Direct measurement of ultrafast carbon-carbon cleavage rates via the subpicosecond charge-transfer activation of pinacols

Article abstract of DOI:10.1021/ja980097d

Highly transient (benzpinacol) cation radicals (D₂^+.) and their ultrafast mesolytic fragmentations to the diarylhydroxymethyl radical (D^.) and cation (D⁺) are directly observed on the early picosecond time scale upon the charge-transfer photoactivation of the intermolecular donor-acceptor complexes of pinacol donors with methyl viologen. Ultrashort lifetimes of the cation radicals with τ ~ 10 ps obtain (for the first time) from quantitative kinetics analysis of the time-resolved spectroscopic results. These rapid C-C bond scissions successfully compete with back-electron transfer, which normally predominates on this time scale, and lead to exceptionally high efficiencies for the oxidative fragmentation of pinacols.

Full text of DOI:10.1021/ja980097d

6542
J. Am. Chem. Soc. 1998, 120, 6542-6547
Direct Measurement of Ultrafast Carbon-Carbon Cleavage Rates via
the Subpicosecond Charge-Transfer Activation of Pinacols
T. M. Bockman, S. M. Hubig, and J. K. Kochi*
Contribution from the Chemistry Department, UniVersity of Houston, Houston, Texas 77204-5641
ReceiVed January 9, 1998
Abstract: Highly transient (benzpinacol) cation radicals (D2+•) and their ultrafast mesolytic fragmentations to
the diarylhydroxymethyl radical (D•) and cation (D+) are directly observed on the early picosecond time scale
upon the charge-transfer photoactivation of the intermolecular donor-acceptor complexes of pinacol donors
with methyl viologen. Ultrashort lifetimes of the cation radicals with τ ≈ 10 ps obtain (for the first time)
from quantitative kinetics analysis of the time-resolved spectroscopic results. These rapid C-C bond scissions
successfully compete with back-electron transfer, which normally predominates on this time scale, and lead to
exceptionally high efficiencies for the oxidative fragmentation of pinacols.
Introduction
Chart 1. Pinacol Donors (D2)
The unusually facile cleavage of benzpinacols to afford
quantitative yields of various benzophenones in the presence
of organic and inorganic oxidants1,2 has been attributed to a
labile intermediatesthe pinacol cation radical3 prone to me-
solytic fragmentation.4 However, relatively little quantitative
information is available on the rate of this process, and the direct
and unambiguous observation of the cleavage remains an
experimental challenge. In this study, we demonstrate how the
methodology that we previously employed to determine the rates
of ultrafast reactions5 can be utilized (a) to unambiguously
identify the reactive intermediate and (b) to quantitatively
measure its cleavage rate. The time-resolved fs spectroscopic
method exploits charge-transfer activation to effect the instan-
taneous oxidation6 on the subpicosecond time scale of the
different pinacols identified in Chart 1. As such, this unique
approach allows us now to determine the ultrafast rate constants
for the C-C scission of transient pinacol cation radicals with
lifetimes of τ < 10 ps by direct kinetic measurements as
follows: The pinacols in Chart 1 form donor-acceptor com-
plexes with electron-poor substrates such as quinones, cyano-
arenes, and pyridinium cations.3d,e Charge-transfer activation
of the EDA complex effects the instantaneous transfer of an
electron from donor to acceptor, and thus generates the pinacol
cation radical on a time scale that is rapid compared to its
cleavage.6 Time-resolved spectroscopy is then utilized to
observe and quantify the ultrafast follow-up reactions. As the
reactive acceptor in this reaction, we choose the methyl viologen
dication (4,4′-dimethyl-1,1′-bipyridinium) owing to its well-
defined (redox) spectral changes. Charge-transfer (CT) excita-
tion of the electron donor-acceptor or EDA complexes of
(1) (a) Penn, J. H.; Duncan, J. H. J. Org. Chem. 1993, 58, 2003. (b)
Han, D. S.; Shine, H. J. J. Org. Chem. 1996, 61, 3977. (c) Lopez, L.; Mele,
G.; Mazzeo, C. J. Chem. Soc., Perkin Trans. 1 1994, 779.
(2) For a general review of oxidative cleavages of pinacols, see: House,
H. O. Modern Synthetic Reactions, 2nd ed.; Benjamin: New York, 1972;
p 353f.
(3) (a) Reichel, L. W.; Griffin, G. W.; Muller, A. J.; Das, P. K.; Ege, S.
N. Can. J. Chem. 1984, 62, 424. (b) Davis, H. F.; Das, P. K.; Reichel, L.
W.; Griffin, G. W. J. Am. Chem. Soc. 1984, 106, 6968. (c) Chen, L.; Farahat,
M. S.; Gaillard, E. R.; Farid, S.; Whitten, D. G. J. Photochem. Photobiol.
A 1996, 95, 21. (d) Perrier, S.; Sankararaman, S.; Kochi, J. K. J. Chem.
Soc., Perkin Trans. 2 1993, 825. (e) Sankararaman, S.; Kochi, J. K. J. Chem.
Soc., Chem. Commun. 1989, 1800. (f) Albini, A.; Spreti, S. J. Chem. Soc.,
Perkin Trans. 2 1987, 1175.
various pinacolic donors (D2) with methyl viologen (MV2+
)
)
+•
leads to the ultrafast co-generation of the cation radical (D2
and the reduced acceptor (MV+•) upon the 200-fs laser
excitation, i.e.
(4) (a) Maslak, P. Top. Curr. Chem. 1993, 168, 1. (b) Maslak, P.;
Chapman, W. H., Jr. J. Org. Chem. 1996, 61, 2647. (c) Wayner, D. D. M.;
Parker, V. D. Acc. Chem. Res. 1993, 26, 287.
(5) (a) Hubig, S. M.; Bockman, T. M.; Kochi, J. K. J. Am. Chem. Soc.
1996, 118, 3842. (b) Bockman, T. M.; Hubig, S. M.; Kochi, J. K. J. Org.
Chem. 1997, 62, 2210. (c) Bockman, T. M.; Hubig, S. M.; Kochi, J. K. J.
Am. Chem. Soc. 1996, 118, 4502.
(6) Charge-transfer excitation of electron donor-acceptor complexes
effects the transfer of an electron from the donor to the acceptor in less
than 500 fs to produce the donor cation radical in the ground state. See:
Wynne, K.; Galli, C.; Hochstrasser, R. M. J. Chem. Phys. 1994, 100, 4797.
Thus the CT excited state (as the precursor to the ion-radical pair) must
have a very short lifetime, and it is not chemically significant on the
picosecond time scale.
The subsequent disappearance of the pinacol cation radical and
the reduced methyl viologen may then be observed spectro-
scopically on the early picosecond time scale.
Results and Discussion
Spectral Identification of Highly Transient (Pinacol)
Cation Radicals. Charge-transfer excitation of the EDA
S0002-7863(98)00097-3 CCC: $15.00 © 1998 American Chemical Society
Published on Web 06/18/1998

Direct Measurement of Ultrafast C-C CleaVage Rates
J. Am. Chem. Soc., Vol. 120, No. 26, 1998 6543
Table 1. Rates of Carbon-Carbon Scission of Pinacol Cation
Radicals as Transient Intermediates in the Charge-Transfer
Activation of EDA Complexesa
pinacolb kc (1010 s-1
)
Rd
kCe,g (1010 s-1
)
kBETf,g (1010 s-1
)
1
2
3
4
5
6
7
6.0
4.8
3.8
0.30
0.11
0.09
0.35
0.12
0.11
0.16
1.4h
0.5
0.3
8.0
4.9
5.3
6.0
4.6h
4.3
3.5
15
36
44
23
41
50
38
32
a In acetonitrile at 25 °C. b Identified in Chart 1. c Observed rate
constant for the decay of the ion radicals D2 and MV+•
.
d Relative
+•
Figure 1. Transient absorption spectrum (- - -) observed upon the 400-
nm (200-fs) laser excitation of the EDA complex of MV2+ and pinacol
2 in acetonitrile, and the absorption band of 2+• (s) obtained by spectral
subtraction of the authentic absorption of MV+• (‚‚‚). The insets show
amounts of residual MV+• after the complete decay of D2+• (see text).
f
e Rate constant for C-C bond cleavage in eq 2. Rate constant for back
electron transfer in eq 2. g Calculated from the values of R and k (see
ref 18), unless otherwise indicated. h Obtained from the digital simula-
tion in Figure 3.
the decays of the 605- and 730-nm absorptions of MV+• and 2+•
,
respectively, fitted to first-order kinetics (solid lines) both with k ) 5
× 1010 s-1
.
Direct Observation of the Ultrafast (Pinacol) Cleavage.
The predominant pathway for the decay of ion radicals generated
by charge-transfer excitation of EDA complexes is back-electron
transfer within the initial ion-radical pair.12 Since complete
back-electron transfer restores the original starting components
of the EDA complex, it results in a zero baseline of the transient
spectra. On the other hand, the observation of residual
absorbances (R) pointed to a more complex reaction scheme in
which at least one other decay pathway competed with the
ubiquitous back-electron transfer.13 As such, we now identify
complex of pinacol 2 with MV2+ in acetonitrile generated
(within 500 fs) a transient spectrum with a broad absorption
centered at 605 nm and extending beyond 750 nm (see Figure
1). Digital subtraction of the authentic spectrum of the reduced
methyl viologen (MV+•)7 in Figure 1 revealed an absorption
band centered at 730 nm, which was readily assigned to the
cation radical of 2.8 Quantitative comparison of the absorption
bands of MV+• and 2+• initially upon laser excitation established
the requisite 1:1 stoichiometry for the production of the two
ion radicals as described in eq 1.9 The assignment of the 730-
nm band to the pinacol cation radical, 2+• was further confirmed
by its independent generation via the charge-transfer irradiation
of the EDA complex of 2 with dimethyl dicyanofumarate.10 In
this experiment, the band of the cation radical (2+•) was
observed without interference from the absorptions of the
reduced acceptor, which was transparent in this spectral region.
+•
the mesolytic cleavage (kC) of the pinacol cation radical D2
into its fragments D• and D+ as the additional decay pathway
that generates residual amounts of reduced methyl viologen in
accord with reaction 2 (Scheme 1). To establish the validity
of the bond scission (kC) in Scheme 1, we utilized time-resolved
Scheme 1
The absorbances of D2 at 730 nm and MV+• at 605 nm
+•
both decreased concomitantly on the picosecond time scale with
identical (first-order) rate constants of k ) (5 ( 0.5) × 1010
s-1. Whereas the absorption band of D2+• fully decayed to the
spectral baseline within 50 ps, the simultaneous decay of MV+•
resulted in a residual intensity at 50 ps of ∼15%, as shown in
the insets of Figure 1; and this absorption remained unchanged
over nanoseconds. The same kinetics was observed upon the
subpicosecond excitation of the EDA complexes of the pinacols
4-7 in Chart 1si.e., the complete decay of the cation radical
D2+• 11 in less than 30 ps and the partial decay of MV+• on the
same time scale to the residual absorbances (R) listed in Table
1. In all cases, the kinetic evaluation of the decays of both ion
radicals afforded identical (first-order) rate constants, which
permitted us to select the common spectral peak at λmax ) 605
nm as the general monitoring wavelength for the decay of all
the pinacol cation radicals in Chart 1.
spectroscopy to detect the characteristic cleavage product D•
of the pinacol cation radicals. Most importantly, the formation
of D• on the picosecond time scale was simultaneous with the
decays of D2 and MV+•. For example, charge-transfer
+•
excitation of the EDA complex of pinacol 1 with MV2+ in
acetonitrile led initially to a transient absorption spectrum that
was identical with the authentic spectrum of reduced methyl
viologen7 (see Figure 2A). At later times (50-200 ps after
excitation), additional absorptions progressively grew in on the
blue edge of the 605-nm band of MV+• (see Figure 2B). Digital
subtraction of the authentic spectrum of MV+• 7 from the
composite spectrum in Figure 2B revealed the new absorption
band centered at 535 nm, which was readily assigned to the
diphenylhydroxymethyl radical.14 This radical (D•) was clearly
(7) (a) Watanabe, T.; Honda, K. J. Phys. Chem. 1982, 86, 2617. (b)
Bockman, T. M.; Kochi, J. K. J. Org. Chem. 1990, 55, 4127.
(8) The absorption spectra of 2+• and 3+• are essentially the same as
those of the cation radicals of the methylbiphenyls, which were generated
independently by CT excitation of their EDA complex with tetracyanoben-
zene. Compare: Ojima, S.; Miyasaka, H.; Mataga, N. J. Phys. Chem. 1990,
94, 7534.
(11) The cation radicals of the methoxy-substituted pinacols absorb in
the same wavelength region (450-470 nm) as the anisole cation radical.
See: Takamaku, S.; Komitsu, S.; Toki, S. Radiat. Phys. Chem. 1989, 34,
553.
(12) See: (a) Asahi, T.; Mataga, N. J. Phys. Chem. 1989, 93, 6575. (b)
Asahi, T.; Mataga, N. J. Phys. Chem. 1991, 95, 1956. (c) Peters, K. S.
AdV. Electron-Transfer Chem. 1994, 4, 27. (d) Hilinski, E. F.; Masnovi, J.
M.; Kochi, J. K.; Rentzepis, P. M. J. Am. Chem. Soc., 1984, 106, 8071. (e)
Fox, M. A. AdV. Photochem. 1986, 13, 237.
(13) For examples, see: Bockman et al. in ref 5b and Masnovi et al.
(Masnovi, J. M.; Kochi, J. K.; Hilinski, E. F.; Rentzepis, P. M. J. Am. Chem.
Soc. 1986, 108, 1126). See also Ojima et al. in ref 8.
(9) Based on ꢀ605 ) 14 100 M-1 cm-1 as the extinction coefficient of
reduced methyl viologen at 605 nm.7 The extinction coefficient of 2+• at
730 nm is taken as that of the 4-methylbiphenyl cation radical (ꢀ730 ≈ 10 000
M-1 cm-1 ).8
(10) Dimethyl dicyanofumarate has been identified as a versatile electron
acceptor in thermal electron-transfer reactions. See: Gotoh, T.; Padias, A.
B.; Hall, H. K., Jr. J. Am. Chem. Soc. 1986, 108, 4920.
(14) See: Hayon, E.; Ibata, T.; Lichtin, N. N.; Simic, M. J. Phys. Chem.
1972, 76, 2072, and ref 5b.

6544 J. Am. Chem. Soc., Vol. 120, No. 26, 1998
Bockman et al.
Figure 2. (A) Transient absorption spectrum of only MV+• (- - -) obtained at 20 ps following the 355-nm (25-ps) laser excitation of the EDA
complex of MV2+ and pinacol 1 in acetonitrile in comparison with authentic MV+• (- - -). (B) Composite spectrum (- - -) obtained 200 ps later, and
spectrum of D• (s) obtained by subtraction of the absorption band of MV+• (shown in part A).
the direct product of the mesolytic fragmentation of the
benzpinacol cation radical (D2+•) according to eq 3:
values for the transient absorptions of these species are
superimposed in Figure 3, and the excellent agreement between
the computer simulation and the experimental points confirms
the kinetics formulation in Scheme 1. On the basis of this
scheme, the absolute values of kC and kBET for all the pinacol
donors were computed from the overall rate constants of the
ion-radical decay (k) and the residual (R) amounts of MV+•
The characteristic absorption band of the radical D• with λmax
) 535 nm was also observed upon charge-transfer excitation
of the EDA complexes of 1 with other acceptors such as 1,2,4,5-
tetracyanobenzene (TC) and 4-cyano-N-methylpyridinium (CP+).
For example, laser activation of the EDA complex [1, TC] gave
rise on the early picosecond time scale to the spectral absorption
bands of the same radical D• as well as the anion radical of the
new acceptor TC-•.15 The spectral absorption band of D•
observed upon irradiation of the complex of 1 with the cationic
acceptor CP+ was unencumbered by the absorption of the
reduced neutral acceptor (CP•), which does not absorb visible
light.16 In other words, the one-electron oxidation of benzpi-
nacol with a variety of acceptors, including neutral (TC),
monocationic (CP+), and dicationic (MV2+) types, always led
to the same cleavage of the pinacol cation radical 1+• (see eq
3).
according to the equations R ) kC/(kC + kBET) and k ) kC +
18
kBET
,
and the results are listed in Table 1.
Enhanced Efficiency of Pinacolic Cleavage Wia Charge-
Transfer Activation. Charge-transfer photoactivation sponta-
neously generates D2+• as the ion-radical pair within 500 fs, in
accord with the Mulliken formulation in eq 1.6 Previous
investigations have shown that the initial photoinduced electron
transfer (hνCT) is usually followed by an ultrafast reverse
electron transfer (kBET) to regenerate the pinacol D2 as the
ground-state EDA complex.12 However, the persistence of the
residual absorption of MV+• in Figures 1 and 2 points to another
process that must compete with the back-electron transfer. Since
the formation of the radical D• as the direct product of C-C
bond cleavage was observed simultaneously with the rise and
partial decay of MV+• in Figure 3, the fragmentation of the
Picosecond Rates for the Carbon-Carbon Bond Cleav-
ages. Quantitative comparison of the absorption bands of D•
and MV+• indicated that the amount of reduced methyl viologen
(residual after 200 ps) was equivalent to the amount of radical
D• formed by the cleavage of the pinacol cation radical D2
+•
within the same time span, and thus confirmed the stoichiometry
in eq 2.17 [Note that the diphenylhydroxymethyl cation (D+)
does not absorb visible light.17b] Accordingly, the complex
kinetics in Scheme 1 was digitally simulated by taking into
account the finite rise time of the laser pulse (hνCT); and Figure
3 typically shows the kinetic profiles of both the cation radical
MV+• and the radical D•, calculated with the rate constants kC
) 1.4 × 1010 s-1 and kBET ) 4.6 × 1010 s-1. The experimental
(15) Ojima, S.; Miyasaka, H.; Mataga, N. J. Phys. Chem. 1990, 94, 4147.
(16) Itoh, M.; Nagakura, S. Bull. Chem. Soc. Jpn. 1966, 39, 369.
(17) (a) The extinction coefficient of the ketyl radical is ꢀ535 ) 5800
Figure 3. Formation and decay of MV+• (b) and the formation of D•
(O) upon excitation of [MV2+, 1] presented in Figure 2. The solid lines
correspond to the simulated changes of MV+• and D• calculated
according to the kinetics formulation in Scheme 1.
M-1 cm-1 14
. For the extinction coefficient of the reduced methyl viologen,
see ref 9. (b) Ireland, J. F.; Wyatt, P. A. H. J. Chem. Soc., Faraday Trans.
1 1973, 69, 161.
(18) Asahi, T.; Ohkohchi, M.; Mataga, N. J. Phys. Chem. 1993, 97,
13132.

Direct Measurement of Ultrafast C-C CleaVage Rates
J. Am. Chem. Soc., Vol. 120, No. 26, 1998 6545
Chart 2
pinacol cation radical (kC in eq 3) must effectively compete with
back-electron transfer in Scheme 1.
The magnitudes of the cleavage rate constants for the various
pinacol donors in Table 1 are strikingly insensitive to the
substituents (hydrogen, phenyl, or methoxy) on the phenyl rings.
In particular, we note in Chart 2 that the structurally related
series of pinacols 2, 3, and 6 follow a Hammett correlation, in
which both log kC and log kBET decrease linearly with the σ
constant. In other words, the mesolytic fragmentation and
reverse electron transfer respond to substituent effects in
precisely the same way, as quantitatively reflected in the values
of the Hammett parameter, F ) -3.8 and -3.5, respectively.
As a result, the overall efficiency of cleavage Vis a Vis back-
electron transfer remained more or less the same with all
substituents. The rather invariant ratio between kC and kBET
for the various substituted pinacols in Table 1 thus accounts
for the previously puzzling observation that the quantum yields
are singularly independent of the substituents.3d
Figure 4. Comparison of the experimental data for the formation of
the radical D• (O) with the computer simulations of its rise with varying
values of the separation rate constant in eq 4 (see text). The solid lines
(top to bottom) are simulated with values of kS ) 0, 4 × 109, 6 × 109,
8 × 109, and 10 × 109 s-1, and the error bars represent the experimental
uncertainty of (10%.
which the latter may complicate the formulation in Scheme 1,
let us consider the competition for the contact ion pair in Scheme
1 as:
where kS represents the (composite) rate constant for all such
(first-order) processes. Such a competition(s) is readily evalu-
ated by the digital simulation of Scheme 1 as modified by the
inclusion of kS to convert the initial contact ion pair to a species
that does not undergo back-electron transfer in the ps time
domain. Importantly, kinetics analysis (as detailed in the
Experimental Section) predicts a distinct time lag between the
back-electron transfer (as monitored by the decay of MV+•) and
the cleavage (as monitored by the rise of D•). Figure 4 shows
that such a time lag is not observed, and we must conclude that
kS is more than a factor of 10 slower than the cleavage rate
constant, kC.24 In other words, the mesolytic cleavage of
pinacols not only competes with ultrafast back-electron transfer,
but it clearly outruns all other ion-dynamical processes (such
as solvent separation and recombination) that are generally
involved in photoinduced electron transfers.25 As a result, the
elementary chemical steps of C-C bond cleavage and back-
electron transfer of the ion-radical pair within its initial solvent
cage stand alone from the usual (extramural) physicochemical
processes of ion-pair relaxation, diffusion, and migration that
complicate photoinduced electron transfers on longer time scales.
We believe that the positive trend in kC with donor substit-
uents relates mainly to the increasing stability of the fragmenta-
tion products, D• and D+, especially the latter, as the substituent
is changed.4b Indeed, the cleavages of all the pinacol cation
radicals in Table 1 are ultrafast, with unimolecular rate constants
in the range kC ) 1010-1011 s-1 and activation energies of less
than 5 kcal mol-1 19
Thus the dissociation energy of the central
.
carbon-carbon bond in the benzpinacol cation radical is reduced
by about 35 kcal mol-1 compared to that of the neutral pinacol
prior to oxidation.20 In fact, the fast rate constants (kC) which
are comparable to those of the competing back-electron transfer
(kBET) result in unusually high photochemical efficiencies for
the charge-transfer activated cleavages of pinacols.21 The high
quantum yields are particularly noteworthy, since photochem-
istry via charge-transfer activation is commonly limited by the
efficient (energy-wasting) back-electron transfer.22
Ultrafast Cleavages That Obviate Ion-Pair Microdynam-
ics. The successful analysis of the transient (ps) spectral changes
in Figure 3 derives from the remarkably straightforward
mechanism in Scheme 1. As such, it is especially noteworthy
that Scheme 1 is singularly devoid of other (physical) processes
that are usually encountered in photoinduced (electron-transfer)
microdynamics such as ion-pair dissociation/reassociation, sol-
vation, intersystem crossing, etc.23 To establish the limits to
(23) See, e.g.: (a) Kavarnos, G. J.; Turro, N. J. Chem. ReV. 1986, 86,
401. (b) Gould, I. R.; Farid, S. Acc. Chem. Res. 1996, 29, 522. (c) Mauzerall,
D. C. In Photoinduced Electron Transfer: Part A; Fox, M. A., Chanon,
M., Eds.; Elsevier: Amsterdam, 1988; p 228.
(24) (a) Such a rate constant of kS < 109 s-1 is consistent with simple
diffusive separation of the ions in contact ion pairs. See: (b) Knibble, H.;
Rehm, D.; Weller, A. Ber. Bunsen-Ges. Phys. Chem. 1968, 72, 257. (c)
Ojima, S.; Miyasaka, H.; Mataga, N. J. Phys. Chem. 1990, 94, 7534. For
the separation of cation-cation pairs, see: (d) Hubig, S. M. J. Phys. Chem.
1992, 96, 2903.
(19) Estimated on the basis of the Eyring equation. See: Moore, J. W.;
Pearson, R. G. Kinetics and Mechanism, 3rd ed.; Wiley: New York, 1981;
p 173.
(20) The dissociation energies for the central C-C bonds in the neutral
pinacols are taken as those of the corresponding dimethyl ethers (38-46
kcal mol-1). See: Birkhofer, H.; Beckhaus, H.-D.; Peters, K.; von Schnering,
H.-G.; Ru¨chardt, C. Chem. Ber. 1993, 126, 1693.
(25) Since the critical intermediate in CT photoactivation is the contact
ion pair [D2+•, MV+•], the competing processes of back-electron transfer
and C-C bond scission occur within the solvent cage, and do not involve
solvent-separated or free species. However, such a formulation does not
preclude the involvement of a variety of contact ion pairs (relaxed or
unrelaxed) with different rates of back electron transfer and cleavage. It is
notable, however, that the identical competition between cleavage and back
ET governs the reactions of both positive-positive ion pairs [D2+•, MV+•],
as well as oppositely charged ion pairs [D2+•, TC-•].
(21) Note that the subsequent oxidation of D• by MV2+ to complete the
stoichiometry for the retropinacol reaction occurs by second-order kinetics
and is too slow to be observed on the picosecond time scale.
(22) The photochemical quantum yields of CT-activated photoreactions
are generally very low (<10-2). See, e.g.: Jones, G., II In Photoinduced
Electron Transfer: Part A; Fox, M. A., Chanon, M., Eds.; Elsevier:
Amsterdam, 1988; p 245.

6546 J. Am. Chem. Soc., Vol. 120, No. 26, 1998
Bockman et al.
1,2-ethanediol (benzpinacol):29 yield 78%, mp 187-188 °C. 1,2-
Diphenyl-1,2-bis(4′-methoxyphenyl)-1,2-ethanediol: yield 64% as a
1:1 mixture of isomers. 1H NMR (CDCl3): δ 7.34 (m, 4H), 7.20 (m,
10H), 6.73 (m, 4H), 3.83 (s, 3H, isomer 1), 3.81 (s, 3H, isomer 2).
The bis-trimethylsilyl ether, 2,3-bis-(4′-methoxyphenyl)-2,3-bis(tri-
methylsiloxy)butane, was prepared by the reductive coupling of 4′-
methoxyacetophenone with magnesium in hexamethylphosphoric tri-
amide in the presence of chlorotrimethylsilane and was recrystallized
as a single isomer from ethyl acetate and hexane.30 1H NMR
(CDCl3): δ 7.42 (m, 4H), 6.83 (m, 4H), 3.83 (s, 6H), 1.37 (s, 6H),
0.02 (s, 18H).
Instrumentation. The time-resolved spectroscopic measurements
on the femtosecond and on the picosecond and nanosecond/microsecond
time scales were acquired with laser spectroscopic systems that have
been described previously.31 1H and 13C NMR spectra were recorded
on a General Electric QE-300 NMR spectrometer and the chemical
shifts are reported in ppm units downfield from tetramethylsilane.
Melting points were recorded on a Mel-Temp apparatus and are
uncorrected.
Preparation of Samples for Time-Resolved Spectroscopy. Solu-
tions of the acceptors (0.04-0.5 M) and the pinacols (0.02-0.1 M)
were prepared in acetonitrile for experiments on the femtosecond and
on the picosecond time scales. Preliminary experiments indicated that
laser irradiation generated large amounts of reduced acceptor, which
interfered with the time-resolved measurements. Accordingly, the
solutions were prepared in the open air atmosphere and were not
degassed. Since the problem was particularly acute for the methyl
viologen acceptor, it necessitated the solutions to be saturated with pure
oxygen prior to spectroscopic measurements. The sample cell for the
experiments with methyl viologen was a 1.0 cm quartz flow cell, and
the sample solution was circulated with a Labline 5514 peristaltic pump
to minimize problems arising from secondary photolysis of the products.
Connections were made with Teflon and Fluoran tubing. The UV-
vis spectra of the circulating solution were acquired before and after
laser photolysis to ensure that no significant thermal or photochemical
changes had occurred. The experiments with the other acceptors were
carried out under nonflow conditions, and the sample was changed after
each spectral acquisition to minimize problems of secondary photolysis.
The samples were further checked by acquisition of spectra prior to
the laser pulse, and the solution was changed if there was any indication
of long-lived spectral features. Each spectrum was an average of 100
laser shots, and the spectra were smoothed with a 5-point adaptive
smoothing routine.
The yields of residual reduced methyl viologen (R) upon excitation
of the pinacol/MV2+ complexes were measured by the method of
transient actinometry described earlier.5b Solutions of methyl viologen
ditriflate and the pinacol donor were prepared in a 1-cm quartz cuvette
equipped with a Teflon valve and the absorbance at 355 nm was
adjusted to match that of the actinometer solution. The solutions were
then degassed by four freeze-pump-thaw cycles. The transient
absorbance of MV+• was determined at 605 nm. The trace was linearly
extrapolated to zero time to ensure that the secondary formation of the
reduced acceptor, which was formed on the 50-500 ns time scale,
was not included. The values of R so obtained did not vary with the
energy of the laser pulse (5-25 mJ).
We hope that further studies on the femtosecond and picosecond
time scales will bring to solution-phase chemistry the elegance
and simplicity that characterize the dynamics in the gas phase.
Summary and Conclusions
Charge-transfer (laser) excitation of the EDA complexes of
various substituted benzpinacols with methyl viologen effects
ultrafast cleavage of the pinacol, and time-resolved (fs/ps)
spectroscopy offers the opportunity to quantitatively monitor
the retropinacol reaction step by step. Scheme 1 presents
carbon-carbon bond scission and back-electron transfer as the
two dominant pathways for the decay of the pinacol cation
radical, and the rate constants for both competing processes can
thus be determined from the temporal decay of the transients.
Direct kinetic measurements reveal for the first time a series of
ultrafast (1010 to 1011 s-1) cleavage rates of pinacol cation radical
which can readily compete with back-electron transfer within
the ion-radical pair. As a result, efficient pinacol cleavage can
be achieved by simple photoactivation of the electron donor-
acceptor complexes of pinacols with various acceptors.3d
Experimental Section
Materials. Benzophenone (Aldrich) was recrystallized from ethanol
before use. 4-Methoxyacetophenone, 4-methoxybenzophenone, 4,4′-
dimethoxybenzophenone, 4-acetylbiphenyl, and 1,2,4,5-tetracyanoben-
zene (Aldrich) were used as received. Samarium diiodide was used in
the form of a 0.1 M solution in tetrahydrofuran as supplied by Aldrich.
The solvents acetonitrile and dichloromethane were spectrophotometric
grade and used as received. Tetrahydrofuran was distilled from sodium
and benzophenone under an inert argon atmosphere. Dimethyl dicyano-
fumarate was prepared by the method reported in the literature.26
Synthesis of the pinacol donors. The pinacols substituted with
methyl groups on the bridge were synthesized by reductive coupling
of the corresponding acetophenones, promoted by samarium diiodide,
as follows. Under an inert atmosphere of argon, the acetophenone (0.01
mol) was dissolved or suspended in 50 mL of tetrahydrofuran at room
temperature. A solution of samarium diiodide (100 mL; 0.1 M) in
tetrahydrofuran was added through a stainless steel cannula by using a
positive pressure of argon. The mixture was stirred magnetically as
the dark blue-green solution became pale yellow in color. The mixture
was stirred at room temperature for 2 h more, and 30 mL of saturated
sodium bicarbonate was then added. The reaction mixture was filtered
through a 5-cm pad of Celite and washed with a total of 200 mL of
diethyl ether. The organic fraction of the filtrate was separated, and
the aqueous layer was extracted with diethyl ether (3 × 50 mL). The
combined organic layers were washed sequentially with 5% aqueous
sodium bisulfite and water, and then dried with anhydrous magnesium
sulfate. The solvent was removed by rotary evaporation, and the crude
product was adsorbed on silica gel. The pinacols were isolated by
chromatography on a silica gel column by using mixtures of hexane
and ethyl acetate as eluants and were purified by recrystallization from
the same solvent mixture. 2,3-Bis(4′-methoxyphenyl)-2,3-butanediol:
yield 60%, as a 3:1 mixture of meso and dl isomers.27 1H NMR
(CDCl3): meso isomer (CDCl3) δ 7.14 (m, 4H), 6.80 (m, 4H), 3.83 (s,
6H), 1.49 (s, 6H). dl isomer (CDCl3) δ 7.16 (m, 4H), 6.76 (m, 4H),
3.80 (m, 6H), 2.2 (s, OH), 1.57 (s, 6H). 2,3-Bis-(4′-biphenylyl) 2,3-
butanediol: yield 48% as a mixture of isomers. 1H NMR (CDCl3): δ
7.78-7.28 (m, 18H), 1.67(s, major isomer), 1.61 (s, minor isomer).
2,3-bis-(3′-biphenylyl) 2,3-butanediol: yield 20% as a single isomer.
1H NMR (CDCl3): δ 7.50-7.28 (m, 18 H), 2.7 (s, 2H), 1.72 (s, 6H).
The other pinacols were synthesized by the photoreduction of the
corresponding benzophenones in isopropyl alcohol and were recrystal-
lized from the same solvent. 1,1,2,2-Tetrakis-(4′-methoxyphenyl)-
1,2-ethanediol:28 yield 20%; mp 180-182 °C. 1,1,2,2-Tetraphenyl-
Kinetics Analysis of the Formation and Decay of the Radical
Pairs. The absorbance of reduced methyl viologen on the picosecond
time scale was analyzed as described earlier,5b except that the temporal
pulse shape of the laser was simulated as a Gaussian (Iabs ) exp[-(t/
τ)2]).32 A value of τ ) 12 ps (corresponding to a 21-ps pulse width)
was obtained by simulating the rise of the signal of reduced methyl
viologen in Figure 3. The kinetics in Scheme 1 were numerically
integrated with a Microsoft QBASIC program that calculated the
(29) Bachmann, W. E. In Organic Syntheses; Blatt, A. H., Ed.; Wiley:
New York, 1943; Collect. Vol. II, p 71.
(30) Perrier, S. Ph.D. Thesis, University of Houston, 1995.
(31) (a) For the flash photolysis experiments, a Ti:sapphire laser system
(200 fs, 400 nm, 2 mJ)5a and a Nd: YAG laser (25 ps, 355 nm, 10 mJ)31b
were used. (b) Bockman, T. M.; Kochi, J. K. J. Chem. Soc., Perkin Trans.
2 1996, 1633.
(26) Ireland, C. J.; Jones, K.; Pizey, J. S.; Johnson, S. Synthet. Commun.
1976, 6, 185.
(27) Sisido, K.; Nozaki, H. J. Am. Chem. Soc. 1948, 70, 778.
(28) Depovere, P.; Devis, R. Bull. Soc. Chim. Fr. 1968, 6, 2470.
(32) Peters, K. S.; Li, B. J. Phys. Chem. 1994, 98, 401.

Direct Measurement of Ultrafast C-C CleaVage Rates
J. Am. Chem. Soc., Vol. 120, No. 26, 1998 6547
integrals by means of the trapezoid rule. The temporal step size was
1 ps and the data were plotted with use of the program Slidewrite Plus.
The best fit for k in Figure 3 was obtained by overlaying the
experimental data points with the simulated rises and decays for various
values of k. A fixed value of kCC ) 0.3kBET was used to ensure the
agreement of the residual absorbance of MV+• with the value of R )
0.3 for the methyl viologen/benzpinacol complex determined on the
nanosecond time scale.
The possible incursion of reactions other than C-C bond cleavage
(which might give rise to residual ion radicals) was tested by digital
simulation. Scheme 1 was modified as follows: A reaction step, with
rate constant kS, was added. This step converted the contact ion-radical
pair to an “inactive” ion pair that was constrained from undergoing
back-electron transfer (kBET). The rate constant for the cleavage of the
pinacol cation radical in the inactive ion pair was taken to be the same
as that in the contact pair (i.e., kC). The sum kC and kS was kept constant
at 1.4 × 1010 s-1 to agree with the experimentally observed quantum
yield for the formation of MV+• (0.3). The simulation generated the
family of working curves shown in Figure 4 which represented the
rise of the radical (D•). All of the rises simulated with kS > 0 showed
a distinct time lag between the rise and decay of MV+• (see Figure 3)
and the formation of D•. Since the experimental results (dots) showed
no time lag and lay on the curve corresponding to kS ) 0.0, we readily
concluded that the “separation” process was not significant on this time
scale. The maximum value of kS estimated from the error in the
datapoints (10%) is about 1 × 109 s-1
.
The relative amounts of the various transients were determined by
comparing their absorbances and quantifying them by using the
following extinction coefficients ꢀmax (λmax): MV+•, 14100 M-1 cm-1
(605 nm);7 diphenylhydroxymethyl (ketyl radical), 5800 M-1 cm-1 (535
nm).14 A value of ꢀmax ) 10000 M-1 cm-1 was used for the biphenyl-
substituted pinacol cation radical in accord with the value for biphenyl
cation radical.8
Acknowledgment. We thank the National Science Founda-
tion and the R. A. Welch Foundation for financial support.
JA980097D