EVidence for a Radical Relay Mechanism
J. Am. Chem. Soc., Vol. 118, No. 9, 1996 2187
enhanced rates for ≈DPP decomposition compared with the inert
spacer molecules.
competitively abstract H from ≈DPP (eq 9) or D from ≈DPM-
d2 (eq 10), respectively, giving a measured PhCH3/PhCH2D
To gain additional evidence for the involvement of ≈DPM
in the novel radical relay process depicted in Figure 3,
thermolysis studies of ≈DPP/≈DPM-d2 were conducted where
the deuterium is located at the active methylene position as
shown in Figure 1. GC-MS analysis of ≈DPP/≈DPM-d2
reactions conducted at <5% conversion gave a deuterium
distribution in the vapor-phase toluene product of 50% PhCH3
and 50% PhCH2D. GC-MS analysis (see Experimental Section)
of the surface-attached toluene product, following liberation
from the surface as cresol and derivatization to the trimethylsilyl
ether, gave the composition 39% ≈PhCH3, 51% ≈PhCH2D, and
•
PhCH + ≈Ph(CH ) Ph f PhCH + 1 or 2
(9)
2
2 3
3
•
•
PhCH + ≈PhCD Ph f PhCH D + ≈PhCD Ph (10)
2
2
2
product ratio of 1.0. This value along with the ≈DPM-d2/≈DPP
surface coverage ratio of 2.9 leads to k≈DPP/k≈DPM-d2 ) 1.4 for
H(D) abstraction on a per active H(D) basis. The relative rate
for hydrogen abstraction by PhCH2 from DPM-h2 versus DPP
has been measured as 3.3 (per active hydrogen basis) at 170
•
2
2
°C. Extrapolation of this value to 375 °C and making the
reasonable assumption that surface immobilization will not have
any significant effect gives k≈DPM-h2/k≈DPP ) 2.2. Hence, a
kinetic isotope effect of k≈DPM-h2/k≈DPM-d2 ) 1.4 × 2.2 ) 3.1 at
375 °C would be required to give the observed PhCH3/PhCH2D
product ratio, which is consistent with the data presented above.
What is the origin of the efficacy of the radical relay process
observed for the ≈DPP/≈DPM system illustrated in Figure 3?
Our current hypothesis centers around the fact that bimolecular
hydrogen transfer steps are occurring between reactants that are
constrained to the surface and are, therefore, preorganized for
reaction. This preorganization can be considered as resulting
in a “reduction in the dimensionality of the reaction space”,
which has been proposed to contribute to enhanced rates of
reactions on the surfaces of layered solids such as clays.2
Enhanced rates for bimolecular reactions under such conditions
then result from enhanced encounter rates. An implication of
this argument is that solid support surfaces with fractal dimen-
sion, D, of ca. 2 should be more likely to exhibit this kinetic
1
0% ≈PhCHD2. No deuterium incorporation could be mea-
sured in the PhCHdCH2 and ≈PhCHdCH2 products. The
results provide direct evidence for the involvement of D (H)
transfer between ≈DPM-d2 (-h2) and both chain-carrying vapor-
phase and surface-attached benzyl radicals. The formation of
≈
PhCHD2 suggests that the surface-attached toluene product
can also participate in the hydrogen transfer, radical relay
process to some extent. This observation is supported by studies
of dimethylbenzene spacer molecules described later. If the
hydrogen transfer, radical relay process described above is the
key reason for the enhanced thermolysis rates for ≈DPP in the
presence of ≈DPM spacers, then one would expect to observe
a kinetic isotope effect for the ≈DPP/≈DPM-d2 case. From a
comparison of the rate data at 375 °C for ≈DPP/≈DPM-h2 and
3,24
≈
DPP/≈DPM-d2 in Table 3, a kinetic isotope effect of 2.7 (
0
.2 can be calculated. In the absence of tunneling, the maximum
deuterium kinetic isotope effect, kH/kD, for hydrogen transfer
between carbon centers based on the zero-point energy differ-
ence for the C-H stretching mode is 2.4 at 375 °C.18 However,
large isotope effects in excess of the classical limits have been
23
enhancement effect than porous solids with D-values near 3.
The fractal dimension for a nonporous, fumed silica support,
Aerosil 200, analogous to the one employed in this work has
been measured to be 2.08 and is consistent with the argument
19
measured for hydrogen abstraction by benzyl and 2-allylben-
18
zyl radicals at 160-170 °C, and the involvement of tunneling
effects on these H-atom transfers has been demonstrated. An
example relevant to the current work is H abstraction from
diphenylmethane, for which isotope effects of 6.6 at 160 °C
2
5
presented above.
Therefore, the pyrolysis mechanism of ≈DPP in the presence
of ≈DPM must be augmented by the inclusion of the additional
propagation steps 11-13 shown below. In this process, the
(
2-allylbenzyl radical)18 and 7.1 at 170 °C (benzyl radical)19
have been reported. Hence, as a consequence of the involve-
ment of tunneling in the H transfer between benzylic sites,
extrapolation of these isotope effect values to higher tempera-
tures is uncertain. However, an isotope effect of at least 3 at
•
•
PhCH (or ≈PhCH ) + ≈PhCH Ph f
2
2
2
•
PhCH (or ≈PhCH ) + ≈PhCH Ph (11)
3
3
4
20 °C has been invoked to explain the H(D) incorporation
•
•
≈
PhCH Ph + ≈PhCH Ph f ≈PhCH Ph + ≈PhCH Ph
(
2
2
selectivity in the toluene product from thermolysis of PhCD2-
CH2Ph under nitrogen.20 Furthermore, in our recent study of
the thermolysis of liquid-phase PhCD2CH2OPh, a kinetic isotope
effect of 2.6 at 375 °C was observed compared to the protium
analog.21 Hence, the measured isotope effect of 2.7 at 375 °C
for the thermolysis of ≈DPP in the presence of ≈DPM-h2 (d2)
is consistent with the proposed radical relay process in which
H-atom transfers from the DPM spacer molecules play a key
role in controlling the rate of ≈DPP thermolysis.
12)
•
≈
PhCH Ph + ≈Ph(CH ) Ph f ≈PhCH Ph + 1 or 2 (13)
2
3
2
diphenylmethane molecules serve as catalysts for translocating
radical centers across the surface by a hydrogen transfer process
that does not involve conventional diffusion. A related phe-
nomenon, termed “hydrogen or radical hopping”, has been
invoked to explain the radiation-induced chemistry of certain
2
6
The amount of deuterium incorporation in the vapor-phase
toluene product generated from thermolysis of ≈DPP/≈DPM-
polymers.
For example, Clough has reported that the γ
irradiation at room temperature of a cocrystallized mixture of
C24H50 and C24D50 resulted in extensive isotopic exchange in
the solid state that was not observed for a similar isotropic
-1
d2 at a surface coverage of 0.12/0.35 mmol g is also consistent
with available kinetic data. The vapor-phase benzyl radicals
2
6
mixture irradiated in the liquid phase. Hence, this type of
(
18) Franz, J. A.; Alnajjar, M. S.; Barrows, R. D.; Kaisaki, D. L.;
Camaioni, D. M.; Suleman, N. K. J. Org. Chem. 1986, 51, 1446.
19) Bockrath, B. C.; Bittner, E. W.; Marecic, T. C. J. Org. Chem. 1986,
1, 15.
20) (a) Guthrie, R. D.; Shi, B.; Rajagopal, V.; Ramakrishnan, S.; Davis,
(22) Bockrath, B.; Bittner, E.; McGrew, J. J. Am. Chem. Soc. 1984, 106,
135.
(23) Laszlo, P. Acc. Chem. Res. 1986, 19, 121.
(24) For a recent theoretical analysis of the effect of reduction in
dimensionality on bimolecular reaction rates for interfacial reactions, see:
Mandeville, J. B.; Kozak, J. J. J. Am. Chem. Soc. 1992, 114, 6139.
(25) Pines-Rojanski, D.; Huppert, D.; Avnir, D. Chem. Phys. Lett. 1987,
139, 109.
(
5
(
B. H. J. Org. Chem. 1984, 59, 7426. (b) Rajagopal, V.; Guthrie, R. D.;
Shi, B.; Davis, B. H. Prepr. Pap.-Am. Chem. Soc. DiV. Fuel Chem. 1993,
3
8, 1114.
(21) Britt, P. F.; Buchanan, A. C., III; Malcolm, E. A. J. Org. Chem.
1
995, 60, 6523.
(26) Clough, R. L. J. Chem. Phys. 1987, 87, 1588 and references therein.