Photoreduction of CCl3F
J. Phys. Chem. B, Vol. 105, No. 40, 2001 9745
•
-
illustrated in Figure 4 is not unexpected because CO2 can
The impact of step 5 varies with [H+] because electron
•
-
reduce CCl3F easily (E0 ) 1.36 V), whereas the flat-band
injection by CO2 radicals present in the double layer is less
probable at pH 8, where the TiO2 surface is negatively charged,
but increases below pzc as the surface turns positively charged.
This interpretation accounts for the falling rates of Cl- formation
during photolysis at pH g 5.9 shown in the inset of Figure 4.
The larger decreases in r(Cl-) noticed at pH 5.9 result from a
higher efficiency of step 5 combined with the lower driving
force for step 6 compared with the E6 value for pH 8 (0.22 V).
7
potential of ecb is pH-dependent, Efb ) E0 - (0.059)(pH).35
-
fb
The pH-corrected potential for step 6 (E′6 ) Efb - 0.44 V) can
then be utilized to estimate nonstandard potentials employing
the Nernst equation, E6 ) E′6 - (RT/F)ln Q, with Q )
[Cl-][•CCl2F]ss/([ecb-][CCl3F]). Electrons inside TiO2 are known
to persist for some time even when O2 is present,36 and in
suspensions of P-25 containing formate, their decay is completed
in about 1.5 s.37 Thus, E6 can be evaluated from the data
gathered immediately after the induction period. Since a higher
driving force is required to reduce the CFC, E0(O2/O2-) ) -0.16
-
On the other hand, HCO2 is bound to the oxide below pzc,
where step 4 is mainly a surface reaction. Thus, most of the
-
•CO2 radicals remain adsorbed on the oxide; even radicals
formed in the double layer are attracted to the particles, making
step 5 far more probable. This process is particularly detrimental
to the photoreaction at pH < 5 because of the lower driving
force of step 6 and is probably the reason for the slower Cl-
formation at pH 4 shown in the inset of Figure 4. Step 5 is
expected to yield some trapped electrons, which are known to
raise the Fermi level of TiO2,39 augmenting the driving force
of step 6. This effect may explain the increases in r(Cl-) during
extended illuminations at pH 4 depicted in the inset of Figure
4.
The efficient conversion of the undesirable CFC 11 to HCFC
21 using semiconductor-initiated chain reductions has been
demonstrated in this study. CCl3F is transformed via reductions
much faster than those of CFC 113, with long chains occurring
at high light intensities. Despite the coexistence of aqueous and
CFC phases, the kinetic data can be understood in terms of
simple steady-state assumptions combined with well-known
interfacial processes of TiO2. Subsequent reports will show that
these concepts are useful for rationalizing the improvements in
reaction efficiency induced by O2 and by the presence of the
CFC phase.
-
V,23 the fraction of photogenerated ecb that decay within 1 s
of illumination is probably small. During this time [ecb-] )
0.5úiI0 ) 9.9 × 10-8 M (the factor 0.5 accounts for the similar
contributions of ecb- and •CO2- to úi). As Figure 1 indicates, at
the beginning of the fast step [Cl-] ≈ 1 × 10-4 M for 0.3 M
HCO2-, which together with E′6 ) -0.05 V results in E6 )
0.12 V for pH 5.9; when E′6 ) -0.11 V and a similar value of
Q are used, the resulting potential is 0.06 V at pH 5.
These rough estimates suggest that E6 is even smaller below
pH 5, which translates to a lower reactivity of ecb- toward CCl3F
and a decrease in efficiency of step 6. This favors termination
via step 10 because the driving force needed for reducing
trihalomethyl radicals is less than that for the reduction of the
parent compounds.27 Hence, the P.E.(F-)i value measured at
pH 4 represents the combined contributions of steps 9 and 10.
In addition, the contributions of these reactions and of step 11
to the overall yield of Cl- can no longer be neglected,
invalidating the steady-state assumptions made to derive eq 12.
Differences in the driving force of step 6 can account, in part,
for the change in P.E.(Cl-)i at pH > 4 illustrated in Figure 4,
where CCl3F is reduced by both ecb and CO2-, but step 7 is
probably faster than step 6. At pH < 5 the Freon is reduced
primarily via step 7 whereas reaction 6 must compete with steps
2 and 10. Diversion of electrons into the latter reaction channels
lowers the efficiency of step 6, limiting the chain length and
decreasing P.E.(Cl-)i while P.E.(F-)i increases.
-
•
Acknowledgment. We thank Degussa for a gift of TiO2 P-25
samples, and G. Goodloe for his help with GC-MS measure-
ments. We are grateful to the U.S. Navy for supporting R.L.C.
through the CIVINS program and to NTC for partial support
of this research.
Analysis of the initial rate data using steady-state methods
assumes that Cl- and F- are formed largely via propagations
and terminations that proceed in the double-layer region.
However, the declines in r(Cl-) at longer times (Figures 1 and
4) are incompatible with a mechanism based only on homoge-
neous steps. The results of Figure 5 suggest that the concentra-
tions of chain carriers decrease as photolysis proceeds because
References and Notes
(1) Li, Y. In Molecular and Supramolecular Photochemistry 1 (Organic
Photochemistry); Ramamurthy, V., Schanze, K. S., Eds.; Marcel Dekker:
New York, 1997; p 295.
(2) Mills, A.; Le Hunte, G. J. Photochem. Photobiol., A 1997, 108, 1.
(3) Kamat, P. V.; Vinodgopal, K. In Molecular and Supramolecular
Photochemistry 2 (Organic and Inorganic Photochemistry); Ramamurthy,
V., Schanze, K. S., Eds.; Marcel Dekker: New York, 1998; p 307.
(4) Hoffmann, M. R.; Martin, S. T.; Choi, W.; Bahnemann, D. W.
Chem. ReV. 1995, 95, 69.
(5) Serpone, N.; Terzian, R.; Minero, C.; Pelizzetti, E. In PhotosensitiVe
Metal-Organic Systems; Kutal, C., Serpone, N., Eds.; Advances in Chemistry
Series 238; American Chemical Society: Washington DC, 1993; p 281.
(6) (a) Micic, O. I.; Zhang, Y.; Cromak, K. R.; Trifunac, A. D.;
Thurnauer, M. C. J. Phys. Chem. 1993, 97, 7277. (b) Lawless, D.; Serpone,
N.; Meisel, D. J. Phys. Chem. 1991, 95, 5166.
(7) (a) Wada, Y.; Yin, H.; Kitamura, T.; Yanagida, S. J. Chem Soc.,
Chem. Commun. 1998, 2683. (b) Kuhler, R. J.; Santo, G. A.; Caudill, T.
C.; Betterton, E. A.; Arnold, R. G. EnViron. Sci. Technol. 1993, 27, 2104.
(c) Bahnemann, D. W.; Mo¨nig, J.; Chapman, R. J. Phys. Chem. 1987, 91,
3782.
(8) (a) Choi, W.; Hoffmann, M. R. EnViron. Sci. Technol. 1997, 31,
89. (b) Choi, W.; Hoffmann, M. R. J. Phys. Chem. 1996, 100, 2161. (c)
Choi, W.; Hoffmann, M. R. EnViron. Sci. Technol. 1995, 29, 1646.
(9) (a) Calza, P.; Minero, C.; Pelizzetti, E. EnViron. Sci. Technol. 1997,
31, 2198. (b) Hilgendorff, M.; Hilgendorff, M.; Bahnemann, D. W. J. AdV.
Oxid. Technol. 1996, 1, 35.
(10) (a) Huang, Z.-Y.; Barber, T.; Mills, G.; Morris, M.-B. J. Phys.
Chem. 1994, 98, 12746. (b) Popovic, I. G.; Katsikas, L.; Mu¨ller, U.;
Velickovic, J. S.; Weller, H. Macromol. Chem. Phys. 1994, 195, 889. (c)
Hoffman, A. J.; Yee, H.; Mills, G.; Hoffmann, M. R. J. Phys. Chem. 1992,
lowering [•CO2 ]ss and [•CCl2F]ss favors generation of Cl- over
-
F-. Tests with data collected at 10 min of photolysis (the longest
time for which the derived expressions yield self-consistent
results) confirm this suspicion. Using the úi, r(Cl-)i, and r(F-)i
values obtained at pH 5.9 in eqs 13, 15, 16, and 17 (5.6 ×
10-3, 1.9 × 10-6 M s-1, and 6.1 × 10-9 M s-1, respectively)
yields k7 ) 9.6 × 104 M-1 s-1, k8 ) 2.6 × 103 M-1 s-1, and
[•CO2 ]ss ) [•CCl2F]ss ) 2.5 × 10-9 M. The large change in úi
-
from 0.18 to 5.6 × 10-3 after 10 min indicates that step 2
becomes more significant as irradiation continues. Faster charge
carrier recombinations are expected when electrons accumulate
in the TiO2 particles.38 Electron accumulation takes place under
conditions favoring injection of electrons into the oxide via step
5, which has an adverse cascading effect on the Freon reduction
•
-
since the strongly reducing CO2 radical transforms into the
weaker reductant ecb-. Acceleration of step 2 lowers the rate of
step 4 and also the amount of radicals able to participate in
step 7.