Lifetimes and UV-visible absorption spectra of benzyl, phenethyl, and cumyl carbocations and corresponding vinyl cations. A laser flash photolysis study

Article abstract of DOI:10.1139/v99-210

Benzyl (4-MeO, 4-Me, and 4-methoxy-1-naphthylmethyl), phenethyl (4- Me₂N, 4-MeO, 3,4-(MeO)₂, 4-Me, 3-Me, 4-F, 3-MeO, 2,6-Me₂, parent, and 4- methoxy-1-naphthylethyl) and cumyl (4-Me₂N, 4-MeO, 4-Me, parent) cations have been studied by laser flash photolysis (LFP) in 2,2,2-trifluoroethanol (TFE) and 1,1,1,3,3,3-hexafluoroisopropanol (HFIP). In most cases styrene or α-methylstyrene precursors were employed for the phenethyl and cumyl ions, the intermediate being obtained by solvent protonation of the excited state. Benzyl cations were generated by photoheterolysis of trimethylammonium and chloride precursors. While a 4-MeO substituent provides sufficient stabilization to permit observation of cations in TFE, cations with less stabilizing substituents usually require the less nucleophilic HFIP. Even in this solvent, the parent benzyl cation is too short-lived (lifetime <20 ns) to be observed. When generated in HFIP, phenethyl cations can be seen to react with unphotolyzed styrene, giving rise to dimer cations that are observed to grow in as the initial phenethyl cation decays. The dimer cations, in common with the oligomer cations seen in cationic styrene polymerization, have a κ(max) 15-20 nm higher than the monomer and react with both solvent and styrene several orders of magnitude more slowly. This stabilization relative to the phenethyl may reflect an interaction with the aryl group present at the λ-carbon. Cations 4-MeOC₆H₄C⁺(R)-CH₃ (R = Me, Et, i-Pr, t-Bu, cyclopropyl, C₆H₅, 4-MeOC₆H₄) were generated in TFE via the photoprotonation route. The alkyl series shows that steric effects are important in the decay reaction. The cation with R = cyclopropyl is a factor of 1.5 less reactive than the cation where R = phenyl. Several vinyl cations have also been generated by photoprotonation of phenylacetylenes. ArC⁺=CH₂ has a reactivity very similar to that of its analog ArC⁺H-CH₃, the vinyl cation being slightly (factors of 2-5) shorter-lived. For the various series of cations, including vinyl, substituents in the aryl ring have a consistent effect on the κ(max), a shift to higher wavelength relative to hydrogen of 15 nm for 4-Me, 30 nm for 4-MeO, and 50 nm for 4-Me₂N.

Full text of DOI:10.1139/v99-210

2069
Lifetimes and UV-visible absorption spectra of
benzyl, phenethyl, and cumyl carbocations and
corresponding vinyl cations. A laser flash
photolysis study
Frances L. Cozens, V.M. Kanagasabapathy, Robert A. McClelland,
and Steen Steenken
Abstract: Benzyl (4-MeO, 4-Me, and 4-methoxy-1-naphthylmethyl), phenethyl (4-Me2N, 4-MeO, 3,4-(MeO)2, 4-Me,
3-Me, 4-F, 3-MeO, 2,6-Me2, parent, and 4-methoxy-1-naphthylethyl) and cumyl (4-Me2N, 4-MeO, 4-Me, parent)
cations have been studied by laser flash photolysis (LFP) in 2,2,2-trifluoroethanol (TFE) and 1,1,1,3,3,3-
hexafluoroisopropanol (HFIP). In most cases styrene or α-methylstyrene precursors were employed for the phenethyl
and cumyl ions, the intermediate being obtained by solvent protonation of the excited state. Benzyl cations were
generated by photoheterolysis of trimethylammonium and chloride precursors. While a 4-MeO substituent provides
sufficient stabilization to permit observation of cations in TFE, cations with less stabilizing substituents usually require
the less nucleophilic HFIP. Even in this solvent, the parent benzyl cation is too short-lived (lifetime <20 ns) to be
observed. When generated in HFIP, phenethyl cations can be seen to react with unphotolyzed styrene, giving rise to
dimer cations that are observed to grow in as the initial phenethyl cation decays. The dimer cations, in common with
the oligomer cations seen in cationic styrene polymerization, have a λmax 15–20 nm higher than the monomer and react
with both solvent and styrene several orders of magnitude more slowly. This stabilization relative to the phenethyl may
reflect an interaction with the aryl group present at the γ-carbon. Cations 4-MeOC6H4C+(R)-CH3 (R = Me, Et, i-Pr,
t-Bu, cyclopropyl, C6H5, 4-MeOC6H4) were generated in TFE via the photoprotonation route. The alkyl series shows
that steric effects are important in the decay reaction. The cation with R = cyclopropyl is a factor of 1.5 less reactive
than the cation where R = phenyl. Several vinyl cations have also been generated by photoprotonation of
phenylacetylenes. ArC+=CH2 has a reactivity very similar to that of its analog ArC+H-CH3, the vinyl cation being
slightly (factors of 2–5) shorter-lived. For the various series of cations, including vinyl, substituents in the aryl ring
have a consistent effect on the λmax, a shift to higher wavelength relative to hydrogen of 15 nm for 4-Me, 30 nm for
4-MeO, and 50 nm for 4-Me2N.
Key words: photogenerated carbocations, carbocation lifetime, styrene, photoprotonation. 2082
Résumé : Faisant appel à la photolyse éclair au laser (« LFP »), on a étudié les cations de type benzyles (4-MeO et 4-
Me ainsi que 4-méthoxy-1-naphtylméthyle), phénéthyle (4-Me2N, 4-MeO, 3,4-(MeO)2, 4-Me, 3-Me, 4-F, 3-MeO, 2,6-
Me2 et le fondamental ainsi que le 4-méthoxy-1-naphtyléthyle) et cumyle (4-Me2N, 4-MeO, 4-Me et le fondamental)
dans le 2,2,2-trifluoroéthanol (TFE) et le 1,1,1,3,3,3-hexafluoroisopropanol (« HFIP »). Dans la plupart des cas, des
précurseurs de type styrène ou α-méthylstyrène ont été utilisés pour les ions de type phénéthyle et cumyle, obtenus par
solvatation de l’état excité à l’aide d’un solvant. Les cations de type benzylique ont été générés par photohétérolyse de
précurseurs triméthylammonium et chlorure. Un substituant 4-MeO fournit assez de stabilisation pour permettre
d’observer les cations dans le TFE; toutefois, pour observer des cations portant des substituants moins stabilisants, il
faut utiliser le « HFIP » qui est moins nucléophile. Même dans ce solvant, le temps de vie du cation benzyle
fondamental est trop court (<20 ns) pour qu’on puisse l’observer. Quand on les génère dans le « HFIP », on peut
observer la réaction des cations phénéthyles avec le styrène qui n’a pas réagi; cette réaction donne lieu à la formation
de cations dimères dont on observe la croissance avec la décroissance du cation phénéthyle initial. Les cations dimères,
comme les cations oligomères observés lors de la polymérisation cationique du styrène, ont un λmax de 15 à 20 nm
plus élevé que celui du monomère et leurs vitesses de réaction avec le solvant et le styrène sont de plusieurs ordres de
grandeur plus faibles. La stabilisation relative du phénéthyle peut être un reflet d’une interaction avec le groupe aryle
Received April 19, 1999.
F.L. Cozens, V.M. Kanagasabapathy, and R.A. McClelland.1 Department of Chemistry, University of Toronto, Toronto, ON
M5S 3H6, Canada.
S. Steenken. Max-Planck-Institut fur Strahlenchemie, D-45413 Mulheim, Germany.
1Author to whom correspondence may be addressed. Telephone: (416) 978-3592. Fax: (416) 978-3592.
e-mail: rmcclell@alchemy.chem.utoronto.ca
Can. J. Chem. 77: 2069–2082 (1999)
© 1999 NRC Canada

2070
Can. J. Chem. Vol. 77, 1999
présent sur le carbone γ. On a généré les cations 4-MeOC6H4C+(R)-CH3 (R = Me, Et, i-Pr, t-Bu, cyclopropyle, C6H5,
4-MeOC6H4) dans le TFE par le biais d’une réaction de photoprotonation. Les séries alkyles montrent que les effets
stériques sont importants dans la réaction de désactivation. La réactivité du cation avec R = cyclopropyle est 1,5 fois
moins grande que celle du cation dans lequel R = phényle. On a aussi généré plusieurs cations vinyliques par
photoprotonation de phénylacétylènes. La réactivité d’ion ARC+=CH2 est très semblable à celle de son analogue
ArC+H-CH3, le temps de vie du cation vinylique étant légèrement plus court (de 2 à 5 fois plus court). Pour les
diverses séries de cations, y compris le cation vinyle, les substituants du cycle aromatique apportent un effet régulier
sur le λmax, soit des déplacements d’environ 15 nm (pour le 4-Me), 30 nm (pour le 4-MeO) et de 50 nm (pour le 4-
Me2N) vers une longueur d’onde plus grande par rapport au produit fondamental, portant un hydrogène.
Mots clés : carbocations photogénérés, temps de vie d’un carbocation, styrène, photoprotonation.
[Traduit par la Rédaction] Cozens et al.
Laser flash photolysis (LFP) has been employed for a
number of years for the observation of reactive intermediates
and for the direct measurement of rate constants of their de-
cay reactions. Photolyzing some precursor with a laser pulse
initiates a rapid photochemical transformation to the reactive
species of interest. This is detected by some method, in most
cases absorption spectroscopy, and the decay directly moni-
tored. In recent years, a number of applications involving
carbocations have appeared, with the intermediate being
generated and directly studied under normal solvolytic con-
ditions, i.e., with water or an alcohol as solvent (1, 2). In this
respect, detailed studies have been reported involving spe-
cies such as xanthylium (3, 4), triarylmethyl (5–8),
diarylmethyl (6–11), and fluorenyl cations (12–14). These
ions are relatively conjugated, with the formal cationic cen-
ter bearing at least two aromatic rings. Carbenium ions bear-
ing a single stabilizing aryl ring are inherently more
reactive, and thus in principle more difficult to study. In
spite of this, such cations are accessible by LFP under ap-
propriate conditions, and in two preliminary communica-
tions we have outlined aspects of this work (15, 16). This
paper provides the full details, with a discussion of the pho-
tochemical methods employed to generate such cations in
LFP experiments, and a comparison of the effect of solvent,
cation structure, and aryl substituent. Comparisons with
analogous vinyl cations are also presented.
Scheme 1.
tonation, where it can be used, is the better method. This possi-
bility was suggested through the observation of Markovnikov
hydration upon irradiation of styrenes in aqueous acid, with
the proposed mechanism involving protonation of the ex-
cited styrene to give intermediate phenethyl cations (17–19).
With TFE (and HFIP) as solvent, addition proceeds
smoothly without added acid, as illustrated in Scheme 1 for
4-methoxystyrene 1 where the ether 3 is the only product
(Scheme 1). The protonation must be occurring with the sol-
vent OH proton. This is perhaps not surprising, since the
protonation of excited styrenes is general-acid catalyzed
with a relatively small Bronsted α value (19).
LFP of 4-methoxystyrene in TFE reveals a relatively
strong transient with λmax at 340 nm, coupled with signifi-
cant negative optical density below 300 nm (Fig. 1A). The
bleaching signal is the negative of the spectrum of 4-meth-
oxystyrene and arises from use of the precursor solution as
the optical density reference. The absorbance changes mea-
sured immediately after the laser pulse at both 340 and
255 nm are linear in laser intensity, indicating the occurrence
of a monophotonic process. The absorbance above 300 nm
decays to essentially zero at all wavelengths with the same
rate constant of (3.9 ± 0.3) × 105 s–1. This implies a single
transient, and we assign this to the 1-(4-methoxy-
Results and discussion
In total, 32 cations have been investigated. Table 1 pro-
vides their structure, absorption maxima, and decay rate
constants in 2,2,2-trifluoroethanol (TFE) and (or)
1,1,1,3,3,3-hexafluoroisopropanol (HFIP). These solvents
are necessary with most of the ions in Table 1, since their
lifetimes in more nucleophilic solvents such as methanol,
ethanol and water are less than 20 ns, the limit of detection
with our LFP apparatus. Even then a 4-methoxy substituent
(or 4-Me2N) is usually required for the cation to be observed
in TFE, and the less stabilized ones require HFIP. Changing
from TFE to HFIP increases the lifetime by a factor of 103–
104. Water, on the other hand, is three orders of magnitude
more reactive than TFE.
4-Methoxyphenethyl cation
Two general photochemical routes were employed, sty-
rene photoprotonation and photoheterolysis, and the results
with the 4-methoxyphenethyl cation are typical. Photopro-
© 1999 NRC Canada

Cozens et al.
2071
Table 1. Absorption maxima and rate constants (20°C) for cumyl, phenethyl, benzyl and 1-phenylvinyl cations obtained by 248 nm
laser flash photolysis.
Ar-subst.
R1, R2
λmax,a (nm)
kTFE (s–1)
kHFIP (s–1)
knuc (M–1 s–1)
Cumyl and phenethyl cations ArC+(R1)-CH3
b
4-Me2Nc
4-Me2N
4-MeO
Me
H
4-MeOC6H4
380
360
330 (w), 490
380
320 (w), 440
365
365
320, 420
360
(1.5 × 102)d
(4.0 × 101)e
1.3 × 102
3.3 × 102
5.1 × 102
3.0 × 103
9.0 × 103
7.5 × 103
1.6 × 104
(4 × 107)d,k
1.6 × 104
1.5 × 105
2.1 × 105
3.9 × 105
1.0 × 107
>5 × 107
>5 × 107
>5 × 107
>5 × 107
>5 × 107
>5 × 107
>5 × 107
<102
<102
<102
<102
<102
<102
<102
<102
<102
2.4 × 107 (N3 )
– d
– e
1.6 × 107 (N3 )
– f
8.2 × 108 (N3 )
– f
4-MeO
4-MeO
4-MeO
4-MeO
4-MeONapj
4-MeO
cycPrg
C6H5
i-Pr
Et
H
Me
2.5 × 109 (N3 )
– f
2.7 × 109 (N3 )
5.0 × 107 (Br–)h,i
1.2 × 108 (Br–)h,i
– f
5.1 × 109 (N3 )
1.1 × 108 (Br–)h,i
4-MeO
H
3,4-(MeO)2
t-Bu
Ph
H
370
<102
2.2 × 108 (Br–)h,i
2.1 × 109 (Br–)h
2.5 × 109 (Br–)h
4.5 × 109 (Br–)h
1.3 × 109 (Br–)l
6.7 × 109 (Br–)l
9.8 × 109 (Br–)l
7.1 × 109 (Br–)l
8.9 × 109 (Br–)l
9.7 × 109 (Br–)l
8.2 × 109 (Br–)l
5.8 × 109 (Br–)l
310 (w), 420
345, 430 (w)
340
340, 380 (sh)
325, 400 (w)
315, 475 (w)
325, 390 (w)
320
3 × 102
2 × 102
2 × 102
6 × 102
9 × 103
4-MeO
H
4-Me
H
2,6-Me2
Me
Me
H
3 × 104
m
7 × 104
m
4-Me
H
4-F
H
2 × 105
m
3-Me
H
320, 430 (w)
315, 410 (w)
335, 390 (sh)
2 × 105
m
H
H
4 × 105
3-MeO
H
5 × 105
+
Benzyl cations ArCH2
4-MeONapj,n
4-MeOn
o
– f
305, 400
320, 360 (sh)
310, 410 (w)
1.8 × 105
4.3 × 106
>5 × 107
—
4.9 × 109 (N3 )
3 × 102
2 × 106
4.8 × 109 (Br–)l
1.0 × 1010 (Br–)l
4-Mep
1-Phenylvinyl cations ArC+=CR1R2
4-MeOq
4-MeOq
4-MeOr
Hq
2,4,6-Me3
4-Mer
Hr
Hr
Ph, Ph
Me, Me
H, H
Ph, Ph
H, H
H, H
H, Ph
H, H
340
330
335
310, 410 (w)
305, 420 (w)
315, 350 (sh)
310, 390 (w)
295, 360 (w)
1.4 × 104
2.3 × 105
1.3 × 106
>5 × 107
>5 × 107
>5 × 107
>5 × 107
>5 × 107
<1 × 102
<1 × 102
3 × 102
2 × 104
4 × 104
3 × 105
6 × 105
2 × 106
6 × 106 (Br–)h
2.2 × 108 (Br–)h
2.3 × 109 (Br–)h
8.2 × 108 (Br–)l
4.8 × 109 (Br–)l
9.2 × 109 (Br–)l
1.1 × 1010 (Br–)l
1.0 × 1010 (Br–)l
r,s
aw = weak; sh = shoulder.
bPrecursor unless otherwise noted is ArC(R)=CH2.
cPrecursor is 4-Me2NC6H4CMe2CMe2C6H5.
dSolvent is 2:1 (v:v) water:acetonitrile. Note that a similar value for ks is expected in pure water, since kdecay in water:acetonitrile does not vary a great
deal with solvent composition (ref. 6).
eSolvent is 50:50 water:TFE.
fSolvent is 10:90 water:TFE.
gCyclopropyl.
hIn TFE.
iAddition of bromide is reversible.
j4-Methoxy-1-naphthyl.
kPrecursor is (4-MeOC6H4CMe2)2.
lIn HFIP.
mBased on extrapolation to zero styrene concentration.
–
nPrecursor is trimethylammonium BF4 salt.
oNot examined.
pPrecursor is 4-MeC6H4CH2Cl.
qPrecursor is ArC(Br)C=CH2.
rPrecursor is ArCϵCR1.
sIdentical behavior from ArCϵCH and ArC(Cl)C=CH2.
© 1999 NRC Canada

2072
Can. J. Chem. Vol. 77, 1999
Fig 1. Transient spectra following 248 nm laser irradiation of 0.1 mM 4-methoxystyrene (A) and 1 mM 1-(4-methoxyphenyl)ethyl
acetate (B – D) in TFE. Spectra B – D were obtained on the same solution; B, at a laser dose of 6 mJ and with oxygen saturation, C,
6 mJ and with argon, and D, 75 mJ and with oxygen saturation. Spectrum A was obtained at 25 mJ and with argon saturation. This
spectra was unchanged on saturating with oxygen. The spectra in C and D, and the initial spectra in A and B were obtained at 50–100
ns after the laser pulse. The final spectra in A and B were obtained at 25–30 µs.
Scheme 2.
characteristic of carbocations (2), being quenched by nucleo-
philes such as bromide and azide, while oxygen has no effect.
Additional evidence is the observation of the same inter-
mediate from a different precursor, 1-(4-methoxyphenyl)ethyl
acetate 5 (Scheme 2). The photochemical reaction here is a
photoheterolysis, and, as is typical of this route (2, 6), there
are competing photo-reactions. Thus, as shown in Fig. 1C,
irradiation of the ester results in two transients, the one with
λmax at 340 nm and a second at 275 nm. The former is the
same as the transient from the styrene precursor. It has an
identical spectrum, decays in TFE alone with the same rate
constant, (4.2 ± 0.4) × 105 s–1, and is quenched in an identi-
cal manner by bromide and azide ions. The 275 nm transient
does not decay by exponential kinetics in argon-saturated so-
lution and, while nucleophiles have no effect, oxygen is a
very effective quencher. In fact, in the experiment in oxy-
gen-saturated TFE, the 275 nm signal is no longer seen
(Fig. 1B). This intermediate can be assigned as the 1-4-
methoxyphenethyl radical 6, the result of a competing
photohomolysis. At higher laser doses additional
absorbances appear: a band with λmax near 295 nm and a
second smaller broad peak at 400–500 nm. The 295 signal
overlaps that of the radical but can be seen clearly in experi-
ments with oxygen saturation (Fig. 1D). The two new bands
appear to represent the same intermediate, since they decay
with the same rate constant, with k ~ 1 × 105 s–1. A clue to
the identity is the square dependence on the laser dose of the
phenyl)ethyl cation 2 implicated by the Markovnikov ether
photo-product 3. The transient does exhibit kinetic behavior
© 1999 NRC Canada

Cozens et al.
Table 2. Effect of water on the irradiation of 4-methoxystyrene.
2073
%(v/v) TFE
kdecay (s–1)
% Ether product
Initial OD (340 nm)a
100
95
90
80
70
60
4.0 × 105
1.5 × 106
3.6 × 106
6.3 × 106
9.6 × 106
1.3 × 107
100
90
80
40
Trace
0
1.00
0.85
0.70
0.60
0.50
0.40
aRelative to 100% TFE. This is equivalent to a relative quantum yield, since the measurements involve solutions of the
same concentration of styrene being irradiated at 248 nm under identical conditions.
optical densities measured at 420 nm. This points to a
biphotonic process, and suggests that this intermediate is the
cation radical 7, the result of a two-photon ionization. In fact
the spectrum is that expected for the cation radical of an
anisole derivative (20). The 340 nm transient on the other
hand follows a linear dependence on laser intensity. Thus, as
summarized in Scheme 2, the initial excited state, presum-
ably the singlet, reacts to produce the cation and radical and,
in competition, is further excited resulting in electron ejec-
tion. These experiments do not distinguish mechanisms for
the formation of the cation and radical. There is the sugges-
tion with benzyl carboxylates, including 4-methoxybenzyl
acetate, that the primary event is photohomolysis with the
cation being subsequently produced by electron transfer
within the so-formed radical pair (21, 22).
A important feature with the styrene precursor is that
there is no return of the absorbance below 300 nm as the
cation decays back to the baseline above 300 nm. This is
consistent with studies of thermal alkene additions that show
that protonation in general is irreversible (23). In other
words, the cation reacts predominantly by solvent addition,
with little styrene being regenerated by elimination of a pro-
ton at the β-position.
As expected, addition of water significantly increases the
rate constant for the decay of the cation. In addition, the al-
cohol 4 is produced at the expense of the ether (Table 2). In-
terestingly, the initial increase in rate constant is much
greater than predicted on the basis of the amount of alcohol
product. For example, 10% water results in only 20% of the
alcohol product, but the rate constant for the decay of the
cation is increased by almost 10-fold. This is consistent with
a mechanism where, in addition to its direct nucleophilic
combination, water accelerates the decay of the cation by
acting as a general base to assist the TFE addition (24). Wa-
ter also decreases the quantum yield for photoprotonation, as
shown in the last column of Table 2. Coupled with the rate
accelerating effect, this makes it increasingly difficult to
study the cation in highly aqueous mixtures.
Substituted phenethyl cations by styrene protonation
In the same way, other phenethyl and cumyl cations are
observed upon irradiation of the appropriate substituted sty-
rene in TFE and (or) HFIP. Spectra of three cumyl cations
and a naphthalene derivative are shown in Figs. 2 and 3A re-
spectively. The other examples are available as supplemen-
tary material.2 These show the same features discussed
above for 4-methoxystyrene — a positive absorbance above
300 nm representing a single intermediate as shown by com-
mon decay kinetics across the entire spectrum,3 effective
quenching by bromide4 and (or) azide ion, and bleaching
due to depletion of the precursor. Spectra under stable ion
conditions are available for the cumyl cation, λmax = 326 and
390 nm (ε = 11 000 and 1400 respectively) and the 1,1-
diphenylethyl cation, λmax = 312 and 422 nm, (ε = 11 050
and 37 000 respectively) (25). The essentially identical spec-
tra obtained here provide further confidence in the assign-
ment of cation intermediates (see the spectrum for the cumyl
ion in Fig. 2).
A further feature in common with 4-methoxystyrene is the
lack of return of the bleaching as the absorbance of the cat-
ion decays. Thus the cations in general react by solvent ad-
dition. One exception is the 2,6-dimethylphenethyl cation
where a competing deprotonation is indicated by the signifi-
cant return of the precursor absorbance at the same rate con-
stant as that obtained for cation decay. This aspect of 2,6-
dimethyl systems has been recently demonstrated in a study
of solvolysis of cumyl derivatives (26). While solvent addi-
tion is retarded by the steric effect of the ortho substituents,
the rate constants for elimination are actually increased rela-
tive to analogs without the 2,6-dimethyl groups (26).
Styrene dimerization
The other exception occurs in HFIP with styrene and its
3- and 4-substituted derivatives. One indication of the differ-
ence is that the rate constants depend on the precursor con-
centration. This dependence is linear (eq. [1]), implicating a
reaction of unphotolyzed precursor with the cation. With the
2 Supplementary material may be purchased from: The Depository of Unpublished Data, Document Delivery, CISTI, National Research
Council Canada, Ottawa, Ontario, Canada, K1A 0S2.
3 The one exception is α-tert-butyl-4-methoxystyrene, where the decay requires fitting to a double exponential, with rate constants of 7 × 105
and 1.6 × 104 s–1. The spectrum of the intermediate responsible for the slow decay, obtained at longer times when the faster one is complete,
shows a λmax of 370 nm. This is assigned to the α-tert-butyl-4-methoxyphenethyl cation on the basis of the close similarity with the spectra
of the other α-alkyl-4-methoxyphenethyl derivatives. The faster decaying transient has its λmax between 330 and 340 nm and has not been
identified.
4 With some of the cations, bromide addition shows reversible behavior. This has been seen previously (5) and occurs when the ionization of
the bromide product has a rate constant similar to the quenching rate constant. With the very stable cations where ionization of the C—Br
bond becomes faster than nucleophile–cation combination, bromide does not react at all. These cations are still quenched by azide ion.
© 1999 NRC Canada

2074
Can. J. Chem. Vol. 77, 1999
Fig. 3. Transient spectra following 248 nm irradiation of
~0.0001 M 4-methoxy-1-vinylnaphthalene (A) and 4-methoxy-1-
napthylmethylammonium acetate (B) in argon-saturated TFE.
Data points were obtained immediately after the laser pulse and
at the completion of the exponential decay.
Fig. 2. Spectra of cumyl cations 4-XC6H4C+(CH3)2 obtained
upon 248 nm laser irradiation of ~0.1 mM solutions of
4-XC6H4C(CH3)=CH2 — (A) X = H in HFIP, (B) X = Me in
HFIP, and (C) X = MeO in TFE. ∆OD values have been
adjusted so that the maximum is 0.1 in each case.
Table 3. Rate constants (20°C) for the reaction of phenethyl
cations X-C6H4CH+-CH3 with the styrene X-C6H4CH=CH2 in
HFIP.
X
k1 (M–1 s–1)
sty
H
3-Me
4-F
4-Me
4-MeO
1.0 × 109
9.4 × 108
4.3 × 108
3.4 × 108
1.3 × 106
assumption (see Experimental) that the concentration of sty-
rene is effectively unchanged in the laser pulse, the slopes of
these plots (Table 3) are the second-order rate constants for
the dimerization. The intercepts are the rate constants for the
solvent reaction (the numbers in Table 1 for these cations).
only approximate because of detector instability at very long
times).
A similar situation has been observed for cations gener-
ated in strong acid solution. For example, when styrene is
added to 98–100% sulfuric acid, the species observed is the
bis-styryl cation, not the initially-formed phenethyl cation
(27, 28). With α-methylstyrene, the cumyl cation is suffi-
ciently stable to be seen, but it is not stable, reacting to give
a new, long-lived, species with a higher λmax identified as the
bis-cumyl cation (27, 28). In a similar manner, we propose
that phenethyl cations 9, when generated by styrene
photoprotonation in HFIP, react in part with residual styrene
forming much more stable bis-styryl cations 10 (Scheme 3).
The dimerization is obviously the first step in the propagation
sequence of the acid-initiated polymerization of styrenes.
The dimer ions 10 in fact appear similar to the oligomeric 11
that have been detected during polymerization. With styrene,
for example, the propagating cation has a λmax of 340 nm
(29, 30), the same as the cation 10 (X = H) observed in this
work. One interesting observation is that the monomer ions
9 have a slightly slower λmax. A possible explanation is that
[1]
k
obs/s–1 = kHFIP + k1sty[ArCH=CH2]
Spectral changes provide further evidence. Figure 4 shows
an example where dimerization dominates and a high laser
intensity is employed to provide strong signals. After the ini-
tial “prompt” bleaching at 240 nm, there is a further expo-
nential decay occurring at the same rate as at the higher
wavelengths. Thus additional styrene is reacting. Moreover
as the cation decays, a new species appears. There is consid-
erable overlap, but it is clear that this absorbs at a slightly
higher wavelength, with a λmax around 345 nm for the prod-
uct in Fig. 4. Similar observations are made with the other
ArCH=CH2, with a shift of ~20–30 nm in each case. The
new absorbance is much more stable than the original one
but there is decay at very long times, for example, with sty-
rene (at 340 nm) with a rate constant of ~0.1 s–1. (This is
© 1999 NRC Canada

Cozens et al.
2075
Fig. 4. Transient spectra following 248 nm irradiation of
0.0004 M 4-methylstyrene in HFIP. The open circles were
obtained immediately after the laser pulse and the closed circles
at the completion (40 µs) of the exponential process. The other
two traces were obtained at intermediate times. The Inset shows
the absorbance change at the three indicated wavelengths. The
rate constants (s–1) are 1.6 × 105 (240 nm), 2.0 × 105 (320 nm)
and 1.85 × 105 (360 nm).
Scheme 3.
steric effect, since the reactivities of both the cation and the
nucleophile are not that different from the those in the
ArCH=CH2 series. There is no dimerization with any of the
cations in TFE. Here the more nucleophilic solvent can suc-
cessfully compete with the low concentrations of precursor.
Benzyl cations
Styrene photoprotonation cannot be applied for the gener-
+
ation of ArCH2 ; therefore we were forced to return to
photoheterolysis. For 4-methoxybenzyl, both chloride and
–
trimethylammonium (BF4 salt) precursors give a transient
in 10 and 11 there is an interaction of the aryl group at the γ
carbon with the positive charge at the α-carbon, as shown in
structures in Scheme 3. Another interesting feature is that
the dimer cation is much more stable than the monomer.
that can be assigned to the cation, although the former is
limited to TFE solutions containing greater than 50%
acetonitrile because of rapid ground state solvolysis. The
cation is identified by several criteria. Its λmax (320 nm,
Fig. 5A) fits in a comparison with the 4-methoxyphenethyl
(340 nm) and 4-methoxycumyl (360 nm) cations. The decay
kinetics are typical of cations, a very efficient quenching by
bromide, and a large difference between HFIP and TFE (Ta-
ble 1). The same transient is observed with the two precur-
sors, with the same decay rate constant, 9.0 × 106 s–1 and 8.6
× 106 s–1 for the chloride and ammonium salt respectively in
1:3 TFE:acetonitrile. Finally, product analysis following ir-
radiation of the ammonium salt in TFE shows the presence
of 4-methoxybenzyl 2,2,2-trifluoroethyl ether, the product
derived from trapping of the cation by solvent. The other
product formed in about the same amount is 4-methylani-
sole. This indicates the presence of a radical intermediate as
well (31, 32), and, in argon-saturated solvents, LFP of both
precursors shows a large band with λmax at 290 nm, presum-
ably due to the 4-methoxybenzyl radical. With 4-methoxy-
benzyl acetate in TFE, only the 290 absorbance of the
radical is seen at low laser doses, while, like the phenethyl
analog, absorbance for a cation radical dominates at higher
intensities. Thus, this precursor results in so little cation that
the intermediate cannot be detected in competition with the
other pathways. In fact, 4-methoxybenzyl acetate is known
to result in only low yields of photosolvolysis product when
irradiated in alcohol solvents (22, 33).
This is also true of the oligomer. The rate constant for sty-
pr
rene propagation (k of Scheme 3) is ~105 M–1 s–1 at 20°C
sty
(29, 30), four orders of magnitude slower than the rate con-
stant k1sty for the phenethyl cation reacting with styrene. It is
interesting that the difference in stability between phenethyl
and dimer/oligomer cations was suggested in 1968, and the
interaction with the adjacent phenyl group in the latter pro-
posed as the explanation (27).
We can be confident that it is the dimer cation 10 being
observed at the end of the exponential process in the LFP
experiments, and not an oligomer. One argument is based on
the observation (Fig. 4) that the “slow” decay of styrene fol-
lows clean exponential kinetics, moreover reaching a stable
absorbance at the completion of the change. If styrene were
continually adding to a growing oligomeric cation, one
would have expected a downward drifting absorbance. The
other argument is that subsequent reactions with styrene are
much slower than the first one, as discussed above. In other
words, the reaction of the dimer cation with further styrene
occurs on a much longer time scale than the initial reaction.
Experimental limitations unfortunately preclude observation
at long times at the lower wavelengths, i.e., where further
depletion of styrene would be observed.
Cumyl cations do not show any sign of reacting with the
residual α-methylstyrene in HFIP. This would appear to be a
© 1999 NRC Canada

2076
Can. J. Chem. Vol. 77, 1999
Fig. 5. Transient spectra following 248 nm irradiation of
~0.001 M 4-methoxybenzyl chloride (A) and 4-methylbenzyl
chloride (B) in oxygen-saturated 1:3 TFE: acetonitrile and HFIP
respectively. Triangles are the ∆OD immediately after the laser
pulse, and squares, ∆OD at the completion of the exponential
decay that is observed. The crosses are the difference in the two.
Scheme 4.
Table 4. Rate constants at 25°C for the solvolysis of
4-methylbenzyl bromide (20 µM) in HFIP in the
presence of tetrabutylammonium bromide.
[Bu4N+ Br–]
ksolv (s–1)
a
0
2.01 × 10−4
8.26 × 10–5
5.15 × 10–5
3.92 × 10–5
2.94 × 10–5
2.32 × 10–5
0.0003
0.0006
0.0009
0.0012
0.0015
aBased on initial rate.
with a ratio obtained from the kinetics of ground state
solvolysis of 4-methylbenzyl bromide in HFIP. As shown in
Table 4, added bromide results in significant common ion
inhibition. By fitting to ksolv = kion:(1 + kBr:kHFIP[Br]),
kBr:kHFIP is found to be (4.8 ± 0.2) × 103, the same number
obtained from the transient spectroscopy.
As shown in Fig. 3B, the 4-methoxy-1-naphthylmethyltri-
methylammonium ion gives a strong signal that can be as-
signed to the primary cation. The evidence here is the
efficient quenching by nucleophiles and the close similarity
with the secondary cation obtained by photoprotonation.
There is little indication of the presence of radical intermedi-
ate, as shown by very similar spectra being obtained under
argon and oxygen.
For the 4-methylbenzyl cation, the chloride could be used
as a precursor in HFIP. Spectra in oxygen-saturated solvent
are shown in Fig. 5B. Below 300 nm a large non-decaying
band of uncertain origin is observed. A shoulder above
300 nm does decay in an exponential fashion. By taking the
difference between the initial and final absorbances, this spe-
cies is found to have a λmax of 310 nm with a weaker
absorbance extending out above 400 nm. As will be dis-
cussed, this fits the expectation for this cation. The
absorbance at 310 is quenched by bromide ion, with the ra-
tio kBr:kHFIP equal to 5.0 × 103 M–1. This allows for confir-
mation that this is due to the cation, through a comparison
All attempts to generate detectable amounts of the parent
benzyl cation in HFIP failed to provide any evidence of the
cation. Precursors of benzyl chloride, benzyl bromide and
benzyl acetate were employed. There were transients, but
these could be attributed to radicals, or products from the
radicals. A comparison of the 4-H and 4-Me derivatives in
the phenethyl and cumyl series (Table 1) suggests that the
benzyl cation is at least an order of magnitude more short-
+
lived than 4-methylbenzyl. Thus the lifetime of PhCH2 in
HFIP is less than 50 ns. Even if this were the lifetime, the
cation would be very difficult to detect unless the photo-
chemical route to it was very efficient.
Radical cation fragmentation
By using photoprotonation, the 4-methoxycumyl cation
can be studied in aqueous TFE until about 70% water is
© 1999 NRC Canada

Cozens et al.
2077
Fig. 7. Spectra of vinyl cations 4-XC6H4C+=CH2 obtained upon
248 nm laser irradiation of ~0.1 mM solutions of 4-XC6H4CϵCH
— (A) X = H in HFIP, (B) X = Me in HFIP and (C) X = MeO
in TFE. ∆OD values have been adjusted so that the maximum is
0.1 in each case.
Fig. 6. Transient spectra following 248 nm irradiation of 4-
dimethylaminobiscumene 13 in oxygen-saturated 2:1
water:acetonitrile. Closed circles obtained immediately after the
laser pulse are attributed to the cation radical 13+!. Open circles
obtained near the end of the completion of the exponential decay
are due to the 4-dimethylaminocumyl cation 16. Rate constants
measured at 275, 380 and 500 nm are respectively 8.8 × 104,
1.12 × 105, and 1.03 × 105 s–1, respectively.
Scheme 5.
present. At this point the cation decays with a rate constant
of 7.9 × 106 s–1, still below the 20 ns limit. However, the
amount of intermediate being produced upon irradiation is
so small that further increases in water content are not possi-
ble. To obtain a water rate constant, we turned to bis-4-
methoxycumene 12 as a precursor (Scheme 4). The photo-
chemistry here involves a two photon ionization to the cat-
ion radical, which fragments within the laser pulse to cation
14 and radical 15 (34). With 2:1 water:acetonitrile as sol-
vent, the cation can just be detected, with a decay of ~4 ×
107 s–1. While the acetonitrile is required in this experiment
to dissolve the precursor, a similar decay is expected in pure
water, because rate constants in water:acetonitrile do not
vary a great deal with solvent composition (6).
The very stable 4-dimethylaminocumyl cation is gener-
ated from the biscumene 13 in the same water:acetonitrile
mixture. In this case the cation radical is much more stable
so that it is this species that is present after the laser pulse,
and the fragmentation occurs on the microsecond time scale
(Fig. 6). This observation has been reported previously (35),
although the fragmentation of 13+· in the aqueous solvent of
the present study is significantly faster than that for the same
cation radical in solvents previously used. The 4-dimethyl-
aminocumyl cation can be observed to grow in as the cation
radical decays. This has a λmax at 380 nm (382 nm, ref. 35),
and a lifetime in the aqueous solvent of 7 ms.
Phenylacetylenes as precursors of vinyl cations
Laser flash photolysis of p-methoxyphenylacetylene in both
TFE and HFIP, and p-methylphenylacetylene, 2,4,6-
trimethylphenylacetylene, diphenylacetylene, and phenyl-
acetylene itself in HFIP results in transient spectra (see
Fig. 7) that can be assigned to vinyl cations produced by
photoprotonation (Scheme 5). With each precursor, there is a
single intermediate, as shown by the absorbance above
300 nm decaying to near zero with the same rate constant at
all wavelengths. Nucleophiles such as bromide ion have a
pronounced quenching effect, evidence that these are cat-
ions. Vinyl cations, usually with two substituents on the β-
carbon, have been produced in the past by photoheterolysis
of halide precursors (36–39). As seen in Table 1, the 1-(4-
methoxyphenyl)vinyl cation produced from the acetylene has
a similar λmax to that of the 2,2-dimethyl analog produced
from the bromide. Even more compelling is the case of the
2,4,6-trimethylphenyl derivative, where the same cation was
produced in the two different ways. Like styrenes, substi-
tuted phenylacetylenes have been shown to undergo photo-
© 1999 NRC Canada

2078
Can. J. Chem. Vol. 77, 1999
Fig. 8. Substituent effect on the λmax for benzyl, phenethyl,
cumyl, 1-phenylvinyl, and 2,2-diphenyl-1-phenylvinyl cations.
See text for explanation of x axis. Lines have been drawn with a
slope of unity.
higher wavelength is usually not visible for the 4-MeO and
4-Me2N substituted cations.
There has been some question as to the spectrum of the
parent benzyl cation. An early theoretical calculation
showed bands at 400–500 nm and 300 nm, with the latter
being significantly stronger (44). This paper also suggested
that a report of the benzyl cation in strongly acidic solutions
(45) was incorrect. In agreement with the more intense
absorbance, a band at 290 nm assigned to the cation was ob-
served in an experiment involving photodissociation in a
mass spectrometer (46). A more recent pulse radiolysis ex-
periment in neat benzyl chloride is also in agreement, sug-
gesting a strong band at 303 ± 2 nm and a weak band near
500 nm (47). In contrast, there have been several assign-
ments to the benzyl cation where the λmax is around 360 nm.
These are based upon experiments by Dorfmann and co-
workers who claimed to have generated the benzyl cation by
pulse radiolysis in chlorocarbon solvents (48–50). The basis
of this assignment was quenching of the observed transient
by nucleophiles, and the fact that the approach does work
for the triphenyl- and diphenylmethyl cations. A similar
band was then observed and assigned to the benzyl cation
following photoionisation of benzyl bromide in an argon
matrix (51), and upon photolysis of dibenzyl ketone in vac-
uum-dried Nafion membranes (52).
While the parent cation has not being observed in this
work, the data in Table 1 clearly shows that it must have
λmax around 300 nm with a weaker absorbance extending
into the visible spectrum. This is in agreement with the first
set of results discussed in the previous paragraph. The
~360 nm transient cannot be the benzyl cation. The pulse
radiolysis approach requires fragmentation of a cation radi-
cal and we suspect that under the conditions employed by
Dorfman and coworkers this is not occurring. In fact this is
exactly what happens when precursors of the 9-fluorenyl
cation are employed; there is no fragmentation and the inter-
mediate that is observed is the cation radical of the precursor
(13). Based on estimated pKR values, the 9-fluorenyl cation
is more stable than benzyl (53). Thus, if the approach does
not work for 9-fluorenyl it is unlikely to have worked for
benzyl. It can also be noted that the nucleophilic reactivities
of the 360 nm transient have been criticized as not being ap-
propriate for the benzyl cation (54). As a final comment, we
note that even in those cases where the ~300 nm absorbance
for the benzyl cation has been observed, this is clearly not
the only species present. In other words there has not yet be-
ing a study showing a clean spectrum of this cation.
hydration in aqueous acids, with the ultimate product an
acetophenone (17, 18). Flash photolysis investigations have
revealed the presence of the enol intermediate (40–42), the
immediate product of the hydration of the vinyl cation. The
cations themselves were not detected. As discussed above,
they would be too short-lived in water for detection by ns LFP.
Absorption maxima
Substituents affect λmax in the order H < 4-Me < 4-MeO <
4-Me2N. The effect is essentially the same for a given type
of cation with, on average, a shift relative to H of 15 nm for
4-Me, 30 nm for 4-MeO, and 50 nm for 4-Me2N. This is
demonstrated in Fig. 8 where the observed λmax are plotted
against these factors. Recognizing that λmax is only deter-
mined to ±5 nm, the correlations are excellent. Also appar-
ent from this figure are the similar λmax values for the vinyl
and benzylic series. This was previously commented upon in
a comparison of the 4-methoxy-substituted ions (15). In an
arylvinyl cation, the π systems of the aryl ring and the
alkene are orthogonal to one another (43). Thus the aryl ring
interacts with the formally empty p orbital on the α-carbon
Substituent effects on rate constants
(a) Vinyl vs. phenethyl cations
As was noted previously in the case of the 4-methoxy de-
rivatives (15) there is remarkably little difference in the life-
time of a vinyl cation ArC+=CH2 relative to that of a
phenethyl cation ArC+H-CH3 with the same aryl group. The
vinyl cation is slightly more reactive, but the differences are
only 3.3 (TFE) and 1.4 (HFIP) for Ar = 4-MeOC6H4, 4.4 for
4-MeC6H4 in HFIP, and 2.5 for C6H5 in HFIP. As with the
explanation for the similar spectra, this is consistent with the
idea that the aryl ring interacts with the p-orbital on the α-
carbon in exactly the same way in the two series.
in exactly the same way as it does in a benzylic cation.
+
For the same aryl group, the order of λmax is ArCH2
<
+
ArCHMe+ < ArCMe2 , with about a 20 nm difference be-
tween each pair. A weaker absorbance at higher wavelength
is also apparent with some of the systems. The trend appears
to be exactly opposite that of the larger absorbance at lower
wavelength, and this is true both in terms of the aryl
substituent and the α-methyl substituents. In fact the band at
© 1999 NRC Canada

Cozens et al.
2079
Fig. 9. Correlation of log kTFE for 4-MeOC6H4C+(R)CH3 vs.
Es(R).
Table 5. Values of ρr, ρn, and r+ obtained by fitting rate constants
for solvent capture of 1-arylethyl cations to eqs. [2] and [3].
Parameter
HFIP
TFE:watera
ρr
ρn
r+
4.7
2.7
1.7
6.1
2.7
2.3
aReference 58, as corrected in ref. 14.
Scheme 6.
(b) α-Alkyl effects
LFP in TFE (57). For a given Ar, the reactivities are similar.
However the phenyl series shows a greater effect of chang-
ing substituents in Ar. The result is that the cyclopropyl cat-
ion is slightly more stable (kinetically) when Ar is phenyl,
but less stable when Ar is 4-methoxyphenyl.
For a given aryl group the rate constant for solvent addi-
tion follows the expected order cumyl < phenethyl < benzyl.
Interestingly, the second methyl group exerts a greater ki-
netic stabilization than the first. In the 4-methoxy series in
TFE, for example, there is a 11-fold difference between
benzyl and phenethyl, but this increases to 24-fold for
phenethyl and cumyl. The same comparisons for the 4-Me
cations in HFIP lead to ratios of 28 and 120. This suggests
that it is unlikely that it is electron donation alone that is im-
portant. A steric effect would account for the difference, the
second methyl group increasing the steric bulk around the α-
carbon to such a degree that nucleophilic addition is hin-
dered to an extent greater than that of the first methyl substi-
tution.
While there is insufficient data to perform a detailed
breakdown into steric and electronic contributions, steric ef-
fects are clearly important. As shown in Fig. 9, there is a
good correlation of the data for 4-methoxyphenethyl cations
in TFE against the Taft steric parameter Es, and this includes
the point for the secondary cation. The correlation against
the polar parameter σ* is marginally poorer. In particular,
this requires a difference in rate constant smaller than the
one observed between i-Pr and Et.
The difference in the latter comparison is a factor of 10
(21 vs. 22 in Scheme 6). Interestingly, the methylated
analogs 23 and 24 studied in this work show a slightly
greater stabilization (1.5-fold) for the cyclopropyl system.
This can be traced to the substitution of α-hydrogen with
methyl resulting only in a threefold rate constant diminution
in the diarylmethyl cation, but a 42-fold diminution in the
arylcyclopropyl pair. We attribute this to steric inhibition of
resonance in the former. In these cations, maximum stabili-
zation occurs when both aromatic rings are fully planar with
the plane of the α-carbon. We suspect that methyl substitu-
tion results in additional twisting, so that while there is elec-
tronic and steric stabilization due to the methyl group, this is
compensated by some loss of resonance interaction. The
steric problem is not as severe in the cyclopropyl series,
since the cyclopropyl is perpendicular to the π system.
(d) Aryl substituent effects
Since the initial studies of Richard et al. (58) it has been
known that rate constants for solvent capture of benzylic
carbocations do not correlate with σ+, as might have been
expected. This has been observed now with several systems
(2). What occurs is that σ+ considerably underestimates the
stabilizing effect of para π-donors like 4-methoxy. This is
revealed by fitting to the two parameter Yukawa–Tsumo
equation (eq. [2]), where r+ values significantly greater than
1 result. Fitting to the modified eq. [3] (18), ρr is found to be
larger than ρn.
An additional steric effect is probably responsible for the
deviation of the point for the tert-butyl derivative. In this
case, unfavorable interactions between the ortho hydrogens
and the tert-butyl group twist the aryl ring, diminishing the
resonance interaction with the p orbital on the α carbon. In
fact MM2 calculations indicate a twist of 80 degrees. This
argument is reinforced by the slow rate of solvolysis of 3,3-
dimethyl-2-phenyl-2-chlorobutane, which has also been at-
tributed to steric inhibition of resonance (55).
[2]
[3]
log (k/ko) = ρ{σ + r+(σ+ – σ)}
log (k/ko) = ρnσn + ρr(σ+ – σn)
(c) Cyclopropyl versus phenyl
While an α-cyclopropyl group on a carbocation is
strongly stabilizing, there has been some dispute whether the
effect is greater or less than that of phenyl (56). In terms of
the reactivities of fully-formed cations, a series ArCH+-R
where R = phenyl and cyclopropyl have been compared by
The study of Richard et al. (58) actually involved 1-arylethyl
cations, with the rate constants being obtained through use
of the azide-clock method in 1:1 TFE:water as solvent. Six
© 1999 NRC Canada

2080
Can. J. Chem. Vol. 77, 1999
α
-Isopropyl-4-methoxystyrene was prepared by reacting
such cations were studied directly in HFIP in this work (Ta-
ble 5). A comparison of the parameters obtained by fitting to
the above equations is given in Table 5. Since the two series
do not involve the same substituents, and, moreover, since
there are a limited number of substituents in each, the two
sets of parameters can be regarded as being equivalent. This
is perhaps surprising since the cations are 106-fold less reac-
tive in HFIP.
1
Tebbe’s reagent with 2-methyl-4′-methoxyacetophenone. H
NMR (CDCl3) δ (ppm): 1.10 (d, 6H), 2.82 (sept., 1H), 3.82
(s, 3H), 4.97 (dd, 1H), 5.10 (dd, 1H), 6.86 (d, 2H), 7.31 (d,
2H). Elem. anal. calcd.: C 81.77, H 9.15; found: C 81.44, H
9.02.
4-Methoxybenzyl 2,2,2-trifluoroethyl ether was prepared by
adding 4-methoxybenzyl chloride to 2,2,2,-trifluoroethanol
containing pyridine. After removal of the solvent, the resi-
due was taken up in ether, washed with water, dilute HCl,
and water. The ether was dried over MgSO4, and after filter-
(e) Styrene dimerization
For the reaction of a substituted phenethyl cation with its
precursor styrene, there is a decrease in rate constant with
increased electron donation (Table 3). Because of the man-
ner in which this experiment is performed, the substituent is
changing in both the electrophile and the nucleophile, and
thus there will be compensating effects. Increased electron
donation will stabilize the cation but at the same time in-
crease nucleophilicity of the styrene. What the data show is
that the former is more important. This is consistent with
studies of benzylic-type cations where data are available for
the reaction in both the formation and decay directions, in
particular the finding that there is imbalance in the reaction
coordinate (2, 58). In the direction where the cation is the
reagent, the resonance interaction has decreased at the tran-
sition state to a greater extent than bond making. The reverse
is true in the direction forming the cation, where the devel-
oping resonance interaction in the cation lags. In the present
case, it can be seen that a benzylic-type cation (9 in
Scheme 3) is being replaced by a benzylic-type cation (10)
where the positive charge is stabilized by the same aryl
group. What we argue is that in the transition state, the reso-
nance interaction in the reagent cation has diminished to a
greater extent than the resonance interaction in the product
cation that has developed. Thus, in changing from styrene to
4-methoxystyrene, the stabilizing effect of the 4-methoxy
group is more important in the starting cation than it is in on
the developing cation in the transition state. Thus, the initial
state is stabilized more than the transition state and the reac-
tion is slower.
1
ing to remove the drying agent, the solvent removed. H
NMR (CDCl3) δ (ppm): 3.75 (q, 2H), 3.79 (s, 3H), 6.90 (d,
2H), 7.28 (d, 2H). EIMS (70 ev), m/z: 220 (30%, M+), 189
(8%, M+ – OCH3), 135 (12%, M+ – CH2CF3), 121 (100%,
M+ – OCH2CF3).
4-Methoxyphenethyl 2,2,2-trifluoroethyl ether was prepared
by irradiating a solution of 4-methoxystyrene in 2,2,2-
trifluoroethanol until there was no more styrene left. The
ether was isolated as described in the previous paragraph. 1H
NMR (CDCl3) δ (ppm): 1.45 (d, 3H), 3.65 (q, 2H), 3.81 (s,
3H), 4.56 (q, 1H), 6.90 (d, 2H), 7.28 (d, 2H). EIMS (70 ev),
m/z: 234 (32%, M+), 219 (100%, M+ – CH3), 135 (80%, M+
– OCH2CF3).
Laser flash photolysis experiments were carried out with
equipment in Mulheim and in Toronto. The carbocation pre-
cursors were photolyzed with 20 nanosecond pulses, typi-
cally of 30 mJ, of 248 (KrF*) or 308 nm (XeCl*)
monochromatic light. A pulsed Xenon light provided the
monitoring beam. This was passed through the cuvette con-
taining the substrate, then through a monochromator onto a
photomultiplier tube. The signal was digitized by Tektronix
7612 and 7912 transient recorders, and the digitized signal
sent to various computers for data analysis. Stock solutions
of the substrates of concentrations 0.1–0.01 M were pre-
pared in acetonitrile or dichloromethane. A small amount of
this solution was then added to the appropriate TFE, HFIP,
or aqueous solution. This was then introduced into a flow
cell apparatus, where, if desired, the solution was saturated
with argon or oxygen by bubbling with the appropriate gas
for 5–10 min. Absorption spectra were constructed point-by-
point, with the solution being flowed through the cuvette so
that there was unreacted substrate of the same concentration
at each wavelength. The average optical density over a small
time period was then determined at each wavelength and
plotted against the latter to generate the spectrum. Rate con-
stants were obtained by fitting the decay data to the expo-
nential equation. five to seven kinetic runs were performed
with each solution and the rate constants averaged. Observed
rate constants refer to 20 ± 2°C, and were reproducible to
±2% in TFE and ±10% in HFIP. The nucleophile quenching
experiments involved 5–6 solutions of tetrabutylammonium
bromide or sodium azide varying in concentration from 5 ×
105 to 0.001 M. Plots of the observed rate constant were lin-
ear in the concentration of the added nucleophile. The sec-
ond-order rate constants were obtained as the least-squares
slopes. The rate constants for the styrene quenching in
Table 3 were obtained with initial styrene concentrations of
5 × 10–5 to 5 × 10–4 M. A low laser dose (5–10 mJ) was em-
Experimental section
TFE and HFIP were dried over 4 Å molecular sieves and
distilled under nitrogen through a 40 cm Vigreux column
prior to use. Water was purified with a Millipore-Q system.
The nucleophiles tetrabutylammonium bromide and sodium
azide were used as received.
Table 6 provides the precursors that were employed and
their sources. In most cases the compounds were commer-
cially available or were prepared by a literature method.
Where the compounds were solids the mp was in agreement
1
with values in the literature. The H NMR spectra were in
all cases fully consistent with the structure.
4-Methoxy-1-vinylnapthalene was prepared from 4-methoxy-
1-napthaldehyde by reacting Tebbe’s reagent following stan-
dard procedures (77–79). 1H NMR (CDCl3) δ (ppm): 3.99 (s,
3H), 5.33–5.39 (dd, 1H), 5.64–5.74 (dd, 1H), 6.79 (dd, 1H),
7.32–7.57 (m, 2H), 8.04–8.32 (m, 4H). HRMS: 184.0886
(calcd.: 184.0888). Anal. calcd.: C 84.75, H 6.56; found: C
84.70, H 6.60.
© 1999 NRC Canada

Cozens et al.
2081
Table 6. Cation precursors.
Cation
Precursor
Referencea
4-Me2NC6H4C+(CH3)2
4-Me2NC6H4CH+CH3
(4-MeOC6H4)2C+CH3
4-Me2NC6H4C(CH3)2C(CH3)2C6H5
4-Me2NC6H4CH=CH2
(4-MeOC6H4)2C=CH2
4-MeOC6H4C+(cycPr)=CH2
4-MeOC6H4C+(C6H5)=CH2
4-MeOC6H4C+(i-Pr)=CH2
4-MeOC6H4C+(Et)=CH2
4-MeONapCH=CH2
4-MeOC6H4C(CH3)=CH2
{4-MeOC6H4C(CH3)2}2
4-MeOC6H4C+(t-Bu)=CH2
Ph2C=CH2
3,4-(MeO)2C6H3CH=CH2
4-MeOC6H4CH=CH2
4-MeOC6H4CH(OCOCH3)CH3
4-MeC6H4C(CH3)=CH2
C6H5C(CH3)=CH2
59
60
61
62
64
New
65
New
66
67
4-MeOC6H4C+(cycPr)CH3
b
4-MeOC6H4C+(C6H5)CH3
4-MeOC6H4C+(i-Pr)CH3
4-MeOC6H4C+(Et)CH3
4-MeONapCH+CH3
c
4-MeOC6H4C+(CH3)2
4-MeOC6H4C+(t-Bu)CH3
Ph2C+CH3
3,4-(MeO)2C6H3CH+CH3
4-MeOC6H4CH+CH3
68
Comm.
Comm.
Comm.
69
Comm.
Comm.
Comm.
Comm.
Comm.
Comm.
Comm.
Comm.
32
4-MeC6H4C+(CH3)2
C6H6C+(CH3)2
2,6-Me2C6H3CH+CH3
4-MeC6H4CH+CH3
4-FC6H4CH+CH3
3-MeC6H4CH+CH3
C6H5CH+CH3
2,6-Me2C6H3CH=CH2
4-MeC6H4CH=CH2
4-FC6H4CH=CH2
3-MeC6H4CH=CH2
C6H5CH=CH2
4-MeOC6H4CH+CH3
4-MeOC6H4CH=CH2
4-MeONapCH2N+Me3·–OAc
+c
4-MeONapCH2
4-MeOC6H4CH2
4-MeOC6H4CH2N+Me3·BF4
70
+
–
4-MeOC6H4CH2Cl
4-MeC6H4CH2Cl
4-MeOC6H4CBr=CPh2
4-MeOC6H4CBr=CMe2
4-MeOC6H4C/CH
C6H5CBr=CPh2
2,4,6-Me3C6H2C/CH
2,4,6-Me3C6H2CCl=CH2
4-MeC6H4C/CH
Comm.
Comm.
71
72
73
74
75
75
76
+
4-MeC6H4CH2
4-MeOC6H4C+=CPh2
4-MeOC6H4C+=CMe2
4-MeOC6H4C+=CH2
C6H5C+=CPh2
2,4,6-Me3C6H2C+=CH2
4-MeC6H4C+=CH2
C6H5C+=CHPh
C6H5C+=CH2
C6H5C/CC6H5
C6H5C/CH
Comm.
Comm.
aComm. = commercially available.
bcycPr = cyclopropyl.
cNap = naphthyl.
4. R.E. Minto and P.K. Das. J. Am. Chem. Soc. 111, 8858 (1989).
5. R.A. McClelland, N. Banait, and S. Steenken. J. Am. Chem.
Soc. 108, 7023 (1986).
ployed to minimize depletion of the precursor. This was ver-
ified by observing that the rate constants did not depend
significantly on laser dose.
6. R.A. McClelland, V.M. Kanagasabapathy, N. Banait, and S.
Steenken. J. Am. Chem. Soc. 111, 3966 (1989).
7. R.A. McClelland, V.M. Kanagasabapathy, N. Banait, and S.
Steenken. J. Am. Chem. Soc. 113, 1009 (1991).
8. R.A. McClelland, V.M. Kanagasabapathy, N. Banait, and S.
Steenken. J. Am. Chem. Soc. 114, 1816 (1992).
9. R.A. McClelland, V.M. Kanagasabapathy, and S. Steenken. J.
Am. Chem. Soc. 110, 6913 (1988).
Acknow ledgements
Financial support from the Natural Sciences and Engi-
neering Research Council of Canada is gratefully acknowl-
edged. We also thank Jim Pincock (4-methoxy-1-
naphthylmethylammonium acetate) and Joggi Wirz (4-
dimethylaminostyrene) for initial gifts of compounds.
10. J. Bartl, S. Steenken, H. Mayr, and R.A. McClelland. J. Am.
Chem. Soc. 112, 6918 (1990).
11. E.O. Alonso, L.J. Johnston, J.C. Scaiano, and V.G. Toscano. J.
Am. Chem. Soc. 112, 1270 (1990).
References
12. S.L. Mecklenburg and E. Hilinski. J. Am. Chem. Soc. 111,
5471 (1989).
13. R.A. McClelland, N. Mathivanan, and S. Steenken. J. Am.
Chem. Soc. 112, 4857 (1990).
1. P.K. Das. Chem. Rev. 93, 119 (1989).
2. R.A. McClelland. Tetrahedron, 52, 6823 (1996).
3. R.A. McClelland, N. Banait, and S. Steenken. J. Am. Chem.
Soc. 111, 2929 (1989).
© 1999 NRC Canada

2082
Can. J. Chem. Vol. 77, 1999
14. F.L. Cozens, N. Mathivanan, R.A. McClelland, and S.
Steenken. J. Chem. Soc. Perkin Trans. 2, 2083 (1992).
15. R.A. McClelland, F. Cozens, and S. Steenken. Tetrahedron
Lett. 31, 2821 (1990).
46. D.A. McCrery and B.S. Freiser. J. Am. Chem. Soc. 100, 2902
(1978).
47. N. Fujisaki, P. Comte, and T. Gaumann. J. Chem. Soc. Chem.
Commun. 848 (1993).
16. R.A. McClelland, C. Chan, F. Cozens, A. Modro, and S.
Steenken. Angew. Chem. Int. Ed. Engl. 30, 1337 (1991).
17. P. Wan, S. Culshaw, and K. Yates. J. Am. Chem. Soc. 104,
2509 (1982).
48. R.L. Jones and L.M. Dorfman. J. Am. Chem. Soc. 96, 5717
(1974).
49. L.M. Dorfman, R.J. Sujdak, and B. Bockrath. Acc. Chem. Res.
9, 352 (1976).
18. P. Wan and K. Yates. J. Org. Chem. 48, 869 (1983).
19. J. McEwan and K. Yates. J. Am. Chem. Soc. 109, 5800 (1987).
20. P. O’Neill, S. Steenken, and D. Schulte-Frohlinde. J. Phys.
Chem. 79, 2773 (1975).
21. D.P. DeCosta and J.A. Pincock. J. Am. Chem. Soc. 111, 8948
(1989); 115, 2180 (1989).
50. Y. Wang and L.M. Dorfman. Macromolecules, 13, 63 (1980).
51. L. Andrews and B.W. Keelan. J. Am. Chem. Soc. 103, 99
(1981).
52. V. Wintgens and J.C. Scaiano. Can. J. Chem. 65, 2131 (1987).
53. D.D.M. Wayner, D.J. McPhee, and D. Griller. J. Am. Chem.
Soc. 110, 132 (1988).
22. J.A. Hilburn, E. MacKnight, J.A. Pincock, and P.J. Wedge. J.
Am. Chem. Soc. 116, 3337 (1994).
54. H. Mayr, R. Schneider, B. Irrgang, and C. Schade. J. Am.
Chem. Soc. 112, 4460 (1990).
23. T.H. Lowry and K.S. Richardson. In Mechanism and theory in
organic chemistry. 3rd ed. Harper and Row, New York. 1987.
pp. 571–573.
24. T.W. Bentley and Z.H. Ryu. J. Chem. Soc. Perkin Trans 2, 761
(1994).
25. G.A. Olah, C.P. Pittman, Jr., R. Waack, and M. Doran. J. Am.
Chem. Soc. 88, 1488 (1966).
26. T.L. Amyes, T. Mizerski, and J.P. Richard. Can. J. Chem. 5,
922 (1998).
27. V. Bertoli and P.H. Plesch. J. Chem. Soc. B, 1500 (1968).
28. S. Bywater and D. J. Worsfold. Can. J. Chem. 44, 1671 (1966).
29. M. Villesange, A. Rives, C. Bunel, J.-P. Vairon, M. Froeyen,
M. Van Beylen, and A. Persoons. Makromol. Chem.
Makromol. Symp. 47, 271 (1991).
55. H.C. Brown, K.J. Morgan, and F.J. Chloupek. J. Am. Chem.
Soc. 87, 2137 (1965).
56. G.A. Olah, V.P. Reddy, and G.K.S. Prakash. Chem. Rev. 92, 69
(1992).
57. W. Kirmse, B. Krzossa, and S. Steenken. J. Am. Chem. Soc.
118, 7473 (1996).
58. J.P. Richard, M.E. Rothenberg, and W.P. Jencks. J. Am. Chem.
Soc. 106, 1361 (1984).
59. P. Maslak and S.L. Asel. J. Am. Chem. Soc. 110, 8260 (1988).
60. H.K. Hall, L.C. Dunn, and A.B. Padias. J. Am. Chem. Soc. 45,
835 (1980).
61. C.D. Hurd and T. Iwashige. J. Org. Chem. 24, 1321 (1959).
62. R.R. Kostikov, A.P. Molchanov, and S.M. Nagi. J. Org. Chem.
U.S.S.R. (Engl. Trans.) 19, 1291 (1983).
30. T. Kunitake and K. Takarabe. Macromolecules, 12, 1061, 1067
(1979).
63. D.H.R. Barton, L. Bohe, and X. Lusinchi. Tetrahedron, 46,
5273 (1990).
31. M.A. Ratcliff and J.K. Kochi. J. Am. Chem. Soc. 93, 3112
(1971).
32. B. Foster, B. Gaillard, N. Mathur, A.L. Pincock, J.A. Pincock,
and C. Sehmbey. Can. J. Chem. 65, 1599 (1987).
33. H.E. Zimmerman and V.R. Sandel. J. Am. Chem. Soc. 85, 915
(1963).
64. C.D. Hurd and C.N. Webb. J. Am. Chem. Soc. 49, 546 (1927).
65. J.T. Gupton and W.J. Layman. J. Org. Chem. 52, 3683 (1987).
66. D. Seymour and K.B. Wolfstirn. J. Am. Chem. Soc. 70, 1177
(1948).
67. S. Perrier, S. Sandkararaman, and J.K. Kochi. J. Chem. Soc.
Perkin Trans. 2, 825 (1993).
34. J.L. Faria and S. Steenken. J. Phys. Chem. 97, 10869 (1992).
35. P. Maslak, W.H. Chapman, Jr., T.M. Vallombroso, Jr., and B.A.
Watson. J. Am. Chem. Soc. 117, 12380 (1995).
36. W. Schnabel, I. Naito, T. Kitamura, S. Kobayashi, and H.
Taniguchi. Tetrahedron, 36, 3229 (1980).
68. G.A. Olah, M.B. Comisarow, and C.J. Kim. J. Am. Chem.
Soc. 91, 1458 (1969).
69. V.M. Potopov, V.M. Dem’yanovich, V.A. Khlebnikov, and S.S.
Koval’skaya. J. Org. Chem. U.S.S.R. (Engl. Trans.) 21, 1602
(1985).
37. S. Kobayashi, T. Kitamura, H. Taniguchi, and W. Schnabel.
Chem. Lett. 117, (1983), 2101 (1984).
70. E. Hilhurst, A.S. Iskander, T.B.R.A. Chen, and U.K. Pandit.
Tetrahedron, 50, 8863 (1994).
38. F.I.M. Van Ginkel, R.J. Visser, C.A.G.O. Varma, and G.
Lodder. J. Photochem. 30, 453 (1985).
71. L.L. Miller and D.A. Kaufman. J. Am. Chem. Soc. 90, 7282
(1968).
39. S. Kobayashi, Q.Q. Zhu, and W. Schnabel. Naturforsch. Teil
B, 42, 825 (1988).
40. Y. Chiang, A.J. Kresge, M. Capponi, and J. Wirz. Helv. Chim.
Acta, 69, 1331 (1986).
41. Y. Chiang, A.J. Kresge, J.A. Santabello, and J. Wirz. J. Am.
Chem. Soc. 110, 5506 (1988).
42. T. Hochstrasser, A.J. Kresge, N.P. Schepp, and J. Wirz. J. Am.
Chem. Soc. 110, 7875 (1988).
72. Z. Rappoport and J. Kaspi. J. Am. Chem. Soc. 96, 4518 (1974).
73. M. Hanack and E. Weber. Chem. Ber. 116, 777 (1983).
74. R.E. Buckles. J. Am. Chem. Soc. 71, 1157 (1949).
75. T.H. Vaughn and J.A. Nieuwland. J. Am. Chem. Soc. 56, 1207
(1934).
76. T.H. Vaughn. J. Am. Chem. Soc. 56, 2064 (1934).
77. F.N. Tebbe, G.W. Parshall, and G.S. Reddy. J. Am. Chem. Soc.
100, 3611 (1978).
43. P.J. Stang, Z. Rappoport, M. Haneck, and L.R. Subramanian.
In Vinyl cations. Academic Press, New York. 1979. p. 16.
44. I. Hanazaki and S. Nagakura. Tetrahedron, 21, 2441 (1965).
45. J.A. Grace and M.C.R. Symons. J. Chem. Soc. 958 (1959).
78. S.H. Pine, R. Zahler, D.A. Evans, and R.H. Grubbs. J. Am.
Chem. Soc. 102, 3270 (1980).
79. S.H. Pine, R.J. Pettit, G.B. Geib, S.G. Cruz, C.H. Gallego,
T. Tijerna, and R.D. Pine. J. Org. Chem. 50, 1212 (1985).
© 1999 NRC Canada