2 1973, 617.
(20) E.g., see: Rappoport, Z.; Ta-Shma, R. J . Chem. Soc. B 1971,
871, 1461.

Nucleophilic Vinylic Substitution
J . Org. Chem., Vol. 61, No. 24, 1996 8541
the case with substitution of mildly activated systems.1,17
This is consistent with the lower field Me group of both
4b and 10 than those of 5b and the geometric isomer of
10. The formation of 16% of the inverted product in the
mainly retained substitution of 5b is ascribed to post-
isomerization since the sample was analyzed after a long
reaction time.
Ta ble 5. Rela tive Activa tion of Nu cleofu ga l
Su bstitu en ts LG in th e Rea ction of
p-O2NC6H4C(LG)dC(CO2Et)2 w ith Am in o Nu cleop h iles
In MeCN
morpholine piperidine
In THF
morpholine piperidine
303
K
313
K
303
K
313
K
303
K
313 303 313
substituent
K
K
K
Cl
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
OTs
OMs
OBs
OTf
17.0
19.1
27.7
14.4
17.4
25.1
15.5
17.6
26.9
12.8 12.9 10.8 11.8 10.6
15.4 12.3 11.5 10.9 10.4
24.2 28.9 24.9 25.7 24.1
Exp er im en ta l Section
Meth od s a n d P r ecu r sor s. For the kinetic studies THF
(spectroscopic grade, Merck) was distilled from benzophenone/
Na before use. Spectrocopic grade EtOH (Merck) and MeCN
(Frutarom) were used without further purification. CCl4 and
CH2Cl2 were dried over CaCl2 before use, and piperidine and
morpholine were distilled from NaOH before use. Diethyl
2-[(methylsulfonyl)oxy]-2-(p-nitrophenyl)ethylene-1,1-dicar-
boxylate (1a ) and diethyl (p-nitrobenzoyl)malonate were pre-
pared according to Rappoport and Topol.8 Diethyl benzoyl-
malonate was prepared according to Lund.25 The
triethylammonium salt of benzoylmalononitrile (10) was pre-
pared according to Fleury and Libis,26a and the enolate salt
was converted to the enol according to Dornow.26b R-Benzoyl-
propionitrile was prepared according to Hassner.27 The sub-
stitution products diethyl 2-morpholino- (and 2-piperidino)-2-
(p-nitrophenyl)ethylene-1,1-dicarboxylate are known from a
previous work.8
124.1 118.3 133.8 103.2 75.0 63.3 89.4 64.5
of p-cyanoaniline with LGCHdC(CN)2 (LG ) Cl, OMs)
k
OMs/kCl ) 1 in MeCN.21
Are the new results compatible with this interpreta-
tion? The σ constants of OMs and OTf are σm 0.39 and
0.56, σp 0.33, 0.47, σI 0.61 and 0.84, and σR -0.28 and
-0.36, respectively.5 If one uses a two-point Taft polar
substituent equation log(kOTf/kOMs) ) FI (σI(OTf) - σI(OMs
)
and the average of all kOTf/kOMs ratios from Table 3, i.e.,
6.5 ( 1.3, then FI ) 3.5, i.e., an appreciable sensitivity to
the polar effect of the substituent. The kBr/kCl ratios of
ca. 11 in related systems are consistent with σI(Br) ) σI(Cl)
) 0.47.22 The kOMs/kCl ratios will give FI ) 8 ( 1. The
expected kOTf/kOMs ratios for a rate-determining C-OS2R
bond cleavage are much higher than those observed,
being 104-(5 × 105) in solvolysis of neutral species.5 The
ratios will be, however, smaller if the precursor in the
rate-determining bond cleavage will be the carbanion, as
in the k2 step. Since we find no experimental way to
distinguish between kobs ) k1 or kobs ) k1k2/k-1 we
conclude that the leaving group tool, even if it could
properly be studied and in spite of the significant ratios,
will be a problematic mechanistic probe to distinguish
the single and multistep routes.
Dieth yl 2-(p-Nitr op h en yl)-2-[[(tr iflu or om eth yl)su lfo-
n yl]oxy]eth ylen e-1,1-d ica r boxyla te (1b). To a solution of
diethyl (p-nitrobenzoyl)malonate (2 g, 6.5 mmol) in CH2Cl2 (60
mL) was added pyridine (1.05 mL, 13 mmol). After being
stirred for 15 min, the solution was cooled in an ice-salt bath
and trifluromethanesulfonic anhydride (1.5 mL, 9 mmol) was
added. The mixture was stirred for 3 days in the dark and
filtered, and the solvent was evaporated. The residue was
extracted with 1:1 ether-water (100 mL), the aqueous phase
washed with ether (50 mL) and dried (MgSO4), and the solvent
was evaporated, leaving an orange oil (2.1 g). Chromatography
on silica with 85:15 petroleum ether (40-60 °C): ether eluent
gave 1b admixed with acetophenone and diethyl malonate
(according to GC-MS) as the main fraction (1.2 g). Chroma-
tography of this fraction on silica with 1:1 petroleum ether:
CH2Cl2 eluent gave 1b (700 mg, 24%) as a colorless oil, which
then crystallized to a solid, mp 58-9 °C. UV λmax (MeCN):
271 nm (ꢀ 22 000 M-1 cm-1). IR νmax (neat): 1742 (s, COOEt),
We see no trend in the kOTf/kOMs ratios as a function of
the solvent and the amine, and we ascribe no significance
to the difference between the numbers.
The effect of LG ) Cl, OMs, OTs, and OBs on the rates
of reaction of p-O2NC6H4C(LG)dC(CO2Et)2 with piperi-
dine and morpholine was previously studied.8 The
present work adds LG ) OTf, and the extended data,
which are the only data available for these groups in
vinylic substitution, are given in Table 5. The differences
between the Cl and OTf extremes are significant: 63-
138. Still, these are much smaller than the ratio of 109
for these groups where they serve as nucleofuges in SN1
reaction.23
A F ew Mor e Mech a n istic Com m en ts. As for other
vinylic substitutions,24 piperidine is a better nucleophile
than morpholine: kpip/kmorp ratios are 5.4-20.8 (Table 4),
being twice as high in MeCN than in THF and higher
for 1a than for 1b, but more similar for the 1b/2b pair.
The kMeCN/kTHF ratios are 1.5-5.3 and are consistently
higher for piperidine than for morpholine (Table 4). This
is reasonable for a reaction forming a highly polar
zwitterionic-like transition state, which is slightly later
for piperidine than for morpholine.
1654 (w, Ph) cm-1
.
1H NMR (CDCl3) δ: 1.17, 1.38 (2 × 3H,
2t, J ) 7.1 Hz), 4.15, 4.41 (2 × 2H, 2q, J ) 7.1 Hz), 7.71, 8.32
(4H, AA′BB′q, J ) 8.8 Hz). Mass spectrum m/z (relative
abundance, assignment): 441 (2, M), 396 (4, M - OEt), 292
(10, M - CF3SO3), 150 (B, p-O2NC6H4CO), 69 (28, CF3). Anal.
Calcd for C15H14F3NO9S: C, 40.82; H, 3.20; N, 3.17. Found:
C, 40.68; H, 3.07; N, 3.16.
Dieth yl 2-[(Meth ylsu lfon yl)oxy]-2-p h en yleth ylen e-1,1-
d ica r boxyla te (2a ). To a solution of diethyl benzoylmalonate
(1 g, 4 mmol) in CH2Cl2 (30 mL) was added triethylamine (0.64
mL, 4.6 mmol), and the mixture was stirred for 24 h. Meth-
anesulfonic anhydride (0.8 g, 4.6 mmol) was then added, the
mixture was stirred for 8 h and filtered, and the solvent was
evaporated. The remainder was extracted with 1:1 ether-
water (100 mL), the aqueous phase washed with ether (50 mL)
and dried (MgSO4), and the solvent was evaporated, leaving
an orange oil (1.1 g). Chromatography on silica with 1:3 ether:
petroleum ether (40-60 °C) eluent gave an oil (300 mg) impure
by GC-MS. Additional chromatography on silica with 2:3
ether:petroleum ether gave 2a (250 mg, 18%) as an oil, which
on standing gave a colorless solid, mp 46-7 °C. UV λmax
(MeCN): 382 nm (ꢀ 15 700 M-1 cm-1). IR νmax (Nujol): 1730
The single isomer that was obtained from 4b with
MeS- was ascribed to the retained E-product as is usually
(s, COOEt), 1654 (w, Ph) cm-1 1H NMR (CDCl3) δ: 1.06, 1.34
.
(21) Rappoport, Z.; Topol, A. J . Chem. Soc., Perkin Trans. 2 1972,
1823.
(22) Charton, M. Prog. Phys. Org. Chem. 1981, 13, 119.
(23) Bentley, W. T.; Christl, M.; Kemmer, R.; Llewellyn, G.; Oakley,
J . E. J . Chem. Soc., Perkin Trans. 2 1994, 2531.
(2 × 3H, 2t, J ) 7.1 Hz), 3.13 (3H, s), 4.10, 4.32 (2 × 2H, 2q,
J ) 7.1 Hz), 7.42-7.58 (5H, m). Mass spectrum m/z (relative
(25) Lund, H. Chem. Ber. 1934, 67, 935.
(24) For collection of data, see: Rappoport, Z. In Nucleophilicity;
Harris, J . M., McManus, S. P. , Eds.; Advances in Chemistry Series
215; American Chemical Society: Washington, D.C., 1987; p 399.
(26) (a) Fleury, J . P.; Libis, B. Compt. Rend. 1962, 254, 2020. (b)
Dornow, A.; Grabhofer, H. Ber. 1958, 91, 1824.
(27) Rasmussen, J . K.; Hassner, A. Synthesis 1973, 11, 682.

8542 J . Org. Chem., Vol. 61, No. 24, 1996
Schottland and Rappoport
abundance, assignment): 342 (3, M), 297 (2, M - OEt), 109
(B, PhCO), 77 (20, Ph). Anal. Calcd for C15H18F3O7S: C, 52.62;
H, 5.30. Found: C, 52.80; H, 5.16.
M), 159 (88, M - MeSO3), 105 (B, PhCO), 77 (70, Ph). Anal.
Calcd for C11H11NO3S: C, 55.68; H, 4.67; N, 5.90. Found: C,
55.90; H, 4.80; N, 6.19.
Dieth yl 2-[[(Tr iflu or om eth yl)su lfon yl]oxy]-2-p h en yl-
eth ylen e-1,1-d ica r boxyla te (2b). To a solution of diethyl
benzoylmalonate (1.5 g, 5.7 mmol) in CCl4 (50 mL) was added
triethylamine (1 mL, 7 mmol). After 3 days of stirring
trifluoromethanesulfonic anhydride (1.2 mL, 7.2 mmol) was
added at ice-salt bath temperature. TLC analysis after 2 days
showed that the precursor did not completely disappear,
additional amine and anhydride (7 mmol each) were added,
and the mixture was stirred for 2 additional days. The solvent
was evaporated, the remainder was extracted with 1:1 ether-
water mixture (100 mL), and the aqueous phase was washed
with ether (50 mL), dried (MgSO4), and evaporated. The
yellow liquid obtained was chromatographed with 1:4 CH2Cl2-
petroleum ether (40-60 °C), giving 2b (1 g, 44%) as a colorless
oil that on cooling solidified to a colorless solid, mp 32-3 °C.
UV λmax (MeCN): 259 nm (ꢀ 12 000 M-1 cm-1). IR νmax
(E)- a n d (Z)-2-Cya n o-2-p h en yl-2-[[(tr iflu or om eth yl)-
su lfon yl]oxy]eth ylen e (4b a n d 5b). To a solution of R-ben-
zoylpropionitrile (1.5 g, 10 mmol) in CH2Cl2 (50 mL) was added
triethylamine (1.5 mL, 11 mmol), and after the mixture was
cooled in an ice-salt bath trifluoromethanesulfonic anhydride
(2.5 mL, 15 mmol) was added. After the mixture was stirred
overnight, the reaction was incomplete, and more amine and
anhydride (0.4 mL each) were added. After an additional 18
h of stirring, the solvent was evaporated, the remainder was
extracted with 1:1 ether-water (60 mL), the aqueous phase
washed with more ether (30 mL), and the combined ethereal
phase was dried (MgSO4) and evaporated, leaving an oil (2 g).
Chromatography on silica with 1:9 ether:petroleum ether (40-
60 °C) eluent gave two fractions.
(a) 4b (700 mg, 24%) was a colorless oil. The assignment
as the E isomer is based on analogy with the formation of 4a
as the major isomer. UV λmax (MeCN): 257 nm (ꢀ 11 000 M-1
cm-1). IR νmax (neat): 2230 (s, CtN), 1648 (s, Ph or CdC)
(Nujol): 1740 (s, CO2Et), 1660 (m, Ph) cm-1 1H NMR (CDCl3)
.
δ: 1.06, 1.37 (2 × 3H, 2t, J ) 7.1 Hz), 4.11, 4.38 (2 × 2H, 2q,
J ) 7.1 Hz), 7.43-7.53 (5H, m). Mass spectrum m/z (relative
abundance, assignment): 396 (4.5, M), 351 (5, M - OEt), 247
(13, M - CF3SO3), 105 (B, PhCO), 77 (25, Ph). Anal. Calcd
for C15H15F3O7S: C, 45.45; H, 3.81. Found: C, 45.52; H, 3.77.
1,1-Dicya n o-2-[(m et h ylsu lfon yl)oxy]-2-p h en ylet h yl-
en e (3). To a solution of NaOEt, prepared from Na (0.15 g,
6.5 mmol) in EtOH (75 mL), was added benzoylmalononitrile
(1.19 g, 7 mmol). The mixture was stirred for 2 days, and the
solvent was evaporated, leaving the sodium enolate as a white
solid. To a solution of the sodium enolate (1.06 g, 5.5 mmol)
in MeCN (35 mL) was added methanesulfonic anhydride (1.22
g, 7 mmol), the mixture was refluxed for 2 h and filtered, and
the solvent was evaporated. The remainder was extracted
with ether (30 mL) and water (20 mL), the aqueous phase was
washed with ether (20 mL) and dried (MgSO4), and the solvent
was evaporated, leaving an oil (1.3 g). Chromatography on
silica with 45:55 ether:petroleum ether (40-60 °C) gave 3a
(650 mg, 48%) as an oil, which on cooling gave a solid, mp
42-3 °C. UV λmax (MeCN): 298 nm (ꢀ 16 000 M-1 cm-1); IR
cm-1
.
1H NMR (CDCl3) δ: 2.21 (3H, s), 7.48-7.67 (5H, m).
Mass spectrum m/z (relative abundance, assignment): 291 (67,
M), 142 (9, M - CF3SO3), 105 (B, PhCO), 77 (58, Ph). Anal.
Calcd for C11H8F3NO3S: C, 45.36; H, 2.77; N, 4.81. Found:
C, 45.12; H, 3.05; N, 5.13.
(b) An oil (290 mg), which was a mixture of two compounds
1
according to H NMR, was obtained. Chromatography with
3:7 CH2Cl2:petroleum ether (40-60 °C) eluent gave 5b (120
mg, 4%) as a colorless oil, which on cooling gave a colorless
solid, mp 35 °C. UV λmax (MeCN): 256 nm (ꢀ 19 000 M-1 cm-1).
1
IR νmax (Nujol): 2235 (s, CtN), 1654 (s, Ph or CdC) cm-1. H
NMR (CDCl3) δ: 2.09 (3H, s), 7.46-7.54 (5H, m). Mass
spectrum m/z (relative abundance, assignment): 291 (27, M),
142 (9, M - CF3SO3), 105 (B, PhCO), 77 (68, Ph). Anal. Calcd
for C11H8F3NO3S: C, 45.36; H, 2.77; N, 4.81. Found: C, 45.43;
H, 2.48; N, 4.79.
Dieth yl 2-P ip er id in o-2-p h en yleth ylen e-1,1-d ica r boxy-
la te (9a ). A solution containing diethyl 2-phenyl-2-[[(trifluo-
romethyl)sulfonyl]oxy]ethylene-1,1-dicarboxylate (2b) (200 mg,
0.5 mmol) and piperidine (0.12 mL, 1.2 mmol) in MeCN (5 mL)
was stirred for 3 h. The solvent was evaporated, and the
remainder was extracted with ether (40 mL) and filtered.
Evaporation of the solvent left a yellow oil that solidifed on
cooling. Attempted crystallizaion of the oil from EtOH, CHCl3,
or petroleum ether had failed due to the high solubility of the
compound. Chromatography on silica with EtOH eluent,
followed by evaporation of the solvent, gave an oil that after
addition of petroleum ether gave yellow grains (40 mg, 25%),
mp 88 °C. UV λmax (MeCN): 240 nm (ꢀ 18 000 M-1 cm-1), 332
nm (12 000 M-1 cm-1). IR νmax (Nujol): 1687, 1660 (s,
ν
max (Nujol): 2240 (s, CN), 1600 (m, CdC), 1563 cm-1 1H NMR
.
(CDCl3) δ: 3.47 (3H, s), 7.54-7.95 (5H, m). Mass spectrum
m/z (relative abundance, assignment): 248 (16, M), 170 (8,
PhCOCH(CN)2), 105 (73, PhCO), 79 (B, MeSO2), 77 (38, Ph).
Anal. Calcd for C11H8N2O3S: C, 53.22; H, 3.25; N, 11.28.
Found: C, 53.39; H, 3.21; N, 11.30.
(E)- a n d (Z)-2-Cya n o-1-[(m eth ylsu lfon yl)oxy]-1-p h e-
n ylp r op en e (4a a n d 5a ). To a solution of R-benzoylpropi-
onitrile (1 g, 6.3 mmol) in CH2Cl2 (50 mL) were added
triethylamine (1 mL, 7.2 mmol) and methanesulfonic anhy-
dride (1 g, 7.2 mmol). After the solution was refluxed for 26
h, 30% of the precursor was still unreacted, additional amounts
(7.2 mmol each) of the amine and anhydride were added, and
reflux continued for 17 h. The solvent was evaporated, and
the remainder was extracted with a 1:1 ether-water mixture
(100 mL). The aqueous phase was further washed with ether
(2 × 50 mL), and the combined organic phase was dried
(MgSO4) and evaporated, leaving an oil (1.4 g). Two consecu-
tive chromatographies on silica with 3:7 ether:petroleum ether
and 3:7 CH2Cl2:petroleum ether gave two fractions.
N-CdCCO2Et), 1640 (w) cm-1 1H NMR (CDCl3) δ: 1.02 (2
.
× 3H, t, J ) 7 Hz), 1.69 (6H, br), 3.14 (4H, br), 3.98 (2 × 2H,
q, J ) 7 Hz), 7.42 (5H, m). Mass spectrum m/z (relative
abundance, assignment): 331 (17, M), 286 (16, M - OEt), 258
(B, M - CO2Et), 240 (24, M - CO2Et - H2O), 212 (66, M -
CO2Et - EtOH).
Dieth yl 2-Mor p h olin o-2-p h en yleth ylen e-1,1-d ica r box-
yla te (9b). A solution containing morpholine (0.1 mL, 1.1
mmol) and diethyl 2-phenyl-2-[[(trifluoromethyl)sulfoxyl]oxy]-
ethylene-1,1-dicarboxylate (2b) in MeCN (5 mL) was stirred
for 2 h. The solvent was evaporated, and the oil obtained was
chromatographed on silica with 3:7 ether:petroleum ether
eluent. A yellow oil that crystallized to a solid (70 mg, 85%),
mp 92 °C, was obtained. UV λmax (MeCN): 237 nm (ꢀ ) 16 000
M-1 cm-1), 328 nm (ꢀ ) 11 000 M-1 cm-1). IR νmax (Nujol):
(a) 4a (600 mg, 40%) was colorless oil that solidified on
cooling to a colorless solid, mp 60 °C. Identified as the E
isomer by X-ray crystallography. UV λmax (MeCN): 258 nm
(ꢀ 13 000 M-1 cm-1). IR νmax (neat): 2223 (s, CtN), 1653 (s,
Ph or CdC) cm-1 1H NMR (CDCl3) δ: 2.19 (3H, s), 2.85 (3H,
.
s), 7.47-7.69 (5H, m). Mass spectrum m/z (relative abun-
dance, assignment): 237 (48, M), 159 (69, MH - MeSO3), 105
(B, PhCO), 77 (80, Ph). Anal. Calcd for C11H11NO3S: C, 55.68;
H, 4.67; N, 5.90. Found: C, 55.71; H, 4.68; N, 5.87.
1688, 1672 (s, N-CdCCO2Et), 1580 (w) cm-1
.
1H NMR
(CDCl3) δ: 1.03 (2 × 3H, t, J ) 7 Hz), 3.18 (4H, t, J ) 5 Hz),
3.76 (4H, t, J ) 5 Hz), 4.00 (2 × 2H, 2q, J ) 7 Hz), 7.41 (5H,
m). Mass spectrum m/z (relative abundance, assignment): 333
(53, M), 288 (74, M - OEt), 260 (77, M - CO2Et), 242 (66, M
- CO2Et - H2O), 214 (B, ?).
(b) 5a (100 mg, 6%) was a colorless oil that was chromato-
graphed on a PLC plate with 3:7 CH2Cl2:petroleum ether (40-
60 °C), giving 50 mg (3%) of an oil that on cooling gave a
colorless solid, mp 70 °C. UV λmax (MeCN): 256 nm (ꢀ 17 000
(E)-2-Cya n o-1-(m eth ylth io)-1-p h en ylp r op en e (10). To
a solution of (E)-2-cyano-1-[[(trifluoromethyl)sulfonyl]oxy]-1-
phenylpropene (4b) (200 mg, 0.86 mmol) in ethanol (5 mL)
was added sodium methylthiolate (0.3 g, 1.1 mmol). A rapid
M-1 cm-1). IR νmax (Nujol): 2220 (s, CtN), 1640 (s, Ph) cm-1
.
1H NMR (CDCl3) δ: 2.07 (3H, s), 3.21 (3H, s), 7.49 (5H, s).
Mass spectrum m/z (relative abundance, assignment): 237 (65,

Nucleophilic Vinylic Substitution
J . Org. Chem., Vol. 61, No. 24, 1996 8543
exothermic reaction took place. After 5 min the solvent was
evaporated and the remainder was extracted with ether (20
mL) and water (20 mL). The organic phase was washed with
water (3 × 20 mL), dried (Na2SO4), and evaporated. The
remaining yellow ether (110 mg) was chromatographed on
silica with 95:5 ether:petroleum ether eluent, giving a yellow
oil (80 mg, 49%). UV λmax (MeCN): 279 nm (ꢀ 17 000 M-1
A similar experiment of 4a with NaOMe in MeOH gave a
similar result.
Dim eth yl 2-Mesityl-2-(tosyloxy)eth ylen e-1,1-d ica r box-
yla te (13). A mixture of dimethyl mesitoylmalonate (1.6 g, 6
mmol) and tosyl chloride (1.2 g, 6 mmol) in pyridine (5 mL)
was kept in the refrigerator for 20 h. A solid (presumably
pyridine‚HCl) was obtained. The mixture was poured on 1 N
HCl (50 mL), extracted with CH2Cl2 (2 × 100 mL), washed
with dilute HCl solution and with water, and dried (MgSO4),
and the solvent was evaporated at 50 °C. A white solid,
sparingly soluble in EtOH or hexane, was obtained. Crystal-
lization from benzene-hexane gave a white solid, mp 120 °C
(0.89 g, 32%). UV λmax (EtOH): 243 nm (ꢀ 8000 M-1 cm-1),
284 (6000), 293 (6000). IR νmax (CHCl3): 1730 (s, CO2R), 1640
cm-1). IR νmax (neat): 2210 (s, CtN) cm-1 1H NMR (CDCl3)
.
δ: 1.91 (3H, s), 2.12 (3H, s), 7.31-7.44 (5H, m). Mass
spectrum m/z (relative abundance, assignment): 189 (B, M),
174 (45, M - Me), 147 (24, M - Me - HCN), 140 (20), 115
(55, PhCtCMeH). Anal. Calcd for C11H11NS: C, 69.80; H,
5.86; N, 7.40. Found: C, 70.10; H, 6.11; N, 7.18.
Su bstitu tion of (Z)-2-Cya n o-1-[[(tr iflu or om eth yl)su l-
fon yl]oxy]-1-p h en ylp r op en e (5b) w ith MeS-. MeSNa
(0.03 g, 0.26 mmol) was added with stirring to (Z)-2-cyano-1-
[[(trifluoromethyl)sulfonyl]oxy]-1-phenylpropene (5b) (30 mg,
0.13 mmol) in ethanol (5 mL). The solvent was evaporated
after 1 h, the remainder was extracted with ether (10 mL) and
water (10 mL), and the etheral phase washed with water (3 ×
10 mL) and dried (Na2SO4). The oil obtained (20 mL) was
chromatographed on silica with 9:1 ether:petroleum ether (40-
60 °C), giving a few mg of a yellow oil. The material was
sufficient only for measuring a mass spectrum [m/z (relative
abundance, assignment): 189 (B, M), 174 (41, M - Me), 147
(CdO) cm-1
.
1H NMR (CDCl3) δ: 2.15 (6H, s), 2.23 (3H, s),
2.38 (3H, s), 3.58 (3H, s, CO2Me cis to Mes), 3.87 (3H, s, CO2-
Me trans to Mes), 6.69 (2H, s), 7.12, 7.39 (4H, AA′BB′ q, J )
8.4 Hz). Mass spectrum m/z (relative abundance, assign-
ment): 432 (5, M), 277 (72, M - SO2Tol), 246 (B, M - SO2Tol
- MeO), 228 (29, M- SO2Tol - OMe - H2O), 218 (19, M -
SO2Tol - COOMe), 201 (23, M - SO2Tol - COOMe - H2O),
173 (65, MesCHdCHCO), 155 (77, Tos), 147 (95, MesCO), 119
(86, Mes). Anal. Calcd for C22H24SO7: C, 61.11; H, 5.55; S,
7.41. Found: C, 61.11; H, 5.84; S, 6.88.
1
(24, M - Me - HCN), 140 (31), 115 (62, PhCtCMeH)] and H
Rea ction s of 2-Mesityl-2-(tosyloxy)eth ylen e-1,1-d ica r -
boxyla te w ith Nu cleop h iles. (a ) With Sod iu m p-Tolu -
en eth iola te. To the tosylate 13 (0.43 g, 1 mmol) in DMF (5
mL) was added solid sodium p-toluenethiolate (0.19 g, 1.3
mmol). No color was developed. After the mixture was stirred
for 48 h at rt and worked up as above, the product was
analyzed by NMR. A 55:45 mixture of 13 to dimethyl R-mesi-
toylmalonate (12) was observed, with no trace of the substitu-
tion product. Chromatography on silica gave an oil (0.15 g,
NMR spectrum. The latter showed the presence of the
retained substitution product [δ (CDCl3): 1.92 (3H, s), 1.94
(3H, s), 7.2-7.5 (5H, m)] as the major product, admixed with
16% of the isomeric product.
Rea ction s of 1b w ith Sod iu m p-Tolu en eth iola te.
A
solution containing sodium p-toluenethiolate (80 mg, 0.55
mmol) and 1b (220 mg, 0.5 mmol) in EtOH (10 mL) was stirred
for 2 h. The solvent was evaporated, and the remainder was
extracted with ether, which was dried (MgSO4) and evapo-
rated, giving an oil (180 mg, 77%) with an 1H NMR [δ
(CDCl3): 0.96, 1.37 (2 × 3H, 2t, J ) 7.1 Hz), 2.19 (3H, s), 3.94,
4.36 (2 × 2H, 2q, J ) 7.1 Hz), 6.87, 7.05 (2 × 2H, 2AA′BB′q,
J ) 8.2 Hz), 7.19, 7.94 (2 × 2H, 2AA′BB′q)] identical with that
reported for the substitution product, diethyl 2-(p-tolylthio)-
2-(p-nitrophenyl)ethylene-1,1-dicarboxylate.8
1
55%), which according to the H NMR and TLC is 12, and an
oil (0.15 g), which is a 4:1 mixture of 13 to 12.
(b) With Sod iu m p-Cr esola te. To a solution of 13 (0.22
g, 0.51 mmol) in DMF (3 mL) was added sodium p-cresolate
(0.1 g, 0.8 mmol). The colorless clear solution was stirred for
47 h at rt. After the usual workup a viscous oil was obtained,
1
which according to the H NMR consists of a 15:85 mixture of
Rea ction s w ith Th io-n u cleop h iles. In a typical experi-
ment, 1 equiv of the vinyl mesylate (1a or 4a ) was mixed with
2 equiv of NaSMe in ethanol, and the reaction was followed
by TLC. When all the precursor had disappeared the EtOH
was evaporated and the remainder was extracted with 1:1
ether-water. The organic phase was separated, dried (Mg-
SO4), and evaporated, leaving very small quantities of uni-
dentified compound. Acidification of the organic phase gave
a white emulsion that was extracted with ether. The organic
phase was dried (MgSO4) and evaporated, giving an oil that,
according to its 1H NMR, was identical with that of the
precursor enol for the preparation of the mesylate.
13 to 12. The expected substitution product was not obtained.
Chromatography over silica gave 12 (90 mg, 65%), which was
identified by NMR and TLC.
Ack n ow led gm en t. This work was supported by the
United States-Israel Binational Science Foundation
(BSF), J erusalem, Israel. We thank Dr. A. Gazit for the
experiments with compound 13.
J O961108T