Production and infrared absorption spectrum of ClSO2 in matrices

Article abstract of DOI:10.1021/jp993792w

A new species, ClSO₂, is produced and identified with infrared (IR) absorption spectra when an argon or krypton matrix containing Cl₂ and SO₂ is irradiated with laser emission at 355 nm. Lines at 1311.0, 1309.6, 1099.8, 1098.2, 497.7, and 455.8 cm^-1 (or 1309.5, 1098.5, 497.0, and 454.2 cm^-1) are assigned to ClSO₂ isolated in solid Ar (or Kr). Assignments of IR absorption lines are based on results of ³⁴S and ¹⁸O isotopic substitution (in solid Kr) and theoretical calculations. Theoretical calculations using density-functional theories (B3LYP and B3P86 with an aug-cc-pVTZ basis set) were performed to predict the geometry, energy, vibrational frequencies, and infrared intensities of possible isomers of ClSO₂: (pyramidal) ClSO₂, cis-ClOSO, (nonplanar) ClOSO, and cis-ClSOO. Results predicted for pyramidal ClSO₂ agree with observed experimental data. This is the first identification of ClSO₂, which is presumably an important intermediate during photolysis of Cl₂SO₂ and in the reaction of Cl with SO₂, especially at low temperatures. In addition to ClSO₂, irradiation of the Cl₂/SO₂/Ar matrix sample with laser light at 308 nm produces Cl₂SO₂. Possible mechanisms of formation are discussed.

Full text of DOI:10.1021/jp993792w

J. Phys. Chem. A 2000, 104, 3613-3619
3613
Production and Infrared Absorption Spectrum of ClSO2 in Matrices†
Mohammed Bahou, Shih-Fung Chen, and Yuan-Pern Lee*
Department of Chemistry, National Tsing Hua UniVersity, 101, Sec. 2, Kuang Fu Road,
Hsinchu, Taiwan 30013
ReceiVed: October 26, 1999; In Final Form: January 31, 2000
A new species, ClSO2, is produced and identified with infrared (IR) absorption spectra when an argon or
krypton matrix containing Cl2 and SO2 is irradiated with laser emission at 355 nm. Lines at 1311.0, 1309.6,
1099.8, 1098.2, 497.7, and 455.8 cm-1 (or 1309.5, 1098.5, 497.0, and 454.2 cm-1) are assigned to ClSO2
isolated in solid Ar (or Kr). Assignments of IR absorption lines are based on results of 34S and 18O isotopic
substitution (in solid Kr) and theoretical calculations. Theoretical calculations using density-functional theories
(B3LYP and B3P86 with an aug-cc-pVTZ basis set) were performed to predict the geometry, energy, vibrational
frequencies, and infrared intensities of possible isomers of ClSO2: (pyramidal) ClSO2, cis-ClOSO, (nonplanar)
ClOSO, and cis-ClSOO. Results predicted for pyramidal ClSO2 agree with observed experimental data. This
is the first identification of ClSO2, which is presumably an important intermediate during photolysis of
Cl2SO2 and in the reaction of Cl with SO2, especially at low temperatures. In addition to ClSO2, irradiation
of the Cl2/SO2/Ar matrix sample with laser light at 308 nm produces Cl2SO2. Possible mechanisms of formation
are discussed.
I. Introduction
The reaction of Cl with SO2
to predict geometry, energy, vibrational frequencies, and IR
intensities of these species. They found that ClSO2 is nonplanar
with S at the top of the pyramid; the Cl-S bond distance (2.13-
2.15 Å) is relatively large. There seems to exist no experimental
report of ClSO2, even though it might be an important
intermediate during photolysis of Cl2SO2 and in reaction of Cl
with SO2.
Cl + SO2
9
M8 products
(1)
may play an important role in coupling of chlorine and sulfur
cycles in the atmosphere. Kaufman and co-workers1,2 reported
the only kinetic investigation of this reaction; small termolecular
rate coefficients, k1 ) (2.2 ( 1.8) × 10-34 exp[(1240 ( 1050)/
RT] and (9.9 ( 8.1) × 10-34 exp[(1770 ( 970)/RT] cm6
molecule-2 s-1 with Ar and SO2, respectively, as third bodies
were determined with a fast-flow reactor coupled with a
modulated molecular-beam mass spectrometer. Although such
values of k1 imply that reaction 1 is probably unimportant in
linking chlorine and sulfur cycles in the terrestrial stratosphere,
a possibility remains that reaction 1 may be important in the
vicinity of a volcanic eruption or in the atmosphere of Venus
in which sulfur compounds are present at large concentrations.3
DeMore et al.3 employed a Fourier transform infrared (FTIR)
spectroscopy to find that SO2 is quantitatively converted to
sulfuryl chloride (Cl2SO2) during photolysis of a mixture
containing Cl2 and SO2. With their modified model for the
stratospheric chemistry of Venus, it appears that Cl2SO2 is a
key reservoir species for chlorine and that reaction 1 contributes
to an important cycle for destroying O2 and converting SO2 to
H2SO4.
We have demonstrated the advantages of matrix isolation as
a technique in producing novel isomeric species by making use
of selective laser photolysis and matrix cage effect.6-13 In this
paper, we report production and identification of ClSO2 in
matrices with infrared spectra according to such techniques, and
quantum-chemical calculations of possible isomers of ClSO2.
II. Experiments
The experimental setup is similar to one described previ-
ously.6,10 Matrix samples were prepared on co-depositing
mixtures of Cl2/Rg and SO2/Rg (Rg ) Ar or Kr) with a typical
volumetric ratio of Cl2/SO2/Rg = 2/1/400 onto a platinum-plated
copper mirror maintained at 13 K. Approximately 4-10 mmol
of mixture was deposited over a period of 3 h.
A XeCl excimer laser (308 nm, 10 Hz with energies ∼5 mJ/
pulse) and a frequency-tripled Nd:YAG laser (355 nm, 10 Hz,
∼5 mJ/pulse) were employed to photodissociate Cl2 to facilitate
the reaction of Cl + SO2. Dye lasers (pumped by a XeCl
excimer laser or a frequency-doubled Nd:YAG laser) were used
to perform secondary photolysis at 415 and 640 nm. IR
absorption spectra were recorded after each stage of photo-
irradiation with an FTIR spectrometer equipped with a KBr
beam splitter and a Hg-Cd-Te detector (77 K). For the far-
infrared region, a Mylar beam splitter and a pyroelectric DTGS
detector were used. Typically, 900 scans were collected at a
Li4,5 performed ab initio calculations on XSO and XSO2 (X
) F, Cl) in their electronic ground and first excited states with
Mφller-Plesset perturbation theory with second-order correction
of correlation energy (MP2) and with quadratic configuration
interaction including single and double substitution (QCISD)
resolution of 0.3 cm-1
.
* To whom correspondence should be addressed. Jointly appointed by
the Institute of Atomic and Molecular Sciences, Academia Sinica, Taipei,
Taiwan. E-mail: yplee@chem.nthu.edu.tw. Fax: 886-3-5722892.
† Part of the special issue “Marilyn Jacox Festschrift”.
SO2 (99.98%, Matheson), Cl2 (99.9%, Solkatronic), Ar
(99.999%, Spectra Gases), and Kr (99.995%, Scientific Gas
Products) were used without further purification except degas-
10.1021/jp993792w CCC: $19.00 © 2000 American Chemical Society
Published on Web 03/17/2000

3614 J. Phys. Chem. A, Vol. 104, No. 16, 2000
Bahou et al.
Figure 1. Difference spectra of Cl2/SO2/Rg (2/1/400) matrix samples
after irradiation: (A) Rg ) Ar, irradiation at 308 nm for 20 min; (B)
Rg ) Kr, irradiation at 355 nm for 20 min; (C) after annealing of the
matrix described in (B) at 20 K for 3 min.
Figure 2. Difference spectra of various isotopically labeled Cl2/SO2/
Kr matrix samples after irradiation at 355 nm for 20 min: (A) SO2;
(B) 34SO2; (C) scrambled 16O- and 18O-substituted SO2, 16O:18O = 1:1;
(D) scrambled 16O- and 18O-substituted SO2, 16O:18O = 1:6.
sing of SO2 and Cl2 at 77 K. 34SO2 and S18O2 (Cambridge
Isotope Laboratories) have nominal isotopic purity of 93.5%
and 90%, respectively.
shown in Figure 1C. As intensities of new lines (set “a”) increase
after annealing, the intensity of the ν1 + ν3 line of SO2 decreases,
indicating that these new lines may be associated with a species
produced on reaction of Cl with SO2. Lines at 1309.5, 1098.5,
and 497.0 cm-1 (in Kr) are close to those of SO2 (1350.9,
1149.5, and 519.0 cm-1), indicating that the new species may
contain a SO2 moiety. Observation of a fourth line at 454.4
cm-1 indicates that the species contains more than three atoms.
Lines in set “b” observed after photolysis of the matrix sample
at 308 nm are readily assigned to absorption of Cl2SO2.
Literature values for IR absorption of gaseous Cl2SO2 are 1434,
1205, 586, and 577 cm-1 (ref 26), or 1437 and 1212 cm-1 (ref
3). We also recorded an IR absorption spectrum of Cl2SO2
isolated in solid Ar (Cl2SO2/Ar ) 1/1700) and observed lines
at 1425.9, 1200.5, 585.6, and 564.8 cm-1, consistent with
observed values for set “b”.
A far-IR spectrum of the irradiated matrix sample was also
recorded in the spectral region 200-700 cm-1; no new line other
than those already detected at 497.0 and 454.2 cm-1 was
observed.
B. 34S and 18O Isotopic Experiments. Because there are
multiple matrix sites for the new species (set “a”) in solid Ar,
we perform isotopic experiments in a Kr matrix in order to avoid
interference due to site splitting.
III. Computational Method
The equilibrium structure, vibrational frequencies, IR intensi-
ties, and single point energies were calculated with the Gaussian
98 program.14 We used three methods: MP215 and density
functional theory (DFT) of two types, B3LYP and B3P86. The
B3LYP method uses Becke’s three-parameter hybrid exchange
functional, which includes the Slater exchange functional with
corrections involving a gradient of the density, and a correlation
functional of Lee, Yang, and Parr, with both local and nonlocal
terms.16,17 The B3P86 method uses Becke’s three-parameter
hybrid exchange functional with Perdew’s gradient-corrected
correlation functional.18,19 The standard 6-31+G* basis set were
used in MP2 calculations and Dunning’s correlation-consistent
polarized valence triplet-zeta basis set, augmented with s, p, d,
and f functions (aug-cc-pVTZ)20 were applied in DFT calcula-
tions. Analytic first derivatives were utilized in geometry
optimization, and vibrational frequencies were calculated ana-
lytically at each stationary point.
IV. Results and Discussion
Similar experiments with photolysis of a Cl2/34S16O2/Kr (2/
1/400) matrix sample at 355 nm yield lines at 1291.6, 1091.4,
491.8, and 445.6 cm-1 (marked “x”), as shown in Figure 2B.
A spectrum recorded after photolysis of the Cl2/32S16O2/Kr
sample is shown in Figure 2A for comparison. When 32S16O2
was replaced with a scrambled mixture of 32SO2 with 16O:18O
= 1:1, the resultant spectrum (Figure 2C) recorded after
irradiation at 355 nm shows two additional lines (marked “y”
and “z”) for each of the three high-frequency lines in Figure
2A; loss of the reactant isotopomer S18O2 interferes with lines
associated with natural isotopomer of the new species. Isotopic
lines near 452 cm-1 are unresolved; the broadened line width
is ∼6.2 cm-1. To investigate further this broad feature, we
employed a second scrambled mixture of SO2 with a ratio 16O:
18O = 1:6; the resultant spectrum is shown in Figure 2D. In
this case the line at 448.4 cm-1 exhibits a line width similar to
the corresponding member at 454.2 cm-1 (from the Cl2/S16O2/
Kr sample) with wavenumber decreased by 5.8 cm-1. The broad
feature in Figure 2C is thus deconvoluted into three lines with
two lines fixed at 454.2 and 448.4 cm-1 to yield the peak
position of the central line at 452.4 ( 0.3 cm-1. Wavenumbers
of observed new lines that are associated with each isotopic
species are listed in Table 1.
A. Production of ClSO2. The IR spectra of SO2 in various
matrices are well characterized.11,21-23 Our observation of SO2
lines at 1355.2, 1351.1, 1152.2, 1147.1, 519.7, and 517.3 cm-1
(in solid Ar) and 1350.9, 1149.8, and 519.0 cm-1 (in solid Kr)
are consistent with previous reports. An additional set of lines
at 1348.2, 1146.6, and 516.7 cm-1 were observed for SO2 in
solid Kr at concentrations SO2/Kr = 1/200. Although absorption
lines of Cl2 at 549.2 and 554.3 cm-1 (in solid Ar)24 and at 547.1
cm-1 (in solid Ar, due to perturbation by O3)25 are reported,
we observed no such absorption in our Cl2/SO2/Ar (Kr) matrix.
Irradiation of an Ar matrix sample containing Cl2 and SO2
with laser emission at 308 nm for 20 min produced lines in
two sets at 1309.6, 1098.2, 497.7, and 455.8 cm-1 (marked “a”),
and at 1425.3, 1202.3, 586.3, and 566.5 cm-1 (marked “b”), as
shown in the difference spectrum Figure 1A. Additional broad
lines at 1311.6 and 1099.8 cm-1 in set “a” are observed; they
are due to matrix site splitting. Because the fundamental lines
of SO2 are saturated, weak lines due to ν1 + ν3 combination
near 2480 cm-1 (Figure 1) are used to monitor the variation in
concentration of SO2. Irradiation of a Kr matrix sample
containing Cl2 and SO2 with laser emission at 355 nm produced
similarly four lines at 1309.5, 1098.5, 497.0, and 454.2 cm-1
,
as shown in Figure 1B. Lines corresponding to those marked
“b” in Figure 1A are not produced at this photolysis wavelength.
Annealing of the irradiated sample at 20 K for 3 min followed
by cooling to 13 K enhanced the intensities of these lines, as
Upon 34S substitution, the four lines shift toward smaller
wavenumbers by 17.9, 7.1, 5.2, and 8.6 cm-1, respectively,
indicating that all four modes involve the motion of the S atom.

Production of ClSO2 in Matrices
J. Phys. Chem. A, Vol. 104, No. 16, 2000 3615
TABLE 1: Observed Vibrational Wavenumbers/cm-1 of
Various Isotopic Species of ClSO2 Isolated in Matrices
ClSO2
in Ar
Cl34SO2 ClS(16O)18O ClS18O2
mode
in Kr
in Kr
in Kr
in Kr
ν1
ν2
ν3
ν4
1309.6 (1311.0)a 1309.5 1291.6
1290.6
1071.6
488.8
1267.6
1049.5
480.1
1098.2 (1099.8)a 1098.5 1091.4
497.7
455.8
497.0
491.8
445.6
454.2b
∼452.4c
448.4
a Broad features. b This feature is deconvoluted to two lines at 455.3
and 453.4 cm-1, corresponding to 35ClSO2 and 37ClSO2, respectively;
see text. c Deconvoluted from a broad feature containing ClS16O2,
ClS(16O)18O, and ClS18O2; see text.
Two high-frequency lines at 1309.5 and 1098.5 cm-1 shift
substantially (41.9 and 49.0 cm-1) upon 18O substitution (Figure
2D). The isotopic ratios (defined as ratio of wavenumber of a
specific isotopomer to that of the natural species) of the fully
18O-substituted species, 1267.6/1309.5 ) 0.9680 and 1049.5/
Figure 3. Geometries of four isomers of ClSO2 calculated with the
B3P86/aug-cc-pVTZ method.
1098.5 ) 0.9554 for lines at 1309.5 and 1098.5 cm-1
,
C. Theoretical Calculations. 1. Geometry and Energy.
According to theoretical calculations (B3LYP and B3P86 with
an aug-cc-pVTZ basis set), there are four stable isomers of
ClSO2, as illustrated in Figure 3. Predicted geometry and
energies of each isomer are listed in Table 2 and compared with
previous results on pyramidal ClSO2 using MP2/6-311G(2d)
and QCISD/6-31G* methods.4,5
respectively, are similar to ratios 0.9680 and 0.9560 for
asymmetric stretching (ν3) and symmetric stretching (ν1) modes
of S18O2 (Table 4 of ref 11). A single component for the central
line of the triplet that was observed for scrambled 18O and 16
O
species (Figure 2C) indicates that motion of two O atoms is
equivalent in these modes. The isotopic ratios of 1290.6/1309.5
) 0.9856 and 1071.6/1098.5 ) 0.9755 for lines at 1309.5 and
1098.5 cm-1, respectively, are similar to the ratios 0.9860 and
0.9760 for corresponding modes (ν3 and ν1) of 16OS18O. Hence
lines at 1309.5 and 1098.5 cm-1 are assigned to asymmetric
stretching and symmetric stretching modes within the SO2
moiety of the new species. The line at 497.0 cm-1 is red-shifted
by 8.2 and 16.9 cm-1 upon 18O substitution. Observed isotopic
ratios of 0.9835 and 0.9660 differ slightly from values 0.9790
and 0.9580 for the bending (ν2) mode of 16OS18O and S18O2,
respectively (ref 11), indicating a small change of OSO bond
angle or mode mixing for this vibration. That only a single
component is observed for the central line of the triplet in the
scrambled 16O and 18O species indicates that motion of the two
O atoms in this mode is equivalent. Hence, the fourth (Cl) atom
of this new species must be bonded with the S atom. The 18O
isotopic shift (5.8 cm-1) of the line at 454.2 cm-1 is smaller
than that (8.6 cm-1) for the 34S species, indicating that motion
of the S atom is greater than that of O atoms in this mode. As
the S atom is likely central, absorption at 454.2 cm-1 might be
associated with an umbrella motion.
Although the presence of chlorine is required to produce the
new species, no observed line shows a distinct pattern of splitting
due to 35Cl and 37Cl isotopes. Observed lines are thus not directly
associated with the S-Cl stretching mode because 37Cl in natural
abundance is expected to yield resolvable line splitting, esti-
mated to be ∼6.5 cm-1 for a S-Cl stretching mode near 500
cm-1. The observed width (∼3.8 cm-1) of the line at 454.2
cm-1 is larger than those (∼2.4 cm-1) of the other three lines,
indicating a small but unresolvable 37Cl shift. After deconvo-
lution of the ν4 line at 454.2 cm-1 with two lines each having
a shape similar to those of ν1-ν3 and an intensity ratio of 3:1,
line positions at 455.3 ( 0.5 and 453.4 ( 0.5 cm-1 are derived.
The 37Cl isotopic shift ∼1.9 cm-1 in this mode is consistent
with the assignment of an umbrella motion to this line.
On the basis of our experimental results and the consideration
that the reaction Cl + SO2 takes place upon photolysis as well
as after annealing, we tentatively assign observed new lines (set
“a”) to absorption of ClSO2. As shown in the following section,
theoretical calculations provide further support of such an
assignment.
The species of the least energy has a pyramidal structure with
the S atom on top (designated as ClSO2); it is bound by -10.6
(B3LYP) or -15.9 kcal mol-1 (B3P86) compared with Cl +
SO2. A planar cis isomer (designated cis-ClOSO) is predicted
to lie ∼9.2 and 14.2 kcal mol-1 above ClSO2 based on B3LYP
and B3P86 methods, respectively. It is slightly lower in energy
than Cl + SO2; thus it might be formed after photolysis of the
matrix. cis-ClOSO has a relatively elongated Cl-O bond; hence
it can be taken as a Cl atom loosely attached to the O atom of
SO2. However, the B3P86 method also predicts a second
conformer with a nonplanar geometry having a dihedral angle
of 69.1° (designated ClOSO), similar to that predicted by the
MP2/6-31+G* method. The nonplanar ClOSO has a terminal
SdO bond length (1.469 Å) similar to that of SO2 and a central
S-O bond length (1.622 Å) similar to that (∼1.62 Å) of SOO.11
The Cl-O bond length of 1.761 Å is similar to that for ClOOCl
(∼1.74 Å).27,28
A fourth isomer (cis-ClSOO) is planar with an elongated S-O
bond (1.860 Å) and a short O-O bond (1.254 Å); it lies ∼50
kcal mol-1 above ClSO2. cis-ClSOO is unlikely to be produced
in our experiment as reaction from Cl + SO2 involves
rearrangement of O atoms and is endothermic by ∼50 kcal
mol-1. Similarly, ClOSO is unlikely to be produced in the
matrix.
The geometry of ClSO2 predicted with various theoretical
methods is consistent except for the length of the Cl-S bond.
Both B3LYP and QCISD methods predict a Cl-S bond ∼2.150
Å, whereas the B3P86 method yields 2.107 Å. The SO2 moiety
in ClSO2 resembles SO2 itself, with nearly identical bond length
but a greater bond angle ∼121.5°, compared with 118.7° for
SO2.
2. Vibrational WaVenumbers and Mode Assignments. Vibra-
tional wavenumbers and infrared intensities of four isomers of
ClSO2 predicted with various theoretical methods are listed in
Table 3. For pyramidal ClSO2, variations in wavenumbers for
lines above 400 cm-1 among QCISD, B3LYP, and B3P86
methods are within 8.7% with an average of 3.9 ( 2.8%.
Observed wavenumbers of the new species fit well with those
predicted for ClSO2, with maximum and averaged deviations

3616 J. Phys. Chem. A, Vol. 104, No. 16, 2000
Bahou et al.
TABLE 2: Geometry and Energies of Four Isomers of ClSO2 Predicted with Various Theoretical Methods
MP2a/
6-31+G*
QCISDb/
6-31G*
B3LYPa/
aug-cc-pVTZ
B3P86a/
aug-cc-pVTZ
species
parameters
E/hartree
rClS/Å
rSO/Å
Cl + SO2
ClSO2
-1007.252477
2.118
1.468
-1008.878761
2.150
1.456
-1009.775300
2.107
1.450
2.149
1.465
ClSO (deg)
OSO (deg)
108.38
123.40
43.80
-1007.241724
6.8
107.8
122.4
108.30
121.35
46.65
108.23
121.59
46.43
Cl-S off-plane angle
46.0
E/hartree
-1007.23965c
-1008.895590
-1009.800674
∆Ed/kcal mol-1
-10.6
-15.9
cis-ClOSO
r
r
r
ClO/Å
OS/Å
SO/Å
2.740
2.644
1.459
1.449
1.455
1.444
ClOS (deg)
OSO (deg)
108.23
117.10
180.0
-1008.881057
-1.4
103.76
116.81
180.0
-1009.777992
-1.7
1.761
dihedral angle (deg)
E/hartree
∆Ed/kcal mol-1
ClOSO
r
r
r
ClO/Å
OS/Å
SO/Å
1.728
1.711
1.465
1.622
1.469
ClOS (deg)
OSO (deg)
114.25
112.97
42.15°
-1007.208201
27.8
2.024
1.738
116.69
111.28
69.11°
-1009.758839
10.3
1.979
1.857
1.254
99.17°
116.16°
-0.27°
-1009.700115
47.2
dihedral angle (deg)
E/hartree
∆Ed/kcal mol-1
cis-ClSOO
r
r
r
ClS/Å
SO/Å
OO/Å
ClSO
SOO
2.000
1.884
1.262
99.59°
116.38°
-0.01°
-1008.798979
50.1
1.327
100.77°
112.79°
82.55°
-1007.131244
76.1
dihedral angle
E/hartree
∆Ed/kcal mol-1
a This work. b References 4 and 5. The MP2/6-311G(2d) calculations on ClSO2 were also reported. c (E + 1007)/hartree ) -0.63060 from
MP4SDTQ/6-311G(2df, 2pd), -0.63432 from PMP4SDTQ (2df, 2pd), -0.58291 from QSISD/6-311G(2df, 2pd); see refs 4 and 5. d Relative to the
energy of Cl + SO2.
-4.3% and -2.5 ( 1.2% for the B3LYP method, and 0.9%
and 0.3 ( 0.5% for the B3P86 method, respectively. Calculated
relative IR intensities also agree satisfactorily with experimental
observations, as listed in parentheses in Table 3. IR intensities
predicted for two low-frequency modes (ν5 and ν6) are at most
one-tenth that of the weakest line (454.2 cm-1) observed in this
work, consistent with our lack of observation of absorption line
34S isotopic shift is greater than the 18O isotopic shift for the
line at 454.2 cm-1. The ν5 mode involves Cl-S stretching
motion mixed with SO2 rocking, whereas the ν6 mode shows a
SO2-wagging motion with accompanying motion of the Cl atom.
3. 34S, 18O, and 37Cl Isotopic Ratios. Predicted and observed
isotopic ratios for 34S-, 18O-, and 37Cl-substituted isotopomers
of ClSO2 are listed in Table 4. Isotopic ratios predicted with
B3LYP and B3P86 methods are similar; they both agree with
experimental observations with maximum deviations of 0.10%
and 0.42%, respectively, for ν1-ν3. The largest deviations
between experimental values and those predicted with B3LYP
and B3P86 methods are 0.32% and 0.45%, respectively, for ν4.
As error in deconvolution of lines due to ClS(16O)18O is likely
for this mode, the agreement is satisfactory. A 37Cl isotopic shift
of ∼1.3 cm-1 predicted for the ν4 mode is consistent with
experimental observation of a line further broadened by ∼1.4
cm-1, as compared with those of ν1-ν3 lines. When the
deconvoluted value (455.3 cm-1) is used for ν4(35Cl32S(16O)16O),
agreements in isotopic ratios are improved, as also listed in Table
4.
Predicted isotopic ratios for 34S-, 18O-, and 37Cl-substituted
isotopomers of two conformers of ClOSO and cis-ClSOO are
listed as Supporting Information; no experimental data are
available for comparison. Predicted isotopic ratios clearly
eliminate the possibility of observed new lines being assigned
to ClOSO or cis-ClSOO.
D. Formation Mechanism, Secondary Photolysis, and
Atmospheric Implication. At 355 nm, SO2 cannot be dissoci-
ated. Photolysis of Cl2 at 355 nm produces predominantly
Cl(2P3/2), with only 1.6% of Cl(2P1/2).29 Possibility of exit from
of ClSO2 below 400 cm-1
.
Wavenumbers predicted for ClOSO (either in a cis planar or
a nonplanar form) disagree with those of lines observed in this
work. For cis-ClOSO, predicted wavenumbers of three vibra-
tional modes of the SO2 moiety are similar to those predicted
for ClSO2, but the relative intensities differ greatly. Our results
do not match those predicted for cis-ClOSO. For the nonplanar
form, the central S-O bond is weak and has a vibrational
wavenumber ∼660 cm-1. We searched carefully but failed to
locate weak features that might fit with predicted absorption
lines of ClOSO or cis-ClSOO.
Displacement vector diagrams of each vibrational mode of
ClSO2 calculated with the B3P86 method are illustrated in
Figure 4. For convenience, modes are numbered according to
vibrational wavenumbers rather than symmetry. A side-view
and a top-view diagram placed below each of the three-
dimensional diagram help to illustrate the motion of each atom.
On the basis of predicted displacement vectors, ν1-ν3 are readily
assigned to modes for asymmetric stretching, symmetric stretch-
ing, and bending motions of the SO2 moiety, in agreement with
assignments based on experimental results. The ν4 mode is an
umbrella or inversion motion with displacement of S atom being
greatest, consistent with the experimental observation that the

Production of ClSO2 in Matrices
J. Phys. Chem. A, Vol. 104, No. 16, 2000 3617
TABLE 3: Vibrational Wavenumbers/cm-1 and Infrared Intensities of Four Isomers of ClSO2 Predicted with Various
Theoretical Methods
QCISDa/
6-31G*
B3LYPb/
aug-cc-pVTZ
B3P86b/
aug-cc-pVTZ
species
ClSO2
νi
sym
mode
SO2 a-str
expt Kr
expt Ar
ν1
a′′
1299
(100)c
1081
(74)
489
(71)
420
1273.4
(100)c
1085.5
(67)
488.62
(59)
434.66
(51)
261.24
(5)
1305.9
(100)c
1108.6
(70)
497.28
(80)
456.51
(39)
288.15
(3)
1309.5
(100)c
1098.5
(78)
497.0
(60)
1309.6
(100)c
1098.2
(72)
497.7
(59)
ν2
ν3
ν4
ν5
ν6
ν1
ν2
ν3
ν4
ν5
ν6
ν1
ν2
ν3
ν4
ν5
ν6
ν1
ν2
ν3
ν4
ν5
ν6
a′
a′
a′
a′
a′′
a′
a′
a′
a′
a′
a′′
a′
a′
a′
a′
a′
a′′
a′
a′
a′
a′
a′
a′′
SO2 s-str
OSO bend
ClSO2 umbrella
Cl-S str
454.2
(45)
456.1
(45)
(69)
247
SO2 wag
214
(11)
248.49
(1)
260.10
(1)
cis-ClOSO
SO2 a-str
1317.6
(100)c
1144.9
(11.2)
516.7
(7.9)
102.4
(4.3)
45.5
1340.8
(100)c
1163.7
(12.0)
522.1
(8.0)
119.3
(5.4)
58.5
SO2 s-str
OSO bend
mixed Cl-O str
ClOS bend
out-of-plane
SdO str
(0.2)
41.2
(2.3)
(4.9)
41.9
(2.5)
ClOSO
1185.5
(100)c
663.2
(42.0)
445.5
(7.3)
378.1
(44.0)
210.5
(9.6)
S-O str
OSO bend
mixed
Cl-O str
torsion
92.2
(1.7)
cis-ClSOO
O-O str
1236.1
(100)c
589.9
(14.6)
509.3
(16.5)
282.6
(15.6)
162.2
(1.8)
1274.6
(100)c
612.2
(15.8)
528.6
(17.2)
309.4
(15.8)
170.6
(2.2)
S-O str
SOO bend
mixed
mixed Cl-S str
out-of-plane
104.9
(0.0)
113.9
(0.0)
a Reference 5. b This work. c Relative infrared intensities are listed in parentheses. IR intensities (in km mol-1) for ν1 are calculated to be 114.4
and 117.9 for ClSO2, 327.6 and 330.3 for cis-ClOSO, 155.3 and 150.9 for cis-ClSOO, respectively, with B3LYP and B3P86 methods, and 88.9 for
ClOSO with B3P86 method.
a matrix cage remains less than 10-6 for excitation wavelength
greater than 203 nm; at wavelengths between 177 and 203 nm,
a dissociation probability of approximately 2% was reported.30
Hence, ClSO2 is presumably formed during photolysis via
reaction of Cl atom with SO2 within the same matrix cage.
Annealing of the matrix sample also facilitates reaction of Cl
with SO2, resulting in increased yield of ClSO2.
atom of SO2, consistent with our observation of only ClSO2 as
a reaction product.
Photolysis of Cl2 at 308 nm also produces mainly Cl(2P3/2
)
with only 0.95% of Cl(2P1/2).29 Observation of Cl2SO2 at this
photolysis wavelength rather than 355 nm is confusing. It
indicates that either there is a small barrier for the reaction
Cl + ClSO2 f Cl2SO2
(2)
The absence of cis-ClOSO suggests that either there is a
barrier for Cl atom to attack the terminal O atom or cis-ClOSO
is unstable. Considering possible errors, the small exothermicity
predicted by theory does not guarantee the stability of cis-
ClOSO. Calculations using the MP2-full/6-31+G* method
indicate no barrier for formation of ClSO2 from Cl approaching
the central S atom of SO2 but a barrier ∼5 kcal mol-1 for
formation of cis-ClOSO from Cl approaching the terminal O
or excitation of SO2 to a long-lived electronically excited triplet
state plays a role in the formation of Cl2SO2. As annealing of
the matrix after irradiation at 355 nm produces no Cl2SO2
despite an expectation that Cl and ClSO2 might coexist
appreciably within the same matrix cage, a small barrier for
reaction 2 is needed to explain why Cl2SO2 was observed upon
photolysis at 308 nm rather than 355 nm. At smaller photolysis

3618 J. Phys. Chem. A, Vol. 104, No. 16, 2000
Bahou et al.
Figure 4. Displacement vector diagrams of vibrational modes of ClSO2 calculated with the B3P86/aug-cc-pVTZ method. The modes are ordered
by wavenumbers.
TABLE 4: Comparison of Observed Isotopic Ratiosa of
ClSO2 with Predictions from Two Theoretical Methods
Secondary photolysis of ClSO2 at 308, 355, 415, and 640 nm
diminishes substantially the absorption of lines in set “a” with
concomitant increase of SO2 lines.
vib
mode
B3LYP/
B3P86/
species
aug-cc-pVTZ aug-cc-pVTZ
expt
(1309.5)b
The small exothermicity (10-16 kcal mol-1) of reaction 1
implies that its rate coefficient is small at room temperature
because ClSO2 decomposes readily to form Cl and SO2.
However, when one considers the equilibrium constant of the
reaction, an equilibrium between Cl + SO2 and ClSO2 might
be attained at a temperature less than 273 K. Hence, it is
important to investigate both the gaseous reaction at low
temperature and reactions of the adduct ClSO2 with important
species such as O2, NO2, and NO. Such information will help
to assess the importance of reaction 1 in the atmosphere.
ν1
ν2
ν3
ν4
ν5
ν6
35Cl32S(16O)16O (1273.4)b
37Cl32S(16O)16O 1.0000
35Cl34S(16O)16O 0.9871
35Cl32S(16O)18O 0.9861
35Cl32S(18O)18O 0.9681
35Cl32S(16O)16O (1085.5)
37Cl32S(16O)16O 1.0000
35Cl34S(16O)16O 0.9941
35Cl32S(16O)18O 0.9750
35Cl32S(18O)18O 0.9547
35Cl32S(16O)16O (488.6)
37Cl32S(16O)16O 0.9993
35Cl34S(16O)16O 0.9894
35Cl32S(16O)18O 0.9825
35Cl32S(18O)18O 0.9650
35Cl32S(16O)16O (434.7)
37Cl32S(16O)16O 0.9971
35Cl34S(16O)16O 0.9807
35Cl32S(16O)18O 0.9931
35Cl32S(18O)18O 0.9859
35Cl32S(16O)16O (261.2)
37Cl32S(16O)16O 0.9849
35Cl34S(16O)16O 0.9984
35Cl32S(16O)18O 0.9918
35Cl32S(18O)18O 0.9768
35Cl32S(16O)16O (248.5)
37Cl32S(16O)16O 0.9947
35Cl34S(16O)16O 0.9957
35Cl32S(16O)18O 0.9773
35Cl32S(18O)18O 0.9619
(1305.9)b
1.0000
0.9871
0.9860
0.9681
(1108.6)
1.0000
0.9941
0.9751
0.9547
(497.3)
0.9986
0.9883
0.9848
0.9702
(456.5)
0.9972
0.9816
0.9919
0.9827
(288.2)
0.9857
0.9988
0.9894
0.9748
(260.1)
0.9947
0.9957
0.9789
0.9622
1.0000
0.9863
0.9856
0.9680
(1098.5)
1.0000
0.9935
0.9755
0.9554
(497.0)
1.0000
0.9895
0.9835
V. Conclusion
0.9660
We have produced and identified ClSO2 for the first time on
irradiating at 355 nm a low-temperature matrix containing Cl2
and SO2. Lines at 1309.5, 1098.5, 497.0, and 454.2 cm-1 (or
1309.6, 1098.2, 497.7, and 455.8 cm-1) are assigned to ClSO2
isolated in solid Kr (or Ar) based on experiments with 34S and
18O isotopes and theoretical calculations. Present information
indicates that Cl can react with SO2 at low temperature to form
ClSO2, but not ClOSO. Further investigation of photochemistry
and reactions of ClSO2 is needed to assess the importance of
the reaction Cl + SO2 in the atmosphere.
(454.2)/(455.3)c
-/0.9958
0.9811/0.9787
∼0.9963d/0.9936
0.9872/0.9848
Acknowledgment. The authors thank the National Science
Council of the Republic of China (grant no. NSC89-2119-M-
007-001) for support and the National Center for High-
performance Computing for providing the facilities to perform
part of the calculations.
a Ratios of ν(isotopomer)/ν(35Cl32S(16O)16O). b Calculated or observed
ν(35Cl32S(16O)16O) in cm-1 are listed in parentheses. c Derived from
deconvoluted lines; see text. d When the deconvoluted line at 455.3
cm-1 is used for ν(35Cl32S(16O)16O); see text.
Supporting Information Available: Predicted isotopic
ratios for 34S-, 18O-, and 37Cl-substituted isotopomers of two
conformers of ClOSO and cis-ClSOO are listed in Table IS and
IIS, respectively. This material is available free of change via
the Internet at http://pubs.acs.org.
wavelength, the Cl atom may have still greater excess energy
after relaxation via collisions with matrix hosts to overcome
the barrier. Excitation of SO2 at 308 nm might lead to formation
of the triplet state with lifetimes in the millisecond range;
dissociation of Cl2 and formation of triplet SO2 in the same
matrix cage might form Cl2SO2. The possibility that the absence
of Cl2SO2 is due to secondary photolysis at 355 nm (rather than
at 308 nm) is eliminated because the UV absorption spectrum
of Cl2SO2 shows an absorption continuum beginning ∼360 nm
and becoming increasingly intense toward smaller wavelengths.
References and Notes
(1) Strattan, L. W.; Eibling, R. E.; Kaufman, M. Atmos. EnViron. 1979,
13, 175.
(2) Eibling, R. E.; Kaufman, M. Atmos. EnViron. 1983, 17, 429.

Production of ClSO2 in Matrices
J. Phys. Chem. A, Vol. 104, No. 16, 2000 3619
(3) Demore, W. B.; Yang, Y. L.; Smith, R. H.; Leu, M. T. Icarus 1985,
63, 347.
W.; Andres, J. L.; Head-Gordon, M.; Replogle, E. S.; Pople, J. A. Gaussian
Inc.: Pittsburgh, PA, 1998.
(15) Frisch, M.-J.; Head-Gordon, M.; Pople, J. A. Chem. Phys. Lett.
1990, 166, 275, 281.
(4) Li, Z. Chem. Phys. Lett. 1997, 269, 128.
(5) Li, Z. J. Phys. Chem. A. 1997, 101, 9545.
(6) Kuo, Y.-P.; Wann, G. H.; Lee, Y.-P. J. Chem. Phys. 1993, 99, 3272.
(7) Lo, W.-J.; Lee, Y.-P. J. Chem. Phys. 1994, 101, 5494.
(8) Lo, W.-J.; Lee, Y.-P.; Tsai, J.-H. M.; Beckman, J. S. Chem. Phys.
Lett. 1995, 242, 147.
(9) Lo, W.-J.; Shen, M.-y.; Yu, C.-h.; Lee, Y.-P. J. Chem. Phys. 1996,
104, 935.
(10) Jou, S.-H.; Shen, M.-y.; Yu, C.-h.; Lee, Y.-P. J. Chem. Phys. 1996,
104, 5745.
(11) Chen, L.-S.; Lee, C.-I.; Lee, Y.-P. J. Chem. Phys. 1996, 105, 9454.
(12) Lee, C.-I.; Lee, Y.-P.; Wang, X.; Qin, Q.-Z. J. Chem. Phys. 1998,
109, 10446.
(16) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(17) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. 1988, B37, 785.
(18) Becke, A. D. Phys. ReV. 1988, A38, 3098.
(19) Perdew, J. P. Phys. ReV. 1986, B33, 8822.
(20) Dunning, T. H. J. Chem. Phys. 1989, 90, 1007. Woon, D. E.;
Dunning, T. H. Jr. J. Chem. Phys. 1993, 98, 1358.
(21) Kugel, R.; Taube, H. J. Phys. Chem. 1975, 79, 2130.
(22) Maillard, D.; Allavena, M.; Perchard, J. P. Spectrochim. Acta 1975,
31A, 1523.
(23) Suzuki, E. M.; Nibler, J. W.; Oakes, K. A.; Eggers, D. J. Mol.
Spectrosc. 1975, 58, 201.
(13) Bahou, M.; Lee, Y.-C.; Lee, Y.-P., submitted for publication in J.
Am. Chem. Soc.
(24) Cheng, B.-M.; Lee, Y.-P. J. Chem. Phys. 1989, 90, 5930.
(25) Bondybey, V. E.; Fletcher, C. J. Chem. Phys. 1976, 64, 3615.
(26) Shimanouchi, T. Tables of Molecular Vibrational Frequencies
Consolidated Volume 1; National Bureau of Standards: Washington, DC,
1972.
(27) McGrath, M. P.; Clemitshaw, K. C.; Rowland, F. S.; Hehre, W. J.
J. Phys. Chem. 1990, 94, 6126.
(28) Birk, M.; Friedl, R. R.; Cohen, E. A.; Pickett, H. M.; Sander, S. P.
J. Chem. Phys. 1989, 91, 6588.
(29) Matsumi, Y.; Tonokura, K.; Kawasaki, M. J. Chem. Phys. 1992,
97, 1065.
(14) Gaussian 98 (Revision A.1); Frisch, M. J.; Trucks, G. W.; Schlegel,
H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Zakrzewski, V. G.;
Montgomery, J. A.; Stratmann, R. E.; Burant, J. C.; Dapprich, S.; Millam,
J. M.; Daniels, A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.;
Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.;
Clifford, S.; Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.;
Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman,
J. B.; Cioslowski, J.; Ortiz, J. V.; Stefanov, B. B.; Liu, G.; Liashenko, A.;
Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith,
T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.;
Challacombe, M.; Gill, P. M. W.; Johnson, B. G.; Chen, W.; Wong, M.
(30) McCaffrey, J. G.; Kunz, H.; Schwentner, N. J. Chem. Phys. 1992,
96, 155.