A laser flash photolysis study of nitrous acid in the aqueous phase

Article abstract of DOI:10.1016/j.cplett.2004.12.054

The OH quantum yield from the photolysis of nitrous acid in the aqueous phase by the 355 nm light was measured to be 0.25 ± 0.03. OH radical thus formed reacted readily with HNO₂ to produce NO₂, which sequentially reacted with HNO₂ to form the HNO₂-NO ₂ adduct. The NO₂ + HNO₂ reaction was reversible with a forward rate constant of 3.76 × 10⁷ dm ³ mol^-1 s^-1 and a backward rate constant of 1.06 × 10⁵ s^-1. Decay of the HNO₂-NO ₂ adduct would most likely generate HNO₃ and NO at a rate constant of 3.0 × 10³ s^-1.

Full text of DOI:10.1016/j.cplett.2004.12.054

Chemical Physics Letters 402 (2005) 306–311
www.elsevier.com/locate/cplett
A laser flash photolysis study of nitrous acid in the aqueous phase
*
Bin Ouyang, Wenbo Dong, Huiqi Hou
Department of Environmental Science and Engineering, Institute of Environmental Science, Fudan University, Handan Road 220,
Shanghai 200433, PR China
Received 30 August 2004; in final form 5 November 2004
Available online 5 January 2005
Abstract
The OH quantum yield from the photolysis of nitrous acid in the aqueous phase by the 355 nm light was measured to be
0.25 0.03. OH radical thus formed reacted readily with HNO2 to produce NO2, which sequentially reacted with HNO2 to form
the HNO2–NO2 adduct. The NO2 + HNO2 reaction was reversible with a forward rate constant of 3.76 · 107 dm3 molꢀ1 sꢀ1 and
a backward rate constant of 1.06 · 105 sꢀ1. Decay of the HNO2–NO2 adduct would most likely generate HNO3 and NO at a rate
constant of 3.0 · 103 sꢀ1
.
ꢀ 2004 Elsevier B.V. All rights reserved.
1. Introduction
was also proposed by Fischer to explain the apparent
formation of NOꢀ3 and phenol. Arakaki et al. [5] re-
ported a value of 3.1 · 10ꢀ4 sꢀ1 for the OH formation
rate constant from the photolysis of HNO2 at the
34ꢁN under the vernal equinox solar irradiation and
suggested it to be one of the potent OH sources in
the dew.
NOꢀ2 and HNO2 are two forms of N (III) existing
in aqueous solution. For NOꢀ2 , Treinin and Hayon
[6] applied the flash photolysis technique in 1970
to investigate its photochemical reactions and assigned
the observed transient species as N2O3, N2O4 and
NO2, respectively. However, no literature data
was available concerning the photo-initiated micro-
scopic reactions of HNO2 in solution. Such investiga-
tion is important in order to understand the
photochemical behavior of HNO2 in the aqueous
phase.
The
photolysis
of
nitrous
acid
(HNO2
(g) + hv ! OH (g) + NO (g)) is an important source
of OH radicals in the atmosphere. A great deal of
research has been focused on the atmospheric chemis-
try of HNO2 [1]. Cox [2] found that the OH quantum
yield from the photolysis of HNO2 by the 300–380 nm
light is near 0.9 in the gas phase. Cox [2], and later
Jenkin and Cox [3] reported two values, i.e.,
1 · 109 dm3 molꢀ1 sꢀ1 and 2.8 · 109 dm3 molꢀ1 sꢀ1
respectively, for the gas phase rate constant of the
OH + HNO2 reaction.
There are also some literature data on the aqueous-
,
phase
photolysis
of
nitrous
acid
(HNO2
(aq) + hv ! OH (aq) + NO (aq)). Fischer and War-
neck [4] used benzene as an OH radical scavenger
and derived an OH quantum yield of 0.35 from
photolyzing HNO2 by the 280–390 nm light in the
aqueous phase. The corresponding reaction mechanism
HNO2 has a strong absorption towards the UV irra-
diation in the 300–400 nm region. In this Letter, we used
the third harmonic of Nd:YAG laser (i.e. 355 nm) as the
photolyzing light source and directly monitored the
transient species. The reaction mechanism leading to
the formation of these transient species has been
postulated.
*
Corresponding author. Fax: +86 21 6564 3849.
E-mail addresses: oybboy@yahoo.com.cn, fdesi@fudan.edu.cn
(H. Hou).
0009-2614/$ - see front matter ꢀ 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.cplett.2004.12.054

B. Ouyang et al. / Chemical Physics Letters 402 (2005) 306–311
307
2. Experimental section
The output from a Quanta-Ray GCR-150 Nd:YAG
laser operating at its third harmonic mode (355 nm,
HWFM 5–6 ns, repetition rate 10 Hz, maximum output
energy 150 mJ/pulse, radius of the laser beam cross sec-
tion 0.40 cm, Q-switched, Spectra Physics) was used to
photolyze the HNO2 solution. The laser-excited area
was approximately 0.45 cm2. The laser pulse energy
was measured by a Nova PE25BB-SH-V2 pyroelectric
head (Ophir Optronics Ltd.). Transient spectra were re-
corded by the LKS.60 spectrometer (Applied Photo-
physics Ltd.). The analyzing light was emitted from a
150 W xenon arc lamp and entered the 1 · 1 cm quartz
cuvette at a normal angle to the laser beam with the ana-
lyzing optical length fixed at 0.65 cm. A programmable
Fig. 1. Transient absorption spectra recorded at 20 ns following the
laser flash photolysis of 0.01 mol dmꢀ3 HNO2 solution at pH 1.5.
other photolysis fragment, i.e., NO, does not exhibit
absorption at k > 240 nm [6], the absorption spectrum
in Fig. 1 could only be attributed to that of OH. This
assignment is confirmed by the resemblance between
Fig. 1 and the OH absorption spectrum recorded by Jay-
son et al. [8] using the pulse radiolysis technique. More-
over, we acquired the same absorption spectrum as Fig.
1 after the 266 nm laser flash photolysis of the H2O2
solution, which again verified our assignment. OH
should not have reacted with HNO2 dramatically at
the measurement time frame due to the absence of an
absorption band of NO2 near 400 nm [6,7] in Fig. 1.
The concentration of OH following photolysis was
f/3.4 grating monochromator with
a symmetrical
Czerny-Tuner optical configuration was combined with
a R928 photomultiplier for the transient signal detection
at different wavelengths. Data points were collected by
an Infiniium oscilloscope (model number 54820A, max-
imum sampling speed 2 GSa/s, bandwidth 500 MHz,
output impedance 50 X, Agilent).
All reagents were of analytical grade. Solutions of
HNO2 in triply distilled water were prepared by acidify-
ing the corresponding NaNO2 solution to pH 1.5 so that
more than 98.5% of NOꢀ2 existed in the HNO2 form.
This ensured that HNO2 would account for more than
99% of the 355 nm steady-state absorbance of the sam-
ple solutions when taking the molar extinction coeffi-
cients of HNO2 and NO2ꢀ at 355 nm, i.e., 41.0 and
23.0 dm3 molꢀ1 cmꢀ1, respectively, as determined by a
photodiode array UV–Vis spectrometer (S-3100, Scin-
co), into account. Wherever necessary, the deaeration
was achieved through cal. 15 min vigorous bubbling of
the solution by nitrogen gas (purity > 99.999%).
All experiments were performed under flow condi-
tions. The temperature of the photolysis sample was
15 1 ꢁC.
ꢀꢀ
determined using the Br2 dosimetry (see reaction
(E.1)–(E.3) [9,10]).
k1
Brꢀ þ OH ¢ BrOHꢀ
ðE:1Þ
ðE:2Þ
ðE:3Þ
k
ꢀ1
k2
BrOHꢀ þ H ¢ Br þ H2O
k
ꢀ2
k3
ꢀꢀ
Br þ Brꢀ ¢ Br2
k
ꢀ3
The concentrations of H+, Brꢀ and HNO2 were care-
fully adjusted ([H+] = 0.032 mol dmꢀ3, [Brꢀ] = 5 · 10ꢀ2
mol dmꢀ3, [HNO2] 6 6 · 10ꢀ4 mol dmꢀ3) to make sure
that equilibria (E.1)–(E.3) predominantly shifted to the
right side and that more than 99.8% OH radicals were
3. Results and discussion
scavenged by Brꢀ to form Br2 . It was also noteworthy
ꢀꢀ
3.1. Transient absorption spectrum following flash
photolysis of HNO2 solution and the quantum yield of OH
radical
that under the [HNO2] employed here, the absorbance
of HNO2 towards the 355 nm photolysis laser (denoted
as AHNO ; 355 nm here) was much smaller than 1, which made
2
the expansion 1 ꢀ 10ꢀA
¼ ln 10 ꢁ AHNO ; 355 nm
HNO ; 355 nm
2
2
Fig. 1 shows the transient absorption spectrum re-
corded at 20 ns following the laser flash photolysis of
a 0.01 mol dmꢀ3 aqueous solution of HNO2. Since the
reasonable.
UOH could then be derived using the following
expression:
ꢀ
Br2ꢀ; 360 nm
A
1
h ꢃ c ꢃ NA ꢃ p ꢃ R2
UOH
¼
ꢁ
ꢁ
;
ð4Þ
10ꢀA
ꢃ ln 10 ꢃ ðeBr ; 360 nm ꢀ eHNO ; 360 nmÞ ꢃ eHNO ; 355 nm ꢃ kexc ꢃ Lanal ꢃ 1000
quartz
Elaser
½HNO2ꢂ
ꢀ
ꢀ
2
2
2

308
B. Ouyang et al. / Chemical Physics Letters 402 (2005) 306–311
ꢀꢀ
monitored at 360 nm at 100 ns (the time point at which
ꢀ
ꢀ; 360 nm
where ABr
is the transient absorbance of Br2
have employed. Although the reaction HO2 + HNO2
has not been reported yet (to the best of our knowledge),
the oxidation of HO2 by another weaker but compara-
2
ꢀꢀ
significant decay of Br2 did not occur), Elaser is the laser
energy per pulse in J, h is 6.626 · 10ꢀ34 J s, c is 3.0 · 1010
cm sꢀ1, NA is 6.02 · 1023 molꢀ1, R is the radius of the la-
ser beam cross-section in cm, Aquartz is the absorbance of
a single wall of the quartz cuvette towards the 355 nm
ble oxidant, i.e., NOꢀ2 , at
a rate constant of
5 · 106 dm3 molꢀ1 sꢀ1 was confirmed by Julien [15],
which renders the corresponding oxidation of HO2 by
HNO2 quite reasonable.
ꢀ
ꢀ; 360 nm
laser, eBr
is the molar extinction coefficients (in
With the value UOH = 0.25 0.03, we could calculate
the OH formation rate constant from the HNO2 photol-
ysis in the aqueous phase by actinic radiation to be
(3.8 0.4) · 10ꢀ4 sꢀ1, which agreed well with the exper-
imentally determined value, i.e. 3.1 · 10ꢀ4 sꢀ1 reported
by Arakaki et al. [5].
2
dm3 molꢀ1 cmꢀ1) of Br2 at 360 nm, eHNO ; 355 nm and
ꢀꢀ
2
eHNO ; 360 nm are the molar extinction coefficients (in
2
dm3 molꢀ1 cmꢀ1) of HNO2 in aqueous solution at 355
and 360 nm, respectively, kexc is the wavelength of the
excitation laser in cm and Lanal is the analyzing optical
length in cm.
Good linear correlation was observed between
ꢀ
3.2. Reaction of OH with HNO2
ꢀ
Br2ꢀ; 360 nm
Br2ꢀ; 360 nm
½HNO2ꢂ
A
when [HNO2] was fixed. The linear relationship between
and Elaser, i.e., A
¼ B
ꢁ Elaser
At approximately 0.5 ls after the laser flash photoly-
sis of aqueous solution containing 1 · 10ꢀ2 mol dmꢀ3
HNO2 at pH 1.5, a transient absorption band appeared
with its peak located around 400 nm. Since this band
was quenched completely by excessive ethanol and its
spectral characteristic was identical with that of NO2
as reported previously [6,7], we have assigned it to
NO2 which was generated via reaction (R.5).
ꢀ
ꢀ
ABr
and Elaser suggests the formation of OH from
; 360 nm
2
the photolysis of HNO2 in the aqueous phase is a single
photon process. By varying HNO2 concentration and
measuring the corresponding B
, we got a plot of
against [HNO2], as shown in Fig. 2. Linear fit
½HNO2ꢂ
B
½HNO2ꢂ
of this plot gave B
¼ 860 ꢁ ½HNO2ꢂ, the result of
½HNO2ꢂ
ꢀ; 360 nm
ꢀ
which, together with Aquartz = 0.018, eBr
¼
2
9767 dm3 molꢀ1 cmꢀ1 (the average of literature values
k5
[11–13]) and eHNO ; 360 nm ¼ 41:7 dm3 molꢀ1 cmꢀ1 was
OH þ HNO2 ! NO2 þ H2O
ðR:5Þ
2
later introduced into the expression (4) to get
U
OH = 0.25 ( 0.03).
In previous studies by Fischer and Warneck [4] and
3.3. Reaction of NO2 with HNO2
Arakaki et al. [5], the chain reaction HO2 + N-
O ! OH + NO2 may occur and this complicated their
kinetic analysis. However, Fischer excluded the possibil-
ity of a significant occurrence of this reaction since he
got a similar UOH in the strong-absorption regime of
HNO2 to that in the weak one [4]. Nevertheless this does
not necessarily imply that the reaction between HO2 and
NO is slow (actually it is very fast with a rate constant of
3.2 · 109 dm3 molꢀ1 sꢀ1 according to Goldstein and
Czapski [14]) or that the main reaction product was
not OH and NO2, as claimed by Fischer and Warneck
[4]. The absence of the chain reaction was most likely
to be caused by the interception of HO2 by HNO2 whose
concentration was much larger than that of NO under
the experimental conditions that Fischer or Arakaki
Fig. 3 shows the absorption profiles at 260 and
400 nm as a function of time in the range of 0–20 ls fol-
lowing the laser flash photolysis of 0.01 mol dmꢀ3 aque-
ous solution of HNO2 at pH 1.5. The decay and growth
of the transient absorption traces proceeded at synchro-
nous paces, as shown clearly in Fig. 3. The kinetics was
hardly affected by O2, NOꢀ3 (0–0.5 mol dmꢀ3) or H+
(0.01–0.1 mol dmꢀ3). It was noteworthy that we selec-
tively analyzed the transient absorption curve at
260 nm as a representative of the other ones recorded
at k 6 380 nm since the kinetic trends (both growth
and decay) of these curves were almost identical.
Fig. 3. The transient absorption curve recorded at 260 and 400 nm,
respectively, after the laser flash photolysis of 0.01 mol dmꢀ3 HNO2
solution at pH 1.5.
Fig. 2. The plot of BHNO against [HNO2] with the best linear-fit.
2

B. Ouyang et al. / Chemical Physics Letters 402 (2005) 306–311
309
k6
Interestingly, the transient spectra we recorded at
NO2 þ HNO2 ¢ HNO2 ꢀ NO2
ðE:6Þ
k
k 6 380 nm resembled that of N2O3 reported by Treinin
and Hayon [6], the species generated from the combina-
tion of NO and NO2. Was this species also responsible
for the transient absorption at k 6 380 nm in this work?
Unfortunately, the following experimental results did
not support this assignment.
ꢀ6
For convenience of kinetic analysis, we neglected the
decay of HNO2–NO2 during its building process since
significant decay was observed to occur at time scale
of millisecond. By fitting the growth curve acquired at
different [HNO2] with the single exponential function
y = A · eꢀkt + B and plotting the best-fit value of k
against [HNO2], we got a k–[HNO2] profile as shown
in Fig. 5.
(1) For a fixed [HNO2], the growth rate of the transient
curve at 260 nm hardly varied with the laser output
energy (Elaser) in the range of 8–60 mJ.
(2) The maximum transient absorbance at 260 nm
exhibited excellent linear relations with Elaser as
shown in Fig. 4.
Kinetic calculation revealed that k should equal to
k6 · [HNO2] + k if (E.6) did occur. Correspondingly,
ꢀ6
after linear fit of the k–[HNO2] curve, k6 and k could
ꢀ6
be derived readily by reading the slope and intercept of
(3) For control experiments, i.e., Elaser = 46 mJ,
[HNO2] = 2 · 10ꢀ3 mol dmꢀ3 and Elaser = 8 mJ,
[HNO2] = 1 · 10ꢀ2 mol dmꢀ3, the growth rate of
the transient curve in the latter solution was much
faster than that in the former one. Nevertheless
they were expected to be similar as long as the tran-
sient species was to be assigned to N2O3 since the
initial concentrations of NO and NO2 in the two
solutions have been proved to be identical using
the
fitted
line.
This
yielded
the
result
,
k6 = 3.76 · 107 dm3 molꢀ1 sꢀ1 and kꢀ6 = 1.06 · 105 sꢀ1
and by dividing k6 with kꢀ6, we calculated the equilib-
rium constant of (E.6), denoted as K6, to be
3.55 · 102 dm3 molꢀ1
K6 was also calculated via expression (7)
.
½HNO2 ꢀ NO2ꢂ
½NO2ꢂ ꢀ ½NO2ꢂ
e
0
e
K6 ¼
¼
ð7Þ
½NO2ꢂ ꢃ ½HNO2ꢂ ½NO2ꢂ ꢃ ½HNO2ꢂ
e
e
ꢀꢀ
the Br2 dosimetry, assuming that almost all of
the OH were scavenged by HNO2, as has been ver-
ified by our kinetic calculation.
where the subscript ÔeÕ and Ô0Õ denoted the equilibrium
and starting concentration, respectively, of the species
in the bracket. The values derived at different [HNO2]
were close and by averaging calculation we got
Experimental phenomena listed above gave us a
strong indication that the growth at 260 nm was in itself
a pseudo-first order process instead of a second-order
one. Moreover, the growth rate was accelerated by
increasing [HNO2], indicating that HNO2 may have
been involved into the reaction process.
We were hereby inclined to attribute a new adduct,
i.e., HNO2–NO2, to count for the transient absorption
at 260 nm. If this were true, a newly generated NO2
should sequentially combine with HNO2 rather than
NO. Furthermore, it was observed that the transient
absorbance at 400 nm did not drop to zero in the time
range of 0–20 ls (see Fig. 3), suggesting that an equilib-
rium was reached between the product adduct and the
reactant, as shown by expression (E.6).
K6 = 3.60 · 102 dm3 molꢀ1
, which approximates the
other result of K6 derived in the above paragraph and
again, validated the existence of reaction (E.6).
Cliffton et al. [16] and Spindler et al. [17] both noticed
that NO2 would react with HSOꢀ3 and SO32ꢀ via the pro-
cess of forming adducts instead of direct electron with-
drawing, which supports the rationality of reaction
(E.6). The oxygen atom that connects directly with the
H atom in the HNO2 molecule is among all the others
most feasible to be attacked by electrophilic reagents
e.g., H+ [18], and is thus most likely to join with the
Nd+ atom in the NO2 molecule once the intermediate ad-
duct has been formed.
Fig. 5. The plot of k against [HNO2] with the best linear-fit result; the
error bar represents the standard deviations of k derived from five
parallel experiments at a confidence of 95%.
Fig. 4. The Amax, 260 nm–Elaser profile derived at different HNO2
concentrations.

310
B. Ouyang et al. / Chemical Physics Letters 402 (2005) 306–311
The decay rate of the HNO2–NO2 adduct showed no
sign of concentration-dependence. Fitting the decay
curve at 260 nm by the first-order kinetics yielded a de-
(1) 2NO + O2: Deaeration of the sample solution would
not cause any difference in the transient absorption
curve, showing that reaction 2NO + O2 ! 2NO2
could be neglected over the entire monitoring time
range (20 ls ꢄ 1 ms). It could be explained by the
dmꢀ6 molꢀ2 sꢀ1[4], of this termolecular reaction.
(2) NO + HNO2: No reaction was observed when
injecting NO into nitrogen-saturated HNO2 solu-
tion, showing that the reaction NO + HNO2 is
immeasurably slow or does not happen at all.
(3) NO + NO2: Possibility of a significant occurrence
of this second-order reaction has been excluded,
as stated in Section 3.3. Formation of N2O3 was
suppressed by HNO2 as the efficient scavenger of
NO2. The concentration of HNO2 was by nearly
three orders of magnitude higher than that of NO.
(4) NO + NO2–HNO2: If this reaction occurred signif-
icantly, the decay rate of the NO2–HNO2 adduct
would have been influenced by the laser pulse
energy which determined the concentrations of
NO and NO2–HNO2. However, we hardly wit-
nessed such energy-dependence in the experiment.
cay rate constant of 3.0 · 103 sꢀ1
.
The decay products were unlikely to be ONOOH,
NO3 or HNO because no transient absorption signals
were observed during the decay process of HNO2–NO2
in the wavelength range of 240–800 nm, where strong
absorptions are expected for these species [6,19,20].
The formation of HNO3 and NO was therefore most
likely, as shown by reaction (R.8). The overall reaction
would then be (R.9) if this hypothesis were true, with
impressively
low
rate
constant,
4 · 106
a
standard Gibbs free energy change DrGU9 of
ꢀ24.54 kJ molꢀ1 [21]. The negative value suggests that
(R.9) is at least thermodynamically feasible and may
constitute one source of HNO3 in the photolysis of
aqueous solution of HNO2.
k8
HNO2 ꢀ NO2 ! HNO3 þ NO
ðR:8Þ
k9
NO2ðaqÞ þ HNO2ðaqÞ ! NOðaqÞ þ HNO3ðaqÞ
ðR:9Þ
Fischer and Warneck [4] noticed that the theoretically
predicted rise rate of NO3ꢀ using the reported reactions
was obviously lower than the experimentally observed
one, which to some extent validated (R.9): by adding
(R.9) to the reaction sketch summarized by Fischer, the
rise rate of NOꢀ3 was expected to increase due to a more
efficient conversion of NO2 to NOꢀ3 by (R.9) at the
expense of the reaction NO + NO2 ! N2O3 ! 2HNO2,
which essentially converted NO2 back to HNO2.
It was thus concluded that NO was not observed to be
involved into the chemical reaction process within 1 ms.
In the later time stages, it may dissolve into the solution.
4. Conclusion
According to the reactions (E.6) and (R.8), HNO2ꢀ
NO2 would either decompose to the final product
HNO3 and NO or dissociate back to the reactants once
formed. Interestingly, an opposite reaction as
The photolysis of HNO2 in the aqueous phase was
investigated using the laser flash photolysis technique.
The OH quantum yield was measured to be
0.25 0.03.
A
novel reaction NO2 + HNO2 ()
NOðgÞ þ HNO3ðgÞ ! NO2ðgÞ þ HNO2ðgÞ
ðR:10Þ
HNO2–NO2 was reported with a forward reaction rate
constant of 3.76 · 107 dmꢀ3 mol sꢀ1 and a backward
rate constant of 1.06 · 105 dmꢀ3 mol sꢀ1 at 15 1 ꢁC.
The HNO2–NO2 adduct would probably decompose
to form HNO3 and NO at a rate constant of
3.0 · 103 sꢀ1. These reactions need to be incorporated
into the solution phase chemistry of nitrogen.
has been recently reported by Mochida and Finlayson-
Pitts [22] and Saliba et al. [23]. They found HNO3 and
NO would react on the surface of porous or borosilicate
glass in the presence of water to form HNO2 and NO2,
which was the reverse of (R.9). Judging from the stan-
dard Gibbs free energy change of the reactions, one
would expect that reaction (R.10) is only thermodynam-
ically possible in the gas phase while the one (R.9) only
occurs in the aqueous phase. The direct reason for the
changing of the reaction direction is that NOꢀ3 can be
highly stabilized by its solvation process (DGU =
ꢀ37.8 kJ molꢀ1 [21]), which in turn notably lowers the
standard Gibbs formation free energy of the ion.
Acknowledgements
We thank Dr. Lei Zhu in the State University of New
York at Albany for her constructive comments on the
manuscript.
3.4. Fate of the other photolysis fragment NO
References
NO may react with O2, HNO2, NO2 and the NO2–
HNO2 adduct in the experiment system. However, pos-
sibility of a significant occurrence of the above reactions
has been excluded as following:
[1] B.J. Finlayson-Pitts, J.N. Pitts Jr., Atmospheric Chemistry
Fundamentals and Experimental Techniques, John Wiley & Sons,
1986.
[2] R.A. Cox, J. Photochem. 3 (1974) 175.

B. Ouyang et al. / Chemical Physics Letters 402 (2005) 306–311
311
[3] M.E. Jenkin, R.A. Cox, Chem. Phys. Lett. 137 (1987) 548.
[4] M. Fischer, P. Warneck, J. Phys. Chem. 100 (1996) 18749.
[5] T. Arakaki, T. Miyake, T. Hirakawa, H. Sakugawa, Environ. Sci.
Technol. 33 (1999) 2561.
[15] R. Julien, Int. J. Radiat. Phys. Chem. 7 (1975) 741.
[16] C.L. Cliffton, N. Altstein, R.E. Huie, Environ. Sci. Technol. 22
(1988) 586.
[17] G. Spindler, J. Hesper, E. Bru¨ggemann, R. Dubois, Th. Mu¨ller,
H. Herrmann, Atmos. Environ. 37 (2003) 2643.
[18] J.S. Francisco, J. Chem. Phys. 115 (2001) 2117.
[19] T. Løgager, K. Sehested, J. Phys. Chem. 97 (1993) 6664.
[20] P.H. Wine, R.L. Mauldin III, R.P. Thorn, J. Phys. Chem. 92
(1988) 1156.
[6] A. Treinin, E. Hayon, J. Am. Chem. Soc. 92 (1970) 5821.
[7] M. Ottolenghi, J. Rabani, J. Phys. Chem. 72 (1968) 593.
[8] G.G Jayson, B.J. Parson, J. Chem. Soc., Faraday Trans. I 69
(1973) 1597.
[9] A. Mamou, J. Rabani, D. Behar, J. Phys. Chem. 81 (1977) 1447.
[10] G. Merenyi, J. Lind, J. Am. Chem. Soc. 116 (1994) 7872.
[11] M. DÕAngelantonio, M. Venturi, Q.G. Mulazzani, Radiat. Phys.
Chem. 32 (1988) 319.
[21] A.D. John, LangeÕs Handbook of Chemistry, McGraw-Hill Inc,
1999.
[22] M. Mochida, B.J. Finlayson-Pitts, J. Phys. Chem. 104 (2000)
9705.
[12] P. Morliere, L.K. Patterson, Inorg. Chim. Acta. 64 (1982) 183.
[13] G. Ferraudi, J. Phys. Chem. 97 (1993) 2793.
[14] S. Goldstein, G. Czapski, Free Radical Biol. Med. 19 (1995) 505.
[23] N.A. Saliba, H. Yang, B.J. Finlayson-Pitts, J. Phys. Chem. A 105
(2001) 10339.