Photo-oxidative dealkylation of α-alkylbenzyl methyl ethers induced by titanium dioxide in acetonitrile

Article abstract of DOI:10.1002/poc.991

TiO₂-sensitised photo-oxidation of α-alkylbenzyl methyl ethers in deaerated CH₃CN gives the expected corresponding ketone, whereas in an aerated medium a dealkylation compound (methyl benzoate) is obtained as the principal product, with respect to the ketone, or the exclusive product (when the alkyl group is tert-butyl). The product analysis and distribution, together with a qualitative estimate of the relative reactivity, suggest that the mechanism in deaerated CH₃CN is the same as that previously hypothesized for non-α-alkylated benzyl ethers; in particular, the carbonyl compound should be formed through oxidation of the benzylic radical (obtained by deprotonation of the cation radical) to the corresponding cation. To justify the formation of methyl benzoate in an aerated medium, a reasonable hypothesis is that the α-alkyl-α-methoxybenzylperoxy radical (obtained from the competitive attack of oxygen on the benzylic radical) undergoes a dealkylation process; in particular, this intermediate, as a tertiary peroxy radical, could form a dimer that evolves into the corresponding oxy radical, giving the ester through a β-scission process. The general mechanism suggested in aerated medium was confirmed by evaluating the relationship between the ester/carbonyl (E/C) molar ratio and the adiabatic ionization potential of the benzylic radical intermediate. To evaluate the influence of medium heterogeneity, the E/C ratio data were compared with those obtained from an electron-transfer photosensitised [by 9,10-dicyanoanthracene (DCA)] oxidation in a homogeneous phase. A significant confirmation of the mechanism, in both deaerated and aerated media, was obtained by the reaction performed from 4-methoxy-α-ethylbenzyl methyl ether (as a model) in the presence of H₂¹⁸O, sensitised by either TiO₂ or DCA. Copyright

Full text of DOI:10.1002/poc.991

JOURNAL OF PHYSICAL ORGANIC CHEMISTRY
J. Phys. Org. Chem. 2006; 19: 18–24
Published online 28 September 2005 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/poc.991
Photo-oxidative dealkylation of a-alkylbenzyl methyl
ethers induced by titanium dioxide in acetonitrile
Marta Bettoni,1 Tiziana Del Giacco,1,2 Cesare Rol1* and Giovanni V. Sebastiani3**
1
`
Dipartimento di Chimica, Universita di Perugia, Via Elce di Sotto, 06123 Perugia, Italy
2
`
Centro di Eccellenza Materiali Innovativi Nanostrutturati (CEMIN), Universita di Perugia,Via Elce di Sotto, 06123 Perugia, Italy
3
`
Dipartimento di Ingegneria Civile ed Ambientale, Universita di Perugia, Via G. Duranti, 06125 Perugia, Italy
Received 24 April 2005; revised 5 July 2005; accepted 28 July 2005
ABSTRACT: TiO2-sensitised photo-oxidation of ꢀ-alkylbenzyl methyl ethers in deaerated CH3CN gives the expected
corresponding ketone, whereas in an aerated medium a dealkylation compound (methyl benzoate) is obtained as the
principal product, with respect to the ketone, or the exclusive product (when the alkyl group is tert-butyl). The product
analysis and distribution, together with a qualitative estimate of the relative reactivity, suggest that the mechanism in
deaerated CH3CN is the same as that previously hypothesized for non-ꢀ-alkylated benzyl ethers; in particular, the
carbonyl compound should be formed through oxidation of the benzylic radical (obtained by deprotonation of the
cation radical) to the corresponding cation. To justify the formation of methyl benzoate in an aerated medium, a
reasonable hypothesis is that the ꢀ-alkyl-ꢀ-methoxybenzylperoxy radical (obtained from the competitive attack of
oxygen on the benzylic radical) undergoes a dealkylation process; in particular, this intermediate, as a tertiary peroxy
radical, could form a dimer that evolves into the corresponding oxy radical, giving the ester through a ꢁ-scission
process. The general mechanism suggested in aerated medium was confirmed by evaluating the relationship between
the ester/carbonyl (E/C) molar ratio and the adiabatic ionization potential of the benzylic radical intermediate. To
evaluate the influence of medium heterogeneity, the E/C ratio data were compared with those obtained from an
electron-transfer photosensitised [by 9,10-dicyanoanthracene (DCA)] oxidation in a homogeneous phase. A
significant confirmation of the mechanism, in both deaerated and aerated media, was obtained by the reaction
performed from 4-methoxy-ꢀ-ethylbenzyl methyl ether (as a model) in the presence of H218O, sensitised by either
TiO2 or DCA. Copyright # 2005 John Wiley & Sons, Ltd.
KEYWORDS: titanium dioxide; photo-oxidation; benzyl ethers; ionization potential
INTRODUCTION
substrate structure and the preferential adsorption site of
the molecule at the semiconductor surface. It should be
noted that the mechanistic approach preferentially in-
volves CH3CN as a solvent because, in contrast to water,
it is inert under the experimental conditions used and the
primary oxidation products can usually be detected. This
approach is reasonable because, with the above substrates,
the same primary steps are involved in both solvents.4
In this context, we have become interested in the
photo-oxidation of alkyl benzyl ethers (4-X-PhCH2OR0)
sensitised by TiO2 as powder in CH3CN.5–7 These are
interesting substrates because they can be simplified
model compounds for lignin, a biopolymer whose degra-
dative oxidation has been widely studied.8
The heterogeneous photo-oxidation of organic com-
pounds sensitised by semiconductors such as TiO2, a
process used to decontaminate waste water,1 can also be
used to generate radical cations (electron-transfer process
from the substrate) and radicals (derived from these inter-
mediates) and to study their decomposition pathways.2
This mechanistic investigation also provides interesting
information about the photodegradation pathway of the
investigated organic substrates as pollutants or pollutant
models. In this context, we have pointed out3 that the inter-
mediates generated from benzylic derivatives have dif-
ferent fates depending on the reaction conditions, the
In particular, the mechanistic study concerning these
ethers showed that an ꢀ-OR0 benzyl radical, obtained by
the deprotonation of the cation radical, gives the corre-
sponding carbonyl compound (aldehyde) in deaerated
CH3CN, accompanied by alkyl benzoate when the mix-
ture is bubbled with oxygen (Scheme 1).6
In this paper we extend the mechanistic study to the
TiO2-sensitised photo-oxidation of ꢀ-alkylbenzyl methyl
ethers 1a–d.
`
*Correspondence to: C. Rol, Dipartimento di Chimica, Universita di
Perugia, Via Elce di Sotto, 06123 Perugia, Italy.
E-mail: rol@unipg.it
**Correspondence to: G. V. Sebastiani, Dipartimento di Ingegneria
`
Civile ed Ambientale, Universita di Perugia, Via G. Duranti, 06125
Perugia, Italy.
E-mail: gseb@tech.ing.unipg.it
`
Contract/grant sponsor: Ministero dell’Istruzione, dell’Universita e
della Ricerca; Contract/grant number: COFIN 2003.
Copyright # 2005 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2006; 19: 18–24

TiO2-INDUCED PHOTO-OXIDATIVE DEALKYLATION OF ꢀ-ALKYLBENZYL ETHERS
19
hν
X
X
COOR'
CHO
+
(TiO2)e-
(TiO2)h+
Ag+
TiO2
O2
O2
.
X
CHOR'
+
X
CHO
Ag
(TiO2)e-
TiO2
+
+
R
R
.
+
(TiO2)h+
Scheme 1
X
CHOCH3
X
CHOCH3
X
H
R
CH3
R
-H+
a
b
c
d
X
CHOCH3
X = H, OCH3
R = H, CH3, C2H5, C(CH3)3
OCH3 CH3
OCH3 C2H5
OCH3 C(CH3)3
R
X
C
OCH3
1
.
2
(TiO2)h+
R
O2
Whereas in deaerated CH3CN the expected carbonyl
(b)
(a)
R
derivative (in this case the corresponding ketone) is the
exclusive reaction product, in the aerated medium
the principal or exclusive product is methyl benzoate.
The formation of this ester from any one of the ꢀ-R-
substituted ethers is interesting as it derives from a deal-
kylation process, through a Cꢀ—C bond breaking pathway,
which is a crucial process in the mineralization of a
pollutant containing a moiety structurally similar to 1.
The mechanistic study was carried out (i) through
product analysis, (ii) considering the benzoate/ketone
ratio in the aerated medium and (iii) by comparing the
ratio obtained in this heterogeneous process with that
observed in the homogeneous electron-transfer (ET)
photo-oxidation sensitised by 9,10-dicyanoanthracene
(DCA).
X
C
O
OCH3
X
C
OCH3
+
.
O
3
-H+
4
H2O
-CH3OH
X
COOCH3
X
COR
Scheme 2
and 1f.6 Therefore, the mechanistic pattern reported in
Scheme 2 (path a), involves, in the first step, the oxidation
of the substrate by the photogenerated hole, ðTiO2Þhþ , to
give the corresponding radical cation that, by deprotona-
tion, gives benzylic radical 2; the last intermediate should
then be oxidized to cation 3 by a second hole (as
previously shown from photoelectrochemical experi-
ments with benzyl alcohols9) giving the corresponding
carbonyl products.
In an attempt to confirm this mechanistic hypothesis,
the product yields and the reaction times in Table 1
provide qualitative information on the relative reactivity
of the substrates, taking in account that the material
recovery (substrate þ product) was nearly quantitative.
In line with the suggested mechanism that involves a
kinetically significant electron-transfer step, the introduc-
tion of a 4-methoxy group into the ring increased the
reaction rate, as can be noted by comparing the data for
RESULTS AND DISCUSSION
In the TiO2-sensitised photo-oxidation of a series of ꢀ-
alkylbenzyl methyl ethers (1a–d), in deaerated CH3CN
and in the presence of Ag2SO4 (as an electron acceptor),
the only observed product was the corresponding ketone
(Table 1). As shown, the previously investigated benzyl
methyl ethers 1e (R ¼ H, X ¼ H) and 1f (R ¼ H,
X ¼ OCH3) gave the corresponding aldehyde as product.6
Based on these results, it seems reasonable to suppose
that the mechanism for the photo-oxidation of ꢀ-alkyl-
benzyl methyl ethers is the same as that suggested for 1e
Table 1. Product yield in the TiO2-sensitised photo-oxidation of benzyl methyl ethers in deaerated CH3CN and in the presence
of Ag2SO4
R
TiO2, hν, Ag+
X
CHOCH3
X
COR
CH3CN
Substrate
X
R
Time (h) Unreacted substrate (%)a Product yield (%)a
Ep (V vs SCE)
1a
1b
1c
1d
1eb
1fb
H
CH3
CH3
C2H5
C(CH3)3
H
2.0
1.0
1.0
1.0
2.0
1.0
75
10
26
81
77
30
22
85
64
10
18
62
2.30
1.50
1.53
1.57
2.22b
1.55b
OCH3
OCH3
OCH3
H
OCH3
H
a With respect to the starting material.
b Ref. 6.
Copyright # 2005 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2006; 19: 18–24

20
M. BETTONI ET AL.
Table 2. Product yield and distribution in the TiO2-sensitised photo-oxidation of benzyl methyl ethers in aerateda CH3CN and in
the presence of Ag2SO4
R
TiO2, hν, Ag+
X
CHOCH3
X
COR
+
X
COOCH3
CH3CN
Product yield (%)b
Unreacted
Substrate
X
R
Time (h) substrate (%)b
C
E
E/C molar ratio
IP (eV)c
1a
1b
1c
1d
1e
1f
H
CH3
CH3
C2H5
C(CH3)3
H
2.0
1.0
1.5
1.5
2.0
1.0
80
20
31
80
63
2
2
11
45
35
10
27
66
6
1.8
1.6
>200
14
2.2
6.28
6.06
6.01
6.39
6.45
6.18
OCH3
OCH3
OCH3
H
26
22
—
2
OCH3
H
30
a Equilibrated with atmosphere before and during irradiation.
b With respect to the starting material.
c
Adiabatic ionization potential for the radical 2, calculated at semiempirical levels by using the AM1, PM3 and PM5 Hamiltonians and corrected for the
solvent effect (see Experimental).
1a and 1b. The rate increase in the presence of an
electron-donating group was expected because this struc-
tural modification significantly lowers the substrate re-
dealkylation product derived from the breaking of the
Cꢀ—C bond. Therefore, for all the considered ꢀ-alkyl-
benzyl ethers, we propose the hypothesis that was pre-
viously suggested for 1d only on the basis of the product
analysis,5 where the methyl benzoate is derived from the
Cꢀ—C fragmentation process that occurs in path b in
Scheme 2. In particular, the peroxy radical intermediate
4, as reported for tertiary peroxy radicals,11 could un-
dergo dimerisation and subsequent fragmentation to the
corresponding oxy radical 5 and, in turn, this intermedi-
ate should yield the ester through a ꢁ-scission process
(Scheme 3).12
An alternative hypothesis for this Cꢀ—C bond break-
ing could be the direct fragmentation of the radical
cation,13 but this can be excluded because the corre-
sponding benzaldehyde should be formed through the
homolytic step shown in Scheme 4 (the most probable
Cꢀ—C bond breaking of this intermediate, based on the
high stability of the ꢀ-methoxybenzyl cation13).
The values of the ester/carbonyl compound molar ratio,
E/C (Table 2), can be used to confirm the reaction
mechanism in aerated medium. It should be noted that
these values are reproducible and independent of the
reaction time when the medium is equilibrated with
atmosphere before and during irradiation. In particular,
the distribution of the two products should be associated
with the competition between paths a and b in Scheme 2.
More precisely, path a should depend on the reduction
potential of the cation 3, whereas path b should not be
influenced by this parameter. In previous studies,5,14 the
reduction potentials of these intermediates were used to
evaluate the oxidisability of benzylic radicals. They were
estimated as the electrochemical oxidation half-wave
þꢀ
duction potential (E1
,
determined as Ep by
=1
voltammetric measurements; see Table 1), according to
the Rehm–Weller equation for a photochemical electron
transfer process.10 By comparing the data relative to
ethers with the same substituent (4-CH3O) in the ring,
it can be noted that the introduction of an ꢀ-CH3 or an ꢀ-
C2H5 group does not significantly change the reactivity as
shown by the similar Ep values (compare 1f with 1b or
1c). As an exception, the reactivity of ether 1d is much
lower than that of 1f, 1b and 1c, despite a similar Ep; this
behaviour can be principally ascribed to the very limited
adsorption of 1d at the semiconductor surface, because
this substrate is highly hindered owing to the presence of
the tert-butyl group.
When the reaction is carried out in aerated CH3CN
(equilibrated with atmosphere before and during irradia-
tion), the principal product from 1a–c is the correspond-
ing methyl benzoate with smaller amounts of the ketone,
whereas this ester is the exclusive reaction product for 1d
(Table 2). For the sake of comparison, the reaction of
benzyl methyl ethers 1e and 1f was again performed
under the experimental conditions of this work, to yield
the methyl benzoate accompanied by minor amounts of
aldehyde (Table 2). The relative reactivities of the sub-
strates considered are qualitatively similar to those in
deaerated medium. The results (product analysis and
reactivity data in aerated medium) are in line with the
mechanistic hypothesis in Scheme 2, where radical 2 can
either be oxidized to cation 3 (path a) or, competitively, it
can be trapped by oxygen to yield the ester through the
peroxy radical intermediate 4 (path b), as suggested
previously for 1e and 1f.6
potentials (E ), measured by Wayner et al. using photo-
½
modulation voltammetry.15 These authors also showed a
good correlation between the values of the experimental
It must be observed that, in contrast to ethers 1e and 1f
(both with R ¼ H), the methyl benzoate from 1a–d is a
E
½
in CH3CN and those of the calculated adiabatic
Copyright # 2005 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2006; 19: 18–24

TiO2-INDUCED PHOTO-OXIDATIVE DEALKYLATION OF ꢀ-ALKYLBENZYL ETHERS
21
R
R
R
C
R
X
C
OCH3
.
- O2
O O
.
X
C
O
OCH3
X
O
X
C
O O
.
O
OCH3
2
OCH3
4
5
(R = Alkyl group)
.
- R
X
COOCH3
Scheme 3
.
- R
H2O
- CH3OH, - H
+
.
X
CHOCH3
X
CHO
X
CHOCH3
+
+
Scheme 4
ionization potentials (IP) of an extended series of sub-
stituted benzyl radicals.15 Therefore, in this paper, we
used the IP values as a valid alternative, calculating them
by a similar method (AM1, PM3 and PM5 Hamiltonians,
corrected for the CH3CN solvation effect; see Experi-
mental).
Relative to the ethers without the ꢀ-alkyl group, the
E/C ratio for 1e (Table 2) is very high (14), whereas for
the 4-OCH3 ring-substituted ether, 1f, the ratio is much
lower (2.2). In fact, the insertion of an electron donor into
the ring (such as a 4-OCH3 group) lowers the ionization
potential of the benzylic radical 2 because it stabilizes
the corresponding carbocation 3, which causes the rate
of path a to increase with respect to that of path b
(Scheme 2).
Concerning the ꢀ-alkylbenzyl ethers, the E/C ratio
(6.0) for 1a is lower than that for 1e (14) because the IP
value of radical 2 decreases when an electron donor
(methyl group) is inserted into the ꢀ-position. In
addition, the introduction of a further electron donor
(4-OCH3 group) in the ring of 1a (substrate 1b) further
reduces the E/C ratio (to 1.8), in line with the lower IP
of radical 2b with respect to that of 2a. Moreover, on
going from 1f to 1b (both 4-methoxy derivatives), it can
be noted that the E/C ratio decreases slightly in line
with a slightly higher IP value of the benzylic radical 2f
with respect to 2b; this small E/C ratio decrease is
probably due to the slight stabilizing effect of an ꢀ-
methyl group when two strong electron-donor groups,
ꢀ- and 4-methoxy, are already present in intermediate
2. Moreover, the E/C ratios for 1b and 1c and the
ionization potentials of radicals 2b and 2c show similar
values, in line with the similar electronic effect of the
methyl and the ethyl group in the ꢀ-position of the
benzylic cation 3.
In contrast to 1c and 1b, the only product observed for
the ꢀ-tert-butyl-substituted ether 1d is the methyl benzo-
ate (the E/C ratio is >200, considering the experimental
error of the gas chromatographic method). A significant
contribution to the absence of the corresponding ketone
can be given by the presence of a bulky ꢀ-tert-butyl
group that should destabilize the cation 3d owing to the
steric hindrance to resonance (Fig. 1). This phenomenon
is consistent with a higher IP value for 2d than for 2b or
2c (see Table 1), even if the observed E/C ratio for 1d is
too high in relation to the IP (6.39 eV) of the correspond-
ing radical 2d. A possible explanation could be as
follows. The IPs of the radicals 2a–d corrected for the
CH3CN solvation, as expected, are significantly different
(0.3–0.4 eV lower) to the values in the gas phase. Actu-
ally, the radical 2d should present a significantly higher
IP (similar to the value in the gas phase) owing to steric
inhibition of the tert-butyl group to the 3d cation solva-
tion, but this effect is not evaluated in the IP calculation.
To determine if the results (E/C ratio) obtained with ꢀ-
alkyl ethers in aerated CH3CN depend on the medium
heterogeneity, an electron-transfer photosensitised [by
9,10-dicyanoanthracene (DCA)] reaction in the homoge-
neous phase was carried out on 4-methoxy derivatives
1b–d. Following the previously suggested mechanism for
this process,16 as reported in Scheme 5 the first electron
should be removed from the substrate 1 by the excited
C(CH3)3
H
H3CO
H
OCH
Figure 1. Cation 3d (preferred conformation)
J. Phys. Org. Chem. 2006; 19: 18–24
Copyright # 2005 John Wiley & Sons, Ltd.

22
M. BETTONI ET AL.
hν
2. With DCA, anisaldehyde is obtained from 1d (entry 5
1DCA*
DCA
in Table 3) in addition to the corresponding ester. The
formation of this product can be explained by suppos-
ing that, when R ¼ C(CH3)3, the homolytic Cꢀ—C
fragmentation of 1dþꢀ (Scheme 4) competes with the
thermodynamically favoured deprotonation of this
intermediate (Scheme 5) because of stereoelectronic
requirements (collinearity of the scissible Cꢀ—X
bond and the ꢂ-system). In fact, the steric hindrance
of the bulky C(CH3)3 group in the radical cation
makes the conformation suitable for the Cꢀ—C frag-
mentation much more stable (Fig. 2).13
3
DCA
(a)
1DCA*
+
.
-
- H
- DCA
O2
.
+
1
1
2
Products
.
-
- DCA
(b)
4
O2
.
-
DCA
DCA
.
-
- O
Scheme 5
sensitiser (1DCA*), whereas the benzylic radical 2 should
be oxidized to the cation or captured by oxygen; the
sensitiser should then be regenerated by the reaction of
DCAꢁꢀ with O2 (DCA is used in a catalytic amount).
It has been observed that in this homogeneous medium,
aerated as described for the heterogeneous medium, the
E/C ratio, reproducible and independent of time, for 1b is
similar to that for 1c (6–7; see entries 1 and 3 in Table 3),
but is much lower than that for 1d (entry 5). This similar
trend allowed us to hypothesize that radical 2 undergoes
the same fate in either homogeneous or heterogeneous
phase.
This competition was excluded for the reaction of
1d with TiO2 (see above) owing to the absence of
aldehyde as product (Table 2). This result agrees with
the known basicity of the semiconductor surface,18
which is high enough to favour the radical cation
deprotonation also with this substrate. As a confirma-
tion, in the DCA-sensitised photo-oxidation in the
presence of lutidine (as base), the aldehyde is not
present as product from 1d (Table 3, entry 6).
It can be observed that the introduction of a base did
not change the E/C ratio in any of the substrates
considered in DCA-sensitised photo-oxidation; this
observation therefore confirms that deprotonation is
not involved in the steps responsible for the ester/
ketone distribution.
In addition, the results that are different in the two
media could be consistent with a common mechanism:
1. The E/C ratios with DCA for 1b and 1c are higher than
those (1.8 and 1.6 in Table 2) observed for the same
substrates in the presence of TiO2. This can be
explained by considering that in the homogeneous
phase the oxidant of intermediate 2 (probably the
sensitiser in the ground state or oxygen) should be
less powerful than that in heterogeneous medium,
ðTiO2Þhþ (E ꢂ ¼ 2.4 V17) and therefore path a is dis-
favoured.
3. A final difference observed between the two oxidizing
systems is that with DCA/lutidine (the homogeneous
medium that can be compared with the heterogeneous
one, see previous point) the reactivity of 1d is qualita-
tively similar to that of 1b and 1c (compare, for
example, entries 2, 4 and 6 in Table 3), while with
TiO2 1d is much less reactive than 1b and 1c (Table 2).
This confirms the above suggested hypothesis that
the lower reactivity of 1d in TiO2-sensitised
Table 3. Product yield and distribution in the photo-oxidation of 4-methoxybenzyl methyl ethers (1b–d) sensitised by 9,
10-dicyanoanthracene (DCA) in aerateda CH3CN
R
DCA, hν
+
H3CO
COR
H3CO
COOCH3
H3CO
CHOCH3
CH3CN
C
E
Product yield (%)b
Unreacted
Entry
Substrate
R
Time (h)
substrate (%)b
C
E
E/C molar ratio
1
1b
1c
1d
CH3
4
0.5
3
0.6
0.3
1
62
74
27
79
76
61
4
3
8
2
—
—
24
17
56
14
6
6
7
2c
3
C2H5
4c
5
7
C(CH3)3
7d
>200
>200
6c
30
a Equilibrated with atmosphere before and during irradiation.
b With respect to the starting material.
c In the presence of lutidine (1.1 ꢃ 10ꢁ5 M).
d Anisaldehyde is also present as reaction product (13%).
Copyright # 2005 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2006; 19: 18–24

TiO2-INDUCED PHOTO-OXIDATIVE DEALKYLATION OF ꢀ-ALKYLBENZYL ETHERS
23
.
.
+
+
purged) or aerated (equilibrated with the atmosphere
before and during irradiation), containing the substrate
(0.05–0.10 mmol) and Ag2SO4 (0.10–0.20 mmol) in a
cylindrical flask provided with a water cooling jacket.
After irradiation with stirring at room temperature, the
reaction mixture was filtered through double paper and
analysed by GC (in the presence of durene as an internal
standard). Quantitative analysis of the crude product was
performed by 1H NMR and/or GC with a suitable internal
standard (durene).
C(CH3)3
H
- H+
.
- C(CH3)
OCH3
C(CH3)3
OCH3
H
Figure 2. The conformations of 1dþ. suitable for C—H and
Ca—C fragmentation
heterogeneous photo-oxidation is principally due to a
reduced substrate adsorption at the semiconductor
surface.
Photochemical oxidation sensitised by DCA. The sub-
strate (0.10 mmol) was dissolved in 10 ml of a solution
containing DCA in aerated CH3CN (4 ꢃ 10ꢁ4 M), in
either the absence or presence of lutidine (0.10 mmol).
The resulting solution was irradiated with an Applied
Photophysics Multilamp Photochemical reactor (12
lamps at ꢃ ¼ 355 ꢄ 20 nm).
A significant confirmation of the mechanism, both for
the TiO2- and DCA-sensitised photo-oxidation, was ob-
tained by labelling experiments in the presence of H218O.
In particular, we considered substrate 1c as a model and
we observed that the corresponding ketone (path a in
Scheme 2) shows the complete incorporation of 18O
(after correction of the 18O content of the labelled water),
whereas the methyl benzoate (via b in Scheme 2) did not
show any 18O incorporation. This fact also allows us to
exclude (within the experimental error) the possibility
that the ketone should also be formed through a ꢁ-
scission process from the oxy radical 5c (path b) with
the loss of a methoxy radical.
Reaction products. The products were identified directly
from the crude by comparison of H NMR and GC–MS
1
data with those for commercial samples (aldehydes,
ketones and esters).
Cyclic voltammetry. Ep values were obtained from cyclic
voltammetry experiments, conducted with an AMEL 552
potentiostat controlled by a programmable AMEL 568
function generator (at 100 mV sꢁ1, 1 mm diameter plati-
num disc anode and SCE as reference) in CH3CN–
LiClO4 (0.1 M).
EXPERIMENTAL
1H-NMR spectra were run on a Bruker AC 200 (200 MHz)
spectrometer, with solutions in CDCl3 with TMS as an
internal standard. GC–MS analyses were performed on a
Hewlett-Packard (HP) 6890A gas chromatograph (HP-
Innovax capillary column, 15 m) coupled with an MSD-
HP 5973 mass-selective detector (70 eV). GC analyses
were carried out on an HP 5890 gas chromatograph using
an HP-Innovax capillary column (15 m).
Adiabatic ionization potentials. The ionization potentials
of the substrates were calculated at semiempirical levels by
using the AM1, PM3 and PM5 Hamiltonians (MO-
PAC2002, CAChe 6.1, Fujitsu).21 The Conductor-like
Screening Model (COSMO) was also included in the
calculations to model the effect of the polar solvent on
the molecular properties.22
Photochemical oxidation of 1c in the presence of H218O.
The photo-oxidation of 1c, sensitised by either TiO2 or
DCA, was also carried out in CH3CN–H218O (99.5:0.5, v/
v). The reaction mixture was analysed by GC–MS. The
percentage of 18O incorporated in the two reaction
products (ketone and methyl benzoate) was determined
by the relative abundances of the peaks at m/z 166 and
164 [Mþ ] and at 137 and 135 (for the corresponding
ketone); at m/z 168 and 166 [Mþ ] and at 137 and 135(for
the methyl benzoate). In the ester, in both cases, 18O is
absent, whereas in the ketone the amount of incorporated
18O (corrected considering the 18O content of the labelled
water) is 100 ꢄ 8%.
Materials. TiO2 (anatase) (Aldrich, 99.9%, dried at
110 ꢂC), CH3CN (HPLC grade, water content 0.02%
by Karl Fischer coulometry), Ag2SO4, H218O (Aldrich,
97 atom%), benzaldehyde, 4-methoxybenzaldehyde,4-
methoxyacetophenone, 4-methoxypropiophenone, methyl
benzoate and methyl 4-methoxybenzoate were commer-
cial samples. ꢀ-Methylbenzyl methyl ether (1a),19
4-methoxy-ꢀ-methylbenzyl methyl ether (1b),20 4-meth-
oxy-ꢀ-ethylbenzyl methyl ether (1c),20 ꢀ-tert-butyl-4-
methoxybenzyl methyl ether (1d),5 benzyl methyl ether
(1e)6 and 4-methoxybenzyl methyl ether (1f)6 were pre-
pared and characterized as described in the literature.
Photochemical oxidation sensitised by TiO2. The reac-
tions with benzyl methyl ethers were carried out by
external irradiation (Helios Italquartz 500 W high-pres-
sure mercury lamp, Pyrex filter) of a TiO2 (70 mg)
suspension in 10 ml of CH3CN either deaerated (N2
Acknowledgements
This work was carried out with the financial support of
`
the Ministero dell’Istruzione, dell’Universita e della
Copyright # 2005 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2006; 19: 18–24

24
M. BETTONI ET AL.
10. Rehm D, Weller A. Isr. J. Chem. 1970; 8: 259–271.
11. Howard JA. In Free Radicals, Kochi JK (ed). Wiley: New York,
1973; Part II, 33.
12. (a) Baciocchi E, Bietti M, Salamone M, Steenken S. J. Org. Chem.
2002; 67: 2266–2270; (b) Wietzerbin K, Bernadou J, Meunier B.
Eur. J. Inorg. Chem. 1999; 1467–1477.
Ricerca (COFIN 2003). Thanks are due to Professor
Fausto Elisei for the calculation of adiabatic ionization
potentials.
13. Baciocchi E, Bietti M, Lanzalunga O. Acc. Chem. Res. 2000; 33:
243–251.
REFERENCES
14. Baciocchi E, Rol C, Sebastiani GV, Taglieri L. J. Org. Chem.
1994; 59: 5272–5276.
15. Wayner DDM, Sim BA, Dannemberg JJ. J. Org. Chem. 1991; 56:
4853–4858.
16. (a) Freccero M, Mella M, Albini A. Tetrahedron 1994; 50:
2115–2130; (b) Gould IR, Godleski SA, Zielinski PA, Farid
S. Can. J. Chem. 2003; 81: 777–788; (c) Julliard M, Galadi
A, Chanon M. J. Photochem. Photobiol. A: Chem 1990; 54:
79–90.
17. Kabir-ud-Din RC, Owen RC, Fox MA. J. Phys. Chem. 1981; 85:
1679–1682.
18. Baciocchi E, Del Giacco T, Ferrero MI, Rol C, Sebastiani GV.
J. Org. Chem. 1997; 62: 4015–4017.
19. (a) Nishino T, Nishiyama Y, Sonoda N. Bull. Chem. Soc. Jpn
2003; 76: 635–641; (b) Plesnicar B, Kovae F, Schara M. J. Am.
Chem. Soc. 1988; 110: 214–222; (c) Lillis V, McKenna J,
McKenna JM, Smith MJ, Taylor PS, Williams IH. J. Chem.
Soc., Perkin Trans. 2 1980; 83–86.
1. (a) Helz GR, Zepp RG, Crosby DG (eds). Aquatic and Surface
Photochemistry. Lewis: London, 1994; (b) Serpone N, Pelizzetti E
(eds). Photocatalysis. Fundamentals and Applications. Wiley:
New York, 1989; (c) Bahnemann D. In Environmental Photo-
chemistry, Boule P (ed). Springer: Berlin, 1999; 285–351.
2. (a) Fox MA, Dulay MT. Chem. Rev. 1993; 93: 341–357; (b)
Minero C, Maurino V, Pelizzetti E. In Semiconductor Photo-
chemistry and Photophysics, Ramamurthy V, Schanze KS (eds).
Marcel Dekker: New York, 2003; 211–229.
3. Bettoni M, Del Giacco T, Rol C, Sebastiani GV. J Photochem.
Photobiol. A 2004; 163: 481–487, and references cited therein.
4. Ranchella M, Rol C, Sebastiani GV. J. Chem. Soc., Perkin Trans.
2 2000; 311–315.
5. Baciocchi E, Bietti M, Ferrero M, Rol C, Sebastiani GV. Acta
Chem. Scand. 1998; 52: 160–164.
6. Del Giacco T, Rol C, Sebastiani GV. J. Phys. Org. Chem. 2003;
16: 127–132.
20. Mohan R, Katzenellenbogen JA. J. Org. Chem. 1984; 49: 1238–
1246.
7. Bettoni M, Del Giacco T, Rol C, Sebastiani GV. J. Chem. Res. (S)
2003; 415–417.
21. Zerner MC. In Reviews in Computational Chemistry,
vol. 2, Lipkowitz KB, Boyd DB (eds). VCH: New York, 1991;
313–365.
22. Dolney DM, Hawkins GD, Winget P, Liotard DA, Cramer CJ,
Truhlar DG. J. Comput. Chem. 2000; 21: 340–366.
8. Lanzalunga O, Bietti M. J. Photochem. Photobiol. B 2000; 56:
85–108.
9. (a) Palombari R, Ranchella M, Rol C, Sebastiani GV. Sol. Energy
Mat. Sol. Cells 2002; 71: 359–368; (b) Ranchella M, Rol C,
Sebastiani GV. J. Chem. Res. (S) 2002; 239–241.
Copyright # 2005 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2006; 19: 18–24