J. CHEM. RESEARCH (S), 1998 645
photooxidation of 1, 3 and 6 in the homogeneous phase
�
in the presence of TPP BF4 ; in eect, the expected decreas-
ing reactivity trend as a function of increasing reduction
potential of substrate is observed [k(4-CH )/k(H) 1.8 and
3
k(3-CF
3
)/k(H) 0.024].
It should be noted that the high reduction potential of the
two tri¯uoromethyl substituted derivatives could suggest
the involvement of a mechanism dierent from electron
transfer such as radical benzylic hydrogen transfer, not
depending on reduction potential. Otherwise this hypothesis
is not plausible in CH
substrates; for example the observed higher reactivity of
the primary vs. the secondary alcoholic site in the TiO
3
CN, even with scarcely oxidizable
2
9
sensitized photooxidation of pentane-1,4-diol (an aliphatic
alcohol, a compound with an high reduction potential, see
above) is opposite to the behaviour expected in a radical
process. In eect, the observed trend has been justi®ed
through a preferential electron transfer to the less hindered
and more eciently adsorbed primary group with respect to
the secondary one.
Fig. 1 Plot of log krel vs. E
of X-ring substituted benzyl alcohols in deaerated CH
the presence of Ag SO
p
for TiO
2
photosensitized oxidation
3
CN and in
2
4
electron-donating groups 4-CH
lower reactivity is observed with the electron-withdrawing
substituents 3-CF and 4-CF . A quantitative assessment of
the phenomenon is provided by the log krel/E
In particular
good linear correlation (r 0.992) is
observed for benzyl alcohols with X 4-CH O, 4-CH , 4-
Cl, H, 3-Cl. This behaviour is in line with a single electron
transfer (SET) rate determining step to (TiO where the
electron is removed from the aromatic moiety [eqn. (4)].
3 3
O and 4-CH while a much
The benzylic radical obtained from deprotonation
eqn. (6)] of both cation radicals 8 and 9 should undergo a
[
further oxidation (probably by Ag ) giving the protonated
3
3
7
p
plot (Fig. 1).
2
d
form of the corresponding benzaldehyde [eqn. (7)].
a
3
3
Experimental
2
)
h
+
A solution of benzylic alcohol (0.22±0.24 mmol) in N
CH CN (20 ml, HPLC grade) was externally irradiated (through a
Pyrex ®lter) by a 500 W high pressure mercury lamp, with stirring
at room temperature, in the presence of 130 mg of TiO (Aldrich,
SO . The
powder was then ®ltered through double paper and repeatedly
CN and diethyl ether; the reaction mixture was
2
purged
3
�
1
The low value of the slope (�2.1 V ) could be related to
2
a substrate like transition state, as expected for a slightly
exoergonic step.
9
9.9%, anatase, dried at 110 8C) and 0.30 mmol of Ag
2
4
7
TiO
2
washed with CH
3
poured into water and extracted with diethyl ether. The quantitative
analysis of the crude product was performed by H NMR and/or
1
by GC relative to a suitable internal standard. The same photo-
chemical reactor was used to irradiate a solution of the alcohol
�
(0.021 mmol) in O
(
(
0.5 mmol) and TPP BF
20 ml).
4
2 3
purged CH CN
Competitive experiments were performed at 25 8C by irradiation
multilamp photochemical reactor, l 350230 nm) of mixtures
(
containing equimolar amounts of two substrates and determination
(
by GC with respect to an internal standard) of the amount of
unreacted starting material at dierent times.
values were obtained by cyclic voltammetry (100 mV s
mm diameter platinum disc anode) in CH CN/LiClO (0.1 M).
�
1
E
p
,
1
3
4
This work has been carried out with the contribution of
the Ministry of University and Technological Research
MURST) and the National Council of Research (CNR).
(
In contrast the points for the benzyl alcohols with signi®-
cantly higher reduction potential (X 3-CF and 4-CF ) lie
3
3
Received, 10th March 1998; Accepted, 10th June 1998
Paper E/8/01949E
well above the line ®tted by the other substituents; in par-
ticular the last two compounds show a similar and higher
than expected reactivity. A plausible explanation of this
phenomenon could be the involvement of an alternative one
electron transfer from the OH group [eqn. (5)], a process
much less kinetically in¯uenced by the substituent eect.
The changeover of electron abstraction site is not justi®able
on a thermodynamic basis, as the reduction potential of the
References
1
2
M. A. Fox and M. T. Dulay, Chem. Rev., 1993, 93, 341.
(a) E. Baciocchi, C. Rol, G. C. Rosato and G. V. Sebastiani,
J. Chem. Soc., Chem. Commun., 1992, 59; (b) E. Baciocchi,
C. Rol, G. V. Sebastiani and L. Taglieri, J. Org.Chem., 1994, 59,
5272; (c) E. Baciocchi, T. Del Giacco, M. I. Ferrero, C. Rol and
G. V. Sebastiani, J. Org. Chem., 1997, 62, 4015; (d) E. Baciocchi,
M. Bietti, M. I. Ferrero, C. Rol and G. V. Sebastiani, Acta
Chem. Scand., 1998, 52, 160.
alcoholic site (reasonably similar to that of an aliphatic
8
alcohol, E 13 V ) is higher than that of 4-tri¯uoromethyl-
p
p
benzyl alcohol (E =2.70 V). This behaviour is in line with
3
Aquatic and Surface Photochemistry, ed. G. R. Helz, R. G. Zepp
and D. G. Crosby, Lewis Publishers, London, 1994, p. 343.
the increase of oxidizability of the OH group, explained
on the basis of its preferential adsorption with respect to the
4 M. A. Fox and A. A. Abdel-Wahab, J. Catal., 1990, 126, 693.
5 Photoinduced Electron Transfer, ed. M. A. Fox and M. Chanon,
Elsevier, Amsterdam, 1988, Part A, p. 470.
2
b,4
phenyl ring;
in our case OH could favourably compete,
with electron abstraction, with an aromatic site deactivated
by substituents with a suciently high electron-withdrawing
eect (as 3- or 4-CF ). Further kinetic evidence that the
3
observed phenomenon is connected to the medium hetero-
geneity (probable involvement of preferential adsorption)
could be the relative reactivity values obtained from the
6
Kabir-ud-Din, R. C. Owen and M. A. Fox, J. Phys. Chem., 1981,
5, 1679.
Ref. 5, Part B, p. 220 and references cited therein.
8
7
8
G. Sundholm, Acta Chem. Scand., 1971, 25, 3188.
9 M. A. Fox, H. Ogawa and P. Pichat, J. Org. Chem., 1989, 54,
3847.