Quantum yield for ClOO formation following photolysis of aqueous OClO

Article abstract of DOI:10.1021/ja002148o

The photochemistry of chlorine dioxide (OClO) in aqueous solution was investigated by femtosecond transient absorption spectroscopy. Following the photoexcitation of OClO at 400 nm, the transient absorption dynamics were probed in the spectral range from 400 to 220 nm. As expected from earlier studies, the main photolytic products ClO + O, formed with a quantum yield of ～90%, disappear through fast geminate recombination producing OClO in the electronic ground state. The total quantum yield for chlorine atom production is (Φ_Cl) ～10%, with the chlorine atom production occurring through two competing processes. The dominant channel for chlorine atom production involves the formation of a short-lived intermediate on a ～6 ps time scale with a quantum yield of 8 ± 2%. The remaining 2 ± 1% is formed through the formation and decomposition of ClOO. The lifetime of ClOO was found to be ～0.32 ns, in very good agreement with the result of a recent time-resolved resonance Raman study. Finally, the UV absorption spectrum for aqueous ClOO is reported and compared with previously reported spectra obtained in condensed media.

Full text of DOI:10.1021/ja002148o

J. Am. Chem. Soc. 2000, 122, 12795-12801
12795
Quantum Yield for ClOO Formation following Photolysis of Aqueous
OClO
‡
†
,‡
Carsten L. Thomsen, Philip J. Reid, and Søren R. Keiding*
Contribution from the Department of Chemistry, UniVersity of Aarhus, Langelandsgade 140, DK-8000
Århus C, Denmark, and Department of Chemistry, Box 351700, UniVersity of Washington,
Seattle, Washington 98195
ReceiVed June 15, 2000. ReVised Manuscript ReceiVed September 7, 2000
Abstract: The photochemistry of chlorine dioxide (OClO) in aqueous solution was investigated by femtosecond
transient absorption spectroscopy. Following the photoexcitation of OClO at 400 nm, the transient absorption
dynamics were probed in the spectral range from 400 to 220 nm. As expected from earlier studies, the main
photolytic products ClO + O, formed with a quantum yield of ∼90%, disappear through fast geminate
recombination producing OClO in the electronic ground state. The total quantum yield for chlorine atom
production is (ΦCl) ∼10%, with the chlorine atom production occurring through two competing processes.
The dominant channel for chlorine atom production involves the formation of a short-lived intermediate on a
∼
6 ps time scale with a quantum yield of 8 ( 2%. The remaining 2 ( 1% is formed through the formation
and decomposition of ClOO. The lifetime of ClOO was found to be ∼0.32 ns, in very good agreement with
the result of a recent time-resolved resonance Raman study. Finally, the UV absorption spectrum for aqueous
ClOO is reported and compared with previously reported spectra obtained in condensed media.
Introduction
isomerization to ClOO could contribute significantly to the
catalytic destruction of ozone. The suggested formation of
1
The photochemistry of chlorine dioxide (OClO) is of current
interest due to its participation in the stratospheric reactive
chlorine reservoir as well as its potential role in ozone
depletion.1 An outline of the three distinct photochemical
pathways available to OClO following photoexcitation is
presented in reactions 1-3.
ClOO is still controversial, however, and this species has not
been observed experimentally in the gas phase. The photo-
chemistry of OClO in low-temperature matrixes is completely
different from that in the gas phase. Specifically, the quantum
yield (ΦCl) for chlorine formation in the gas phase is e4%, but
-6
9
,16-19,42
increases to unity in low-temperature matrixes.
Fur-
hV
thermore, the formation of chlorine atoms occurs through the
thermal decomposition of ClOO formed by either direct
photoisomerization of OClO or by geminate recombination of
primary photoproducts as ClO + O or Cl + O2 inside the matrix
2
2
3
OClO(X B )
9
8 ClO( Π) + O( P )
(1)
(2)
(3)
1
g
hV
2
3
3 - 1
g
OClO(X B )
98 Cl( P ) + O ( Σ , ∆ )
1
u
2
g
1
8-23
cage.
It should be noted that the thermodynamically more
hV
9
∆
9
2
2
3
-
g
OClO(X B )
8 ClOO
8 Cl( P ) + O ( Σ )
1
u
2
(6) Sessler, J.; Chipperfield, M. P.; Pyle, J. A.; Toumi, R. Geophys. Res.
Lett. 1995, 22, 687.
(
7) Vaida, V.; Goudjil, K.; Simon, J. D.; Flanders, B. N. J. Mol. Liq.
Since the environmental impact of OClO arises from its ability
to produce atomic chlorine, knowing the quantum yield for
reactions 2 and 3 is essential if the environmental impact of
OClO in both homogeneous and heterogeneous media is to be
predicted. This has served as an inspiration for numerous
experimental and theoretical investigations of OClO in both gas
and condensed phases.1
1
994, 61, 133-152.
(
(
8) R u¨ hl, E.; Jefferson, A.; Vaida, V. J. Phys. Chem. 1990, 94, 4.
9) Davis, H. F.; Lee, Y. T. J. Chem. Phys. 1996, 105, 8142-8163.
(10) Bishenden, E.; Donaldson, D. J. J. Chem. Phys. 1993, 99, 3129.
(
11) Baumert, T.; Herek, J. L.; Zewail, A. H. J. Chem. Phys. 1993, 99,
430-4440.
12) Delmdahl, R. F.; Bakker, B. L. G.; Parker, D. H. J. Chem. Phys.
2000, 112, 5298-5300
13) Kreher, C. J.; Carter, R. T.; Huber, J. R. J. Chem. Phys. 1999, 110,
309-3319
14) Delmdahl, R. F.; Ullrich, S.; Gericke, K. H. J. Phys. Chem. A 1998,
4
(
-37
(
When OClO is photolyzed in the gas phase with near-
ultraviolet light, ClO + O is the dominant product, whereas
3
(
8-15
only a moderate yield of chlorine atoms is observed.
It has
102, 7680-7685
been suggested that photolytic formation of Cl + O2 via
(15) Kreher, C. J.; Carter, R. T.; Huber, J. R. Chem. Phys. Lett. 1998,
2
86, 389-397
*
To whom correspondence should be addressed. E-mail: keiding@
(16) Lawerence, W. G.; Clemitshaw, K. C.; Apkarian, V. A. J. Geophys.
kemi.aau.dk.
Res. 1990, 95, 18591.
(17) Lanzendorf, E. J.; Kummel, A. C. Geophys. Res. Lett. 1996, 23,
1251.
(18) Pursell, C. J.; Conyers, J.; Alapat, P.; Parveen, R. J. Phys. Chem.
1995, 99, 10433-10437.
(19) Adrian, F. J.; Bohandy, J.; Kim, B. F. J. Chem. Phys. 1986, 85,
2692-2698.
(20) Rochkind, M. M.; Pimental, G. C. J. Chem. Phys. 1967, 46.
(21) Arkell, A.; Schwager, I. J. Am. Chem. Soc. 1967, 89, 5999-6006.
(22) Johansson, K.; Engdahl, A.; Nelander, B. J. Phys. Chem. 1993, 97,
9603-9606.
‡
University of Aarhus.
University of Washington.
1) Vaida, V.; Solomon, S.; Richard, E. C.; R u¨ hl, E.; Jefferson, A. Nature
†
(
1
989, 342, 405-408.
(
(
(
2) Vaida, V.; Simon, J. D. Science 1995, 268, 1443-1448.
3) Rowland, F. S. Annu. ReV. Phys. Chem. 1991, 42, 731.
4) Burkholder, J. B.; Talukdar, R. K.; Ravishankara, A. R. Geophys.
Res. Lett. 1994, 21, 585.
5) Solomon, S.; Sanders, R. W.; H. L. Miller, J. J. Geophys. Res. 1990,
5, 13807.
(
9
1
0.1021/ja002148o CCC: $19.00 © 2000 American Chemical Society
Published on Web 12/06/2000

12796 J. Am. Chem. Soc., Vol. 122, No. 51, 2000
Thomsen et al.
stable isomer ClOO is expected to be the terminal product found
within an inescapable cavity at sufficiently low temperature
regardless of its mechanism of formation (i.e., whether it is
formed by concerted isomerization or via recombination of the
primary fragments).24 Hence, OClO is expected to be quanti-
tatively converted into its asymmetrical isomer ClOO in solid
media, which in turn readily decomposes into Cl + O2 at
ambient temperature. The quantum yield of chlorine formation
in aqueous solution is intermediate between the gas-phase and
atoms occurs through a short-lived precursor within 15 ps and
3
4,38
with a total quantum yield of ΦCl ∼ 10%.
Thus, the current
understanding of the production of chlorine atoms from pho-
tolysis of aqueous OClO does not involve any long-lived ClOO
isomer.
Even though the isomer ClOO has been proposed to be a
photoproduct of aqueous OClO for the past decade, it was not
until very recently that its existence was unambiguously
established. By using two-color, time-resolved resonance Raman
(TRRR) spectroscopy, the formation of ClOO following pho-
31,34,39-40
matrixes values, with ΦCl ∼ 10%.
Despite the well-
4
1
determined Cl quantum yield, the formation mechanisms and
time scale of chlorine production in solution phase remains
controversial.
toexcitation of aqueous OClO was confirmed. The integrated
scattering intensity of ClOO was best modeled by a sum of
two exponentials resulting in appearance and decay time
constants of 28 ( 5 ps and 0.40 ( 0.05 ns, respectively. In
addition, a delay of 13 ( 2 ps was included to reproduce the
absence of ClOO scattering at early times. These TRRR results
were the first direct evidence for the formation and decomposi-
tion of ClOO in aqueous solution, and indicate that the current
model for chlorine formation should be extended. The quantum
yield for ClOO formation could not be determined in the TRRR
experiment. Therefore, we have revisited the photolysis of
aqueous OClO with the aim of studying the contribution from
ClOO in the formation of atomic chlorine.
Studies of OClO solution-phase photochemistry have at-
tempted to identify the specific solvent-solute interactions
responsible for the phase-dependent quantum yield of the
photoproducts. The solution-phase photochemistry of OClO was
pioneered by Simon and co-workers using picosecond flash
photolysis.3
0-33
The current understanding of the photolysis of
OClO in solution, however, has been obtained using femtosec-
ond transient absorption and time-resolved resonance Raman
3
4-41
experiments.
It is now well established that the dominant
primary photoproduct channel is the ClO + O, reaction 1,
3
4-41
occurring with a ∼ 90% yield.
These studies also show
In this work, we present femtosecond transient absorption
data of aqueous OClO. The photoproduct formation following
photoexcitation at 400 nm is monitored over the entire spectral
range from 400 to 220 nm on an absolute optical density scale,
with emphasis on the dynamics occurring at long time. In
contrast to earlier studies, this allows the simultaneous monitor-
ing of the decomposition and formation of all photoproducts,
including ClOO. In agreement with earlier femtosecond work
on aqueous OClO, it is observed that the majority (∼90%) of
the photoexcited OClO molecules dissociate into ClO and O,
with geminate recombination of these products resulting in the
formation of vibrationally excited OClO. In addition, it is
observed that most of the chlorine atoms are produced on a ∼6
ps time scale with a quantum yield of 8 ( 2%, whereas 2 (
that the majority of the ClO and O photoproducts undergo
geminate recombination to form vibrationally excited OClO in
the electronic ground state as a result of the solvent caging effect.
The subsequent vibrational relaxation of the hot OClO molecule
on the ground-state potential surface occurs with a time constant
of ∼10 ps.
2
3
2
2
ClO( Π) + O( P ) f OClO*(X B ) f OClO(X B ) (4)
g
1
1
Transient absorption spectroscopy covering the first 50 ps of
the photolysis of OClO indicates that the production of chlorine
(
23) M u¨ ller, H. S. P.; Willner, H. J. Phys. Chem. 1993, 97, 10589-
0598.
24) Brusa, M. A.; Perissinotti, L. J.; Churio, M. S.; Colussi, A. J. J.
Photochem. Photobiol. A 1996, 101, 105-111.
25) Mauldin, R. L., III; Burkholder, J. B.; Ravishankara, A. R. J. Phys.
1
1
% is formed through the formation and decomposition of
(
ClOO. The decomposition of ClOO occurs with a time constant
of ∼0.32 ns in agreement with recent time-resolved resonance
Raman studies. The femtosecond data presented here represent
the first direct measurement of the quantum yield for ClOO
formation in aqueous solution, and significantly clarify the role
of this controversial species in the mechanism for chlorine
formation in aqueous solution.
(
Chem. 1992, 96, 2582-2588.
(
26) Gole, J. L. J. Phys. Chem. 1980, 84, 1333-1340.
(27) Jafi, J. A.; B. H. Bauschlicher, J.; Phillips, D. H. J. Chem. Phys.
1
8
9
1
4
985, 83, 1693-1701.
28) Peterson, K. A.; Werner, H.-J. J. Chem. Phys. 1992, 96, 8948-
961.
(
(
29) Peterson, K. A.; Werner, H.-J. J. Chem. Phys. 1996, 105, 9823-
832.
Experimental Section
(30) Dunn, R. C.; Richard, E. C.; Vaida, V.; Simon, J. D. J. Phys. Chem.
991, 95, 6060-6063.
The details of the ultrafast transient absorption spectrometer have
been presented elsewhere.3 Briefly, a regenerative amplified Tita-
nium:Sapphire laser producing 90-fs, 750-µJ pulses centered at 800
nm with a repetition rate of 1 kHz was employed in this work. The
fundamental beam at 800 nm was frequency doubled in a 0.5-mm-
thick BBO crystal to generate the 400-nm pump pulse used for the
photolysis of OClO. The pump pulse was sent through a variable delay
line and a wave-plate rotated the polarization of the pump pulse 54.7°
relative to that of the probe pulse before it was weakly focused into a
sample flow cell. The pump pulse energy was typically 50 µJ with a
spot diameter of 1.3 mm at the sample and it was modulated at 0.5
kHz, phase-locked to the 1 kHz repetition rate of the regenerative
(31) Dunn, R. C.; Simon, J. D. J. Am. Chem. Soc. 1992, 114, 4856-
8,43
860.
32) Dunn, R. C.; Anderson, E. J. L.; Foote, C. S.; Simon, J. D. J. Am.
Chem. Soc. 1993, 115, 5307.
33) Dunn, R. C.; Flanders, B. N.; Simon, J. D. J. Phys. Chem. 1995,
9, 7360-7370.
34) Philpott, M. J.; Charalambous, S.; Reid, P. J. Chem. Phys. Lett.
997, 281, 1-9.
35) Philpott, M. J.; Hayes, S. C.; Reid, P. J. Chem. Phys. 1998, 236,
(
(
9
1
2
2
(
(
07-224.
(36) Hayes, S. C.; Philpott, M. J.; Reid, P. J. J. Chem. Phys. 1998, 109,
596-2599.
(37) Hayes, S. C.; Philpott, M. J.; Mayer, S. G.; Reid, P. J. J. Phys.
Chem. 1999, 103, 5534-5546.
38) Thøgersen, J.; Jepsen, P. U.; Thomsen, C. L.; Poulsen, J. A.; Byberg,
J. R.; Keiding, S. R. J. Phys. Chem. A 1997, 101, 3317-3323.
39) Poulsen, J. A.; Thomsen, C. L.; Keiding, S. R.; Thøgersen, J. J.
Chem. Phys. 1998, 108, 8461-8471.
40) Thøgersen, J.; Thomsen, C. L.; Poulsen, J. A.; Keiding, S. R. J.
Phys. Chem. A 1998, 102, 4186-4191.
41) Thomsen, C. L.; Philpott, M. P.; Hayes, S. C.; Ried, P. J. J. Chem.
Phys. 2000.
amplifier. The results obtained were independent of the pump pulse
(
energy indicating negligible multiphoton absorption.3
8-40
The probe
(
pulse was generated using a combination of second harmonic generation
and sum-frequency mixing of a supercontinuum generated by focusing
(
(42) Bishenden, E.; Donaldson, D. J. J. Phys. Chem. 1994, 101, 1.
(43) Thomsen, C. L.; Madsen, D.; Keiding, S. R.; Thøgersen, J. J. Chem.
Phys. 1999, 110, 3453-3462.
(

Photochemistry of Chlorine Dioxide in Aqueous Solution
J. Am. Chem. Soc., Vol. 122, No. 51, 2000 12797
3
55 nm.44 The absorption spectrum of ClOO has not been
determined in aqueous solution due to the reactivity of this
molecule. However, an accurate UV spectrum in the gas phase
2
5
has been obtained, and a spectrum extended into the visible
2
3
region has been measured in solid neon. These spectra agree
23
within the limit of error. The spectrum has also been measured
in amorphous ice at 80 K and resembles the gas-phase and
matrix-isolated spectra with only a small (≈7 nm) red shift of
1
8
the absorption band. This indicates the absence of large
solvatochromic effects on the ClOO absorption band. Accord-
ingly, we adopt the spectrum measured in amorphous ice
calibrated with the gas-phase value for the extinction coefficient.
The absorption spectra of aqueous ClO and Cl:H2O have been
well characterized spectroscopically, and the spectra depicted
in Figure 1 are taken from refs 45 and 46. It should be noted
that two of the potential photoproducts, O and O2, do not have
significant absorption bands over the spectral range covered in
these measurements. The spectra of the three potential photo-
products Cl, ClO, and ClOO overlap to a significant extent,
but since ClOO absorbs much stronger than the other products
it is possible to estimate the quantum yield of each of the
competitive reactions pathways 1-3 by analyzing the transient
absorption spectra.
Description of Transient Absorption Data and Assign-
ments. The photoinduced absorption, ∆A(λ,t) of aqueous OClO,
produced by the 400-nm pump pulse was measured at 16
wavelengths in the spectral range from 220 to 400 nm.
Additional transient absorption measurements were performed
in the spectral range from 220 to 310 nm to better resolve the
spectral behavior within the first 100 ps. Figure 2 shows the
transient absorption for representative wavelengths in the region
from 220 to 400 nm.
Figure 1. Absorption spectra of OClO(aq) (s), ClOO(s) (- - -), ClO-
2
(aq) (b), and the Cl:H O charge-transfer complex (O). The absorption
spectrum of OClO(aq) was measured by us and the absolute absorption
scale was taken from ref 44. The absorption spectrum of aqueous ClOO
has not previously been measured, and depicted here is the ClOO
1
8
spectrum measured in amorphous ice calibrated with the gas-phase
25
value for the extinction coefficient. The absorption spectra of aqueous
ClO and Cl:H O are taken from refs 45 and 46.
2
the residual 800 nm beam into a 2 mm ethylene glycol jet. The spectral
range from 400 to 295 nm was covered by frequency doubling the
white light continuum in a 1-mm -thick BBO crystal. To generate the
probe wavelengths from 220 to 317 nm, the white light continuum
was sum-frequency mixed with part of the 400-nm beam in a 0.2-mm-
thick BBO crystal. The probe beam was separated into signal and
reference beams using a beam splitter. The reference portion of the
probe was sent to a detector while the remainder of the beam was sent
through the sample. Both beams were spectrally filtered before they
were detected by two matched photodiodes and boxcar integrators, and
the ratio of the signals from the two boxcar integrators was sent to a
digital lock-in amplifier referenced to the pump beam modulation. The
probe wavelength was measured by a calibrated spectrometer and the
instrument-response time was 300 ( 50 fs (fwhm). The pulse energy
was kept well below 0.1 FJ to avoid photolysis by the probe pulse.
At 400 nm, a strong initial bleaching is observed which
recovers to reveal a small residual depletion in optical density
that persists out to the longest delays investigated. The recovery
of the signal exhibits a double exponential decay with time
constants of 6.0 ( 1.0 ps and 0.33 ( 0.05 ns in addition to a
constant absorption signal. The time constants determined for
this and other probe wavelengths are presented in Table 1. The
spectral region from 400 to 320 nm covers the absorption band
2
2
Aqueous solutions of OClO were prepared by mixing NaClO
K S O as described elsewhere. The OClO solution was exchanged
2 2 8
2
and
of the Cl:H2O complex and the ground-state (X B1 f A A2)
electronic absorption of OClO. The strong initial bleaching
observed at 400 nm is due to the removal of ground-state OClO
by the pump pulse and indicates that the primary photoproducts,
including the pertinent excited states of OClO, all absorb much
less at 400 nm than does ground-state OClO. The nearly
complete, fast recovery of the absorption at 400 nm has
previously been assigned to reformation and subsequently
3
8
frequently during the series of measurements, and the flow was adjusted
to ensure a fresh sample of OClO for each laser pulse. The absorbance
of the OClO solution was 5 at 400 nm, the optical path length was 2
mm, and a reservoir solution of 1 L was employed. A series of
measurements with different OClO concentrations showed no difference
in the dynamics observed indicating no OClO clustering effects at the
1
5
concentrations employed in this work. Absorption spectra obtained
before and after experiments were identical demonstrating that no
sample degradation occurred during the course of the experiment. The
overall uncertainty of the transient absorption measurements is estimated
to be less than (15% for the spectral region from 220 to 400 nm.
2
relaxation of vibrationally excited OClO(X B1) and the forma-
3
4-41
tion of Cl atoms.
The transient absorption signal is very
sensitive to the probe wavelength and, as the wavelength is
decreased, the bleaching at 400 nm is replaced by an induced
absorption caused by the increasing extinction coefficient of
the Cl:H2O complex (λmax ) 310 nm). This is exemplified by
the 320-nm transient absorption data (Figure 2). The evolution
in optical density at this probe wavelength also shows a double
decay with time constants of 6.3 ( 1.0 ps and 0.29 ( 0.07 ns
in addition to a constant absorption signal. To investigate the
spectral behavior of the induced absorption, transient absorption
Results and Discussion
Static Absorption Spectra of the Species Involved. In
accordance with previous work3
1,40
the present results are
interpreted in terms of the equilibrated spectral properties of
the following potential photolytic products OClO, Cl, ClO, and
2
2
ClOO (Figure 1). The near-UV (X B1 f A A2) absorption
(
44) Stritt, F.; Friedlander, S.; Lewis, H. J.; Young, F. E. Anal. Chem.
1954, 26, 1478-1884.
45) Kl a¨ ning, U. K.; Wolff, T. Ber. Bunsen-Ges. Phys. Chem. 1985, 89,
43-245.
46) Buxton, G. V.; Subhami, M. S. J. Chem. Soc., Faraday Trans. 1
1972, 68, 947-957.
spectrum of OClO in aqueous solution was measured on a Cary
(
2
19 spectrophotometer with 2-nm resolution. The spectrum
2
obtained is in good agreement with previous studies where the
(
extinction coefficient was determined to be 1050 M- cm at
1
-1

12798 J. Am. Chem. Soc., Vol. 122, No. 51, 2000
Thomsen et al.
Table 1. Kinetic Parameters Determined from the Analysis of the
Transient Absorption Data of OClO in Aqueous Solution Using the
Function f(t) ) A
1
exp(-t/τ
1
) + A
2
exp(-t/τ
2
) + A
3
Convoluted
with the Instrument-Response Time
probe
(
nm)a
A1b
τ1 (ps)
A2
τ2 (ns)
A3
2
2
2
2
2
2
2
3
3
3
3
4
20
30
40
50
60
70
80
0.92 ( 0.06 2.0 ( 0.5 0.04 ( 0.02 0.08 ( 0.05 0.04 ( 0.03
0.89 ( 0.06 2.5 ( 0.6 0.06 ( 0.02 0.09 ( 0.06 0.05 ( 0.03
0.88 ( 0.07 2.5 ( 0.5 0.06 ( 0.02 0.15 ( 0.06 0.06 ( 0.02
0.84 ( 0.05 3.5 ( 0.5 0.08 ( 0.03 0.19 ( 0.05 0.08 ( 0.02
0.77 ( 0.07 4.0( 0.5 0.13 ( 0.03 0.27 ( 0.05 0.10 ( 0.03
0.74 ( 0.08 5.0( 1.0 0.09 ( 0.03 0.30( 0.05 0.17 ( 0.03
0.70 ( 0.08 9.5 ( 1.8
0.30 ( 0.05
20 -0.63 ( 0.06 6.3 ( 1.0 -0.09 ( 0.05 0.29 ( 0.05 0.28 ( 0.06
40 -0.80 ( 0.04 10.8 ( 1.2 -0.07 ( 0.04 0.31 ( 0.05 0.13 ( 0.04
60 -0.93 ( 0.02 11.9 ( 0.2 -0.02 ( 0.03 0.32 ( 0.05 0.05 ( 0.02
80 -0.98 ( 0.02 10.2 ( 0.8 -0.02 ( 0.02 0.35 ( 0.06 0.00 ( 0.02
00 -0.96 ( 0.02 6.0 ( 1.0 -0.03 ( 0.02 0.33 ( 0.05 -0.01 ( 0.02
a
Wavelength at which the transient change in optical density was
analyzed. The transient change was also measured between 290 and
b
3
10 nm, but not analyzed for multiexponential behavior. Amplitudes
are normalized such that ∑ |A | ) 1. Errors represent one standard
i
deviation from the mean of the number of measurements performed at
a given wavelength.
dominated by the Cl:H2O complex as exemplified in Figure 2
by the measurement at 320 nm. The spectral evolution of the
transient absorption data in Figures 2 and 3 shows that the
majority of the induced absorption originating from the Cl:H2O
complex appears within the first 20 ps, whereas only a very
small and slow increase in the signal is observed for longer
delays (>20 ps). This observation immediately suggests that
Cl atoms are formed through two competing processes, reactions
2
and 3, with the majority of Cl formed within 20 ps. The time
scale for the slow increase in the induced absorption in the entire
spectral region from 400 to 320 nm is about ∼0.32 ns (Table
1
). Time-resolved resonance Raman studies have shown that
this same time scale corresponds to the lifetime of aqueous
4
1
ClOO. This immediately suggests that a small number of Cl
atoms are produced through the formation and decomposition
of ClOO on a ∼0.32 ns time scale.
Interpretation of the transient absorption curves recorded at
shorter wavelengths (λmax < 320 nm) is rather involved since
the absorption cross sections for all photoproducts (ClOO, ClO,
and the Cl:H2O complex) are substantial. Furthermore, each
curve in this spectral region displays an initial sharp peak, which
is due to two-photon absorption by water (one pump and one
4
3
probe photon). If thermal dissociation ClOO is responsible
for the slow production of Cl atoms, a decrease in the absorption
from 270 to 240 nm due to the decay of ClOO concentration
should be observed. This is indeed the behavior observed in
Figures 2 and 3. At 260 nm, an instrument-limited increase in
the absorption is followed by a double exponential decay with
time constants of 4.0 ( 0.5 ps and 0.27 ( 0.05 ns to a constant
absorption level (Figure 2). Furthermore, the transient absorption
spectra depicted in Figure 3 unambiguously demonstrate the
formation and decay of ClOO. The transient spectrum at 50 ps
has a clear absorption maximum centered around 260 nm, which
can be identified as ClOO through comparison with the spectra
presented in Figure 1. In the 0.46-ns spectrum, the 260-nm band
has disappeared, and the absorption between 300 and 400 nm
has increased consistent with Cl:H2O complex formation. The
spectra presented in Figure 3 also indicate that decomposition
of ClOO occurs within 0.5 ns resulting in the production of Cl
and O2. We therefore conclude that most of the Cl atoms are
formed within 20 ps, whereas a small number of Cl atoms are
produced through the formation and decomposition of ClOO
on a 0.32 ns time scale.
Figure 2. The transient absorption data of aqueous OClO measured
in the spectral region from 400 to 220 nm following excitation at 400
nm. The probe wavelength used for each data point is labeled and the
continuous curve is a fit to the measured data using a sum of two
exponentials and a long-time offset convoluted with the instrument
.
response function The time evolution of the transient absorption data
at 400 nm is dominated by the dissociation and reformation of OClO,
the data at 320 nm are dominated by the production of Cl atoms, and
the data at 260 nm are dominated by the decay of the ClOO isomer,
Figure 1.
spectra were constructed from the individual transient absorption
data, Figure 3. At short delays (3.5 ps) the spectral region
between 400 and 320 nm is, as already discussed, dominated
by the induced transient bleaching of the sample, whereas the
transient absorption originating from the Cl:H2O complex is seen
to dominate at longer delays. The time evolution of the Cl
production can be observed by investigating the transient data
in the spectral range 340 to 300 nm where the absorption is

Photochemistry of Chlorine Dioxide in Aqueous Solution
J. Am. Chem. Soc., Vol. 122, No. 51, 2000 12799
Figure 4. Transient absorption spectra measured 50 ps (9) and 0.46
ns (b) after excitation. The quantum yield for each of the reactions
channels is obtained using the spectra depicted in Figure 1 and an initial
concentration obtained from the 400 nm data (see details in the text).
The spectrum at 50 ps is overlade (s) with the spectrum obtained using
a quantum yield for direct Cl and ClOO formation of 8 ( 2% and 2 (
1%, respectively, where a 3 ( 2% quantum yield for ClO was assumed.
The remaining 87 ( 5% ClO + O was assumed to geminately
recombine on a subpicosecond time scale to reform OClO. The spectrum
at 0.46 ns is overlade (- - -) with the spectrum obtained using the same
condition as describe above, but also including the result that ClOO
decomposes to form Cl on a ∼0.32 ns time scale.
(
0.05 and 0.19 ( 0.05 ns, respectively. As the probe pulse is
tuned to shorter wavelengths, the time-dependent signals decay
faster and the long-time absorption signal decreases (Table 1).
From the equilibrium spectra, depicted in Figure 1, it is seen
that ClOO has maximum absorption around 260 nm with a small
absorption between 240 and 220 nm. The slow part of these
transients decays at a decreasing rate when going from 220 to
2
40 nm, and approaches that of the 260-nm transient. This
spectral narrowing of the ClOO transient with time is thus
similar to what is expected for a vibrationally relaxing species
and suggests that ClOO is formed in a vibrationally excited state.
The decay in the transient absorption signal in the spectral region
around the ClOO absorption band might contain contributions
from both vibrational relaxation and thermal decomposition of
ClOO.
The transient absorption data at 280 nm shown in Figure 2
demonstrate that the long-time absorption signal is constant after
5
0 ps, indicating an isosbestic point where the absorptions
originating from ClOO and the Cl:H2O complex are comparable,
in agreement with the spectra shown in Figure 1. The isosbestic
point at this wavelength indicates that the solution-phase
spectrum of ClOO is very similar to the amorphous ice spectrum
depicted in Figure 1.
Quantum Yield for ClOO Formation. To further justify
the assignments made above, the measured absorption spectra
obtained at delays of 50 ps and 0.46 ns after the photolysis are
presented in Figure 4. The quantum yield for each of the four
reactions, eq 1-4, including the two proposed reactions leading
to Cl formation, can be extracted from the measured transient
data using the known photoproduct extinction coefficients
Figure 3. The transient absorption spectra of aqueous OClO measured
in the spectral region from 400 to 220 nm following excitation at 400
nm. The time delay used for each spectrum is labeled. The transient
spectrum at 50 ps has a clear absorption maximum centered at 260
nm, which can be identified as ClOO. In the 0.46-ns spectrum, the
2
4
60-nm band has disappeared, and the absorption between 300 and
00 nm has slightly increased consistent with Cl:H O complex
2
formation. By investigating the time evolution of the transient spectra
in the region from 400 to 300 nm it is observed that the majority of
the induced absorption originating from the Cl:H O complex appears
2
within the first 20 ps, indicating that the main portion of Cl atoms is
not formed through the ground-state ClOO isomer.
(Figure 1). When analyzing the spectra, it is assumed that all
the species have become fully thermalized by 50 ps, so that the
extinction coefficient (ꢀi) of the equilibrated species applies.
Furthermore, it is assumed in this analysis that the initial
The time-dependent signals at 270 and 250 nm show the same
behavior as the 260-nm signal, but with decay constants of 0.30

12800 J. Am. Chem. Soc., Vol. 122, No. 51, 2000
Thomsen et al.
concentration of photoexcited OClO can be found using the
initial bleaching at 400 nm. Taking the maximum bleaching at
of the ClOO extinction coefficient might change when going
from gas to solution phase. However, within the uncertainty of
our data, the quantum yield for ClOO formation is equaled by
the increase in the quantum yield of Cl atoms when going from
50 ps to 0.46 ns. Hence, the reported quantum yields and the
shape of the ClOO spectrum in water can be obtained self-
consistently from the experimental data without any prior
knowledge or assumptions concerning the ClOO spectrum in
ice. The ClOO spectrum obtained is instead critically dependent
on the shape and extinction coefficient of the Cl:H2O spectrum
and we estimate an uncertainty on the ClOO extinction
coefficient on the order of (25%
4
00 nm (∆A(400 nm) ) - 0.008, Figure 2) to indicate the
initial, photoinduced deficit of ground-state OClO, and hence
the initial concentration of photoexcited OClO, we obtain
-
5
∆
[OClO]t)0 ≈ 5 × 10 M. In the absence of significant hole-
burning, the absolute value of ∆[OClO]t)0 is a lower limit since
we have assumed that none of the photoproducts or excited
states of OClO present absorb at 400 nm. The spectra are
analyzed using the constraint that the total concentration of all
the photoproducts equals the residual concentration of OClO
that has not disappeared via fast geminate recombination of ClO
and O:
The analysis of the transient absorption spectra presented here
indicates that the current model for Cl production in aqueous
solution should be extended. Thus, in this model the chlorine
atoms are only produced through a short-lived intermediate with
∆
[OClO](t) ) [Cl](t) + [ClO](t) + [ClOO](t)
(5)
Furthermore, we assume that the transient absorption spectrum
3
4,38-39
a lifetime of 6 ps.
However, the results presented here
∆Aλ,t can be expressed simply as:
indicate that two channels are involved in the Cl production,
with the majority of Cl formed within 20 ps and the remaining
∆
Aλ,t ) dOClO{ꢀ [Cl] + ꢀClO[ClO] + ꢀClOO[ClOO] -
Cl
(
20%) formed through the formation and thermal decomposition
of ClOO.
Formation and Relaxation of ClOO. The possibility of
ꢀOClO∆[OClO]}λ,t (6)
where d is the thickness of the sample and ꢀi is the extinction
coefficient of the equilibrated species involved. The concentra-
tion of all the photoproducts is derived from the concentration
of initially excited OClO molecules and the static spectra shown
in Figure 1. Finally the quantum yield defined as Φ(X)/[X]/∆-
ClOO formation in aqueous solution has been discussed for the
past decade, but it was not until recently that it was observed
41
directly following the photolysis of aqueous OClO. Time-
resolved resonance Raman (TRRR) spectroscopy of OClO
following 390 nm excitation showed the appearance of ClOO
[
OClO]t)0 where X ) Cl, ClO, and ClOO is calculated. From
this analysis, we obtain the quantum yields for ClOO of 2 (
%, Cl of 8 ( 2%, and ClO of 3 ( 2%. The analysis also shows
that the majority of photoexcited OClO (87 ( 5%) reforms
1
3 ( 2 ps after the photolysis pulse. The integrated scattering
intensity of ClOO was best modeled by a sum of two
exponentials resulting in an increase and decay time constant
of 28 ( 5 ps and 0.40 ( 0.05 ns, respectively. The ClOO decay
time constant of ∼0.4 ns was interpreted as thermal dissociation
resulting in the formation of Cl and O2, whereas vibrational
kr ei nl a ext ai ct si o. n most likely contributed to the ∼30 ps appearance
1
34-41
within 50 ps in agreement with previous femtosecond studies.
The induced absorption signal at 0.46 ns is assumed to be due
to the production of Cl and residual ClO that has not
recombined, whereas all ClOO is assumed to have decayed.
The contribution to the total absorption from the small amount
of ClOO that has yet not decayed at this time is insignificant
and has therefore, within experimental error, no influence on
the final quantum yields. Hence, the spectrum at 0.46 ns can
be reproduced by assuming quantum yields for Cl and ClO
formation of 10 ( 3% and 3 ( 2%, respectively. The obtained
quantum yields at 50 ps and 0.46 ns clearly demonstrate the
relationship between the decay of ClOO and the formation of
Cl at long delays. This is in agreement with estimates of the
barrier for thermal dissociation of ClOO along the Cl-O
coordinate of only 0.26 eV.7 In our analysis, diffusive
geminate recombination on a long time scale between the
primary photoproducts has been neglected. Previous studies of
diffusion-limited geminate recombination have shown that the
Valuable information about the possible formation mechanism
of ClOO from photoexcited OClO in aqueous solution may be
obtained from high-level ab inito calculations on the electronic
potential energy surfaces of ClOO, OClO, and their asymptotic
correlation to product states. Arguments based on correlation
diagrams in solution-phase photochemistry might be misleading,
since studies of CS2, for instance, have shown that selection
rules pertaining to gas-phase dynamics become less stringent
49
in solution phase. Thus, the discussion below is only intended
as a guide. The photolysis of OClO in the near-UV region is
,27
2
2
initiated by exciting from the B1 ground state to the A2 state,
2
2
which couples to the nearby A1 and B2 states through spin-
2
8
orbit and vibronic interactions. In the liquid phase these
processes are thought to be faster than the reactive changes, so
primary dynamics happen within the first 30 ps, which is
2
that the photochemistry of OClO originates from the B2 state,
consistent with our model.4
7,48
This assumption is further
where the three competitive reactions 1-3 take place.3
0-33
Cl
supported by the fact that the majority of the ClO and O
2
atoms can thus be produced from the electronically excited B2
of OClO via two allowed reaction channels. The first mechanism
occurs by direct asymmetric photoisomerization of OClO to
34-41
fragments geminately recombine within the first picosecond.
It is important to recall that the spectrum used to represent ClOO
was that obtained in amorphous ice calibrated with the absolute
scale from recent gas-phase measurements. The ability of this
spectrum to completely reproduce the short time (50 ps) optical
density decay observed here implies that the solution-phase
spectrum of ClOO is similar to that obtained in amorphous ice.
This is further supported by the observed isosbestic point at
ClOO. Under Cs symmetry, the lowest excited state of OClO
2
(
B2) only correlates with an excited state of ClOO and not the
2
7,26,28
A′′ ground state.
Therefore, direct isomerization of
photoexcited OClO results in the formation of excited-state
2
1
ClOO which can decay to form Cl( P2) and O2( ∆g) or
3
-
O2( Σ ). However, excited-state internal conversion of ClOO
g
280 nm, indicating the absence of large solvent effects on the
to the ground-state A′′ surface can precede dissociation in which
ClOO absorption band. It should be noted that the absolute value
2
3
-
g
case Cl( P2) and O2( Σ ) are formed by ground-state decom-
(
47) Otto, B.; Schroeder, J.; Troe, J. J. Chem. Phys. 1984, 81, 202-
13.
48) K u¨ hne, T.; V o¨ heringer, P. J. Chem. Phys. 1996, 105, 10788-10802.
2
(49) Thomsen, C. L.; Madsen, D.; Thøgersen, J.; Byberg, J. R.; Keiding,
S. R. J. Chem. Phys. 1999, 111, 703-710.
(

Photochemistry of Chlorine Dioxide in Aqueous Solution
J. Am. Chem. Soc., Vol. 122, No. 51, 2000 12801
position of ClOO.27 The second mechanism involves symmetric
ClOO from aqueous OClO remains unidentified with additional
work required to address this issue.
dissociation along a C2V reaction coordinate producing also Cl-
2
1
3 28
-
(
P2) and O2( ∆g) or O2( Σ ).
g
The femtosecond transient absorption data presented here
Conclusions
demonstrate that atomic chlorine is produced on two separate
time scales: the majority (80%) is observed within 20 ps,
whereas the remaining (20%) is produced on a much longer
time scale as a result of thermal decomposition of ClOO. The
observation of the majority of the Cl atoms within 20 ps is in
agreement with the previously proposed model for production
of Cl atoms, which involved a short-lived unidentified precursor
for the Cl formation.38,39 The short-lived Cl atom precursor
absorbs strongly between 220 and 300 nm, and an extinction
coefficient of this species comparable to that of ClOO has
The femtosecond data presented here represent the first direct
measurement of the quantum yield of the ClOO formation in
aqueous solution, and significantly clarify the role of this species
in the mechanism for chlorine formation following the photolysis
of OClO. The main result of the investigation presented here is
that, in contrast to previously reported results, chlorine atoms
are formed through two competing reaction channels. The
quantum yield for ClOO and Cl formation was estimated using
a transient absorption spectrum obtained at 50 ps to be 2 ( 1%
and 8 ( 2%, respectively, whereas the quantum yield for ClO
was estimated to be only 3 ( 2%. The lifetime of ClOO was
found to be ∼0.32 ns in very good agreement with the lifetime
found for ClOO in aqueous solution using time-resolved
resonance Raman spectroscopy. Furthermore, the solution-phase
spectrum of ClOO was found to be very similar to that obtained
in amorphous ice. The dissociation of ClOO was found to result
in formation of Cl and O2 in agreement with that the barrier
for thermal dissociation along the Cl-O coordinate estimated
to be only 0.26 eV. The spectrum obtained at the longest delay
investigated (0.46 ns) could be reproduced assuming a total
quantum yield for Cl formation of 10 ( 3% and quantum yield
for ClO formation of 3 ( 2%. This result supports the assumed
relationship between the decay of ClOO and the formation
of Cl at long delays. The analysis also shows that most
3
8,39
previously been obtained using a simple kinetic model.
However, the absorption spectrum derived for this component
differs significantly from the ground-state absorption spectrum
of ClOO shown in Figure 1 and the precursor was therefore
not assigned to ClOO. Instead, it was proposed that the precursor
was an excited state of OClO approaching a configuration
3
8
resembling that of ClOO before dissociating into Cl and O2.
The observation of a precursor for the Cl production together
with the correlation diagram discussed above makes it tempting
to propose that the excited state of ClOO is the precursor for
the direct chlorine production. However, ab initio calculations
2
7
indicate that all the excited states of ClOO are dissociative
and will therefore probably dissociate into Cl and O2 on a much
shorter time scale than 6 ps. This directly addresses the next
issue, the formation of ground-state ClOO. The direct formation
(87 ( 5%) of the photoexcited OClO molecules in aqueous
2
of ClOO in the electronic ground state from the excited B2
solution reform via geminate recombination in agreement with
previous studies. The analysis of the time evolution of the
transient observed between 280 and 220 nm indicates that ClOO
might be formed with excess vibrational energy. The observation
of excess vibrational energy is in good agreement with the
proposed mechanism for production of ClOO, with ground-state
ClOO formed via internal conversion from the excited state or
via geminate recombination of Cl and O2.
state of ClO2 is intriguing, because the correlation diagram based
on orbital symmetry suggests that ClOO should be formed in
an excited state, and the excited states of ClOO are thought to
be dissociative.7
,26,28
It is therefore doubtful if excited-state
internal conversion of ClOO to the ground-state A′′ surface can
precede dissociation. Hence, recombination of Cl and O2, which
is known to yield ClOO in the gas phase, would seem a more
likely source of ground-state ClOO.2
2,25
The possibility that
ClOO can be formed by geminate recombination of ClO + O
is excluded, since the time delay for ClOO appearance was
found to be 13 ( 2 ps in the TRRR experiment, substantially
slower than the subpicosecond time scale for ground-state OClO
Acknowledgment. The authors would like to thank Jan
Thøgersen and Jørgen R. Byberg for many stimulating discus-
sions during the course of this work. Furthermore, the authors
are grateful to the Danish National Science Research Council
for financial support. P.J.R. would like to acknowledge the
National Science Foundation for their support of this work
through the CAREER program (CHE-9701717). P.J.R. is a
Cottrell Fellow of the Research Corporation and an Alfred P.
Sloan Fellow.
4
1
formation.
Both mechanisms for ClOO formation, excited-state internal
conversion and geminate recombination of Cl and O2, result in
production of vibrationally excited ground-state ClOO in
agreement with the present observation and the TRRR data. The
proposed interpretation of the data is not conclusive and the
initial mechanism of the photolytic formation of Cl atoms and
JA002148O