Dissociation and recombination in the photochemical decay of carbonyl cyanide CO(CN)2 in cryogenic matrixes

Article abstract of DOI:10.1021/jp065731l

The photochemistry of CO(CN)₂ in cryogenic matrixes has been investigated employing pulsed laser excitation at 193 nm. During irradiation, the parent molecule, the intermediate, and the final photoproducts were monitored by IR spectroscopy. Four new species were identified including the isocyano isomer of the parent NCC(O)-NC, cyanogen NCCN, isocyanogen CNCN, and CO according to spectroscopic features and ab initio calculations. After prolonged irradiation, the only remaining species were CO and the two isomers NCCN and CNCN. A reaction scheme is proposed which is in agreement with the first dissociation step being a branching of the decay path into the radical channel to CN + OCCN and the molecular channel to CO + (CN)₂. The caged radicals of the former reaction either recombine to the parent molecule and its isomer which are both photolyzed again or they react directly to the stable and final products.

Full text of DOI:10.1021/jp065731l

764
J. Phys. Chem. A 2007, 111, 764-769
Dissociation and Recombination in the Photochemical Decay of Carbonyl Cyanide CO(CN)2
in Cryogenic Matrixes
H. U. Suter, R. Pfister, A. Furlan, and J. Robert Huber*
Physikalisch-Chemisches Institut der UniVersita¨t Zu¨rich, Winterthurerstrasse 190, CH-8057 Zu¨rich, Switzerland
ReceiVed: September 4, 2006; In Final Form: NoVember 4, 2006
The photochemistry of CO(CN)2 in cryogenic matrixes has been investigated employing pulsed laser excitation
at 193 nm. During irradiation, the parent molecule, the intermediate, and the final photoproducts were monitored
by IR spectroscopy. Four new species were identified including the isocyano isomer of the parent NCC(O)-
NC, cyanogen NCCN, isocyanogen CNCN, and CO according to spectroscopic features and ab initio
calculations. After prolonged irradiation, the only remaining species were CO and the two isomers NCCN
and CNCN. A reaction scheme is proposed which is in agreement with the first dissociation step being a
branching of the decay path into the radical channel to CN + OCCN and the molecular channel to CO +
(CN)2. The caged radicals of the former reaction either recombine to the parent molecule and its isomer
which are both photolyzed again or they react directly to the stable and final products.
Introduction
Carbonyl cyanide CO(CN)2 (Figure 1, insert) has attracted
attention as a precursor for the OCCN radical and as a simple
model system such as formaldehyde for the photoinduced decay
into radical and molecular products.1,2 Following photolysis of
CO(CN)2 in a molecular beam at 193 nm, the excited molecule
was found to dissociate primarily (≈94%) via an R-cleavage to
the radicals OCCN + CN. Under these conditions, only about
18% of the nascent OCCN fragments are unstable and decom-
pose to CO + CN. The generation of a large fraction of stable
OCCN radicals was thus directly demonstrated in this experi-
ment which was carried out using a photofragment translational
energy spectrometer combined with quadrupole mass detec-
tion.1,2 Furthermore, the nascent CN fragments emerging from
the R-cleavage and those from the spontaneous secondary decay
of sufficiently hot COCN radicals were also probed by laser-
Figure 1. Infrared absorption spectrum of carbonyl cyanide CO(CN)2
in an Ar matrix (dilution ratio 1:5000) at 12 K. Assignments are given
in Table 1.
induced fluorescence (LIF) and were discerned by their different
translational and internal energy distributions.3 A minor decay
channel (≈6%) involves a concerted bond-breaking and bond-
Experimental Methods
making process in a single kinetic step leading to the molecular
products CO + (CN)2.2 In addition to the experimental work,
the energy and the structure of the parent molecule, products,
and transition-state species were calculated by ab initio methods
to elucidate the complex photochemical decay.4 In this context,
it is interesting to note that photolysis of acetyl cyanide CH3-
COCN at 193 nm, which was expected to be an efficient source
of OCCN radicals, showed no traces of this radical.5,6
The details of the apparatus and the method for matrix
preparation have been presented elsewhere.7 Briefly, the cham-
ber was backfilled with a mixture of rare gas (Ar or Xe) and
carbonyl cyanide in mole ratios of 1:1000-5000. Films of 55-
600 µm were deposited on a cold CsI plate at a rate of ≈10-
20 µm/min for 5-30 min. The deposition rate was estimated
by comparison with previous IR interferometric measurements.8
IR spectra in the range 250-4000 cm-1 were recorded using
a Fourier transform infrared (FTIR) spectrometer (BIO-RAD
FTS 175C) equipped with a closed-cycle low-temperature
cryostat (Displex CS-202, Air Products). The cryostat is
equipped with two CsI windows and a quartz window. The
sample was irradiated through the quartz window with UV light
at 193 nm using an ArF excimer laser (Lambda Physik EMG
101) at a fluence of ≈4 mJ/cm2 (pulse duration ≈ 20 ns). After
exposures to usually 100-5000 laser shots at 10 Hz, the
substrate was rotated by 90° for the acquisition of an IR
spectrum. Single spectra were obtained by averaging 100 times
Motivated by these results in the gas phase, we investigated
CO(CN)2 in cryogenic matrixes before, during, and after UV
laser irradiation and probed the emerging photoproducts by
infrared spectroscopy. The matrix environment is expected to
trap and isolate intermediate species not accessible by spectro-
scopic means in the previous gas-phase experiments. The
analysis of the spectra was carried out utilizing the results of
ab initio calculations.
* Author to whom correspondence should be addressed. Tel.: 0041-44-
6354461. Fax: 0041-44-6356838. E-mail: jrhuber@pci.unizh.ch.
10.1021/jp065731l CCC: $37.00 © 2007 American Chemical Society
Published on Web 01/12/2007

Photochemistry of CO(CN)2
TABLE 1: IR Frequencies and Relative Intensities of CO(CN)2 in Ar and Xe Matrixes
Ar matrix Xe matrix gas phase
J. Phys. Chem. A, Vol. 111, No. 5, 2007 765
cm-1
int.a
cm-1
int.a
cm-1
int.
assignmentb
2240
2234
1708
1127
1096
1007
715
0.25
1.00
0.75
0.90
1.55
0.10
0.08
0.05
ν1(a1) CN stretch
2229
1701
1125
1091
1002
711
1.00
0.80
0.83
0.94
0.08
0.10
0.04
2230
1711
1124
1102
1004
712
vs
vs
vs
vs
m
s
ν2(b2) CN stretch
ν3(a1) CO stretch
ν4(b2) CC stretch with
ν5(b2) in Fermi resonance
ν combination
ν6(a1/b1) CC stretch
ν7(a1) CCC scissoring
567
560
553
m
a Intensity integrated over peak area and normalized to ν2. Only peaks with a relative intensity of >0.04 are listed. b Assignments according to
the present calculation and refs 16 and 21.
at a resolution of ∆ν ) 0.25 cm-1, whereas for the kinetic data
50 measurements were averaged at ∆ν ) 2 cm-1. The sample
temperature was held at 12 K during the photolysis and the
spectrometric measurements. Annealing of the fresh and the fully
photolyzed samples was carried out at 35 K for the Ar matrix
and 60 K for the Xe matrix.
The highly toxic carbonyl cyanide CO(CN)2 was synthesized
by reaction of tetracyanoethylene oxide with n-butyl sulfide at
50 °C.9 The yellowish product with a boiling point of 66 °C
was purified by four distillations at reduced pressure. The gas-
phase UV absorption spectrum of CO(CN)2 in the range of 190-
410 nm was recorded previously and is displayed in Figure 1
of ref 2.
Calculation Methods
The vibrational frequencies of the parent molecule and
expected intermediate and product species were calculated with
the harmonic force fields using the Gaussian03 program
package10 with the Becke3LYP density functional11 together
with the 6-31g** AO basis set. The choice of this type of
calculation was given by our experience with previous calcula-
tions which showed a rather monotonic behavior for the scaling
using the scaling factor of 0.963 for the calculated frequencies.12
The normal coordinate analysis was performed by transforming
the output of the Gaussian03 to the GAMESS Program.13 The
vibrational frequencies were further analyzed in terms of the
total energy decomposition (TED) of Pulay and Tarok14 (not
shown).
Figure 2. Photolysis of CO(CN)2 in an Ar matrix (dilution ratio 1:5000)
at 12 K. The difference IR absorption spectrum was recorded after 2000
laser pulses at 193 nm. The negative peaks indicate the disappearance
of the reactant CO(CN)2, and the positive peaks indicate the appearance
of intermediate and final product species.
cross section ꢀ is ≈2300 dm3 mol-1 cm-1 (σ ) 8.7 × 10-18
cm2), about 1000 times greater than at the maximum of the nπ*
transition.
Infrared Spectra of CO(CN)2 and Its Photoproducts. The
IR spectrum of CO(CN)2 in an Ar matrix shown in Figure 1
agrees well with the gas-phase spectrum previously reported16,20
and later studied in detail by calculations.21 Only a small matrix
shift of a few wavenumbers and a splitting of some bands
because of different site structures are the expected differences.
Table 1 lists the results of the two studies. To aid the spectral
analysis, IR spectra were also recorded in the “softer” Xe matrix
where the vibrational frequencies are slightly red-shifted and
the site splittings are different compared to the Ar matrix. A
distinct difference is only observed in the intensity ratio of the
two strong bands ν3 and ν4 which is in clear favor of ν4 in Ar
and almost one in Xe. These two bands belong to a Fermi
resonance involving a strong mixture between the fundamental,
antisymmetric C-C stretch (b2) at ≈1100 cm-1 and the
combination between the C-C-C bend (a1, scissor) at around
560 cm-1 and the CdO wag (b2) according to a proposal by
Miller et al.16 Since the frequencies of these modes are different
in the two matrixes, the degree of mixing changes as manifested
by the band intensities. In the following discussion, we refer
only to the frequencies in the Ar matrix.
Results
UV Spectrum of CO(CN)2. At its low-energy region, the
gas-phase absorption spectrum2,15 starts with a weak band
system showing a maximum at about 330 nm. This band,
belonging to the S0 f S1 (nπ*) transition localized on the CO
group, is structured and exhibits at least 10 evenly spaced
vibronic bands between 270 and 400 nm. This progression with
an initial spacing of 1230 cm-1 was assigned to the CdO
stretching vibration in the S1 state. Each of these bands is further
split into eight narrow peaks separated by ≈128 cm-1. Accord-
ing to our calculation, they are due to the excited-state in-plane
δ(N-C-C-C-N) bending mode (a1) whose corresponding
ground state frequency16 is 127 cm-1
.
Beside the nπ* transition, a strong absorption starts at around
220 nm and extends to below 190 nm (the lower limit of our
previous measurement).2 Although a definitive assignment of
Figure 2 shows a difference spectrum in the IR region
between about 1000 and 2300 cm-1 where only the sections
with clearly discernible changes are displayed. They were
recorded after 2000 laser shots at 193 nm. The disappearance
of the CO(CN)2 vibrational bands upon irradiation, indicated
as negative amplitudes in the spectrum Figure 2, is accompanied
1
this transition is lacking, it probably involves a Rydberg (n,-
3s) state by analogy with other carbonyl compounds17.18 or as
later suggested a πCNπ*CO state.19 The present photolysis
experiments were carried out at 193 nm, where the absorption

766 J. Phys. Chem. A, Vol. 111, No. 5, 2007
Suter et al.
TABLE 2: IR Frequencies and Relative Intensities of Intermediate Product Species
Ar matrix
Xe matrix
calculated
cm-1
int.
cm-1
int.
cm-1
2088
int.
1.0
assignmentaNC-(CO)-NC
2086
2080
1752
1114
716
1.0
0.2
0.7
0.9
0.1
2079
2074
1747
1111
1.0
0.1
0.5
0.4
ν2 NC-isostretch
ν2 site structure
ν3 CO stretch
ν4 C-Co-N stretch
ν6
1749
1105
722
0.5
0.7
<0.1
a According to the calculation.
TABLE 3: IR Frequencies of Final Products
TABLE 4: Calculated IR Spectrum of CO(CN)2 and
Comparison with Measurements in the Ar Matrix and in the
Gas Phase (g)
Ar matrix
cm-1
Xe matrixa
cm-1
Ar matrixb
cm-1
gas phase
cm-1
assignment
speciesa
cm-1
int.b
cm-1 (Ar)
cm-1 (g)c
2296
2156
2149
2056
2291
2152
2141
2052
2294
2153
2150
2054
2302a
2158c
2143d
2060a
CNCN
NCCN
CO
ν1(a1)
ν2(b2)
ν3(a1)
ν4(b2)
ν5(b2)
ν6(b1)
ν7(a1)
ν8(a1)
ν9(b2)
ν10(a2)
ν11(b2)
ν12(b1)
ν13(a1)
2267
2264
1713
1093
18
89
144
239
2240
2234
1708
1127d
1096d
715
2230
2230
1711
1124d
1102d
712
712
553
550
307
CNCN
a Reference 30. b Reference 31. c Reference 28. d Reference 27.
728
706
567
536
316
242
215
124
7
4
4
1
0
13
27
5
708
567
by the appearance of at least nine new bands, the strongest and
most distinct of which are listed in Tables 2 and 3. After
prolonged excitation, only four distinct bands located in the
region 2050-2300 cm-1 survive (Figure 3a) suggesting that
the bands at 2086, 2080, 1752, and 1114 cm-1 as well as a
weak band at 716 cm-1 (not shown) belong to intermediate
species. The frequencies of the final product bands are listed in
Table 3.
245
208
128
a C2V symmetry. b Intensity given in km/mol. c Reference 16. d Fermi
resonance.
The “kinetics” of the IR bands in terms of the growth and
disappearance of the integrated peak intensities as a function
of the number of laser shots shown in Figure 4 enabled us to
discriminate among three product groups. While the CO(CN)2
bands at 2234 and 1127 cm-1 decrease steeply to 10% of their
initial intensity after about 200 shots (Figure 4a, b), the intensity
of the first group of product bands at 2086 and 1114 cm-1, as
well as that at 1752 cm-1 (not shown), is zero before irradiation,
reaches a maximum after 100-200 laser shots, and decreases
thereafter to half of the maximum value within 1000 shots
(Figure 4c, d). This behavior points to the creation and decay
of an intermediate. Because interruption of the irradiation and
subsequent annealing of the matrix does not change the
intensities or frequencies of the three IR bands, we concluded
that this intermediate species is stable in the cryogenic rare-gas
environment and in the dark. It must possess a relatively high
absorption cross section at 193 nm which promotes secondary
photolysis as manifested by the consecutive reaction behavior.
The second group of product peaks centered at 2296 and 2056
(Figure 4e, f) builds up slower than the first group. The band
intensities reach the maximum after 800 shots followed by a
slow decrease. Finally, the third group with peaks at 2156 and
2149 cm-1 (Figure 4 g, h) approaches the maximum intensity
after about 2000 laser shots which afterward remains constant.
In separate experiments, we explored the initial growth of
product bands after a restricted number of laser shots (about
30) for which the UV absorption of the parent molecule is not
significantly changed. Immediately after the start of the laser,
the growth rate rises without any delay and in monotonic fashion
for the bands examined at 2086 and 1752 cm-1 and those at
merely relative intensities; they are scaled to about the same
height for each of the curves.
Calculation of the IR Spectra of Feasible Intermediate
Species. Using ab initio methods, we calculated the IR spectra
of the parent molecule given in Table 4 and potential intermedi-
ates created after photolysis of CO(CN)2. According to experi-
mental and theoretical results, we explored as intermediates the
radicals OCCN and OCNC, both in their stable trans form, and
the isomer of the parent molecule NC-C(O)-NC. The latter
includes one cyanide group and one isocyanide group. Table 5
lists the predicted vibrational frequencies and their intensities
for the two radicals, while Table 6 shows the corresponding
spectral features for carbonyl(cyanide)(isocyanide). The stability
and structure of the OCCN radical have been previously
explored by ab initio calculation.22
Discussion
General Aspects. The experimental results of the present
study are summarized in Figures 1-4 and Tables 1-3. For the
spectral analysis, we used ab initio calculations to predict the
IR frequencies of pertinent vibrational bands of the parent
molecule, potential intermediates, and product species as listed
in Tables 4-6. Together with the results of the previous gas-
phase investigations of the molecule in a cold beam,1-3 we
propose the following reaction scheme for the 193-nm photolysis
of isolated CO(CN)2 in an argon matrix cage.
NC-C(O)-CN + hν f NCCO + CN
f CO + (CN)2
(1)
(2)
2149 and 2056 cm-1
.
Finally, Figure 4 provides only a qualitative picture of the
evolution of the reactant and product concentrations. A quantita-
tive analysis of the data was not feasible because of the lack of
pertinent parameters such as the change of the UV absorption
during extended irradiation and the extinction coefficients of
the transient species. The intensities shown in Figure 4a-h are
NCCO + CN f NC-C(O)-NC
f NC-C(O)-CN
(3a)
(3b)
f CO + CN + CN f CO + (CN)2 (4)
f CO + (CN)2 (5)

Photochemistry of CO(CN)2
J. Phys. Chem. A, Vol. 111, No. 5, 2007 767
Figure 3. Infrared absorption spectra in Ar matrixes at 12 K of (a)
the final products after extensive irradiation of CO(CN)2, (b) NCCN
(dilution ratio 1:1000), and (c) NCCN and the isomer CNCN after
irradiation of the former at 193 nm.
According to the previous gas-phase work,2 eqs 1 and 2 are
the two initial dissociation steps with the radical decay (1)
dominating the molecular decay (2). While in the gas phase,
the two fragments of 1 escape each other, they are confined to
a cage in the cold matrix where they are subject to competing
decays. They can recombine (3a) and (3b), or the sufficiently
hot NCCO radicals undergo secondary dissociation to CO and
CN (4). In a second step, the latter react in the cage with the
CN radical from the primary decay to form (CN)2 (4). Finally,
with eq 5 we consider the radical recombination which proceeds
directly to the final products. Because of the continuous
excitation, the products of steps 3a and 3b, which are the initial
molecule and one of its isomers, are also photolyzed and the
reaction sequence 1-5 starts over again. Equations 2, 4, and 5
are here the “exit channels” leading to the final products CO
and (CN)2 with their possible isomers.
Figure 4. Photolysis of CO(CN)2 in an Ar matrix (dilution ratio 1:5000)
at 12 K. The integrated IR peak intensities as a function of the number
of laser shots at 193 nm show the growth and disappearance of selected
bands of reactant, intermediate, and products. The intensities represent
relative intensities and are arbitrarily scaled to about the same height
for each of the curves.
TABLE 5: Calculated IR Spectra of Potential Intermediates
OC-CN
OC-NC
speciesa
cm-1
int.
speciesa
cm-1
int.
ν1(a′)
ν2(a′)
ν3(a′)
ν4(a′)
ν5(a′′)
ν6(a′)
2148
1896
794
544
263
147
185
5
2
12
7
ν1(a′)
ν2(a′)
ν3(a′)
ν4(a′)
ν5(a′)
ν6(a′′)
1982
1856
877
575
209
216
231
28
6
3
2
Intermediate Species. In light of this reaction scheme, the
results of our experiments and calculations are now discussed.
The difference spectrum Figure 2 reveals eight new IR bands
of which the ones at 2086, 2080, 1752, and 1114 cm-1 show
an intermediate behavior (Figure 4). Since OCCN and CN
appear in the gas-phase photolysis,2,3 we expected these bands
to be associated with these species. The vibration of the CN
radical in the gas phase is found at 2042 cm-1 23 with a small
0-1 transition moment, a factor 2.3 smaller than that of CO24
(see below). Even when taking into account a few wavenumbers
of matrix shift, no signal could be discerned in the region around
210
182
a ν1 represents the CtN stretch, ν2 the CdO stretch, and ν3 the C-C
or C-N stretch.
calculation that could be assigned to the OCCN radical. Thus,
we have to conclude that the two radicals CN and OCCN
produced by the R-cleavage could not be isolated in appreciable
concentration in the Ar matrix under our experimental
conditions.
Since cyanide molecules can isomerize to isocyanide mol-
ecules as has been demonstrated in argon matrixes for XCN
compounds (X ) H, F, Cl, Br, I),25,26 we considered carbonyl-
(cyanide)(isocyanide), NCC(O)NC, a candidate for the inter-
mediate. The calculated IR spectrum of this isomer (Table 6)
reveals three strong vibrational bands at 2088 cm-1 (CN-stretch),
1749 cm-1 (CO-stretch), and 1105 cm-1 (C-Co-N stretch) and
a weak one with still appreciable intensity at 722 cm-1. A
2042 cm-1
.
An experimental IR spectrum of OCCN is lacking. Our
calculated spectra of the two possible structures of this radical,
both in their stable trans forms (Table 5), are each dominated
by two strong bands located at 2148 and 1896 cm-1 and at 1982
and 1856 cm-1, respectively. We found no transient IR bands
nor final product bands within the uncertainty range of the

768 J. Phys. Chem. A, Vol. 111, No. 5, 2007
Suter et al.
TABLE 6: Calculated IR Spectrum of the Potential
irradiation indicating that only a small fraction (<1%) of NCCN
isomerizes to CNCN despite the two bands of the latter
exhibiting a substantial absorption intensity. Even the most
intense IR transitions of NCCN are known to be much weaker
then those of CNCN. From gas-phase spectra, Stroh et al.31 have
determined an intensity ratio for ν2 (CNCN) to the ν3(NCCN)
Intermediate NC-(CO)-NC
speciesa
cm-1
int.b
cm-1 (Ar)
ν1(a′)
ν2(a′)
ν3(a′)
ν4(a′)
ν5(a′)
ν6(a′′)
ν7(a′)
ν8(a′)
ν9(a′′)
ν10(a′)
ν11(a′′)
ν12(a′′)
2273
2088
1749
1105
732
722
572
528
280
217
168
39
483
213
337
12
18
1
7
12
8
2086 (1.2)c
1752 (0.7)c
1114 (0.9)c
absorption (which in the matrix appear at 2056 and 2156 cm-1
,
respectively) of 100:1. Adopting this result for the matrix
spectrum of irradiated CO(CN)2 in Figure 3a, we find a ratio
I(ν2):I(ν3) ≈ 2.5 which implies that the major final photolysis
products are NCCN and CO and that only about 2.5% is CNCN
relative to NCCN.
716 (0.1)c
7
2
Mechanism. After determination of the intermediate NCC-
(O)NC and final products CO + NCCN and CNCN, we proceed
to the mechanism of the photochemistry which is proposed to
involve the reaction steps 1-5 introduced above. The primary
dissociation routes (1) and (2) in the gas phase are expected to
occur also in the matrix environment, and our experimental
findings are consistent with this assumption. Decay route 1
creates the radicals OCCN and CN in a matrix cage. There they
recombine to the parent molecule (3b) or to its isomer (3a).
However, this process is in competition with the secondary
dissociation of OCCN which emerges from the R-cleavage (1)
with high vibrational excitation. In the gas phase, about 18%
of these nascent radicals were found to possess sufficient internal
energy to overcome the dissociation barrier35 to CN + CO (4).
In the matrix, these two radicals share a cage with the CN radical
from the primary dissociation (1) and CN + CN react to (CN)2.
However, it is likely that eq 4 also proceeds in only one step,
that is, OCCN + CN f CO + NCCN or CO + CNCN (5),
without the prerequisite for a high vibrational energy content
of OCCN and without dissociation prior to product formation.
Both reactions (5) are strongly exothermic with ∆H ) ≈ -430
kJ/mol and ≈ -295 kJ/mol,2,4 respectively, and are expected
to proceed at most only over a small barrier as usually found
for radical + radical reactions.
According to the scheme above, eqs 4 and 5 and the molecular
dissociation (2) are the paths leading to the observed final
photoproducts. On the basis of the gas-phase results, the
efficiency of eq 4 is expected to be about 3 times that of eq 2.
While eq 2 is probably not much influenced by the matrix, the
yield of the spontaneous decay of hot OCCN (4) is probably
lower in the matrix than in the gas phase because of vibrational
energy transfer from the hot OCCN to the environment. The
18% in the gas phase thus represents merely an upper limit in
the matrix. Should eq 5 be an efficient decay route for the
OCCN radicals, the overall yield of eqs 4 and 5 to the final
products could be substantial.
116
a Cs symmetry. b Intensity given in km/mol. c Relative intensities
from Table 2.
comparison of these bands with the intermediate vibrational
bands of the difference spectrum Figure 2 shows excellent
agreement with respect to frequency and intensity (Table 2).
Therefore, we assigned the intermediate spectrum to the
isocyanide compound. This is a stable molecule, predicted by
ab initio calculation to be about 75 kJ/mol higher in energy
than carbonyl cyanide.4 The remaining vibrational band at 2080
cm-1 is probably caused by a matrix site structure slightly
different from that of the strong 2086 cm-1 band. The
contribution of the corresponding site structure in the softer Xe
matrix is much smaller (Table 2).
Having assigned the intermediate species, we consider its
growth and decay during irradiation shown in Figure 4c and
4d. The initial transient accumulation of NCC(O)NC between
0 and 300 laser shots indicates the production rate of the
intermediate to exceed initially its photodissociation rate. After
prolonged irradiation, the repeated recombination reactions (3)
die out by competition with the secondary dissociation or
recombination discussed below and the intermediate disappears.
Finally, we also calculated the IR spectrum of the di-isocyano
isomer CNC(O)NC. However, no features of Figure 2 coincide
with these calculated IR transitions.
Final Photolysis Products. The spectral features of the final
products displayed in Figure 3a and listed in Table 3 are
dominated by a vibrational transition at 2149 cm-1. Since this
is very close to the CO transition observed in the gas phase27
at 2143 cm-1, we concluded CO to be one of the final products.
In the absence of an IR signal of trapped CN expected at 2042
cm-1, the other three bands at 2056, 2156, and 2296 cm-1 were
supposed to arise from (CN)2 which shows a growth curve
(Figure 4) similar to CO. The molecule (CN)2 is known to exist
in two stable, well-investigated isomers:28-34 cyanogen NCCN
and isocyanogen CNCN, the latter being about 135 kJ/mol2
(Hartree-Fock value ≈ 115 kJ/mol)33 less stable than the
former. To assign the final product bands in Figure 3a, we
recorded the spectrum of a pure cyanogen sample in an Ar
matrix over the pertinent region of 2000-2300 cm-1. This
spectrum is displayed in Figure 3b and exhibits a single band
at 2156 cm-1 which is in exact agreement with the smallest
peak in Figure 3a. If this sample is subsequently subjected to
irradiation at 193 nm, the molecule NCCN isomerizes partly to
CNCN which becomes clearly evident in the spectrum of Figure
3c where two new bands appear at 2056 and 2296 cm-1. The
same IR bands were previously reported for CNCN in an Ar
matrix31 and, most important for the present study, they are
present in Figure 3a which is the spectrum of the final
photoproducts of CO(CN)2. The IR peak due to NCCN at 2156
cm-1 (Figure 3b and 3c) is only slightly decreased upon
The continuous excitation conditions in our experiment causes
the products of eq 3a and 3b to be also photolyzed, and the
reaction sequence is repeated. Since the C-N bond in NCC-
(O)NC is weaker than the C-C bond in the parent molecule,
R-cleavage is usually expected to favor the former. Therefore,
the carbonyl(cyanide)(isocyanide) is assumed to preferentially
dissociate along this weaker bond and the formation of the
possible di-isocyano compound CNC(O)NC in eq 3a is unlikely
which is consistent with our experiments and IR spectrum
calculations. Besides CO, the final products are the two isomers
NCCN and CNCN generated in a ratio of approximately 40:1.
Their formation can occur by the sequential mechanism via
OCCN + CN (4, 5) or by the concerted eq 2. The radical
recombination is expected to yield to a large extent the
thermodynamically favored product pair CO + NCCN but a
small fraction may be created as CO + CNCN. It is also

Photochemistry of CO(CN)2
J. Phys. Chem. A, Vol. 111, No. 5, 2007 769
(3) Li, Q.; Carter, T.; Huber, J. R. Chem. Phys. Lett. 2000, 323, 105.
(4) Suter, H. U.; Huber, J. R.; Ha, T.-K. Chem. Phys. Lett. 1999, 311,
474.
(5) North, S. W.; Marr, A. J.; Furlan, A.; Hall, G. E. J. Phys. Chem.
A 1997, 101, 9224.
(6) Furlan, A.; Scheld, H. A.; Huber, J. R. J. Phys. Chem. A 2000,
104, 1920.
(7) Mu¨ller, R. P.; Hollenstein, H.; Huber, J. R. J. Mol. Spectrosc. 1983,
100, 95.
conceivable that CNCN is formed by the molecular decay (2)
when the isocyanide intermediate NCC(O)NC is photolyzed.
A preference for the first or the second mechanism cannot be
given. However, we believe to be able to exclude CNCN
production via isomerization of NCCN in the photochemistry
of CO(CN)2 because the absorption cross section of NCCN at
193 nm (σ ) 9.6 × 10-20 cm2) is 2 orders of magnitude smaller
than that of the parent molecule. The appearance rate of the IR
signals at 2296 and 2056 cm-1 (Figure 4) is already substantial
when the absorption of CO(CN)2 is still high and thus the
shielding of NCCN from irradiation is essentially complete.
Finally, a direct photolysis of OCCN by the 193-nm radiation
is considered unlikely and thus has not been included in the
mechanism. Fast recombination reactions prevent the OCCN
radical to reach an appreciable concentration relative to the
parent molecule. The strong absorption of the latter at 193 nm
shields OCCN (except maybe at the end of the prolonged
photolysis) and protects it efficiently from photodissociation.
(8) Mu¨ller, R. P.; Russegger, P.; Huber, J. R. Chem. Phys. 1982, 70,
281.
(9) Linn, W. J.; Webster, O. W.; Benson, R. E. J. Am. Chem. Soc.
1965, 87, 3651.
(10) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Montgomery Jr, J. A.; Vreven, T.; Kudin, K. N.;
Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.;
Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.;
Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.;
Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li,
X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Adamo, C.; Jaramillo, J.;
Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.;
Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.;
Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels,
A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, D. D.;
Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.;
Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz,
P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.;
Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson,
B.; Chen, W; Wong, M. W.; Gonzalez, C.; and Pople, J. A. Gaussian 03,
Revision C.02; Gaussian, Inc.: Wallingford, CT, 2004.
(11) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(12) El-Azhary, A.-A.; Suter, H. U. J. Phys. Chem. A 1996, 100, 15056.
(13) Schmidt, M. W.; Baldridge, K. K.; Boatz, J. A.; Elbert, S. T.;
Gordon, M. S.; Jensen, J. H.; Koseki, S.; Matsunaga, N.; Nguyen, K. A.;
Su, S.; Windus, T. L.; Dupuis, M.; Montgomery, J. A. J. Comput. Chem.
1993, 14, 1347.
Conclusion
In this study, our previous work on the photodissociation of
CO(CN)2 in a supersonic jet was extended to cryogenic matrixes.
The parent molecule and its photoproduct in solid matrices were
recorded upon irradiation by IR spectroscopy. In particular, we
monitored growth and disappearance of the observed IR band
intensities as a function of the number of laser pulses at 193
nm. The four new species NC(CO)NC, NCCN, CNCN, and CO
were identified, whereas the major gas-phase fragments OCCN
and CN turned out to be short-lived in the present matrix
environment and hence could not be isolated. On the basis of
ab initio calculations, the new bands appearing at 1114, 1752,
and 2086 cm-1 were assigned to NC(CO)NC, the isomer of
the parent molecule. Upon further irradiation at 193 nm, the
parent molecule as well as its isomer disappeared completely,
while the stable products CO, cyanogen NCCN, and isocyano-
gen CNCN were formed. The two isomers showed a final ratio
of about 40:1 while CO:(CN)2 ) 1:1. A reaction scheme of
this photochemistry was proposed which is in agreement with
the first dissociation step being a branching of the decay path
into the radical channel (1) and the molecular channel (2). The
caging of the nascent radicals CN and OCCN promotes
recombination reactions among these radicals (3, 5) and among
those from the decay of OCCN (4) which eventually yields the
three final products.
(14) Pulay, P.; Tarok, F. Acta Chim. Acad. Sci. Hung. 1966, 47, 273.
(15) Prochorow, J.; Tramer, A.; Wierzchowski, K. L. J. Mol. Spectrosc.
1966, 19, 45.
(16) Miller, F. A.; Harney, B. M.; Tyrell, J. Spectrochim. Acta 1971,
27A, 1003.
(17) Robin, M.B. Higher Excited States of Polyatomic Molecules;
Academic: New York, 1975; Vol. 2.
(18) Gaines, G. A.; Donaldson, D. J.; Strickler, S. J.; Vaida, V. J. Phys.
Chem. 1988, 92, 2762.
(19) Owrutsky, J. C.; Baronavski, A. P. J. Chem. Phys. 1999, 110, 11206.
(20) Bates, J. B.; Smith, W. H. Spectrochim. Acta 1970, 26A, 455.
(21) Tyrell, J. Mol. Struct. 1991, 231, 87.
(22) Francisco, J. S.; Liu, R. J. Phys. Chem. 1997, 107, 3840.
(23) Phillips, J. G. Astrophys. J. 1973, 180, 617.
(24) Treffers, R. R. Astrophys. J. 1975, 196, 883.
(25) Milligan, D. E.; Jacox, M. E. J. Chem. Phys. 1967, 47, 278.
(26) Fraenkel, R.; Haas, Y. Chem. Phys. Lett. 1993, 214, 235.
(27) Huber, K.P.; Herzberg, G. Molecular Spectra and Molecular
Structure IV. Constants of DiatomicMolecules; Van Nostrand: New York,
1979.
(28) Jones, L. H. J. Mol. Spectrosc. 1973, 45, 55.
(29) North, S. W.; Hall, G. E. J. Chem. Phys. 1997, 106, 60.
(30) Stroh, F.; Winnewisser, M. Chem. Phys. Lett. 1989, 155, 21.
(31) Stroh, F.; Winnewisser, B. P.; Winnewisser, M.; Reisenauer, H.
P.; Maier, G.; Goede, S. J.; Bickelhaupt, F. Chem. Phys. Lett. 1989, 160,
105.
Acknowledgment. This work was supported by the Swiss
National Science Foundation. We thank Dr. Marco Nonella for
program support and many helpful discussions.
(32) Nguyen, M. T. Chem. Phys. Lett. 1989, 157, 430.
(33) Sunil, K. K.; Yates, J. H.; Jordan, K. D. Chem. Phys. Lett. 1990,
171, 185.
References and Notes
(1) Furlan, A.; Scheld, H. A.; Huber, J. R. Chem. Phys. Lett. 1998,
(34) Hochlaf, M. J. Mol. Spectrosc. 2001, 207, 269.
(35) Horwitz, R. J.; Francisco, J. S.; Guest, J. A. J. Phys. Chem. 1997,
101, 1231.
282, 1.
(2) Scheld, H. A.; Furlan, A.; Huber, J. R. J. Chem. Phys. 1999, 111,
923.