Reaction paths in the titanium dioxide photocatalysed degradation of dodecane and some of its derivatives

Article abstract of DOI:10.1016/S0040-4020(02)00200-4

The photocatalytic oxidation of dodecane and some of its derivatives (C₁₂H₂₆, 1; C₁₀H₂₁CH=CH₂, 2; C₁₂H₂₅SH, 3; C₁₂H₂₅OH, 4; C₁₂H₂₅NH₂, 5; C₁₁H₂₃CHO, 6) has been studied in air-equilibrated aqueous TiO₂ suspensions. The distribution of the intermediate products has been monitored during the degradation. C₁-C₁₂ Monocarboxylic acids are formed in all cases, though in a low yield with substrates 1-3. With compounds 4-6, acids are formed in a higher yield (>45%), initially including a large proportion of dodecanoic acid and are accompanied by significant amounts of aldehydes and alcohols. Relative efficiencies vary over a factor of 300 from 1 to 5 and a lower limit for the quantum yield is 1.18 for the latter compound, suggesting participation of a chain process. The data suggest that alkylperoxyl radicals are formed through initial hydrogen abstraction on the photocatalyst surface and oxygen addition and their reactivity with the starting reagent determines the course of the reaction. Activated substrates such as compounds 5 and 6, and to a lesser degree 4, easily transfer a hydrogen from position 1.

Full text of DOI:10.1016/S0040-4020(02)00200-4

TETRAHEDRON
Pergamon
Tetrahedron 58 %2002) 2943±2950
Reaction paths in the titanium dioxide photocatalysed degradation
of dodecane and some of its derivatives
Michela Sturini, Federica Soana and Angelo Albinip
Á
Dipartimento di Chimica Organica, Universita di Pavia, Via Taramelli 10, 27100 Pavia, Italy
Received 4 October 2001; accepted 28 February 2002
AbstractÐThe photocatalytic oxidation of dodecane and some of its derivatives %C12H26, 1; C10H21CHvCH2, 2; C12H25SH, 3; C12H25OH, 4;
C12H25NH2, 5; C11H23CHO, 6) has been studied in air-equilibrated aqueous TiO2 suspensions. The distribution of the intermediate products
has been monitored during the degradation. C1±C12 Monocarboxylic acids are formed in all cases, though in a low yield with substrates 1±3.
With compounds 4±6, acids are formed in a higher yield %.45%), initially including a large proportion of dodecanoic acid and are
accompanied by signi®cant amounts of aldehydes and alcohols. Relative ef®ciencies vary over a factor of 300 from 1 to 5 and a lower
limit for the quantum yield is 1.18 for the latter compound, suggesting participation of a chain process. The data suggest that alkylperoxyl
radicals are formed through initial hydrogen abstraction on the photocatalyst surface and oxygen addition and their reactivity with the starting
reagent determines the course of the reaction. Activated substrates such as compounds 5 and 6, and to a lesser degree 4, easily transfer a
hydrogen from position 1. q 2002 Elsevier Science Ltd. All rights reserved.
1. Introduction
the initial step, viz. charge transfer on the semiconductor
surface, while chemical analysis of the degradation process
of the organic molecule is generally limited either to the
initial stage %by measuring the substrate reacted) or the
Photocatalytic degradation of organic compounds using
semiconductor powders has been largely used for the puri-
®cation of polluted water.1 With respect to other abiotic
degradation methods, this has the advantage that no further
chemicals are added and quite often a complete mineralisa-
tion of the organic substrates is obtained. Commercial
titanium dioxide produced as a pigment requires no treat-
ment or activation and can be directly used as photocatalyst.
This has fostered a large number of studies aiming to
ascertain the applicability of this semiconductor for the
elimination of a variety of pollutants. There is a general
consensus that the chemistry occurring upon irradiation in
air or oxygen equilibrated aqueous suspensions involves
formation of hydroxyl radicals which are then the key active
species in the oxidation of the organic substrates,1 although
there is evidence that good organic donors %in particular
aromatics) can also be directly oxidised on the semi-
conductor surface.1m
Figure 1. Photocatalytic degradation of C12H26, 1 %B); C10H21CHvCH2, 2
%O); C12H25SH, 3 %A); C12H25OH, 4 %V); C11H23CHO, 6 %X).
However, differently from oxidation processes where the
hydroxyl radical is produced in a homogeneous solution,2
the intrinsically more complex photocatalysed degradation
has not been characterized in depth and several intriguing
points in the mechanism of the overall process leading to
mineralisation of organic substrates remain. Some of these
have been extensively addressed, but this mostly concerns
Table 1. Apparent ®rst order rate constants %kap) and quantum yield %Fap
moles per Einstein of incident light, see text) for reagent consumption in the
,
TiO2 photo-catalysed oxidation of dodecane and some of its derivatives
Compound
1
2
3
4
5
6
PhOH
kap£104 s21 0.4
1.1
4
0.0035 0.01 0.036 0.063
7
130
1.18
44
0.4 0.14
Keywords: photochemistry; oxidation; titanium dioxide; alkanes.
Corresponding author. Tel.: 139-382-507316; fax: 139-382-507323;
e-mail: albini@chi®s.unipv.it
p
Fap
0040±4020/02/$ - see front matter q 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020%02)00200-4

2944
M. Sturini et al. / Tetrahedron 58 *2002) 2943±2950
Scheme 1.
®nal stage of the overall reaction %through CO2 formation).
The `organic' chemistry occurring under these conditions
has been investigated to a limited extent. A detailed know-
ledge of the intermediates is required in order both to assess
the applicability of the method, i.e. to check whether more
resistantÐor more toxicÐcompounds are formed during
the course of the mineralisation, and to afford a basis for a
mechanistic proposal.
Table 2. Product distribution upon TiO2 photo-catalysed oxidation of dodecane and some of its derivatives
Substrate
Approximate conversion
Carboxylic acidsa
Approximate conversion
Carboxylic acidsa
1, C12H26
50%
C12
2%
95%
C12
2%
C9±C11
C3±C8
C2
1
2
9
6
C9±C11
C3±C8
C2
2
5
13
4
C1
C1
2, C10H23CHvCH2
50%
50%
C12
4%b
3
7
9
95%
95%
C12
3%
3
9
16
4
C9±C11
C3±C8
C2
C9±C11
C3±C8
C2
C1
6
C1
3, C12H25SH
C12
1%
1
7
14
10
C12
3%c
3
9
26
8
C9±C11
C3±C8
C2
C9±C11
C3±C8
C2
C1
C1
Substrate
Approximate conversion
Carboxylic acids
Aldehydes or ketones
Alcohols
4, C12H25OH
50%
C12
7.5%
1.5
8
16
12
4.5
1
12
26
9
11
1
C12
C9±C11
7%
8.5
C9±C11
C3±C8
C2
C1
C12
C9±C11
C3±C8
C2
C1
C12
C9±C11
C3±C8
C2
C1
C12
C9±C11
C3±C8
C2
C1
C12
C9±C11
5.5
2
95
50%
95
C12
C9±C11
6%
7
C9±C11
5, C12H25NH2
C12
C9±C11
2
,1
C12
C9±C11
1
,1
1
36
14
7
19
1
27
33
14
2.5
1
36
19
14
2
C12
C9±C11
3
9
C12
C9±C11
1d
4
6, C11H23CHO
50%
95
C9±C11
C3±C8
C2
C1
C12
C9±C11
C3±C8
C2
C9±C11
2.5
5
C9±C11
3
5
C9±C11
C9±C11
1
46
15
C1
a
Aldehydes, ketones, alcohols, each ,1% except when noted.
1% Undecanal and 2% 1,2-dodecane epoxide also detected.
53% H2SO4.
70% NH3.
b
c
d

M. Sturini et al. / Tetrahedron 58 *2002) 2943±2950
2945
Figure 2. Formation of %a) formic acid and %b) acetic acid during the photocatalytic degradation of C12H26, 1 %B); C10H21CHvCH2, 2 %O); C12H25SH, 3 %A);
C12H25OH, 4 %V); C11H23CHO, 6 %X).
In particular, photocatalysis is known to be effective also
with inactivated substrates, such as linear alkanes or their
simple derivatives.3 The success with these otherwise
unreactive pollutants opens useful perspectives since this
allows the application of the method in various ®elds,
ranging from oil spill cleaning4a to self-cleaning glasses,4b
or to the elimination of surfactants for recovering industrial
water.5 However, the chemical paths followed in the degra-
dation and the effect of substituents have not been investi-
gated with such aliphatic derivatives, while a detailed
knowledge would be particularly useful in this case,
where photocatalysis may have greater application than
thermal alternatives.
kinetics %kap 4£1025 s21 for c0,50 ppm, see Fig. 1, Table
1; the kinetics tended to zero order for larger values of c0).
The only products detected in a signi®cant amount during
the degradation process were the homologous aliphatic
carboxylic acids from formic to dodecanoic %7, see Scheme
1). When a high conversion was reached %e.g. 8 h for the
95% conversion of 100 ppm), the main product was acetic
acid %13 mol%, see Table 2, Fig. 2). Both formic acid %4%)
and longer chain %C3±C12) carboxylic acids %9%) were
present in lower amounts. In separate experiments it was
shown that formic acid %c0 10 ppm) was consumed 20
times faster than acetic acid under the same conditions of
the experiments above. Neither in this nor in the following
cases were signi®cant amounts of dicarboxylic acids formed
%tested: malonic and succinic, but other derivatives would
be likewise detected by ionic chromatography). At very
long irradiation times %e.g. 15±20 h for the case above),
the carboxylic acids were in turn consumed and the
mineralisation of the organic compounds was completed.
In this and in the following cases the product composition
%mol%) at ca. 50 and 95% consumption of the starting
material has been reported in Table 2.
2. Results and discussion
We deemed it worthwhile to carry out a comparative study
of the TiO2 photocatalysed degradation of dodecane and
some of its derivatives bearing a chemical functionality
and to determine the products distribution during the course
of the degradation. Mechanistic hypotheses could be formu-
lated on this basis. The substrates considered were dodecane
%1) and some of its derivatives having a substituent in
position 1 %CvC double bond, OH, SH, NH2, CHO,
compounds 2±6). Stirred suspensions of TiO2 %0.5 g/L) in
water containing aliphatic derivatives 1±6 %initial concen-
trations in the range 30±220 ppm were used) were irradiated
at 360 nm by phosphor coated lamps. The course of the
reaction was monitored under standardised conditions
both by gas chromatography and by ionic chromatography
in order to determine both non-ionic and ionic products. The
kinetics of degradation as well as the intermediates formed
at various stages of the reaction was determined. The
substrates and the products were transparent in the wave-
length range used and blank experiments omitting TiO2
showed no signi®cant decomposition on the time scale
observed for photocatalysis, ascertaining that no direct
photochemistry was involved.
The photo-oxidation of 1-dodecene %2) took place at a some-
what faster rate %Fig. 1); the product distribution was similar
to the case of 1 and again gave homologous aliphatic acids
as the products %Fig. 2, Table 2). These were accompanied
in this case by small amounts of 1,2-epoxydodecane %8) and
of undecanal %9, nˆ10). Likewise, n-dodecylsul®de %3) was
photo-oxidised to give carboxylic acids, accompanied by
sulfuric acid as the only sulfur derived product.
In the case of n-dodecanol %4) the reaction differed both
in being considerably faster and in giving a detectable
amount of primary alcohols %10), aldehydes %9) and %to a
minor extent, ,20% of the corresponding aldehydes)
ketones along with carboxylic acids %7, see Scheme 1).
Although these products were in turn degraded at a rate
equal or larger than that of the starting substrate, their
concentration reached a measurable value at an intermediate
conversion. This is shown for some long-chain derivatives
in Fig. 3.
Dodecane %1) was slowly oxidised %several hours) under this
condition. The substrate consumption followed ®rst-order

2946
M. Sturini et al. / Tetrahedron 58 *2002) 2943±2950
Figure 3. Photocatalytic degradation of dodecanol, 4 %W, left-hand ordinate scale) and formation of intermediates %®lled symbols, right-hand scale): %a)
aldehydes, %b) alcohols: C12 %X), C11 %B), C10 %O), C9 %V).
The behaviour of n-dodecylamine %5) was similar to that of
4 as far as the product distribution was concerned, with
ammonia as the only nitrogen containing product. With
this substrate, however, the reaction was more than 20
times faster than with dodecanol %Fig. 1, Table 1).
Finally, dodecanal %6) was also rapidly oxidised, giving
carboxylic acids, including a signi®cant amount of the
long-chain acids, as well as aldehydes and alcohols %Fig. 4).
Another product formed was hydrogen peroxide, but alkyl
Figure 4. Photocatalytic degradation of dodecanal, 6 %W, left-hand ordinate scale) and formation of intermediates %®lled symbols, right-hand scale):
%a) carboxylic acids, %b) aldehydes, %c) alcohols: C12%X), C11 %B), C10 %O), C9 %V).

M. Sturini et al. / Tetrahedron 58 *2002) 2943±2950
2947
hydroperoxides did not accumulate in a signi®cant amount,
as shown by the fact that treatment with reducing agents
such as sodium sul®te or triphenylphosphine did not yield
new volatile products.
alcohols are presumably also formed with the more slowly
reacting derivatives 1±3, but then are degraded at a faster
rate than the starting material, and therefore are not
accumulated to a degree allowing identi®cation, though
carboxylic acids are.
The incident ¯ux in these experiments was of ca. 8£1029
Einstein s21 cm22, with a 320±390 nm wavelength range,
peak emission at 360 nm. There is some controversy about
measuring the quantum yield in a photocatalytic process,
since this requires that the amount of light absorbedÐ
rather than re¯ectedÐby the semiconductor powder is
determined. An acceptable alternative is measuring the
`apparent' quantum yield on the basis of the impinging
light ¯ux, provided that this is compared with that of a
well known model taken as standard.6 We thus measured
the photo-oxidation of phenol under the same conditions in
order that the results could be compared with what is
reported in literature for other substrates. Phenol has been
found to be oxidised in air saturated solution by Degussa
P-25 titania with an apparent quantum yield of 0.11 under
conditions comparable to our own.6c We found a similar
value, Fapˆ0.12, in our set up and thus the values of Fap
for substrates 1±6 %reported in Table 1) are directly com-
parable. These are proportional to quantum yields and set a
lower limit for the actual value.
It is generally agreed that irradiation leads to the formation
of hydroxyl radicals %or TiIV±OHz groups) and the super-
oxide anion %or, upon protonation, the hydroperoxyl radical
OOHz) on the catalyst surface %Eq. %1)).
hn
TiO2ꢀH2O; O2† ! TiO2ꢀH1; OHz; O2z2†
ꢀ1†
Although the active species are destroyed through fast
charge recombination %OHz1e2),7 adsorbed substrates may
undergo hydrogen abstraction by such intermediates and
®nally lead to irreversible oxidation.1,8 %Eq. %2))
O2
TiO2ꢀH1; OHz; O2z2† 1 R±H ! R´ ) ROz2
ꢀ2†
However, hydrogen abstraction from C±H bonds by OHz
radicalsÐor equivalent Ti-bonded OHz groupsÐis non-
speci®c, since it is strongly exothermic, and is not per se
the cause of the observed large difference in the rate of
degradation of the investigated alkyl derivatives. This is
rather to be referred to the following reaction of the alkyl-
peroxyl radicals formed by oxygen addition to the alkyl
radicals %Eq. %2)), which are known as less ef®cient, and
thus more selective, hydrogen abstracting species %Eq.
%3))9,10
One of the typical examples of the potential of TiO2 hetero-
geneous photocatalysis for the recovery of polluted water is
its effectiveness for degrading such resistant substrates as
alkanes. Indeed, previous studies had shown that dodecane
%1) is mineralised under such conditions giving ca. 100%
CO2, and the same holds for dodecanol %4).3b The present
work extends the investigation to several aliphatic deriva-
tives and monitors the intermediates formed in the oxidation
process. In the following, the chemical paths occurring are
discussed and compared with the known reactivity of
possible intermediate radicals in order to check the viability
of degradation mechanisms.
ROz2 1 R0±H ! RO2H 1 R0z
ꢀ3†
Among the substrates investigated, amine 5 and aldehyde 6
are distinguished by the much larger rate of degradation and
by the higher proportion of characterised intermediates
%acids, aldehydes and alcohols). These pertain to classes
known for their relatively high reactivity with peroxyl
radicals, e.g. a rate of 21 M21 s21 has been measured for
the reaction of tert-BuOz2 with piperidine %and 380 M21 s21
for pyrrolidine)11 and the rate of reaction with aliphatic
aldehydes is similar.12 This is ®ve orders of magnitude
larger than that with alkanes, where reaction rates
Signi®cant indications are the following. The ef®ciency of
the photo-oxidation process changes by a factor of 300 from
alkane 1 to alkylamine 5. The apparent quantum yield,
which is a lower limit, is somewhat above 1 for the latter
compound and also the oxidation of aldehyde 6 %0.4) is
remarkably more ef®cient than that of phenol %0.11) and
of substituted phenols %which arrives at a value of 0.3 for
some derivatives).6
%per equivalent hydrogen) in the range k 1£1025
±
3£1024 M21 s21 have been measured for linear derivatives
in autooxidation processes.13 Therefore, once a peroxyl
radical has been formed in any part of the chain, ef®cient
hydrogen transfer from the activated position will take
place, whether inter- or intramolecularly.
Importantly, several oxidation intermediates are formed in
easily detectable amounts. Indeed, with the most easily
degraded substrates, i.e. 5 and 6, the sum of identi®ed inter-
mediates is a signi®cant fraction of the consumed starting
material %at limited conversion, often 50% or more),
although with less reactive substrates this percentage is
low and most intermediates are barely detectable. The struc-
ture of the intermediates formed provides evidence that a
stepwise degradation of the carbon skeleton takes place and
involves carboxylic acids, primary alcohols, and aldehydes
%Figs. 3 and 4). The latter two classes of compounds are
detected only with rapidly reacting substrates %4±6) and
their formation in preference to ketones is evidence for a
selective reaction at the functional group at position 1 rather
than at the aliphatic backbone. Aldehydes %or ketones) and
In the case of substrate 5, hydrogen abstraction from
position 1 and oxygen addition leads to a-aminoperoxyl
radicals 11 %XˆNH, Eq. %4))
ꢀ4†
Elimination is known to be ef®cient with this type of radical
%Eq. %5))14 and yields an imine, which would hydrolyse to

2948
M. Sturini et al. / Tetrahedron 58 *2002) 2943±2950
the corresponding aldehyde and ammonia %Eq. %6))
Eq. %10), as an example the rates kˆ4£107 M21 s21 for
z
z
n-BuO2 and 1.5£106 M21 s21 for sec-BuO2 have been
reported).21 The envisageable formation of singlet oxygen
is not relevant in the present case, where an unreactive
substrate such as an alkane or a poorly reacting one such
as a primary alkene are involved
ꢀ5†
ꢀ6†
2R2CHOOz ! R2CO 1 R2CHOH 1 O2ꢀor 2R2CO 1 H2O2†
ꢀ10†
This leads again to aldehydes and ketones as well as
alcohols %via a concerted or non-concerted mechanism),22
but since under these conditions the reactivity of such inter-
mediates far exceeds that of the starting hydrocarbons, their
concentration remains below the detection limit.23 There-
fore, with hydrocarbons 1 and 2, only aliphatic carboxylic
acids, the penultimate products of oxidation before ®nal
mineralisation to CO2, accumulate to a sizeable extent.
A mechanism which proceeds via the imine reasonably
explains the formation of ammonia previously reported as
the main product from the TiO2 photocatalytic decompo-
sition of various nitrogen-containing compounds %nitrate
only being formed at long irradiation times).15
Hydrogen abstraction from dodecanal 6 %whether this is the
initial substrate or is formed during the process, as it is the
case with 5, see Table 2; the corresponding imine would
react in the same way)16 leads to an acyl radical and from
this to an acylperoxyl radical, both known to abstract hydro-
gen much more ef®ciently %Eqs. %7) and %8)). Such steps may
lead to a chain reaction
In the case of dodecene 2 the oxidation is somewhat faster
than with 1. Alkenes are known to undergo both allylic
hydrogen abstraction and addition %to form epoxides) with
peroxyl radicals.13,25 A small amount of dodecene oxide %as
well as of undecanal, Eqs. %12) and %13)) is indeed formed
during the present TiO2 photooxidation of 2 %Table 2)
ꢀ7†
In the case of dodecanethiol 3, thiyl radicals are obviously
formed. It is known that in aqueous oxygenated solutions
these easily add oxygen %though the reaction is an equi-
librium, Eq. %14)) to yield thioperoxyl radicals.26 Exo-
thermic rearrangement and further oxidation lead to
sulfonyl and sulfonylperoxyl radicals %Eq. %14)), which are
ef®cient hydrogen abstracting species27
ꢀ8†
Initially a high proportion of dodecanoic acid is formed
%Table 2, Fig. 4), then degradation pursues. Intermediate
peracids would be at least in part decaboxylated by alkyl
radicals according to the process in Eq. %9), which has been
previously been demonstrated to occur17
RCO3H 1 R0z ! Rz 1 CO2 1 R0±OH
ꢀ9†
The contribution of a chain process %via chain-carrying acyl
and acylperoxyl radicals) accounts for the fact that amine 5
and aldehyde 6 react at a faster rate than phenol, which
ef®ciently traps OHz but forms stabilised radicals,18 that
with 5 the apparent quantum yield %which is a lower limit)
is slightly greater than 1, and for the above mentioned
selective attack at position 1.
ꢀ14†
This explains both the somewhat enhanced rate for the
photo-oxidation in the long-chain thiol and the formation
of sulfate as a major product.28 On the other hand, disul®des
which may be expected from the dimerizations of thiyl
radicals30 were not detected, possibly as a consequence of
their low steady state concentration which makes reaction
%14) predominating in oxygen equilibrated solution.
The reaction with alcohol 4 follows a similar course, with
the formation of a detectable amount of intermediate alde-
hydes and alcohols %Fig. 3), though the process is much
slower. This ®ts with the fact that the rate for hydrogen
abstraction from alcohols by alkylperoxyl radicals %Eq. %4),
XˆO) is three orders of magnitude lower than for amines.19
Similarly to the related amino derivatives, a-hydroxy-
peroxyl radicals cleave to aldehydes %Eq. %5), XˆO)20 and
from these acyl and acylperoxyl radicals are formed as
above.
In conclusion, the analysis of the intermediates formed
during the course of the TiO2 photocatalytic degradation
of long-chain hydrocarbons and of some simple functional
derivatives is in accord with a process with alkylperoxyl
radicals as the key intermediates. The low reactivity of
such radicals at ambient temperature makes the progress
of the reaction slow, except when a reactive group is
present. This is the case with amines and aldehydes, with
which radical chain processes are established and lead to a
fast degradation of the substrates. This product study contri-
butes to the knowledge of the mechanism, which would
further require a positive spectroscopic identi®cation of
With dodecane or dodecene, radical abstraction by peroxyl
radicals is still slower, and it is expected that bimolecular
coupling becomes a signi®cant termination path %see

M. Sturini et al. / Tetrahedron 58 *2002) 2943±2950
2949
the intermediate radicals via spin traps, as has been done for
the OHz reaction,31 as well as the evaluation of the fraction
of the reaction occurring on the catalyst surface or in solu-
tion through the calculation of the kinetics of the adsorption
process.32
samples. Ionic products were determined by means of a
Dionex QIC ionic chromatograph ®tted with a conducti-
metric detector by using an anionic exchange column and
eluting with Na2CO3/NaOH solutions with a 1 ml/min ¯ux.
The reaction of phenol was measured in the same experi-
mental set-up. The light ¯ux was measured by means of
ferrioxalate actinometry according to the procedure by
Hatchard and Parker.
From the point of view of the degradation of organics, it is
remarkable that even alkanes are removed in a not incon-
veniently slow process and the only intermediates result
from the progressive oxidative cleavage of the hydrocarbon
chain up to complete mineralisation, with carboxylic acids
as the only intermediates accumulated to an important
extent. TiO2 photocatalysis is an ef®cient method for the
recovery of polluted water, which in the overall mechanism
is reminiscent of the degradation of organic molecules
by ionising radiation,34 but occurs under extremely mild
conditions.1
Acknowledgements
M. S. and F. S. thanks Fondazione Lombardia per
l'Ambiente, Milan, for a fellowship. Financial support by
MURST, Rome, is gratefully acknowledged.
3. Experimental
References
The aliphatic derivatives 1±6 and the reference compounds
7, 9 and 10 were commercial products. 1,2-Epoxydodecane
%8) was prepared from 2 by peracetic acid oxidation.35
Titanium dioxide %Degussa P25), a known mixture of 80%
anatase and 20% rutile with an average particle size of
30 nm and a reactive surface area of approximately
50 m2 g21, was used as received.
1. %a) Mills, A.; Le Hunte, S. J. Photochem. Photobiol. A: Chem.
1997, 108, 1. %b) Pichat, P. In Handbook of Hetereogeneous
Catalysis, Knoezinger, H., Weitkamp, J., Eds.; Wiley: New
York, 1997; p. 2111. %c) Bahnemann, D.; Cunningham, J.;
Fox, M. A.; Pelizzetti, E.; Pichat, P.; Serpone, N. In Aquatic
and Surface Photochemistry, Helz, G. R., Zepp, R. G., Crosby,
D. G., Eds.; Lewis Pubications: Orlando, 1994; p. 261.
%d) Hoffmann, M. R.; Martin, S. T.; Choi, W.; Bahnemann,
D. W. Chem. Rev. 1993, 93, 671. %e) Legrini, O.; Oliveros, E.;
Braun, A. M. Chem. Rev. 1993, 93, 671. %f) Pichat, P. Catal.
Today 1994, 19, 313. %g) Pichat, P.; Guillard, C.; Maillard, C.;
Amalric, L.; D'Oliveira, J. C. In Photocatalytic Puri®cation
and Treatment of Water and Air, Ollis, D. F., Al-Ekabi, H.,
Eds.; Elsevier: Amsterdam, 1993; p. 207. %h) Serpone, N.;
Pelizzetti, E.; Hidaka, H. In Photocatalytic Puri®cation and
Treatment of Water and Air, Ollis, D. F., Al-Ekabi, H., Eds.;
Elsevier: Amsterdam, 1993; p. 225. %i) Matthews, R. W.
In Photocatalytic Puri®cation and Treatment of Water and
Air, Ollis, D. F., Al-Ekabi, H., Eds.; Elsevier: Amsterdam,
1993; p. 121. %j) Fox, M. A. In Photocatalytic Puri®cation
and Treatment of Water and Air, Ollis, D. F., Al-Ekabi, H.,
Eds.; Elsevier: Amsterdam, 1993; p. 163. %k) Serpone, N.
Kirk±Othmer Encyclopedia of Chemical Technology, Vol.
18; Wiley: New York; 1996 p 820. %l) Serpone, N.; Pelizzetti,
E. Photoconversion of Solar EnergyÐPhotochemical and
Photoelectrochemical Approaches; Archer, M. D., Nozik,
A. J., Eds.; Imperial College Press: London, 2001; Vol. III
Chapter 16. %m) Cermenati, L.; Pichat, P.; Guillard, C.; Albini,
A. J. Phys. Chem. 1998, 101, 2650.
Irradiations were carried out on mechanically stirred, air
equilibrated solutions contained in 4.5 cm diameter cylin-
drical reactors. Samples %10 ml) containing 5 mg TiO2 were
used. After that the appropriate amount of substrate was
added, the suspension was stirred for 30 min and then
irradiated using two 15 W phosphor-coated lamps
%320±390 nm, centre of emission, 360 nm). With com-
pounds 1±4 the aqueous solution was extracted with
2£3 ml ethyl acetate, ®ltered with a 0.2 mm porosity ®lter
under vacuum and concentrated at the rotary evaporator in
the cold. The residue was taken up in ethyl acetate %1 ml
total volume, containing an appropriate internal standard)
and examined by GC/MS. In the case of amine 5, the
suspension was analogously extracted with chloroform
after the addition of hydrochloric acid and sodium chloride.
With aldehyde 6 dichloromethane %3£3 ml) was used for the
extraction. In every case, the aqueous solution was ®ltered
with a water-compatible 0.2 mm porosity ®lter under
vacuum and either as such or after concentration at the
rotary evaporator examined by ionic exchange chroma-
tography.
2. %a) Roberts, B. P.; Steel, A. J. J. Chem. Soc., Perkin Trans. 2
1994, 2155. %b) Baulch, D. L.; Cox, R. A.; Hampson, R. F.;
Ken, J. A.; Troe, J.; Watson, R. T. J. Phys. Chem. Ref. Data
1984, 13, 1259. %c) Bennet, J. E.; Gilbert, B. C.; Lawrence, S.;
Whitwood, A. C.; Holmes, A. J. J. Chem. Soc., Perkin Trans. 2
1996, 1789. %d) Croft, S.; Gilbert, B. C.; Lindsay Smith, J. R.;
Stell, J. K.; Sanderson, W. R. J. Chem. Soc., Perkin Trans. 2
1992, 153.
Gas chromatography analyses were carried out using an HP
5890 apparatus with a 0.3 mm£30 m capillary column with
a ¯ame ionisation detector. Gas chromatography/mass spec-
trometry determinations were performed using an HP
5970B instrument operating at an ionising voltage of
70 eV, connected to an HP 5890 instrument equipped with
the same column as above. The mass spectra were compared
to those of authentic samples. Calibration curves for the
determination of the residual starting material and of the
photo-degradation products were constructed by adding
appropriate amounts of the authentic materials to samples
which were taken in the dark and treated as the irradiated
3. %a) Hayashimoto, K.; Kawari, T.; Sakato, T. J. Phys. Chem.
1984, 88, 4083. %b) Pelizzetti, E.; Minero, C.; Maurino, V.;
Hidaka, H.; Serpone, N. Ann. Chim. *Rome) 1990, 80, 81.
4. %a) Heller, A.; Brock, J. R. In Aquatic and Surface Photo-
chemistry, Helz, G. R., Zepp, R. G., Crosby, D. G., Eds.;

2950
M. Sturini et al. / Tetrahedron 58 *2002) 2943±2950
Lewis Publications: Orlando, 1994; p. 427. %b). Heller, A. Acc.
Chem. Res. 1995, 28, 503.
20. Bothe, E.; Schuchmann, M. N.; Schulte-Frohlinde, D.; von
Sonntag, C. Z. Naturforsch. 1983, 212.
5. %a) Hidaka, H.; Asai, Y.; Zhao, J.; Nohara, K.; Pelizzetti, E.;
Serpone, N. J. Phys. Chem. 1995, 99, 8244. %b) Hidaka, H.;
Nohara, K.; Zhao, J.; Pelizzetti, E.; Serpone, N.; Guillard, C.;
Pichat, P. J. Adv. Oxid. Technol. 1996, 1, 27. %c) Hidaka, H.;
Nohara, K.; Horikoshi, S.; Tanaka, N.; Watanabe, T.; Zhao, J.;
Serpone, N. J. Jpn. Oil Chem. Soc. 1996, 45, 21.
21. Howard, J. A. Organic Free Radicals. ACS Symp. Ser. 1978,
69, 413.
22. %a) Russel, G. A. J. Am. Chem. Soc. 1957, 79, 3871.
%b) Khursan, S. L.; Martem'yanov, V. S.; Denisov, E. T.
Kinet. Catal. *Engl. Transl.) 1990, 31, 899. %c) von Sonntag,
C.; Schuchmann, H. P. Peroxy Radicals, Wiley: Chichester,
1997 Chapter 8.
6. %a) Salinaro, A.; Emeline, A.; Zhao, J.; Hidaka, H.;
Ryabchuck, V. Pure Appl. Chem. 1999, 71, 321. %b) Serpone,
N. J. Photochem. Photobiol. A: Chem. 1997, 104, 1.
%c) Serpone, N.; Sauve, G.; Koch, R.; Tahiri, H.; Pichat, P.;
Piccinini, P.; Pelizzetti, E.; Hidaka, H. J. Photochem. Photo-
biol A: Chem. 1996, 94, 191.
23. Other radical combination reactions may lead to radicals, e.g.
Eq. %11) but this is known to contribute only marginally as a
chain carrying mechanism in view of the ef®cient termination
z2 24
reactions of H2Oz/O2
.
ROz2 1 HO2z ! ROz 1 O2 1 HOz
ꢀ11†
7. Serpone, N.; Lawless, D.; Khairutdinov, R.; Pelizzetti, E.
J. Phys. Chem. 1995, 99, 16665.
24. Schuchmann, H. P.; von Sonntag, C. Z. Naturforsch. 1984,
217.
8. %a) Gray, K. A.; Stafford, U. Res. Chem. Intermed. 1994, 20,
835. %b) Amalric, L.; Guillard, C.; Pichat, P. Res. Chem.
Intermed. 1994, 20, 579. %c) Schwitzgebel, J.; Eckerdt, J. G.;
Gerischer, H.; Heller, A. J. Phys. Chem. 1995, 99, 5633.
%d) Sun, Y.; Pignatello, J. J. Environ. Sci. Technol. 1995, 29,
2065. %e) Pichat, P.; Cermenati, L.; Albini, A.; Mas, D.;
Delprat, H.; Guillard, C. Res. Chem. Intermed. 2000, 26, 161.
9. Alfassi, Z. B. Peroxy Radicals, Wiley: Chichester, 1997.
10. Howard, J. A.; Scaiano, J. C. Landolt±Boernstein Numerical
Data and Functional Relationships in Science and
Technology; Fischer, H., Ed.; Springer: Berlin, 1984; Vol.
13d.
25. %a) Sawaki, Y.; Ogata, Y. J. Org. Chem. 1984, 49, 3344.
%b) Morgan, M. E.; Osborne, D. A.; Waddington, D. J.
J. Chem. Soc., Perkin Trans. 2 1984, 1869.
26. %a) Zhang, X.; Zhang, N.; Schuchmann, H. P.; von Sonntag, C.
J. Phys. Chem. 1994, 98, 6541. %b) Schuchmann, H. P.; von
Sonntag, C. Peroxy Radicals, Wiley: Chichester, 1997.
%c) Tamba, M.; Torregiani, A.; Tubertini, O. Radiat. Phys.
Chem. 1995, 46, 596.
27. %a) Sevilla, M. D.; Becker, D.; Yan, M. Int. J. Radiat. Biol.
1990, 57, 65. %b) Razskazovskii, Y.; Sevilla, M. D. Int. J.
Radiat. Biol. 1996, 69, 75.
11. Griller, D.; Howard, J. A.; Marriot, P. A.; Scaiano, J. C. J. Am.
Chem. Soc. 1981, 103, 619.
28. Formation of sulfate has been previously reported in the
photocatalytic degradation of different sulfur-containing
compounds.29
12. Tavadyan, L. A.; Mardoyan, V. A.; Nalbandyan, A. B. Dokl.
Akad. Nauk SSSR 1981, 259, 1143.
29. %a) Hidaka, H.; Nohara, K.; Ooishi, K.; Zhao, J.; Serpone, N.;
Pellizzetti, E. Chemosphere 1994, 29, 2619. %b) Hidaka, H.;
Zhao, J.; Pellizzetti, E.; Serpone, N. J. Phys. Chem. 1992, 96,
226. %c) Lea, J.; Adesina, A. A. J. Photochem. Photobiol.
1998, 118, 111.
13. %a) Howard, J. A. Isr. J. Chem. 1984, 24, 33. %b) Howard, J. A.
Peroxy Radicals, Wiley: Chichester, 1997 Chapter 10.
14. Pan, X. M.; Schuchmann, M. N.; von Sonntag, C. J. Chem.
Soc., Perkin Trans. 2 1993, 1021.
15. %a) Nohara, K.; Hidaka, H.; Pelizzetti, E.; Serpone, N.
J. Photochem. Photobiol. A: Chem. 1997, 102, 265.
%b) Horikoshi, S.; Serpone, N.; Zhao, J.; Hidaka, H. J. Photo-
chem. Photobiol. A: Chem. 1998, 118, 123.
30. Scrivens, G.; Gilbert, B. C.; Lee, T. C. P. J. Chem. Soc., Perkin
Trans. 2 1995, 955.
Â
31. Clement, J. L.; Gilbert, B. C.; Ho, W. F.; Jackson, N. D.;
Newton, M. S.; Silvester, S.; Timmins, G. S.; Tordo, P.;
Whitwood, A. C. J. Chem. Soc., Perkin Trans. 2 1998, 1715.
32. For a discussion of the adsorption equilibria and the inter-
dependence of the overall ef®ciency of TiO2 photocatalysis
on substrate concentration and light intensity, see Ref. 33
and references therein.
16. Hydrogen abstraction from imines is as ef®cient as from
aldehydes, see e.g. Davies, A. G.; Nedelec, J. Y.; Sucliffe,
R. J. Chem. Soc., Perkin Trans. 2 1983, 289.
17. %a) Fossey, J.; Lefort, D. Tetrahedron 1980, 36, 1023.
%b) Fossey, J.; Lefort, D.; Massoudi, M.; Nedelec, J. Y.;
Sorba, J. J. Chem. Soc., Perkin Trans. 2 1986, 781.
18. Mahoney, L. R. Angew. Chem. Int. Ed. Engl. 1969, 8, 547.
19. %a) Vardanyan, R. L.; Denisov, E. T.; Zozulya, V. I. Izv. Akad.
Nauk SSSR, Ser. Khim. 1972, 611. %b) Kovtun, G. A.;
Kazantsev, A. V.; Aleksandrov, A. L. Izv. Akad. Nauk SSSR,
Ser. Khim. 1974, 2635. %c) Asmus, K. D.; Moekel, H.;
Henglein, A. J. Phys. Chem. 1973, 77, 1218.
33. Emeline, A. V.; Ryabchuk, V. K.; Serpone, N. J. Photochem.
Photobiol. A: Chem. 2000, 133, 89.
34. von Sonntag, C. The Chemical Basis of Radiation Biology;
Taylor & Francis: London, 1987.
35. Swern, D.; Billen, G. N.; Scanlan, J. T. J. Am. Chem. Soc.
1946, 68, 1504.