SRN1 photostimulated reaction of triphenylvinyl bromide with carbanions in DMSO

Full text of DOI:10.1021/jo951144c

J . Org. Chem. 1996, 61, 1125-1128
1125
rather than vinylic hydrogen(s) give substituted allene.9,10
When the vinylic halides, e.g., triphenylvinyl bromide
(Ph2CdC(Ph)Br, 1) lack either a vinylic or allylic hydro-
gen they react with nucleophiles exclusively by the SRN1
mechanism.10 This is a chain process whose propagation
steps are shown in Scheme 1.
SRN1 P h otostim u la ted Rea ction of
Tr ip h en ylvin yl Br om id e w ith Ca r ba n ion s
in DMSO
Ana N. Santiago,† Gonzalo Lassaga,†
Zvi Rappoport,*,‡ and Roberto A. Rossi*,†
Sch em e 1
Departamento de Qu´ımica Orga´nica, Facultad de Ciencias
Quı´micas Universidad Nacional de Co´rdoba, Suc. 16, C.C.
61, 5016 Co´rdoba, Argentina, and Department of Organic
Chemistry, The Hebrew University, J erusalem 91904, Israel
(RX)•- f R• + X-
(1)
R• + Nu- f (RNu)•-
(2)
(3)
Received J une 23, 1995
(RNu)•- + RX f RNu + (RX)•-
The radical nucleophilic substitution, or SRN1, has been
found to be suitable to affect the nucleophilic substitution
of aromatic and aliphatic substrates substituted by
electron-withdrawing groups which carry suitable leaving
groups. It has also been reported that aliphatic sub-
strates having no electron-withdrawing groups such as
cycloalkyl, bicycloalkyl, and neopentyl halides react by
this mechanism.1
Vinyl halides react with nucleophiles through different
polar mechanisms,2,3 but there are few reports on sub-
stitution of vinyl halides by the SRN1 mechanism. Bun-
nett and co-workers investigated the photostimulated
reaction of several vinyl halides with ketone enolate and
thiophenoxide ions in liquid ammonia.4 The photostimu-
lated cobalt carbonylation reaction of aryl and vinyl
halides under phase transfer conditions has been sug-
gested to give the carbonylation products through an
Scheme 1 illustrates a nucleophilic substitution in
which radicals and radical anions are involved as inter-
mediates. This chain process requires an initiation step,
and in a few systems spontaneous electron transfer (ET)
from the nucleophile to the substrate has been observed.11
When the ET does not occur spontaneously, it can be
induced by light in aromatic as well as in aliphatic
systems.1,12 In aromatic systems it has also been initi-
ated by solvated electrons13 or sodium amalgam in liquid
ammonia,14 by cathodically generated electrons,15 or by
certain inorganic ions in aromatic16 SRN1 reactions.
Photostimulation and Fe2+ stimulation were recently
applied in vinylic SRN1 reactions.9,10
We undertook the investigation of the photostimulated
reactions of the vinyl bromide 1 with different carbanions
in DMSO in order to extend the scope of the limited
research so far reported.
S
RN1 mechanism.5 The vinylation of iron porphyrins
under electrochemical conditions has been reported,6 and
the reaction of a vinyl halide moiety of a cyanine dye with
different nucleophiles has been suggested to occur by an
Resu lts a n d Discu ssion
S
RN1-type mechanism.7 The reaction of trans-1,2-dichlo-
Rea ction s of 1 w ith Aceton e (2) a n d P in a colon e
(3) En ola te Ion s. A solution containing 1 and potas-
sium acetone enolate (2) was irradiated for 3 h in DMSO
to give 52% yield of bromide ion. Only mere traces of
the substitution product were found, and several minor
unidentified products were formed (experiment 1, Table
1). A slower reaction in the dark under the same
experimental conditions also produced bromide ions, but
no substitution products were found (experiment 2, Table
1). These results suggest that the triphenylvinyl radical
was formed in the initiation step, presumably by ET from
2 to 1, but that the radical, reminiscent of the behavior
roethene with thiolate ions which gave the trans disub-
stitution product in HMPT was also thought to occur by
the SRN1 mechanism.8
It has been shown more recently that vinyl halides
having a â-vinylic hydrogen undergo substitution by the
elimination-addition route involving an acetylenic in-
termediate, whereas vinylic halides carrying an allylic
† Universidad Nacional de Co´rdoba.
‡ The Hebrew University.
(1) For reviews on SRN1 see: (a) Rossi, R. A.; de Rossi R. H. Aromatic
Substitution by the SRN1 Mechanism; ACS Monograph 178; American
Chemical Society: Washington D.C., 1983. (b) Bowman, W. R. Chem.
Soc. Rev. 1988, 17, 283. (c) Rossi, R. A.; Pierini, A. B.; Palacios, S. M.
In Advances in Free Radical Chemistry; Tanner, D. D., Ed.; J AI
Press: Greenwich, CT, 1990; Chapter 5, p 193; J . Chem. Educ. 1989,
66, 720. (d) Norris, R. K. In Comprehensive Organic Synthesis; Trost,
B. M., Ed.; Pergammon: New York, 1991; Vol. 4, p 451. (e) Rossi, R.
A., Pierini A. B., Pen˜e´n˜ory, A. B. In The Chemistry of Functional
Groups, Suppl. D, The Chemistry of Halides, Pseudo-halides and
Azides; Patai S., Rappoport, Z., Eds.; Wiley: New York, 1995; Chapter
24, p 1395.
(2) Modena, G. Acc. Chem. Res. 1971, 4, 73.
(3) Rappoport, Z. Recl. Trav. Chim. Pays-Bas 1985, 104, 309; Acc.
Chem. Res. 1992, 25, 474.
(4) Bunnett, J . F.; Creary, X.; Sundberg, J . E. J . Org.Chem. 1976,
41, 1707.
(5) (a) Brunet, J -J .; Sidot, C.; Caubere, P. Tetrahedron Lett. 1981,
22, 1013. (b) Brunet, J .-J .; Sidot, C.; Caubere, P. J . Org. Chem. 1983,
48, 1166.
(9) Galli, C.; Gentili, P. J . Chem. Soc., Chem. Commun. 1993, 570.
(10) Galli, C.; Gentili, P.; Rappoport, Z. J . Org. Chem. 1994, 59, 6786.
(11) (a) Scamehorn, R. G.; Bunnett, J . F. J . Org. Chem. 1977, 42,
1449. (b) Swartz, J . E.; Bunnett, J . F. J . Org. Chem. 1979, 44, 340. (c)
Bard, R. R.; Bunnett, J . F.; Traber, R. P. J . Org. Chem. 1979, 44, 4918.
(d) Dell’Erba, C.; Novi, M.; Petrillo, G.; Tavani, C. Tetrahedron 1992,
48, 325.
(12) (a) Rossi, R. A.; Bunnett, J . F. J . Org. Chem. 1973, 38, 1407.
(b) Hoz, S.; Bunnett, J . F. J . Am. Chem. Soc. 1977, 99, 4690. (c) Fox,
M. A.; Younathan, J .; Fryxell, G. E. J . Org. Chem. 1983, 48, 3109.
(13) (a) Kim, J . K; Bunnett, J . F. J . Am. Chem. Soc. 1970, 92, 7464.
(b) Rossi, R. A.; Bunnett, J . F. J . Am. Chem. Soc. 1974, 96, 112.
(14) (a) Austin, E.; Alonso, R. A.; Rossi, R. A. J . Org.Chem. 1991,
56, 4486. (b) Austin, E.; Ferrayoli, C. G.; Alonso, R. A.; Rossi, R. A.
Tetrahedron 1993, 49, 4495.
(15) (a) Save´ant, J . M. Acc. Chem. Res. 1980, 13, 323. (b) Save´ant,
J . M. Adv. Phys. Org. Chem. 1990, 26, 1 and references cited therein.
(16) (a) Galli, C.; Bunnett, J . F. J . Org. Chem. 1984, 49, 3041. (b)
Galli, C.; Gentili, P. J . Chem. Soc., Perkin Trans. 1 1993, 1135. (c)
van Leeuwen, M.; McKillop, A. J . Chem. Soc., Perkin Trans. 1 1993,
2433.
(6) Lexa, D.; Save´ant, J . M. J . Am. Chem. Soc. 1982, 104, 3503.
(7) Strekowski, L.; Lipowska, M.; Patonay, G. J . Org. Chem. 1992,
57, 4578.
(8) Kodomari, M.; Yamamoto, T.; Nomaki, M.; Yoshitomi, S. Nippon
Kagaku Kaishi 1982, 5, 796; Chem. Abstr. 1982, 97, 143959h.
0022-3263/96/1961-1125$12.00/0 © 1996 American Chemical Society

1126 J . Org. Chem., Vol. 61, No. 3, 1996
Notes
2, hν
Ta ble 1. Rea ction s of Tr ip h en ylvin yl Br om id e (1) w ith
Ca r ba n ion s in DMSOa
-
1 + -
8
Ph CdC(Ph)CH COPh
+ Br
CH2COPh
2
2
6
5
1
carbanion
(103 M)
Br-
(yield, %) (yield, %)
product,
(5)
expt (103 M)
conds
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
20
20
20
20
20
20
20
17
17
17
17
17
17
17
17
2 (54)
hν
52
12
95
48
16
6
32
95
33
<5
43
35
93
<10
<10
traces
2 (54)
3 (60)
3 (60)
3 (60)
3 (60)
5 (39)
dark
suggest again that 1 reacts with 5 by the vinylic
hν
4 (93)
4 (25)
4 (15)
S
RN1 mechanism.
hν, p-DNB
dark
The overall reactivity in the SRN1 mechanism depends
dark, p-DNB
hν
on the rate (or efficiency) of the photostimulated initiation
step, the individual rates of the propagation steps, and
the termination steps. The fact that 5 reacted sluggishly
under irradiation with 1, but gave high yields in the
presence of the carbanion 2, suggests that carbanion 5
has low ET reactivity in the initiation step, but once the
vinyl radical is formed, it reacts faster with carbanion 5
than with carbanion 2 (entrainment reaction).1 On the
other hand, carbanion 2 is more effective than carbanion
5, in the rate determining initiating step ET to 1 under
irradiation, thus increasing the rate of the overall reac-
tion. This sequence of irradiation by ET follows roughly
the same order of the pKa of the conjugated acids of the
carbanions, as has been previously proposed in the
photostimulated reaction with haloarenes19 and 1-
iodoadamantane.17b
6b
6 (90)
6 (31)
5 (39) + 2 (20) hν
5 (39) + 2 (20) dark
5 (39) + 2 (20) dark, p-DNB
5 (39) + 2 (20) hν, p-DNB
6 (40)
8b
8 (89)
7 (60)
hν
7 (84) + 2 (80) hν
7 (74) + 2 (64) dark
7 (74) + 2 (64) dark, p-DNB
a
b
Reaction time 3 h. Not quantified.
of aliphatic radicals, such as 1-adamantyl17 or 7-nor-
caranyl18 formed under similar SRN1 conditions, reacts
slower with 2 than other competing reaction routes. We
did not study these reactions further.
In contrast, the photostimulation reaction of 1 with the
enolate ion of pinacolone (3) afforded bromide ion (95%
yield) together with an excellent yield (93%) of the
substitution product 4 (experiment 3, Table 1) (eq 4).
Rea ction s of 1 w ith Nitr om eth a n e An ion (7). The
reaction of 1 with the anion of nitromethane (7) as the
nucleophile takes a different course. No substitution
product is formed after 180 min of irradiation (although
a 35% yield of Br- ions was formed). However, when 2
was added to the reaction mixture, a reaction took place,
but rather than yielding the desirable substitution
product, the product isolated was 1,1,2-triphenylpropene
8 (experiment 13, Table 1) (eq 6).
+
9
hν8
-CH2COR
2 (R ) Me)
3 (R ) CMe3)
Ph2CdC(Ph)CH2COR + Br- (4)
4 (R ) CMe3)
Ph CdC(Ph)Br
2
1
2, hν
1 + -
8
+ Br- (6)
Ph CdC(Ph)CH
3
CH2NO2
2
This reaction was inhibited in part by adding p-
dinitrobenzene (p-DNB), a well-known inhibitor of SRN1
reaction,1 to give only 25% yield of the substitution
product 4 under the same conditions (experiment 4, Table
1). Also, the use of irradiation accelerates the reaction
since in the dark the reaction is sluggish and yields only
15% of the substitution product (experiment 5, Table 1).
In the dark and in the presence of p-DNB there was no
substitution product (experiment 6, Table 1). These
results corroborate the recent suggestion of and SRN1
vinylic route for the reaction of 1 and 3.10
8
7
This reaction did not occur in the dark (experiment 15,
Table 1), and the photostimulated reaction was inhibited
by p-DNB (experiments 2 and 3, Table 1).
On the basis of these and previous results, it follows
that nucleophiles 5 and 7 are unable to initiate the
photostimulated reaction, but once the vinyl radical is
formed it is capable of coupling with them, whereas 2 is
more effective in starting the photostimulation reaction.
Rea ction s of 1 w ith Acetop h en on e En ola te Ion s
5. When the photostimulation reaction (3 h) was carried
out with the enolate ion of acetophenone (5) a low yield
of bromide ion (32%) and low yield of the substitution
product 6 (<5%) were obtained (experiment 7, Table 1).
When the reaction was performed in the presence of
acetone enolate ion (2), we obtained a 95% yield of
bromide ion, 90% yield of the substitution product 6 of
acetophenone, but no substitution product of acetone
(experiment 8, Table 1) (eq 5).
Similar reactions in the dark rendered 33% yield of
bromide ion and a 31% yield of 6, and when they were
carried out in the presence of p-DNB, only ca. 5% Br-
ions were found (experiments 9 and 10, Table 1). These
light-catalyzed reactions and the inhibition by p-DNB
The formation of the substitution product 8 can be
explained according to the SRN1 mechanism considering
that anion radicals of nitro derivatives can fragment.
Consequently, we suggest an initial photostimulated ET
from the nucleophile, probably 2 to the substrate to form
the radical anion 1•- which fragments in the usual way
(cf. eq 1) to the radical intermediate 9 and bromide ion
(eq 7). Radical 9 then couples with 7 faster than with 2
to give the radical anion 10 (eq 8). Radical anion 10
fragments faster to the stabilized allyl radical 11 and
-
NO2 ions (eq 9) than it transfers electron to neutral 1
to give the expected substitution product in a step similar
to eq 3. Finally, radical 11 does not react with 7 or 2 in
a step similar to eq 2, but it reacts faster by hydrogen
abstraction from the solvent SH to give the reduced
(17) (a) Borosky, G. L.; Pierini, A. B.; Rossi, R. A. J . Org.Chem. 1990,
55, 3705. (b) Rossi, R. A.; Pierini, A. B.; Borosky, G. L. J . Chem. Soc.,
Perkin Trans. 2 1994, 2577.
(19) Borosky, G. L.; Pierini, A. B.; Rossi, R. A. J . Org. Chem. 1992,
57, 247.
(18) Nazareno, M. A.; Rossi, R. A. Tetrahedron 1994, 50, 9267.

Notes
J . Org. Chem., Vol. 61, No. 3, 1996 1127
Ta ble 2. Com p etition Exp er im en ts in th e
product 8 (eq 10) (Scheme 2).
P h otostim u la ted Rea ction of Ca r ba n ion s 3 a n d 5 tow a r d
1a
Sch em e 2
products (yield, %)
5
3
reaction
-
[Ph2CdC(Ph)Br]•-
Ph CdC4 Ph
(102 M) (102 M) time (h)
6
4
k5/k3
3.5
4.3
3.4
f
+ Br
(7)
(8)
2
9
1
5.0
3.3
3.3
5.0
6.7
8.3
1
3
3
71
59
49
22
37
49
•-
9 + 7 f
[Ph2CdC(Ph)CH2NO2]
3.7 ( 0.4b
10
a
Reactions carried out in 15 mL of DMSO with 0.017 M of
b
substrate 1. Average value.
-
10 f
• + NO2
(9)
Ph2CdC(Ph)CH2
11
radical anion intermediate formed and with the results
obtained in competition experiments of acetone enolate
ion (2) and acetophenone enolate ion (5) toward iodoben-
zene (k5/k2 ) 7.5)19 and 1-iodoadamantane (k5/k2 ) 11).17
The fact that structurally different substrates (phenyl,
vinyl, and 1-adamantyl halides) show nearly similar
selectivity toward carbanions is consistent with their
reaction by the same mechanism and indicates that for
this limited group of substrates the reactivity of the
derived radicals depends less on their structure than on
the stability of the radical ion products formed in the
coupling process.
11 SH8 8
(10)
This variant of the vinylic SRN1 route is reminiscent
of the results obtained in the photostimulated reaction
of iodobenzene with 7 in DMSO,19 which yielded products
derived from benzyl radicals by fragmentation of the
radical anion intermediate arising from the coupling of
phenyl radicals and 7 (eq 11). The main difference is that
benzyl radicals are not exclusively reduced to toluene (eq
12) but are also able to couple partly to the nucleophile
7 to give the radical anion 12, which gives finally the
product 1-phenyl-2-nitroethane (eq 13).19 We did not find
any products from the reaction of the allylic radical 11
and 7.
Exp er im en ta l Section
P h otostim u la ted Rea ction of Tr ip h en ylvin yl Br o-
m id e (1) w ith Differ en t Ca r ba n ion s in DMSO. The
following procedure is representative of all the reactions.
Into a three-necked, 50-mL, round-bottomed flask
equipped with a cold finger condenser and a magnetic
stirrer was added 25 mL of dry, degassed DMSO under
nitrogen. Then, pinacolone (300 mg, 3.0 mmol) and
t-BuOK (350 mg, 3.1 mmol) were introduced in order to
form potassium pinacolone enolate. To this solution was
added substrate 1 (338 mg, 1.0 mmol), and then the
solution was irradiated for 180 min. The reaction was
quenched by adding ammonium nitrate in excess. The
DMSO was dissolved with 3-fold excess v/v of water and
then extracted with methylene chloride. The bromide ion
in the aqueous solution was determined potentiometri-
cally. The methylene chloride extract was washed twice
with water and dried, and the products were quantified
by GLC using tetraphenyltin as internal standard. In
another experiment, the dry organic solvent was removed
under reduced pressure after the extraction and the
product 2,2-dimethyl-5,6,6-triphenyl-5-hexen-3-one (4)
was isolated as a white solid after recrystallization from
Ph• + 7 f (PhCH2NO2)•- f PhCH2• + NO2 (11)
-
PhCH2 SH8 PhCH3
(12)
•
PhCH2• + 7 f
ET8
•-
(PhCH2CH2NO2)
12
PhCH2CH2NO2 (13)
Radical anions derived from anions of nitroalkanes are
prone to fragment when they give stabilized radicals,
such as benzyl, or in our case allyl radicals. Even the
electrochemically induced SRN1 reaction of 4-bromo-
benzophenone with 2-nitropropane anion gave products
only derived from the fragmentation of the radical anion
intermediate.20 When the fragmentation of the radical
anion gives an unstabilized alkyl radical, as is the case
•
for 12 which would lead the primary PhCH2CH2 radical
or for the radical anions formed in the reaction of
1-adamantyl17 or 7-norcaranyl18 radicals with 7, the ET
rate to the substrate is faster than the fragmentation
rate, and only substituted nitroalkanes are obtained.
Rela tive Rea ctivity of Ca r ba n ion s 3 a n d 5 tow a r d
th e Vin yl Ra d ica l 9. In order to determine the relative
rate constants for the reacting nucleophiles toward the
vinyl radicals (step 2), we carried out the competitive
photostimulation reactions of 1 with excess of the enolate
ions 3 and 5. The relative reactivities were estimated
as in previous work.21 Using this procedure, we deter-
mined that the k5/k3 ratio is 3.7 ( 0.4 in their competitive
coupling with the triphenylvinyl radical 9 (Table 2).
These results are consistent with earlier findings
where the reactivity of ketone enolate ions toward
radicals increased with the increasing stability of the
1
hexane: mp 167-170 °C (lit.10 165-168 °C); H NMR
(CDCl3) δ 7.3-7.2 (m, 5HAr), 7.2-7.1 (m, 5HAr), 7.0-6.9
(m, 5HAr), 3.7 (s, 2HCH ); 0.9 (s, 9HCH ).10
2
3
1,3,4,4-Tetr a p h en yl-3-bu ten -1-on e (6). The general
procedure above was used, except that acetophenone and
t-BuOK were used to form potassium acetophenone
enolate ions. Product 6 was isolated as a white solid after
recrystallization from hexane-acetone: mp 190-194 °C;
(21) The equation used for determining the relative reactivity of
pairs of nucleophiles I and II vs. a radical is
kNu
kNu II
where [Nu I]o and [Nu II are initial concentrations and [Nu I]t and
I
ln [Nu I]o/[Nu I]t
)
ln [Nu II]o/[Nu II
]
t
]
o
[Nu II]t are concentrations at time t of both nucleophiles. This equation
is based on a second-order reaction, first order in each one of the
nucleophiles with vinyl radicals. See: Bunnett, J . F. In Investigation
of Rates and Mechanisms of Reactions, 3rd ed.; Lewis, E. S., Ed.;
Wiley-Interscience: New York, 1974; Part 1, p 159.
(20) Amatore, C.; Gareil, M.; Oturan, M. A.; Pinson, J .; Saveant, J .
M.; Thiebault, A. J . Org. Chem. 1986, 51, 3757.

1128 J . Org. Chem., Vol. 61, No. 3, 1996
Notes
1H NMR (CDCl3) δ 7.4-7.0 (m, 15HAr), 8.0-7.8 (m, 5HAr),
(7), 165 (15), 91 (70), 57 (100). HRMS calcd 270.140 850,
found 270.141 658.
4.2-4.1 (s, 2HCH ); 13C NMR (DCCl3 relative to TMS) δ
2
46.3 (CH2), δ 126.2, 126.4, 127.1, 127.5, 127.8, 128.0,
128.3, 128.4, 129.2, 129.7, 130.7, 132.9, 133.3, 137.0,
142.3, 142.9, 143.3 (CAr and Cv), δ 197.6 (CdO); mass
spectrum, m/ z (relative intensity, assignment) 374 (36),
269 (65), 252 (6), 191 (70), 105 (100), 91 (49), 77 (31);
HRMS calcd 374.167 065, found 374.166 933.
1,1,2-Tr ip h en yleth en e (13). In order to ensure that
8 was not the reduction product (13),10 we synthesized
the latter. To the substrate 1 in anhydrous ethanol
under nitrogen was added sodium with stirrer, and the
product found was different from 8 by cgl.
P h otostim u la ted Rea ction of Tr ip h en ylvin yl Br o-
m id e w ith Differ en t Ca r ba n ion s in th e P r esen ce of
p-DNB. The general procedure was followed, except that
20 mol % of p-DNB was added prior to addition of the
substrate.
1,1,2-Tr ip h en ylp r op en e (8). The general procedure
was followed except that nitromethane was added instead
of the ketone. Product 8 was isolated as a solid: mp 87-
90 °C (lit.22 93 °C); 1H NMR (CDCl3) δ 7.4-7.0 (m, 15 H
Ar); 2.0-1.6 (s, 3 H
); 13C NMR (DCCl3 relative to TMS)
CH3
Ack n ow led gm en t. This work was supported in part
by the Consejo de Investigaciones de la Provincia de
Co´rdoba (CONICOR), Consejo Nacional de Investiga-
ciones Cient´ıficas y Te´cnicas (CONICET) (Argentina),
and Antorchas Foundation.
δ 23.309 (CH3), δ 125.8, 126.2, 126.5, 127.0, 127.4, 127.5,
127.8, 127.9, 128.0, 128.1, 128.2, 129.3, 129.5, 130.0,
130.3, 130.6 (CAr and Cv); mass spectrum, m/ z (relative
intensity, assignment): 270 (20), 255 (5), 191 (66), 178
(22) Campbell, A.; Kenyon, J . J . Chem. Soc. 1947, 436.
J O951144C