A non-outer sphere mechanism for the ionization of aryl vinyl sulfides by triarylaminium salts

Article abstract of DOI:10.1021/ja9706965

Evidence is presented that the formation of aryl vinyl sulfide cation radicals from the corresponding neutral precursors via reaction with tris(4- bromophenyl)aminium hexachloroantimonate in the context of a cation radical Diels-Alder addition to 1,3-cyclopentadiene does not occur via outer sphere electron transfer but by a mechanism involving strong covalent interaction between the aminium salt acting as an electrophile and the aryl vinyl sulfide substrate acting as a nucleophile.

Full text of DOI:10.1021/ja9706965

J. Am. Chem. Soc. 1997, 119, 11381-11389
11381
A Non-Outer Sphere Mechanism for the Ionization of Aryl
Vinyl Sulfides by Triarylaminium Salts
Nathan L. Bauld,* J. Todd Aplin, Wang Yueh, and Ainhoa Loinaz
Contribution from the Department of Chemistry and Biochemistry, The UniVersity of Texas,
Austin, Texas 78712
ReceiVed March 4, 1997. ReVised Manuscript ReceiVed September 29, 1997X
Abstract: Evidence is presented that the formation of aryl vinyl sulfide cation radicals from the corresponding neutral
precursors Via reaction with tris(4-bromophenyl)aminium hexachloroantimonate in the context of a cation radical
Diels-Alder addition to 1,3-cyclopentadiene does not occur Via outer sphere electron transfer but by a mechanism
involving strong covalent interaction between the aminium salt acting as an electrophile and the aryl vinyl sulfide
substrate acting as a nucleophile.
The generation of cation radicals from corresponding neutral
substrates has been accomplished by a variety of physical and
chemical means, and the involvement of these cation radicals
in efficient pericyclic chemistry has been confirmed and studied
extensively.1 The use of chemical agents, in particular triaryl-
aminium salts, to generate cation radicals has proved to be
especially convenient for synthetic purposes. Although many
mechanistic aspects of the pericyclic reactions induced by
aminium salts have been elucidated, relatively little is known
concerning the specifics of the substrate ionization process per
se. In particular, the fundamental question of the mechanism
of these electron transfer (ET) reactions has received little
attention.2 Among the most plausible possibilities are (1) outer
sphere ET, (2) inner sphere ET, and (3) a polar mechanism,
that is actually a limiting form of the inner sphere ET
mechanism. The latter involves the formation of a full covalent
bond Via electrophilic addition of the aminium salt to the
substrate, to form a distonic cation radical, followed by
homolysis of the newly formed covalent bond to yield the
substrate cation radical and the neutral triarylamine (Scheme
1). The occurrence of nucleophilic attack on an aryl ring of a
triarylaminium cation radical has already been well established,3
but the possibility that such reactions could lead, Via homolysis,
to overall electron transfer has only recently received explicit
consideration.4 In a recent study of the ionization of stilbene
by tris(4-bromophenyl)aminium hexachloroantimonate (1•+) in
the context of a Diels-Alder addition to 2,3-dimethyl-1,3-
butadiene, an outer sphere ET mechanism was suggested by
substituent effect studies which reveal a transition state for
ionization in which positive charge is distributed symmetrically
over both aryl rings.4
Brown σ+ parameters as opposed to Hammett σ values (Scheme
2).5 In sharp contrast, substituent effects upon the protonation
of these same substrates at the carbon atom â to the heteroatom
to yield an R-heteroatom stabilized carbocation intermediate
correlate preferentially with σ values. Further, substituent
effects upon electrophilic reactions at the heteroatom of such
substrates are expected to, and have been found to, correlate
with σ.6 Consequently a statistically validated preference for a
correlation with σ represents strong evidence for an electrophilic
reaction of these substrates, whereas a correlation with σ+
signals an electron transfer (ET) reaction producing a cation
radical intermediate. The σ/σ+ criterion is therefore a useful
diagnostic test for distinguishing outer sphere ET from elec-
trophilic reactions of these two classes of electron rich substrates.
Moreover, the utility of the σ/σ+ criterion can be significantly
extended in instances where cation radical formation can be
demonstrated by independent criteria. Specifically, the polar
ET mechanism depicted in Scheme 1 is a stepwise process which
involves initial electrophilic addition to the substrate. Providing
that this step is rate determining, substituent effects should
correlate with σ values. Consequently the σ/σ+ criterion can
presumably be used to render the distinction between outer
sphere ET and a polar ET mechanism. Further, an inner sphere
ET mechanism which involves only a small to modest inner
sphere interaction should presumably yield a much better
correlation with σ+ than with σ, i.e., it should more closely
resemble outer sphere ET than a polar reaction. The distinction
between a stepwise polar ET mechanism and a concerted inner
sphere ET mechanism which involves a very strong covalent
interaction is more subtle but potentially can be accomplished
Via stereochemical studies (Vide infra).
The reaction system studied in the present work is the Diels-
Alder cycloaddition of aryl vinyl sulfides to 1,3-cyclopentadiene
catalyzed by tris(4-bromophenyl)aminium hexachloroantimonate
(Scheme 3). The aminium salt catalyzed Diels-Alder additions
of phenyl vinyl sulfide to a variety of dienes have previously
been confirmed, using several independent criteria, to occur Via
cation radical/neutral cycloaddition and specifically Via addition
of phenyl vinyl sulfide cation radicals to neutral diene
molecules.7-9 Carbocation mediated cycloadditions, in par-
ticular, have been decisively ruled out.10,11 Several of the
Results and Discussion
Extensive experimental and theoretical studies have previ-
ously established that substituent effects upon the ionization of
meta- and para-substituted aryl vinyl sulfides and ethers in
solution and in the gas phase preferentially correlate with the
X Abstract published in AdVance ACS Abstracts, November 15, 1997.
(1) Bauld, N. L. AdVances in Electron Transfer Chemistry; Mariano, P.
S., Ed.; JAI Press: Greenwich, 1992; Vol. 2, pp 1-66. Bauld, N. L.
Tetrahedron 1989, 45, 5307-5363.
(2) Hubig, S. M.; Bockman, T. M.; Kochi, J. K. J. Am. Chem. Soc. 1996,
118, 3842-3851.
(5) Bauld, N. L.; Aplin, J. T.; Yueh, W.; Endo, S.; Loving, A. J. Phys.
Org. Chem. 1997, in press.
(6) Lewis, E. S.; Kukes, S.; Slater, C. D. J. Am. Chem. Soc. 1980, 102,
303-306. For example, reaction of aryl methyl sulfides with trimethyloxo-
nium fluoroborate yields r2 values of 0.993 and 0.902, respectively, for the
σ and σ+ correlations.
(3) Eberson, L.; Olofsson, B.; Svensson, J.-O. Acta Chem. Scand. 1992,
46, 1005-1015.
(4) Yueh, W.; Bauld, N. L. J. Chem. Soc., Perkin Trans. 2 1995, 871-
873; 1995, 1715-1716.
S0002-7863(97)00696-3 CCC: $14.00 © 1997 American Chemical Society

11382 J. Am. Chem. Soc., Vol. 119, No. 47, 1997
Bauld et al.
Scheme 1. Mechanisms for the Formation of Cation Radicals in the Reactions of Neutral Substrates with Triarylaminium Salts
Scheme 2. The σ/σ+ Criterion for the Distinction between
Electrophilic and Outer Sphere Electron Transfer Reactions
of Aryl Vinyl Ethers and Sulfides
salt is omitted from the reaction medium. Finally, the same
cycloadducts are formed in essentially the same diastereoiso-
meric (endo:exo) ratio when the reactions are carried out
electrochemically (Via anodic oxidation of the aryl vinyl sulfide)
and by photosensitized electron transfer. The quantitative
aspects of the electrochemical oxidations are especially signifi-
cant and will be considered in greater detail in a subsequent
section.
Cation radical Diels-Alder additions have been found to
occur predominantly Via dienophile cation radicals (aryl vinyl
sulfide cation radicals in the present system) as opposed to diene
cation radicals.1 The evidence is especially clear in the case of
cycloadditions involving aryl vinyl sulfides, where diene cation
radicals have been found to produce vinylcyclobutane adducts
preferentially.10 The present reaction system was designed to
further assure that the aryl vinyl sulfide component of the
cycloaddition is the ionized component. The oxidation potential
of 1,3-cyclopentadiene is ca. 0.4 eV higher than that of phenyl
vinyl sulfide.12 Further, this diene is essentially inert toward
the aminium salt in the absence of the aryl vinyl sulfide
substrate. The cycloaddition step must therefore involve an
ionized aryl vinyl sulfide molecule rather than an ionized diene
molecule (Scheme 4).
Scheme 3. Diels-Alder Cycloadditions of Aryl Vinyl
Sulfides to 1,3-Cyclopentadiene Catalyzed by
Tris(4-bromophenyl)aminium Hexachloroantimonate
Products. The reaction of phenyl vinyl sulfide with 1,3-
cyclopentadiene in the presence of 1•+ in methylene chloride
solvent at 0 °C cleanly yields the endo and exo Diels-Alder
adducts 3a,b in the ratio 3:1 (endo:exo). Only trace quantities
of vinylcyclobutanes were observed, even at the earliest reaction
times (low conversion). The endo and exo adducts were isolated
1
and purified separately, and each was characterized by 2D H
and 13C NMR spectroscopy and HRMS. The preference for
the endo adduct is characteristic of cation radical Diels-Alder
additions.1
Kinetics. The relative rates of cycloaddition of eight aryl
vinyl sulfides to 1,3-cyclopentadiene in dichloromethane solu-
tion were obtained by competition kinetics, using techniques
which have previously been described (Table 1).13 The plots
of log k/k0 Vs σ and σ+ are given in Figures 1 and 2, respectively.
previously developed mechanistic criteria have been applied to
the specific reaction system under study in order more defini-
tively to exclude a carbocation mediated cycloaddition mech-
anism. The inclusion of a hindered pyridine base in the reaction
medium containing the aminium salt is known to selectively
suppress Brønsted acid catalyzed, carbocation mediated reac-
tions. The reaction of Scheme 3, however, proceeds smoothly
even in the presence of an excess of 2,6-di(tert-butyl)-4-
methylpyridine. Further, added strong acid (triflic acid) fails
to generate any of the Diels-Alder adducts when the aminium
(7) Pabon, R. A.; Bellville, D. J.; Bauld, N. L. J. Am. Chem. Soc. 1983,
105, 5158-5159.
(8) Harirchian, B.; Bauld, N. L. Tetahedron Lett. 1987, 28, 927-930.
(9) Harirchian, B.; Bauld, N. L. J. Am. Chem. Soc. 1989, 111, 1826-
1828.
(10) Kim, T.; Pye, R. J.; Bauld, N. L., J. Am. Chem. Soc. 1987. 112,
6285-6290.
(11) Reynolds, D. W.; Lorenz, K. T.; Chiou, H.-S.; Bellville, D. J.; Pabon,
R. A.; Bauld, N. L. J. Am. Chem. Soc. 1987, 109, 4960.
(12) Gassman, P. G.; Singleton, D. A. J. Am. Chem. Soc. 1984, 106,
7993.

Ionization of Aryl Vinyl Sulfides
J. Am. Chem. Soc., Vol. 119, No. 47, 1997 11383
Figure 1. Hammett plot for the Diels-Alder cycloadditions of aryl
vinyl sulfides to 1,3-cyclopentadiene, catalyzed by 1•+ in dichlo-
romethane at 0 °C (log krel ) 0.20 - 4.26σ; r2 ) 0.986).
Figure 2. Hammett-Brown plot for the Diels-Alder cycloadditions
of aryl vinyl sulfides to 1,3-cyclopentadiene, catalyzed by 1•+ in
dichloromethane at 0 °C (log krel ) -0.20 - 2.68σ+; r2 ) 0.915).
Scheme 4. Mechanism of the Cation Radical Diels-Alder
Cycloaddition
Table 1. Relative Rates of Diels-Alder Cycloadditions of Aryl
Vinyl Sulfides to 1,3-Cyclopentadiene, Catalyzed by
Tris(4-bromophenyl)aminium Hexachloroantimonate
substituent
log k/k0(CH2Cl2)
log k/k0(CH3CN)
4-MeO
4-Me
3-Me
H
4-Br
3-Cl
1.36
0.922
0.450
1.210
0.815
0.327
0.000
-0.652
-1.340
Figure 3. Hammett plot for the Diels-Alder cycloadditions of aryl
vinyl sulfides to 1,3-cyclopentadiene, catalyzed by 1•+ in acetonitrile
at 0 °C (log krel ) 0.12 - 3.85σ; r2 ) 0.995).
0.000
-0.790
-1.480
-2.020
ments of the relative rate data for the competition between the
4-MeO and 4-Me derivatives have yielded the value 2.73 (
0.06. For comparison, the predicted value for k(4-MeO)/k(4-
Me) in a σ+ correlation, assuming a value of F determined by
omitting the 4-MeO data point, is ca. 40! Importantly, the mass
balance in these reactions (i.e., the percent of the reactions
accounted for as either Diels-Alder adducts or recovered
starting material) is quite highsin some cases as high as 99%
(see Experimental Section). Consequently, the possibility that
most of the cation radicals of the 4-MeOAVS substrate are lost
Via side reactions is negated. Further, the four independent
measurements span a 25-fold range of relative cyclopentadiene:
aryl vinyl sulfide concentrations (0.2-5), thus ruling out the
possibility that the ionization step becomes reversible for this
substrate. Note that absolute reaction rate measurements have
confirmed that ionization is rate determining for the 4-Me
derivative. Consequently, ionization must also be rate deter-
mining for the 4-MeO derivative. The foregoing discussion
generates substantial confidence in the experimental value of
the relative rate constant for the 4-MeOAVS, and, importantly,
it also establishes that this substrate is not mechanistically
abnormal. Confidence in the significance of the distinction
4-CF3
3,5-Cl2
-1.95
The correlation with σ is seen to be excellent (r2 ) 0.986), while
that with σ+ is much worse (r2 ) 0.915). Analogous kinetic
studies were also carried out in the more polar solvent
acetonitrile (Table 1). Again the plot of log k/k0 Vs σ (Figure
3, r2 ) 0.995) is far superior to that Vs σ+ (Figure 4, r2 ) 0.924),
and the statistical significance of the distinction is borne out
by F tests for the data in both solvents.14
Although the preference for a σ over a σ+ correlation is
maintained when the data point for the 4-methoxy derivative is
omitted, it is noteworthy that the obtention of a statistically
significant preference for the σ correlation criticially depends
upon the inclusion of this data point for all of the data sets
reported in the paper. Consequently we have paid especially
careful attention to this data point. Four independent measure-
(13) Baltes, H.; Steckhan, E.; Scha¨fer, H. J. Chem. Ber. 1978, 111, 1294-
1314. The peak potential for cyclopentadiene (Vs Ag/AgCl) is given as 1.50
V. The peak potential for phenyl vinyl sulfide is 1.1 V.
(14) Yueh, W.; Bauld, N. L. J. Phys. Org. Chem. 1996, 9, 529-538.

11384 J. Am. Chem. Soc., Vol. 119, No. 47, 1997
Bauld et al.
Figure 6. Hammett-Brown plot for the Diels-Alder cycloadditions
of aryl vinyl sulfides to 1,3-cyclopentadiene in acetonitrile under
electrochemical conditions (log krel ) -0.04 - 2.67σ+ - 1.82(σ+)2; r2
) 0.999).
Figure 4. Hammett-Brown plot for the Diels-Alder cycloadditions
of aryl vinyl sulfides to 1,3-cyclopentadiene, catalyzed by 1•+ in
acetonitrile at 0 °C (log krel ) -0.24 - 2.42σ+; r2 ) 0.924).
Figure 7. Hammett plot for the Diels-Alder cycloadditions of aryl
vinyl sulfides to 1,3-cyclopentadiene in acetonitrile under electrochemi-
cal conditions (log krel ) 0.06 - 2.87σ - 1.33(σ)2; r2 ) 0.966).
Figure 5. Hammett plot for the Diels-Alder cycloadditions of aryl
vinyl sulfides to 1,3-cyclopentadiene cataltyzed by 1•+ in dichlo-
romethane at 0 °C. The line shown is for the regression analysis of the
meta points represented as squares (F ) 4.28, r2 ) 0.994). The data
for the para substituents is plotted both using σ values (circles) and σ+
values (triangles).
transition state resembling a loose complex. For comparison,
the F value for the equilibrium formation of these same aryl
vinyl sulfide cation radicals in acetonitrile was found to be -5.8.
On the other hand, F values for electrophilic additions to phenyl
vinyl sulfide have been found to be in the range -1.78 to
-3.0.15,16 Consequently an outer sphere mechanism for ioniza-
tion can be decisively ruled out. An inner sphere mechanism
involving only a weak covalent interaction would appear to be
similarly excluded. The data, however, are nicely consistent
with an initial electrophilic step, involving reaction of the
aminium salt electrophile either at the unsubstituted vinyl carbon
or at the nucleophilic sulfur atom (Scheme 5) or with a concerted
inner sphere mechanism involving very strong covalent interac-
tion between the two reactants. Further insights into the
specifics of the mechanism are provided by stereochemical data
to be discussed in a subsequent section. That the ionization
stage rather than the cycloaddition stage is rate determining is
strongly indicated by several lines of evidence. First, the value
of F is substantially smaller than the F value for the equilibrium
between the log krel Vs σ and σ+ correlations is provided by
statistical tests (F-tests) which indicate the distinction is
significant at or above the 95% confidence level. This point is
brought home visually in an especially clear way in Figure 5,
where only the points for meta substituted derivatives (for which
σ and σ+ are equal) are used to determine the trend line. In
the correlation with σ values (circles) the points for the para
derivatives are all observed to fall very close to this line.
However, in the correlation Vs σ+ (triangles), p-MeO, p-Me,
and p-Cl are all observed to deviate negatively from the line
defined by the meta points. The deviation is greatest for 4-MeO,
as it should be, but it is substantial for both 4-Me and 4-Cl.
Finally, p-CF3 deviates in a positive sense, as it should because
its σ+ value is more positive than its σ value.
Clearly, the step which determines which aryl vinyl sulfide
is preferentially ionized to the cation radical does not, in fact,
resemble an aryl vinyl sulfide cation radical. Further, the F
values observed (-4.21 in dichloromethane; -3.85 in aceto-
nitrile) are much too large to be consistent with a very early
(15) Shoemaker, D. P.; Garland, C. W.; Nibler, J. W. Experiments in
Physical Chemistry; McGraw-Hill: New York, 1989; pp 811-820.
(16) McClelland, R. A. Can. J. Chem. 1977, 55, 548-551.

Ionization of Aryl Vinyl Sulfides
J. Am. Chem. Soc., Vol. 119, No. 47, 1997 11385
Table 2. Relative Rates of Diels-Alder Cycloadditions of Aryl
Vinyl Sulfides to 1,3-Cyclopentadiene under Electrochemical
Oxidation Conditions
Scheme 5. Two Potential Modes of Electrophilic Addition of
Triarylaminium Ions to Aryl Vinyl Sulfides
substituent
log k/k0
substituent
log k/k0
4-Me
3-Me
H
0.614
0.120
0.000
4-Br
3-Cl
-0.501
-1.27
Scheme 6. The Reaction of Phenyl cis-2-Deuteriovinyl
Sulfide with 1,3-Cyclopentadiene
ionization. In the case where ionization is reversible and
cycloaddition is rate determining, the F value is considered likely
to be much closer to the equilibrium F value. In one such case,
the kinetic F value was found to be 83% of the equilibrium F
value.17 Secondly, added neutral triarylamine (1) is found not
to retard the absolute rate of the cycloaddition. If ionization
were reversible, the triarylamine should sharply depress the
rate.18 Finally, a doubling of the concentration of the diene
leaves the absolute reaction rate essentially unchanged.19
The Reaction under Photochemical Electron Transfer
Conditions. Photosensitized electron transfer is a well estab-
lished method for generating cation radicals and studying their
cycloadditions to neutral molecules.10 Acid is not generated
under these conditions, and electron transfer is of the outer
sphere type. When the subject reaction is studied using 1,4-
dicyanobenzene as the sensitizer and the singlet excited state
of this molecule is the single electron acceptor, the same Diels-
Alder cycloadducts (3a,b) are formed cleanly, and the endo:
exo ratio (2.9:1) is virtually identical to that found in the
aminium salt catalyzed reaction. This result provides strong
independent support for interpreting the latter reaction as a cation
radical/neutral cycloaddition.
The Reaction under Electrochemical Oxidation Condi-
tions. Anodic oxidation provides still another means of
generating substrate cation radicals and studying their cycload-
ditions.20 Using a reticulated vitreous carbon anode and
acetonitrile as the solvent, phenyl vinyl sulfide again adds
cleanly to 1,3-cyclopentadiene to yield 3a and 3b, with an endo:
exo ratio quite similar (2.2:1) to those observed in the chemically
and photochemically promoted reactions. By carrying these EC
oxidations out at a voltage (1.05 V) equal to the oxidation
potential of 1, it was possible to provide oxidizing power
comparable to that of 1•+. Since this voltage is less than the
peak potential of phenyl vinyl sulfide (1.4 Vs SCE) and grossly
less than that of 1,3-cyclopentadiene (1.8 Vs SCE), it can be
confidently assumed that essentially only phenyl vinyl sulfide
cation radicals are generated and give rise to the observed
adducts.
rate competitive with or faster than back electron transfer from
the anode and thus that anodic oxidation might provide for a
kinetically controlled oxidation step in the subject Diels-Alder
reactions under EC conditions. If these assumptions are valid,
competition kinetic studies under anodic oxidation conditions
could provide a contrasting example of outer sphere electron
transfer, which should lead to a correlation of log k/k0 with σ+.
The kinetic data are contained in Table 2, and the plots of log
k/k0 Vs σ and σ+ are illustrated in Figures 6 and 7 (Vs σ+ and
σ, respectively). It is noted that both plots exhibit curvature.
Analogous curvature has been observed in aminium salt
catalyzed cycloadditions where the ionization step is shifting
from kinetic to equilibrium control.17 Significantly, a quadratic
correlation of log k/k0 Vs σ+ is of excellent quality (r2 ) 0.999),
while the similar plot Vs σ is of lower quality (r2 ) 0.976).
These data are at least consistent with the scenario in which, to
the extent that oxidation is kinetically controlled, the mechanism
is of the outer sphere type.
Stereochemistry. Stereochemical studies appear to provide
further insight into the specifics of the ionization process.
Phenyl cis-2-deuteriovinyl sulfide was the substrate selected for
these studies (Scheme 6).21 When the reaction as catalyzed by
1•+ was carried out, using the stereospecifically deuterium
labeled substrate, the endo Diels-Alder adduct 3a was isolated
and found to be formed with 50% suprafacial stereospecificity
and 50% stereorandomization (corresponding to 75% of the
cis,endo adduct and 25% of the trans,endo adduct). The
recovered substrate, however, was nearly completely stereo-
randomized (55% cis, 45% trans). These results suggest that
the stereospecificity of the reaction is actually much higher than
50%, since the trans- deuterio substrate should be roughly
equally as reactive as the cis-compound, and some of the trans-
deuterio 3a must have arisen from this isomer. The existence
of even 50% of a stereospecific component would be surprising
if ionization proceeds exclusively Via electrophilic attack at
carbon, since the most reasonable expectation is that torsional
equilibration in intermediate 4 (Scheme 5) should be very fast
compared to homolytic dissociation. However, electrophilic
attack on sulfur to yield the vinylsulfonium ion 5 appears to
provide a plausible path for the stereospecific formation of the
phenyl cis-2-deuteriovinyl sulfide cation radical. Ample pre-
cedent for a stereospecific Diels-Alder addition step is avail-
Kinetics under EC Conditions. Electrochemical oxidations,
particularly those carried out at a chemically relatively inert
anode such as reticulated vitreous carbon, would appear highly
likely to occur Via an outer sphere mechanism, as opposed to
an electrophilic, inner shell process. Given the extremely rapid
rates of cycloaddition of cation radicals of substrates which have
an unsubstituted vinyl moiety, it appeared possible that cy-
cloadditions might, under at least some conditions, occur at a
(17) Hojo, M.; Masuda, R.; Kamitori, Y.; Okada, E. J. Org. Chem. 1991,
56, 1975-1976.
(18) Yueh, W.; Bauld, N. L. J. Chem. Soc., Perkin Trans. 2 1996, 1761-
1766.
(19) This is a consequence of the formation of neutral 1 as a product of
the ionization reaction: Lorenz, K. T.; Bauld, N. L. J. Am. Chem. Soc.
1987, 109, 1157-1160.
(20) This was verified for three different aryl vinyl sulfides (see
Experimental Section).17
(21) Yueh, W.; Sarker, H.; Bauld, N. L. J. Chem. Soc., Perkin Trans. 2
1995, 1715-1716.

11386 J. Am. Chem. Soc., Vol. 119, No. 47, 1997
Bauld et al.
able.22 A concerted inner sphere mechanism for the electron
transfer reaction is also tenable providing that the transition state
closely resembles 4 or 5, i.e., that a very strong covalent
interaction (polar in character) is developed. This possibility
is regarded as somewhat improbable, since it requires that a
transition state very closely resembling 4 or 5 does not actually
proceed to the full fledged intermediate. The mechanism for
the isomerization of the starting material is presently unknown.
The discovery that the mechanisms of ionization of stilbenes
and aryl vinyl sulfides by the aminium salt are quite different
is not at all surprising. Steric effects which would adversely
affect strong covalent attachment are much more severe in
stilbene than in a vinyl sulfide. Further, concerted ionization
of stilbene takes advantage of the conjugative effect of both
aryl rings, whereas an electrophilic mechanism effectively
utilizes only a single aryl ring. Finally, the presence of a
nucleophilic sulfur atom is considered to sharply enhance the
likelihood of an electrophilic attachment to aryl vinyl sulfides.
It should prove to be of considerable interest to determine
ionization mechanisms for a range of systems and to elucidate
the effects of structure variation upon the mechanism of
ionization.
tadiene (Eopx 1.8 V) was freshly prepared from dicylopentadiene,
stored at -20 °C, and used within 8 h.
General Procedure for the Synthesis of Substituted Aryl Vinyl
Sulfides (2). An aryl Grignard reagent, from the reaction of an aryl
bromide or iodide and excess Mg (concentration ∼ 1 M), was added,
Via cannula, to a solution of 2-chloroethyl thiocyanate24 in THF at 0
°C. After the addition was complete, the reaction was allowed to stir
at room temperature for 1 h, at which time the reaction was cooled
back to 0 °C, and a suspension of KOtBu (4- to 6-fold excess) in THF
was carefully added. The reaction was refluxed for 1-2 h and checked
by GC. If GC analysis indicated that the reaction was not complete,
then ∼2 more equiv of KOtBu were added, and reflux continued for
another hour. The reaction was cooled to room temperature and then
transferred to a separatory funnel containing 100 mL of 10% (NH4)2-
SO4, 100 mL of saturated NaCl, and 200 mL of pentane. The aqueous
layer was removed, and the organic layer was washed with 100 mL of
saturated NaCl. The organic layer was then dried over MgSO4, and
the solvent removed at reduced pressure on a rotary evaporator. The
residue was vacuum distilled at ∼5 Torr, allowing a small forerun of
material before collecting the product. Yields are reported for isolated
material of suitable purity (>95% by GC) for quantitative experiments.
In some cases, multiple distillations were necessary to meet this
requirement.
4-Methoxyphenyl Vinyl Sulfide (4-MeO). Using the above
procedure, the Grignard reagent from 4-bromoanisole (18.7 g) and Mg
(3.0 g) was added to 8.20 g of 2-chloroethyl thiocyanate in 20 mL of
THF. The elimination was carried out with 25 g of KOtBu in 100 mL
of THF. Distillation gave 6.621 g (59.1%) of product determined to
be >98% pure by GC: bp 115-120 °C at 5 Torr, 1H NMR δ 3.76 (s,
3H), 5.1 (m, 2H), 6.43 (dd, 1H), 6.85 (d, 2H), 7.31 (d, 2H); 13C NMR
δ 55.1, 112.4, 114.7, 123.3, 133.5, 133.9, 159.5; LRMS m/e 166 (M+),
151 (base), 135, 121, 107, 77; HRMS m/e calcd for C9H10OS 166.0452;
found 166.0453; Eopx 1.097 V.
Conclusions
Substituent effect studies of the electron transfer reactions
between aryl vinyl sulfide substrates and tris(4-bromophenyl)-
aminium hexachloroantimonate decisively rule out both an outer
sphere ET mechanism and an inner sphere mechanism which
involves only a small to moderate covalent interaction between
the aminium salt and the substrate. The preferred mechanism
involves rate determining electrophilic attack on the substrate,
followed by homolysis to the substrate cation radical, i.e., a
stepwise polar mechanism for ET.
4-Methylphenyl Vinyl Sulfide (4-Me). Using the above procedure,
100 mL of a 1 M solution of 4-methylphenyl magnesium bromide
(Aldrich) in diethyl ether was added to 8.10 g of 2-chloroethyl
thiocyanate in 20 mL of THF. The elimination was carried out with
15.4 g of KOtBu in 50 mL of THF and required additional KOtBu (10
g). Distillation gave 5.39 g (53.9%) of product determined to be >97%
Stereochemical studies indicate sulfur is a more likely site
of the electrophilic attack than the beta vinyl carbon.
A
1
concerted inner sphere mechanism having a Very strong (polar)
covalent interaction is also compatible with the kinetic and
stereochemical data but is considered somewhat less likely on
theoretical grounds.
pure by GC: bp 95-105 °C at 5 Torr, H NMR δ 2.3 (s, 3H), 5.2
(m,2H), 6.45 (dd, 1H), 7.07 (d, 2H), 7.25 (d, 2H); 13C NMR δ 21,
114.1, 129.8, 130, 131.2, 132.6, 137.3; LRMS m/e 150 (M+), 135 (base),
105, 91, 77; HRMS m/e calcd for C9H10S 150.0503; found 150.0498;
Eopx 1.267 V.
4-Phenylphenyl Vinyl Sulfide (4-Ph). Using the above procedure,
the Grignard reagent from 4-bromobiphenyl (18.567 g), and Mg (2.198
g) was added to 9.68 g of 2-chloroethyl thiocyanate in 20 mL of THF.
The elimination was carried out with 35 g of KOtBu in 100 mL of
THF and required additional KOtBu (10 g). Flash chromatography
on silica with hexane and recrystallization from hexane afforded 4.72
g (28%) of product determined to be >97% pure by GC: mp 53-55
Experimental Section
Analysis. Analytic gas chromatographic (GC) analyses were
performed on a Perkin-Elmer Model 8500 using a PE Nelson Model
1020 reporting integrator for data collection or a Varian Model 3700
using an HP 3390A integrator for data collection. Both instruments
were equipped with flame ionization detectors and DB-1 capillary
columns (J&W Scientific, 30 m × 0.25 mm on the PE and 15 M ×
0.25 mm on the Varian with film thickness of 1 mm), using helium as
the carrier gas. Naphthalene was used as internal standard for all the
quantiative analyses. In the aryl vinyl sulfide (AVS) series the
difference in response factors for phenyl vinyl sulfide and its cross
adduct with 1,3-cyclopentadiene was small, and this was assumed to
be the case for the entire series. The response factors calculated for
the substituted aryl vinyl sulfides were therefore used for the respective
cross adducts. Electrochemical measurements were performed using
a BAS 100 electrochemical analyzer in the differential pulse volta-
mentary (DPV) mode scanning in the range of 500 to 1900 mV at a
scan rate of 4 mV/s with a pulse amplitude of 50 mV, a pulse width of
50 ms, a pulse period of 1000 ms, and a sensitivity of 1 × 10-6. The
Eopx measurements Vs Ag/Ag+ were converted to Vs SCE by adding 0.3
V to each value.
1
°C, H NMR δ 5.36 (m, 2H), 6.5 (dd, 1H), 7.37 (m, 5H), 7.56 (m,
4H); 13C NMR δ 115.7, 126.9, 127.4, 127.7, 128.8, 130.7, 131.7, 133.3,
140; LRMS m/e 212 (M+, base), 197, 185, 176, 167, 152, 135, 115,
77; HRMS m/e calcd for C14H12S 212.0656; found 212.0651; Eox
p
1.249 V.
3-Methylphenyl Vinyl Sulfide (3-Me). Using the above procedure,
the Grignard reagent from 3-bromotoluene (17.122 g) and Mg (4.0 g)
was added to 8.10 g of 2-chloroethyl thiocyanate in 20 mL of THF.
The elimination was carried out with 25 g of KOtBu in 70 mL of THF.
Distillation gave 10.864 g (72%) of product determined to be >98%
pure by GC: bp 93-94 °C at 5 Torr, 1H NMR δ 2.32 (s, 3H), 5.3 (m,
2H), 6.5 (dd, 1H), 7.0 (m, 1H), 7.16 (m, 3H); 13C NMR δ 21.2, 115.2,
127.4, 127.9, 128.9, 131, 132, 133.9, 138.9; LRMS m/e 150 (M+), 135
(base), 105, 91, 77; HRMS m/e calcd for C9H10S 150.0503; found
150.0497; Eopx 1.312 V.
Reagents. The substituted aryl vinyl sulfides were prepared by
modification of a literature procedure23 with the exception of phenyl
vinyl sulfide (Aldrich) and 4-bromophenyl vinyl sulfide.23 Cylopen-
(24) Kim, T.; Pye, J.; Bauld, N. L. J. Am. Chem. Soc. 1990, 112, 6285.
(25) Verboom, W.; Meuer, J.; Brandsma, L. Synthesis 1978, 577.
(26) Parham, W. E.; Herberling, J. J. Am. Chem. Soc. 1955, 77, 1175.
(27) (a) Parham, W. E.; Stright, P. L. J. Am. Chem. Soc. 1956, 81, 4783.
(b) Parham, W. E.; Motter, R. F.; Mayo, G. L. J. Am. Chem. Soc. 1959 81,
3386.
(22) Hojo, M.; Masuda, R.; Takagi, S. Synthesis 1978, 4, 284.
(23) Bellville, D. J.; Wirth, D. D.; Bauld, N. L. J. Am. Chem. Soc. 1981,
103, 718-720.

Ionization of Aryl Vinyl Sulfides
J. Am. Chem. Soc., Vol. 119, No. 47, 1997 11387
4-Chlorophenyl Vinyl Sulfide (4-Cl). Using the above procedure,
the Grignard reagent from 4-chloroiodobenzene (9.65 g) and Mg (2.3
g) was added to 3.21 g of 2-chloroethyl thiocyanate in 20 mL of THF.
The elimination was carried out with 5.29 g of KOtBu in 25 mL of
THF. Distillation gave 1.01 g (22.4%) of product determined to be
elution with hexane and CH2Cl2), and it was found that the products
were not separated from 1. Next, column chromatography using silica
gel (60 g, gradient elution with 500 mL of hexane INS followed by
250 mL of 9:1 hexane and ethyl acetate) was performed and gave 0.024
g (5.6%) of the exo isomer (90% pure by GC) and 0.103 g (24%) of
the endo isomer (95% pure by GC). The isolated reaction products
1
>95% pure by GC: bp 105-110 °C at 5 Torr, H NMR δ 5.38 (m,
2H), 6.45 (dd, 1H), 7.28 (s, 4H); 13C NMR δ 116.2, 129.2, 131.2, 131.6,
132.4, 132.6; LRMS m/e 170 (M+), 135 (base), 108, 91, 75; HRMS
were characterized using TLC, GC, GC/MS, HRMS, H NMR, 13C
1
1
NMR and for the endo isomer 2D H-1H and 13C-1H NMR were
m/e calcd for C8H6ClS (M-H) 169.9957; found 169.9958; Eox 1.392
collected to make a full spectral assignment: Rf(exo) ) 0.5 and Rf-
p
1
(endo) ) 0.38 on silica using pentane; Exo H NMR (250.1 Mhz) δ
V.
1.5 (m, 2H), 1.75 (m, 2H), 2.85 (s, 1H), 2.93 (s, 1H), 3.08 (m, 1H),
6.06 (dd, 1H), 6.18 (dd, 1H), 7.15 (m, 1H), 7.2-7.4 (m, 4H); 13C NMR
(62.9 MHz) δ 33.8, 41.7, 44.4, 45.9, 47.3, 125.6, 128.7, 129,1, 134.7,
137.8, 138.1; LRMS m/e 202 (M+), 136 (base), 109, 91, 65. HRMS
m/e calcd for C13H14S 202.0816; found 202.08222; Endo 1H NMR
(500.1 MHz) δ 0.9 (m, 1Ha), 1.25 (d, 1Hb), 1.48 (m, 1Hc), 2.2 (m,
1Hd), 2.85 (br s, 1 He), 3.02 (br s, 1Hf), 3.6 (m, 1Hg), 6.05 (dd, 1Hh),
6.2 (dd, 1Hi), 7.08 (m, 1Hj), 7.19 (m, 2Hk), 7.28 (m, 2Hl); 13C NMR
(125.8 Mhz) δ (a) 34.5, (b) 52.6, (c) 45.7, (d) 46.5, (e) 48.8, (f) 125.7,
(g) 128.7, (h) 129.5, (i) 132.9, (j) 137.4, (k) 137.7; LRMS m/e 202
(M+), 136 (base), 109, 91, 65; HRMS m/e calcd for C13H14S 202.0816;
found 202.08221.
3-Chlorophenyl Vinyl Sulfide (3-Cl). Using the above procedure,
the Grignard reagent from 3-chloroiodobenzene (14.507 g) and Mg
(2.27 g) was added to 4.94 g of 2-chloroethyl thiocyanate in 20 mL of
THF. The elimination was carried out with 9.12 g of KOtBu in 25
mL of THF. Distillation gave 4.098 g (58.7%) of product determined
1
to be >96% pure by GC: bp 105-110 °C at 5 Torr; H NMR δ 5.4
(m, 2H), 6.45 (dd, 1H), 7.18 (s, 3H), 7.33 (s, 1H); 13C NMR δ 117.5,
126.9, 127.8, 129.3, 130, 130.4, 135.5, 136.7; LRMS m/e 170 (M+),
135 (base), 108, 91, 75; HRMS m/e calcd for C8H8ClS (M + H)
171.0035; found 171.0034; Eopx 1.467 V.
4-Trifluoromethylphenyl Vinyl Sulfide (4-CF3). Using the above
procedure, the Grignard reagent from 4-bromotrifluoromethylbenzene
(11.123 g) and Mg (4.07 g) was added to 6.643 g of 2-chloroethyl
thiocyanate in 20 mL of THF. The elimination was carried out with
31 g of KOtBu in 100 mL of THF. Distillation gave 6.034 g (59.9%)
of product determined to be >97% pure by GC: bp 80-83 °C at 5
1
Torr, H NMR δ 5.52 (m, 2H), 6.51 (dd, 1H), 7.38 (d, 2H), 7.55 (d,
2H); 13C NMR δ 119, 125.8, 125.9, 128.7, 129.5, 140.1; LRMS m/e
204 (M+), 183, 159, 135 (base), 91, 69; HRMS m/e calcd for C9H7F3S
24.04221; found 1204.0215 Eopx 1.604 V.
3,5-Dichlorophenyl Vinyl Sulfide (3,5-DCl). Using the above
procedure, the Grignard reagent from 1-bromo-3,5-dichlorobenzene
(22.62 g) and Mg (3.25 g) was added to 11.78 g of 2-chloroethyl
thiocyanate in 20 mL of THF. The elimination was carried out with
43.79 g of KOtBu in 100 mL of THF. Distillation gave 5.427 g (26.5%)
of product determined to be >95% pure by GC: bp 110-115 °C at 3
General Procedure for the Synthesis of Substituted Phenylth-
ionorbornenes. To 0.2 mmol of substituted aryl vinyl sulfide in 1
mL of CH2Cl2 at 0 °C was added a 5-fold excess of 1,3-cyclo-pentadiene
1
Torr, H NMR δ 5.5 (m, 2H), 6.48 (dd, 1H), 7.18 (s, 3H); 13C NMR
δ 119.6, 126.6, 126.9, 129.1, 135.3, 138.5; LRMS m/e 204 (M+), 169,
134 (base); HRMS m/e calcd for C8H5,Cl2S (M-H) 203.9567 found
203.9565; Eopx 1.612 V.
followed by 10% of 1•+
. The reaction was stirred for 2 min and
quenched with 0.5 mL of saturated metanolic potassium carbonate. The
quenched reaction mixture was diluted with 5 mL of pentane and
washed once with 5 mL of water, and the organic layer dried over
MgSO4, followed by solvent removal at reduced pressure on a rotary
evaporator. The percent conversion was determined by GC, and the
product identity was confirmed by GC/MS. The endo/exo ratio
determined by GC was approximately 3:1 in all cases.
5-(4-Methoxyphenylthio)norbornene. Conversion 35.9%: LRMS
m/e 232 (M+), 166 (base).
5-(4-Methylphenylthio)norbornene. Conversion 38.4%: LRMS
m/e 216 (M+), 150 (base).
5-(4-Phenylphenylthio)norbornene. Conversion 56.8%: LRMS
m/e 278 (M+), 212 (base).
5-(3-Methylphenylthio)norbornene. Conversion 28.8%: LRMS
m/e 216 (M+), 150, 135 (base).
5-(4-Bromophenylthio)norbornene. Conversion 6.9%: LRMS m/e
282 (M+), 216, 135 (base).
5-(3-Chlorophenylthio)norbornene. Conversion 5.7%: LRMS m/e
236 (M+), 170, 135 (base).
5-(4-Trifluoromethylphenylthio)norbornene. Conversion 2.0%:
LRMS m/e 270 (M+), 204, 135 (base).
4-(Methylthio)phenyl Vinyl Sulfide (4-MeS). First, 9.62 g of
thioanisole (77.4 mmol) in 60 mL of pentane was converted to
4-bromothioanisole by the slow addition of a solution of 12.4 g (mmol)
of Br2 in 10 mL of pentane at room temperature. After 30 min, the
solvent was removed, and the crude product was adsorbed onto 20 g
of basic alumina and placed on top of 50 g of basic alumina. The
product was flashed off the column using 9:1 hexane and EtOAc. After
solvent removal, 14.3 g (87.4%) of 4-bromothioanisole was isolated.
GC analysis of the crude product showed it was 93% pure, and it was
used without further purification: LRMS 204 (M+), 202 (M+, base),
158, 156, 108.
Then, using the above procedure, the Grignard reagent from
4-bromothioanisole (14.3 g) and Mg (4.24 g) was added to 8.13 g of
2-chloroethyl thiocyanate in 20 mL of THF. The elimination was
carried out with 35 g of KOtBu in 100 mL of THF. Distillation gave
6.62 g (55.6%) of product determined to be >92.5% pure by GC: bp
1
100-110 °C at 1.5 Torr, H NMR δ 2.4 (s, 3H), 5.25 (m, 2H), 6.45
(dd, 1H), 7.11 (d, 2H), 7.25 (d, 2H); 13C NMR δ 15.8, 114.9, 127.2,
130.1, 131.7, 132.3, 138.3; LRMS m/e 182 (M+), 167, 135 (base), 123,
108, 91; HRMS m/e calcd for C9H11S2 (M+H) 183.0302; found
183.0296; Eopx 1.104 V.
5-3,5-Dichlorophenylthio)norbornene. Conversion 1.9%: LRMS
m/e 270, 204 (base), 169, 134.
Synthesis of endo- and exo-5-(phenylthio)norbornene (3a,b). A
solution of phenyl vinyl sulfide (0.288 g, 2.1 mmol) and 1,3-
cyclopentadiene (1.495 g 29.5 mmol) in 1.0 mL of CH2Cl2 was added
to 0.505 g (30%) of 1•+ in 4 mL of CH2Cl2 at 0 °C. After 10 min, 1
mL of saturated methanolic potassium carbonate was added, and the
quenched reaction mixture was transferred to a separatory funnel
containing CH2Cl2 (50 mL) and washed with water (3 × 50 mL). The
organic phase was dried with MgSO4 and solvent removed at reduced
pressure on a rotary evaporator. GC analysis of the crude product
mixture showed an endo/exo ratio of 2.8:1. Column chromatography
of the resulting residue was performed using neutral alumina (gradient
Reaction of Phenyl Vinyl Sulfide with 1,3-Cyclopentadiene under
Photosensitized Electron Transfer Conditions. To a dry Pyrex test
tube was added 0.11 g (mol) of phenyl vinyl sulfide, 0.02 g (%) of
1,4-dicyanobenzene, 0.44 g (mol) of 1,3-cyclopentadiene, and 20 mL
of CH3CN. The test tube was placed inside a uranium filter in a 20
°C water bath and irradiated with a medium pressure mercury lamp.
The progress of the reaction was monitored by GC. After 3 h, GC
analysis showed 3% of the Diels-Alder cross adduct with an endo/
exo ratio of 3:1. After 24 h, GC analysis showed 20% of the Diels-

11388 J. Am. Chem. Soc., Vol. 119, No. 47, 1997
Bauld et al.
Alder cross adduct with an endo/exo ratio of 2.9:1. The amount of
conversion and the endo/exo ratio was found to be the same at 48 h.
Reaction of Phenyl Vinyl Sulfide with 1,3-Cyclopentadiene using
Triflic Acid in Dichloromethane. To 0.064 g (mol) of phenyl vinyl
sulfide and 0.219 g (mol) of 1,3-cyclopentadiene at 0 °C was added 3
drops (20-30%) of triflic acid. After 5 min, the reaction was quenched
with 5 mL of saturated K2CO3 in MeOH. GC analysis showed two
peaks (1:7 ratio) at retention times slightly longer than those of the
Diels-Alder cross adducts. GC/MS analysis showed that both acid
cross adducts had the same molecular weight and that, unlike the Diels-
Alder cross adducts, the molecular ion was the base peak in the mass
spectrum: LRMS 202 (M+, base), 187, 179, 161, 147, 134, 115, 91,
77.
<10% conversion in 1 min were found. The reactions were quenched
with 0.3 mL of saturated methanolic potassium carbonate. Then 2 mL
of CH2Cl2 was added, followed by washing twice with 5 mL portions
of water and drying over MgSO4. In all cases the results were the
average of four runs analyzed twice each. The calculated mass balances
for all relative rate analyses were comonly over 95% but in a few
instances were as low as 85%. Response factors for all relative rate
analyses were determined by dilution of a portion of the unused aryl
vinyl sulfide stock solution: log krel 4-MeO 1.03, 4-Me 0.922, 3-Me
0.45, 4-Br 0.79, 3-Cl 1.48, 4-CF3 2.02.
Effect of the Concentration of 1,3-Cyclopentadiene upon the
Relative Rate Ratio of 4-Methoxyphenyl Vinyl Sulfide Ws Meth-
ylphenyl Vinyl Sulfide. A stock solution was prepared in a dry 10
mL volumetric flask by adding a known amount of naphthalene
followed by equimolar amounts of 4-methoxyphenyl vinyl sulfide and
4-methylphenyl vinyl sulfide. The flask was then filled to the mark
Reaction of Phenyl Vinyl Sulfide with 1,3-Cyclopentadiene using
1•+ in the Presence of 2,6-Di-tert-butyl-4-methylpyridine (DTBMP)
in Dichloromethane. To 0.051 g (0.375 mmol) of phenyl vinyl sulfide,
0.168 g (mol) of 1,3-cyclopentadiene, and 0.053 g (0.259 mmol) of
with precooled CH
2Cl2. Four stock solutions of 1,3-cyclopentadiene
DTBMP in 10 mL of CH2Cl2 at 0 °C was added 0.122 g (40%) of 1•+
.
were prepared in a similar manner with the following excesses based
on the sum of the molar amounts of the two sulfides: 5, 2.4, 0.54, and
0.2. Aliquots (1 mL) of each stock solution were added to a vial
followed by a solution of 1•+ to give a conversion of less than 10% in
the more reactive compound. Each relative excess was measured in
duplicate. k(4-MeO)/k(4-CH3) for 5× ) 2.76, 2.4× ) 2.74, 0.54× )
2.67, 0.2× ) 2.78; average ) 2.73.
After 15 min the reaction was quenched with 5 mL of saturated K2-
CO3 in MeOH. GC analysis indicated that the reaction had gone to
15% conversion, and the endo/exo ratio was 2.8:1. There were no peaks
observed for the acid cross adducts.
Synthesis of cis-Phenyl Vinyl Sulfide-â-d1. First, 12.9 g (0.133
mol) of cis-1,2-dichloroethene was converted to cis-1,2-bis(phenylthio)-
ethene (BPTE) by treatment with 25.8 g (0.234 mol) of thiophenol
and 20 g (0.357 mol) of KOH in 350 mL of refluxing EtOH for 4 h.27
Recrystallization from hexane gave 19.7 g (57.4%) BPTE which was
General Procedure for the Competitive Cycloadditions of Sub-
stituted Aryl Vinyl Sulfides with 1,3-Cyclopentadiene under Elec-
trochemical Conditions. A stock solution was prepared in a dry 10
mL volumetric flask by adding a known amount of naphthalene,
equimolar amounts (0.6-1.0 mmol) of two aryl vinyl sulfides that were
of similar reactivity, and a 5-fold excess (based on the sum of the molar
amounts of the aryl vinyl sulfides) of 1,3-cyclopentadiene. Then a
0.1 M solution of lithium perchlorate in acetonitrile was added to the
mark. A 1 mL aliquot of the stock solution was added to 9 mL of
electrolyte solution (0.1 M lithium perchlorate in acetonitrile). The
reaction was run at a constant potential of 0.8 V (Vs Ag/Ag+) using a
divided cell equipped with a reticulated vitreous carbon working
electrode (anode), a reticulated vitreous carbon counter electrode, and
a Ag/Ag+ reference electrode (calibrated against ferrocene/ferrocene+).
After passing 1 C of current, the reaction mixture was transferred to a
test tube followed by the addition of 5 mL of H2O and 3 mL of hexane.
The test tube was vigorously shaken, and the layers allowed to separate.
GC analysis was carried out on the hexane layer, and the standard was
treated in a similar way. In all cases the results were the average of
three runs analyzed twice each, and the recovery was >95%. The
average endo/exo ratio was 2.2:1:log krel 4-Me 0.614, 3-Me 0.12, 4-Br
0.501, 3-Cl 1.27.
1
found to be 99% pure by GC: mp 31-32.5 °C; H NMR δ 6.5 (s,
2H), 7.1-7.5 (m, 10H); 13C NMR δ 124.8, 126,8, 129.0, 129.3, 123.1;
Eopx 1.1 38 V.
Next, 8.72 (35.5 mmol) of BPTE in 10 mL of dry Et2O was
converted to phenylthioacetylene (PTA) by slow addition (∼2 min) to
70 mmol of n-butyllithium (7 mL of 10 M solution in hexanes) in 20
mL of dry Et2O at -10 °C.26 Distillation gave 4.0 g of PTA: bp 78-
80 °C at 4 Torr; 1H NMR δ 3.25 (s, 1H), 7.2-7.5 (m, 5H); 13C NMR
δ 71, 86.9, 126.5, 126.7, 129.2, 131.4; LRMS m/e 134 (M+, base),
109, 92, 91, 77.
Finally, the above PTA was dissolved in 5 mL of dry Et2O and
added dropwise to 20 mL of a 1 M solution of LAD at -10 °C.
Following the addition, the reaction was allowed to warm to room
temperature and worked up. Distillation of the crude product afforded
1
2 mL of cis-phenyl vinyl sulfide-â-d1:21 bp 55-56 °C at 2 Torr; H
NMR δ 5.3 (d, 1H), 6.5 (dt, 1H), 7.2-7.4 (m, 5H); LRMS m/e 137
(M+), 109 (base), 91.
Test for Stereospecificity in the Reaction of Phenyl Vinyl Sulfide
with 1,3-Cyclopentadiene Catalyzed by 1•+. cis-Phenyl vinyl sulfide-
â-d1 was reacted with 1,3-cyclopentadiene, and the starting material
and products isolated using the procedure reported for the synthesis
and isolation of exo and endo 5-phenylthionorbornene. The recovered
starting material was found to be ∼45% isomerized by 1H NMR. The
endo isomer (90% pure by GC) was found to be 25% isomerized: 1H
NMR δ 0.9 (br d, 0.25H), 1.3 (d, 1H), 1.55 (m, 1H), 2.4 (br d, 0.75H),
2.9 (br s, 1H), 3.1 (br s, 1H), 3.7 (dd, 1H), 6.1 (m, 1H), 6.3 (m, 1H),
7.1-.4 (m, 5H); 13C NMR δ 34 (t), 42.5, 45.6, 4.4, 48.8, 125.7, 128.7,
129.5, 133, 137.4; LRMS m/e 203 (M+), 137 (base), 109, 92; HRMS
m/e calcd for C13H142HS (M + H) 204.0957; found 204.0959.
Stability of the Deuterated Endo Cross Adduct in the presence
of 1•+ in Dichloromethane. The above deuterated endo cross adduct
(0.025 g) was dissolved in 1 mL of CH2Cl2 at 0 °C, and 0.012 g (12%)
of 1•+ was added. After 2 min, the reaction was quenched and analyzed
by GC. It was found that no isomerization had taken place.
General Procedure for the Competitive Cycloadditions of Sub-
stituted Aryl Vinyl Sulfides with 1,3-Cyclopentadiene Catalyzed
by 1•+ in Dichloromethane. A stock solution was prepared in a dry
10 mL volumetric flask by adding a known amount of naphthalene,
equimolar amounts (0.6-1.0 mmol) of two aryl vinyl sulfides that were
of similar reactivity, and a 5-fold excess (based on the sum of the molar
amounts of the aryl vinyl sulfides) of 1,3-cyclopentadiene. Then CH2-
Cl2, precooled to 0 °C, was added to the mark. A similar stock solution
of 1•+ was prepared, and trial runs were performed by adding small
amounts (0.2-1.0 mL, 0.1-10% 1•+) of this stock solution to 1 mL
aliquots of the aryl vinyl sulfide solution until conditions that gave
General Procedure for the Competitive Cycloadditions of Sub-
stituted Aryl Vinyl Sulfides with 1,3-Cyclopentadiene Catalyzed
by 1•+ in Acetonitrile. A stock solution was prepared in a dry 10 mL
volumetric flask by adding a known amount of naphthalene, equimolar
amounts (0.6-1.0 mmol) of two aryl vinyl sulfides that were of similar
reactivity, and a 5-fold excess (based on the sum of the molar amounts
of the aryl vinyl sulfides) of 1,3-cyclopentadiene. Then acetonitrile,
precooled to 0 °C, was added to the mark. A similar stock solution of
1•+ was prepared, and trial runs were performed by adding small
amounts (0.2-1.0 mL, 0.1-10% 1•+) of this stock solution to 1 mL
aliquots of the aryl vinyl sulfide stock solution until conditions that
gave <10% conversion in 1 min were found. The reactions were
quenched with 0.3 mL of saturated methanolic potassium carbonate.
Then 2 mL of CH2Cl2 was added, followed by washing twice with 5
mL portions of water and drying over MgSO4. In all cases the results
were the average of three runs analyzed twice each, and the recovery
was >95%. The average endo/exo ratio was 2.7:1: log krel 4-MeO
1.21, 4-Me 0.815, 3-Me 0.327, 4-Br 0.652, 3-Cl 1.34, 3,5-DCl 1.95.
General Procedure for the Competitive Cycloadditions of Sub-
stituted Aryl Vinyl Sulfides with 1,3-Cyclopentadiene Catalyzed
by 1•+ in Acetonitrile Containing 0.1 M Lithium Perchlorate.
Equimolar amounts of two aryl vinyl sulfides were weighed into a dry
vial along with 0.25 g of 1,3-cyclopentadiene and 5 mL of a 0.1 M
solution of lithium perchlorate in acetonitrile. The competitive
experiments were conducted in a manner similar to the previous aryl
vinyl sulfide experiments except that the results are not calibrated, and

Ionization of Aryl Vinyl Sulfides
J. Am. Chem. Soc., Vol. 119, No. 47, 1997 11389
two runs analyzed twice each were averaged to obtain each data point.
The average endo/exo ratio was 2.5:1: log krel 4-MeO 0.96, 4-Me 0.64,
3-Me 0.26, 4-Br 0.64, 3-Cl 1.3.
2 min, the reactions were quenched with 0.3 mL of saturated methanolic
potassium carbonate. Then 2 mL of CH2Cl2 was added, followed by
washing twice with 5 mL portions of water and drying over MgSO4.
No change in the rates or selectivity was observed.
Effect of the Concentration of 1,3-Cyclopentadiene upon the
Absolute Reaction Rates of Cycloaddition. Single point measure-
ments were obtained for identical reactions run for 1 min at two
concentrations of 1,3-cyclopentadiene using three different amounts
of 1•+. Doubling the concentration of the diene gave the following
results: 4-Me 0.89, H 1.05, 3-Cl 0.80.
Effect of Added Triarylamine on the Absolute Reaction Rates
of Cycloaddition and the Selectivity of the Cycloaddition of
Substituted Aryl Vinyl Sulfides with 1,3-Cyclopentadiene Catalyzed
by 1•+ in Dichloromethane. A stock solution was prepared in a dry
10 mL volumetric flask by adding a known amount of naphthalene,
equimolar amounts (0.6-1.0 mmol) of two aryl vinyl sulfides that were
of similar reactivity, and a 5-fold excess (based on the sum of the molar
amounts of the aryl vinyl sulfides) of 1,3-cyclopentadiene. Then CH2-
Cl2, precooled to 0 °C, was added to the mark. A similar stock solution
of 1•+ was prepared. Aliquots (1 mL) of the stock solution were added
to a series of vials containing 1 (0, 25, 50, 100, and 200 mol %)
followed by the addition of equal volumes of the 1•+ solution. After
Acknowledgment. The authors thank the Robert A. Welch
Foundation (F-149) for support of this research.
JA9706965