Reaction dynamics of O(1D) with HCN(0 00 v3)

Article abstract of DOI:10.1063/1.470079

The quantum state resolved reaction dynamics of HCN(0 0⁰ υ³) with O(¹D) atoms were investigated by analyzing the complete product state distributions of OH(X ²Π_Ω,υ,J) and CN(X ²Σ⁺,υ,J) using laser induced fluorescence (LIF).The influence of the CH-stretching mode on the reaction dynamics and different branching ratios was inspected by exciting HCN to its first overtone band of the ν³ CH stretch in the 1.5 μm region.The oxygen atom in the ¹D state was generated in a laser photolysis of ozone at a wavelength of 266 nm.The measured rotational and vibrational distributions of the products were compared with statistical results from phase space theory (PST).Nearly statistical rotational and vibrational distributions are obtained for CN(X ²Σ⁺) in υ=0-3.The rotational and vibrational distributions of OH(X ²Π_Ω) are colder than statistically expected.Insertion of O into the CN bond with subsequent hydrogen migration seems to be a better characterization of the the reaction mechanism than an insertion of the oxygen atom into the CH bond.Direct abstraction of hydrogen to form OH is improbable to describe the molecular process.

Full text of DOI:10.1063/1.470079

Reaction dynamics of O(1 D) with HCN(0 00 v 3)
Christoph Kreher, Robert Theinl, and K.H. Gericke
Citation: The Journal of Chemical Physics 103, 8901 (1995); doi: 10.1063/1.470079
View online: http://dx.doi.org/10.1063/1.470079
View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/103/20?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
A quantum time-dependent wave-packet study of intersystem crossing effects in the O(3P0, 1, 2) + D2(v = 0, j =
0) reaction
J. Chem. Phys. 138, 134309 (2013); 10.1063/1.4795497
Theoretical Studies of O(1D) + HD (v = 0, j = 0, 1, 2, 3) → OD(H) + H(D) Reaction
Chin. J. Chem. Phys. 23, 310 (2010); 10.1088/1674-0068/23/03/310-312
Dynamics of the H+D2→HD+D reaction: Dependence of the product quantum state distributions on collision
energy from 0.98 to 1.3 eV
J. Chem. Phys. 82, 1323 (1985); 10.1063/1.448455
Dynamics of O(1 D) reactions with bifunctional substrates: HCN
J. Chem. Phys. 81, 1785 (1984); 10.1063/1.447850
The dynamics of the O(1 D 2)+CO(X 1Σ+, v=0) reaction
J. Chem. Phys. 64, 4076 (1976); 10.1063/1.432017
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
130.113.86.233 On: Wed, 10 Dec 2014 04:48:23

1
0
Reaction dynamics of O( D) with HCN(0 0 v )
3
Christoph Kreher, Robert Theinl, and K.-H. Gericke
Institut f u¨ r Physikalische und Theoretische Chemie der Johann Wolfgang Goethe-Universit a¨ t,
Marie-Curie-Strasse 11, D-60439 Frankfurt/Main, Germany
͑
Received 24 May 1994; accepted 15 August 1995͒
0
1
The quantum state resolved reaction dynamics of HCN͑0 0 v ͒ with O͑ D͒ atoms were investigated
3
2
2
ϩ
by analyzing the complete product state distributions of OH(X ⌸ ,v,J) and CN(X ⌺ ,v,J)
⍀
using laser induced fluorescence ͑LIF͒. The influence of the CH-stretching mode on the reaction
dynamics and different branching ratios was inspected by exciting HCN to its first overtone band of
1
the ␯ CH stretch in the 1.5 ␮m region. The oxygen atom in the D state was generated in a laser
3
photolysis of ozone at a wavelength of 266 nm. The measured rotational and vibrational
distributions of the products were compared with statistical results from phase space theory ͑PST͒.
2
ϩ
Nearly statistical rotational and vibrational distributions are obtained for CN͑X ⌺ ͒ in vϭ0–3. The
2
rotational and vibrational distributions of OH͑X ⌸ ͒ are colder than statistically expected. Insertion
⍀
of O into the CN bond with subsequent hydrogen migration seems to be a better characterization of
the the reaction mechanism than an insertion of the oxygen atom into the CH bond. Direct
abstraction of hydrogen to form OH is improbable to describe the molecular process. © 1995
American Institute of Physics.
I. INTRODUCTION
H͑2S ͒ϩNCO͑X ⌸͒ ⌬ H298ϭϪ197 kJ/mol,
˜ 2
→
1
/2
r
͑
4͒
Measuring nascent internal state distributions of prod-
ucts is a highly informative means of inspecting molecular
reaction dynamics. The reaction of O with HCN is of interest
because not only does it play an important role in combus-
1
1
ϩ
→NH͑a ⌬͒ϩCO͑X ⌺ ͒
⌬r
H
298ϭϪ195.3 kJ/mol,
͑5͒
͑6͒
͑7͒
1
tion systems, but also having several product reaction chan-
1
ϩ
1
ϩ
→
NH͑b ⌺ ͒ϩCO͑X ⌺ ͒
nels with products accessible by LIF or REMPI. The reaction
can be divided into reactions involving ground-state oxygen
⌬rH298ϭϪ92.0 kJ/mol,
3
atoms, O͑ P͒, and electronically excited oxygen atoms in the
2
2
ϩ
1
→OH͑X ⌸͒ϩCN͑X ⌺ ͒
D state.
⌬rH
298ϭϪ99.6 kJ/mol,
The spin allowed reactions of oxygen in the triplet state
are2
2
2
→
OH͑X ⌸͒ϩCN͑A ⌸͒ ⌬ H298ϭϩ10.9 kJ/mol.
r
˜
1
ϩ
3
HCN͑X ⌺ ͒ϩO͑ P͒
͑
8͒
There are several studies on the dynamics of the reaction
→
→
H͑2S1/2͒ϩNCO͑X ⌸͒ ⌬ H298ϭϪ25 kJ/mol,
˜ 2
r
1
͑
1͒
O͑ D͒ϩHCN, examining the kinetics as well as the product
5
state distributions. Carpenter et al. have shown that the
3
Ϫ
1
ϩ
NH͑X ⌺ ͒ϩCO͑X ⌺ ͒
⌬rH298ϭϪ146 kJ/mol,
1
main channels for deactivation of O͑ D͒ by HCN are reac-
tions ͑5͒ and ͑7͒, of which reaction ͑5͒ is much more domi-
nant with a ratio of the rate coefficients k /k Ϸ9. The pos-
͑2͒
OH͑X ⌸͒ϩCN͑X ⌺ ͒ ⌬ H298ϭϩ75 kJ/mol.
5
7
2
2
ϩ
→
r
sible spin allowed reaction channels producing NOϩCH or
CϩHNO are significantly endothermic and are not further
discussed.
͑
3͒
Previous gas phase studies3,4 have determined that reaction
1͒ is the dominant channel, followed by channel ͑2͒. Chan-
nel ͑3͒ becomes important at higher collision energies. The
reaction dynamics is characterized by an intermediate of
HC͑O͒N, arising from O͑ P͒ addition to the CN–␲ bond of
HCN, with subsequent dissociation to H and NCO. In the
case of channel ͑2͒, a further intermediate, HNCO, is pre-
dicted. The total rate constant for the reaction system
2
2
In this paper the products OH͑X ⌸͒ and CN͑X ⌺͒ of
the reaction channel ͑7͒ were monitored by the laser induced
fluorescence technique ͑LIF͒.6 Furthermore, the influence
of vibrational motion in HCN on the reaction dynamics is
investigated by ir excitation of the first overtone band of the
C–H stretching mode8
͑
,7
3
–13
HCN͑Jϭ8,v ϭ0,v ϭ0,v ϭ0͒ϩh␯͑1.53 ␮m͒
3
1
2
3
O͑ P͒/HCN at 300 K was calculated from kinetic data by
3
Ϫ18
2
ϪE/RT
Perry to be k ϭ 9.4 ϫ 10
T e
with Eϭ28.45 kJ/
→HCN͑JЈϭ9,vЈϭ0,vЈϭ0,vЈϭ2͒.
͑9͒
1
2
3
mol.
Spectroscopic studies of hydrogen cyanide acid have
been performed by many research groups because of its sig-
The possible spin allowed reaction channels for
O͑ D͒ϩHCN are
1
5
0
nificance in chemical physics and astrophysics. The ͑0 0
˜
1
ϩ
1
0
HCN͑X ⌺ ͒ϩO͑ D͒
2͒←͑0 0 0͒ overtone band of HCN has a large band inten-
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
J. Chem. Phys. 103 (20), 22 November 1995
0021-9606/95/103(20)/8901/9/$6.00
© 1995 American Institute of Physics
8901
1
30.113.86.233 On: Wed, 10 Dec 2014 04:48:23

1
8902
Kreher, Theinl, and Gericke: Reaction dynamics of O( D) with HCN
planoconvex lens. The beam diameter in the focal region was
Ϸ0.4 mm. At a pulse energy of 15 mJ a photon flux of
Ϸ9.0ϫ1019 photons/cm is achieved.
2
We excited the ͑0←0͒, ͑1←1͒, ͑2←2͒, ͑0←2͒, and
2
ϩ
2
͑
1←3͒ vibrational bands of the A ⌺ ←X ⌸ transition of
OH. The LIF-probe beam at wavelengths around 305–320
nm was generated using the frequency doubled output of a
Nd-YAG pumped dye laser ͑Quantel TDL IV͒ operated with
DCM at a pulse energy of ϳ100 ␮J. The CN product state
distribution was probed by an excimer pumped dye laser
͑Lambda Physik EMG 101 MSC and FL 2002͒. The dye
laser was operated with QUI in the region of 375 to 390 nm.
All LIF spectra were detected under saturated conditions.
The lasers were operated with a repetition rate of 10 Hz. The
time delay between the photolysis laser—which is in time
with the ir probe laser—and the LIF-probe laser was set to
1
50 ns for CN and 300 ns for OH. The LIF signal was
monitored with a photomultiplier ͑THORN-EMI 9781B͒. To
reduce scattered light from the laser beams, the LIF signal
was spectrally filtered by an interference filter of known
transmission. Furthermore the reaction chamber was
equipped with 50 cm sidearms which contained baffles.
A separate photoacoustic cell, 12 cm in length, was used
for controlling the spectral overlap between the narrow ab-
FIG. 1. Schematic drawing of the experimental setup. ͑RCϭreaction cham-
ber, PMϭphotomultiplier, PAϭphotoacoustic cell, MICϭmicrophone,
Lϭlens, SHGϭsecond harmonic generator, Mϭmirror, BSϭbeam splitter,
DMϭdicroitic mirror, PDϭphotodiode, DYEϭdye laser, MIXϭfrequency
mixing.
Ϫ1
sorption lines of HCN ͑⌬␯Ϸ0.016 cm ͒ and the ir laser
Ϫ1
9
light ͑⌬␯ Ϸ0.1 cm ͒. The cell was equipped with CaF2
l
Brewster windows at both ends. We used a commercially
available microphone to detect the photoacoustic signal. The
spectral width of the ir light was obtained by measuring the
linewidth ͑FWHM͒ of a HCN transition. The cross section ␴
4
10
sity of 8.5ϫ10 cm/mol. Using the Avogadro number we
calculate an integrated absorption cross section of
Ϫ21
for excitation of HCN via the R͑8͒ line is ␴Ϸ␴ /
␴0ϭ5.0ϫ10
cm for the chosen molecular transition.
0
Ϫ20
2
⌬
␯lϷ5.0ϫ10
cm . The high intensity of the ir laser in the
focal region, Iϭ0.9ϫ10 photons/cm , guarantees satura-
20
2 2
II. EXPERIMENT
tion of the HCN transition.
HCN gas was prepared by reaction of NaCN with
1
4,15
H SO .
Subsequent distillation over CaCl and P O re-
2
4
2 2 5
III. RESULTS
1
moved possible water residues. Oxygen atoms in the D state
were generated from ozone in a pulsed laser photolysis. By
firing the fourth harmonic of a Nd-YAG laser ͑Quanta-Ray
DCR 1A͒ into the cell, the focused output generated
O͑ D͒ϩO ͑ ⌬ ͒.
charge in O ͑99.95% purity͒.
O with HCN prior to photolytic initiation, the reactants were
mixed in the reaction chamber by separate nozzles. The
chamber was evacuated by an oil diffusion pump, reaching a
base pressure of ϳ10
A. OH product state distribution
The LIF-spectra observed via reaction ͑7͒ are shown in
2
Figs. ͑2͒ ͑OH͒ and ͑3͒ ͑CN͒. OH͑X ⌸͒ was detected by
1
1
16,17
2
ϩ
2
Ozone was prepared in a silent dis-
2
g
exciting the 0–0 and 1–1 band of the A ⌺ ←X ⌸ transi-
tion. No LIF signal was observed, when the vibrational
bands ͑2–2͒, ͑0–2͒, and ͑1–3͒ of OH were probed in the
wavelength region 318–322 and 390–394 nm. Thus
18,19
To avoid any reactions of
2
3
2
OH͑X ⌸͒ is only generated in the two lowest vibrational
Ϫ2
Pa. The partial pressures were PO3
states. The vibrational distribution of OH is P(vϭ1)/P(v
ϭ0)ϭ0.14Ϯ0.02, thus only a small quantity of OH is pro-
duced in vϭ1.
ϭ 4 Pa and PHCNϭ6 Pa. The experimental setup is shown in
Fig. 1.
0
0
Probing the ͑0 0 2͒←͑0 0 0͒ overtone band of HCN in
the infrared region was performed via difference frequency
mixing of a dye laser and a 1.064 ␮m Nd-YAG laser beam
Figure 4 represents the rotational distribution of
2
OH͑X ⌸ , vϭ0,1͒ as a Boltzmann plot. The distributions
⍀
can approximately be characterized by Boltzmann tempera-
tures. We found rotational temperatures of ͑3090Ϯ90͒ K for
⌸ ͑vϭ0͒ and ͑3270Ϯ140͒ K for ⌸ ͑vϭ0͒. For the vibra-
͑
Continuum YG 680, TDL 60, IRP͒. The dye laser operated
with DCM at an output energy of 80 mJ at 627 nm. The
beams were mixed in a lithium niobate crystal and an auto-
matic wavelength tracking system was used to maintain the
optimum conversion efficiency if the laser was tuned. The ir
output at 1.5 ␮m was separated from the residual radiation at
3/2
1/2
tionally excited state of OH the rotational distribution can be
described by the temperatures of ͑2000Ϯ170͒ K and ͑2310
Ϯ140͒ K for the ⌸3/2 and ⌸1/2 states, respectively. Although
the temperatures are slightly different for the two spin orbit
systems, no deviation from a statistical population of the two
6
27 nm and 1.064 ␮m by a CaF Pellin Broca prism and
focused into the interaction region of the cell with a 65 cm
2
2
2
spin states ⌸3 and ⌸ is observed. The observed popu-
/2 1/2
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
J. Chem. Phys., Vol. 103, No. 20, 22 November 1995
130.113.86.233 On: Wed, 10 Dec 2014 04:48:23

1
Kreher, Theinl, and Gericke: Reaction dynamics of O( D) with HCN
8903
FIG. 4. Boltzmann plot of the rotational distribution, characterizing the
ϭ3/2 and ⍀ϭ1/2 spin orbit systems of OH(X 2⌸⍀ ,vϭ0,1). The lines
represent least square fits to a Boltzmann distribution. Open squares repre-
sent vϭ0 and solid squares refer to vϭ1. Units of the Y axis are relative.
The axis is broken ͑Ϸmark͒ for a better comparison of the rotational distri-
bution.
⍀
FIG. 2. Nascent LIF spectra of the 0–0 and 1–1 band of the
2
ϩ
2
OH͑A ⌺ ͒—OH͑X ⌸͒ transition between 306 and 317 nm. The OH(X)
1
products are generated in the reaction of thermal HCN with O͑ D͒.
lation ratio is P⌸3/2 /P⌸1/2 ϭ 1.09 Ϯ 0.23 for vϭ0 and
P⌸3/2 /P⌸1/2 ϭ 0.94 Ϯ 0.24 for vϭ1. Appropriate statistical
B. CN product state distribution
The rotational state distributions of CN ͑vϭ0–3͒ from
reaction ͑7͒ are depicted in Fig. 6 as Boltzmann plots. The
distributions for vϭ0 and 1 are characterized by a low tem-
perature ͑Tvϭ0ϭ390 K, Tvϭ1ϭ400 K͒ component at low
rotational states and a high temperature ͑Tvϭ0ϭ5000 K
JϾ17, Tvϭ1ϭ4200 K JϾ10͒ component at higher rotational
levels. These observations can be explained by effects of
rotational relaxation. Because of the small energy gap be-
tween the rotational levels at low J, the rotational relaxation
should be very effective for these levels. The fraction of CN
fragments that are relaxed by collisions is ϳ6%. Assuming a
gas kinetic collision rate we calculated a value of ϳ4% of
those CN fragments, which have collided with HCN at least
once within the time delay of 150 ns. This is in agreement
with the experiment. The rotational temperatures for
weights have been considered.
The second electronic fine structure of OH molecules is
the ⌳ doubling which arises from the interaction between
nuclear rotation and orbital angular momentum. At high an-
gular momentum the two levels have opposite symmetry,
⌸͑AЈ͒ and ⌸͑AЉ͒, with respect to reflection in the plane of
rotation. Since each of the two ⌳ components of OH is
probed by different rotational branches, ⌳-doublet partition-
ing is obtained by a comparison of the LIF intensities for
P/R lines, probing the ⌸͑AЈ͒ component, with those of cor-
responding Q lines, probing the ⌸͑AЉ͒ component. No sig-
nificant preference in the population of the ⌳ doublets is
observed ͑Fig. 5͒. No accurate determination of the ⌳ dou-
blet distribution in OH͑vϭ1͒ was possible due to spectral
overlap of some relevant transitions.
2
ϩ
CN͑X ⌺ ͒ in vϭ2 and vϭ3 are ͑2990Ϯ100͒ K and ͑1570
Ϯ80͒ K.
The CN products are formed in the vibrational states
vϭ0–3 with population ratios of P(vϭ0)/P(vϭ1)/P(v
ϭ2)/P(vϭ3)ϭ0.45:0.29:0.18:0.08. We estimate the upper
FIG. 3. Nascent LIF spectra of the B 2
⌺
ϩ
–X 2
⌺
ϩ
transition of CN. The
FIG. 5. ⌳-state distribution for the OH fragment in vϭ0. Different symbols
for the two spin states are used.
1
CN(X) products are generated in the reaction of thermal HCN with O͑ D͒.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
J. Chem. Phys., Vol. 103, No. 20, 22 November 1995
130.113.86.233 On: Wed, 10 Dec 2014 04:48:23

1
8904
Kreher, Theinl, and Gericke: Reaction dynamics of O( D) with HCN
FIG. 7. Photoacoustic spectra of the X 1⌺ (vЈ ϭ 2, JЈ)←X ⌺ (v
˜
ϩ
˜
1
ϩ
3
3
ϭ 0, J) transition of HCN. Since HCN is a linear molecule, the ir spectrum
consists of a P and R branch.
FIG. 6. Boltzmann plot of the rotational distribution of
2
ϩ
1
CN(X ⌺ ,vϭ0Ϫ3) from the reaction of thermal HCN with O͑ D͒. The
lines represent the best fits to Boltzmann distributions. Units of the Y axis
are relative. The axis is broken ͑‘‘Ϸ’’͒ for a better comparison of the rota-
tional distribution.
be excited by the ir-probe beam. With the rotational constant
Ϫ1
for HCN, Bϭ1.478 cm , and in case of complete saturation
of the transition we calculated the highest amount of excited
HCN molecules ͑R line͒ according to
limit of the population of CN in vϭ4 from the noise level to
be Ͻ0.02.
C. H؉NCO channel
n͑J͒
1
2
2Jϩ3 2Jϩ1
ϭ
•
•
•eϪ͓BhcJ͑Jϩ1͔͒/͓kT͔.
͑10͒
˜
2
2
N
2Jϩ1
qrot
The possible production of NCO͑X ⌸͒ from reaction
˜
˜ 2
͑
4͒ was investigated by probing the B ⌸←X ⌸ transition
For (JЈϭ9←Jϭ8) only 4% of all HCN molecules are in the
of NCO in the region 300–320 nm. No laser induced fluo-
rescence was detected. From the different partitioning func-
tions of NCO and OH and from the S/N ratio of the spectra
we estimate an upper limit for the branching ratio,
k /k р0.5.
0
excited ͑0 0 2͒ state. We also tried to monitor OH in the
second vibrational state where no products are generated by
the ‘‘thermal reaction.’’ However, no lines in the region of
2
ϩ
2
the A ⌺ (vЈϭ0)←X ⌸(vϭ2) transition could be ob-
4
7
served.
0
D. Influence of vibrationally excited HCN(0 0 2)
The excitation of HCN into its first overtone band of v3
by an ir laser beam at 1.53 ␮m was checked by using pho-
IV. DISCUSSION
˜
1
ϩ
toacoustic spectroscopy. The X ⌺ (vЈ
ϭ
2, JЈ)
The rotational state distribution of CN observed in our
experiments shows considerable differences to the results of
Carpenter et al. Although their O photolysis was performed
3
˜
1
ϩ
←
X ⌺ (v ϭ 0, J) spectra of HCN shows P and R lines
3
according to JЈϭJϪ1 and JЈϭJϩ1. Because of the low
3
˜
1 ϩ
symmetry of HCN͑X ⌺ ͒, the lines can easily be assigned
at 248 nm—the available energy Eav for the products is
about 15.9 kJ/mol higher—the observed rotational distribu-
tions of CN in vϭ0, 2, and 3 were significantly colder. Fur-
thermore, Carpenter observed in the vibrational state vϭ1 of
CN a rotational distribution, which could be described by
two temperatures; a lower temperature of ϳ1000 K and a
temperature near 6000 K of higher rotational levels. The
break in the observed population distribution was explained
by a minor channel leading to the production of vibrationally
excited OH͑vϾ1͒. Nascent OH product state distributions
were not obtained due to strong background chemilumines-
cence. In the present experiment we observe no chemilumi-
nescence and the complete OH distribution is determined.
Due to the facts that we did not observe OH products in vϾ1
and that the rotational relaxation of CN products is very fast
it may be possible that the break of population could be an
to their transitions. Figure 7 shows the high resolution spec-
0
0
tra of the ͑0 0 2͒←͑0 0 0͒ overtone band in the wavelength
Ϫ1
region of 6488–6564 cm . In order to maximize the
amount of excited HCN, we used the JЈϭ9←Jϭ8 rota-
Ϫ1
tional transition at 6544.31 cm . The excitation was per-
formed under saturation conditions.
The experiments with vibrationally excited HCN were
performed by using a trigger-suspending device which trig-
gered the Q switch of the Nd-YAG laser used in the ir pack-
age on every second pulse. Within a single scan, we obtained
two data sets, one with HCN͑v ϭ0͒ and another with
3
HCN͑v ϭ2, JЈϭ9͒. A very precise evidence on the influence
3
of HCN͑v ϭ2͒ is given by the difference of both data sets.
3
0
No effects of additional excitation of HCN into its ͑0 0
͒ vibrational state on the reaction dynamics of O͑ D͒ϩHCN
1
2
1
was observed. Although the available energy is increased by
ϳ79 kJ/mol, no increase of the LIF signal was observed. A
reason for that could be the low portion of HCN which can
effect of relaxation. It is unlikely that the different O͑ D͒
translational energy is responsible for a bimodal rotational
distribution in only one vibrational state.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
J. Chem. Phys., Vol. 103, No. 20, 22 November 1995
130.113.86.233 On: Wed, 10 Dec 2014 04:48:23

1
Kreher, Theinl, and Gericke: Reaction dynamics of O( D) with HCN
8905
TABLE I. Vibrational product state distribution, mean rotational and vibrational energies of the reaction prod-
ucts.
Vibrational
level
P(v)
PST
P(v)
Experiment
͗Erot(v)͘
kJ mol
͗Evib(v)͘
kJ mol
Ϫ1
Ϫ1
Product
OH
0
1
sum
0
1
2
3
sum
0.69
0.25
0.94
0.47
0.28
0.15
0.07
0.97
0.88
0.12
1.0
0.45
0.29
0.18
0.08
1.0
23.17
1.27
24.44
17.36
7.20
3.01
0.76
28.33
0
4.85
4.85
0
7.08
8.74
5.79
21.61
CN
A. Product energies
and ͗Evib͘ϭ21.61 kJ/mol. The translational center-of-mass
energy of the products, Etrans , is obtained from the conser-
vation of energy according to
The available energy Eav is partitioned among the de-
grees of freedom of the reaction products OH and CN. Eav is
given by
͗
͘ ͗ ͘ ͗ ͘ ͗ ͘
Etrans ϭE Ϫ E ͑OH͒ Ϫ Evib͑OH͒ Ϫ E ͑CN͒
av rot rot
0
E ϭϪ⌬ H ϩE ͑HCN͒ϩ⌬E
.
͑11͒
Ϫ
͗
Evib͑CN͒
͘
.
͑16͒
av
r
int
coll
At room temperature E ͑HCN͒ includes only the rotational
47.77 kJ/mol or 38% of the available energy is allocated to
translational motion of the products, OH and CN. The results
int
energy of HCN; E ͑HCN͒ϭRTϭ2.5 kJ/mol. The collision
int
energy ⌬Ecoll is calculated by
and the respective partition numbers f ϭ
͗
Ei
͘
/Eav are shown
i
in Tables I and II.
1
2
1
2
2
2
2
Different reaction pathways have to be considered in or-
der to explain the experimental observations: The attacking
oxygen atom may simply abstract the hydrogen atom, the
OH and CN products may be formed by an insertion of
⌬
Ecoll
ϭ
␮O,HCN•v ϭ ␮O,HCN͑vO3ϩv
ϩv ͒.
rel
HCN
Ocm
2
͑12͒
2
O
2
can be calculated by the equation v2
v
and v
HCN
1
1
3
O͑ D͒ into the H–C bond or by an insertion of O͑ D͒ into
2
Ocm
ϭ 3RT/m and v
by the equation
the CN bond with subsequent migration of the H atom.
A direct abstraction of an H atom by O͑ D͒ has already
been observed in reactions between O͑ D͒ and alkanes
1
1
2
Ocm
1
v
ϭ
2E ͑O,O ͒␮ .
O,O
2
͑13͒
2
O
t
2
m
21
͓
C H , CH͑CH ͒ ͔. However, in that case the CN product
3
8
3 3
1
The translational center-of-mass energy, E ͑O,O ͒, of O͑ D͒
should show very low internal excitation because it acts as a
spectator during the reaction process. On the other hand, the
OH product should be highly vibrationally excited, but no
high rotational levels should be populated. Excitation of the
v3 stretch in HCN should even prefer the formation of OH in
high vibrational states. For a sensitive detection of OH in
t
2
and O from ozone photolysis at 266 nm was determined by
2
2
0
Sparks et al. to be 54.4 kJ/mol. With the reaction enthalpy
0
of ⌬ H ϭϪ99.6 kJ/mol the average available energy for the
products OH and CN is ϳ127 kJ/mol. For the reaction of ir
selected HCN͑JЈϭ9, v ϭ2͒ the available energy is ϳ206
r
3
2
ϩ
2
kJ/mol.
vϭ2 A ⌺ (vЈϭ0)←X ⌸(vϭ2) transition around 392
nm is excited using an excimer pumped dye laser with a QUI
dye solution and observing the fluorescence of the
vЈϭ0→vϭ0 band. Scattered light from the probe laser is
discriminated by an interference filter. We could not find any
The mean rotational energies for OH and CN were cal-
culated by the equation
͗
Erot
͘
ϭ
͚
͗
E
͘
ϭ
͚
P ͑J͒E ͑J͒,
͑14͒
rot v
v
rot
2
ϩ
2
v
v,J
lines according to the A ⌺ (vЈϭ0)←X ⌸(vϭ2) transi-
tion of OH. The sensitivity of the OH͑vϭ2͒ detection was
where P (J) is the rotational state distribution and E (J) is
v
rot
the rotational energy of a rotational state in a selected vibra-
tional state. The mean vibrational energy is given by
TABLE II. Partitioning of the available energy ͑E ϭ127 kJ/mol͒ into the
av
vibrational, rotational and translational degrees of freedom of OH and CN.
͗
Evib
͘
ϭ
͚
P͑v͒Evib͑v͒,
͑15͒
v
Energy ͑kJ molϪ1
͒
f ϭE /E
fi PST
i
i
av
where P(v)ϭ͚ P (J) is the population of each vibrational
Erot ͑CN͒
Evib ͑CN͒
Erot ͑OH͒
Evib ͑OH͒
Etrans ͑OH,CN͒
sum
28.33
21.61
24.44
4.85
47.77
127.0
0.22
0.17
0.19
0.04
0.38
1.0
0.19
0.17
0.25
0.12
0.27
1.0
J
v
state v, ͚ P(v)ϭ1, and Evib(v) is the energy of the vibra-
v
tional state.
The average rotational and vibrational energies of OH
from Eqs. ͑14͒ and ͑15͒ are ͗E ͘ϭ24.44 kJ/mol and
rot
͗
Evib͘ϭ4.85 kJ/mol. For CN we obtain ͗E ͘ϭ28.33 kJ/mol
rot
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
J. Chem. Phys., Vol. 103, No. 20, 22 November 1995
130.113.86.233 On: Wed, 10 Dec 2014 04:48:23

1
8906
Kreher, Theinl, and Gericke: Reaction dynamics of O( D) with HCN
checked by adding an equivalent amount of O /H O into the
3
2
1
reaction cell. The reaction of O͑ D͒ and H O produces vi-
2
brational excited OH with 9% of the OH in vϭ2. From the
S/N ratio of the spectra when using H O or HCN, we esti-
2
1
0
mate that in the reaction of O͑ D͒ with HCN͑0 0 2͒ less
than 4% of the OH products are generated in vϭ2. Although
only 4% of HCN are vibrationally excited, the signal to noise
ratio is sufficient to observe OH͑vϭ2͒ if it were generated
͑17͒
1
via HCN(v)ϩO͑ D͒ in
a
significant amount [P(v
and OHϩCN ͓reaction ͑7͔͒
ϭ2)у0.04] because the reaction of ‘‘thermal’’ HCN reac-
tants leads only to OH in vϭ0,1 state. Since all experimental
observations contradict the expectations of an abstraction
mechanism, this type of reaction can be ruled out.
.
͑18͒
1
Also a ‘‘head on’’ collision of O͑ D͒ on the H atom of
Such a collision complex has to live longer than a vibrational
period before it can decay into OH and CN products because
a migration of the hydrogen atom is necessary.
A long living complex implies randomization of energy.
Thus it is very useful to investigate the reaction using phase
HCN seems to be very unlikely because the rotational exci-
tation of CN should be low and no significant vibrational
excitation of the stiff CN bond can be expected.
1
Quite often an insertion of the O͑ D͒ atom into an H–R
space theory ͑PST͒.2
3–25
In the following we will compare
bond is observed which leads to OH products via the
H–O–R intermediate. A prototype of this reaction is the pro-
our experimental results with completely statistical distribu-
tions. The phase space theory assumes that the reaction com-
plex will dissociate with equal probability into each of the
accessible product channels, therefore enabling a statistical
partitioning of the available energy into the product states.
Accessible channels are those that satisfy the laws of conser-
vation of energy and angular momentum between reactants
and products. The PST is therefore more precise than the
prior distribution, which only satisfies the law of conserva-
tion of energy. A detailed discussion about these statistical
theories has been carried out by Ben-Saul.26
1
18,22
cess O͑ D͒ϩH →OH(v,J)ϩH,
where the collision
2
complex is H O with high excitation of the bending motion.
2
When sufficient energy flows into a H–OH bond so that this
bond will break, then the remaining energy will essentially
be released as rotation.
1
In case of an insertion of O͑ D͒ into the H–CN bond
followed by a direct decay into HO and CN one would also
expect highly rotationally excited OH products. The CN ro-
tation should also be excited, but significantly lower than
that of OH. The OH vibrational excitation is expected to be
low and that of the CN product should be even lower. Al-
though OH is formed in high rotational states with low vi-
brational excitation, the CN product state distribution dis-
The conservation of angular momentum for the reaction
1
O͑ D͒ϩHCN→OHϩCN can be written by JϭJ ϩJ
O
HCN
ϩLϭJ ϩJ ϩLЈ, where J is the total angular momentum
OH
CN
and L ͑or LЈ͒ is the orbital angular momentum of the reac-
tants ͑or products͒. PST assumes that all accessible states
have equal weight and the probability to find the products in
the rotational states JOH and JCN with given J is equal to the
number of allowed values of LЈ. We calculate this number by
counting over all allowed angular momentum vectors LЈ,
that are given by the so-called triangle condition,
1
agrees with the picture of a direct reaction where O͑ D͒ is
inserted into the H–C bond followed by a fast break of the
HO–CN bond.
In general, one assumes that a direct and fast reaction
leads to a vibrationally excited ‘‘new’’ bond ͑OH͒ while the
‘‘old’’ bond ͑CN͒ already existing in the reactant ͑HCN͒ re-
͉
JЈϪJCN͉рLЈр͑JЈϩJCN͒ with JЈϭJCNϩLЈ. Conservation
mains essentially unaffected, i.e., a low vibrational excitation
of CN is expected. Since CN is generated with a significant
of angular momentum implies the restriction of JЈ analogous
to LЈ according to ͉JϪJOH͉рJЈр͑JϩJOH͒. For given values
of J, JOH , and JCN , the probability P:ϭP͑J,JOH ,JCN͒ is
obtained by summing over LЈ and JЈ
1
amount of vibration, a direct reaction of O͑ D͒ϩHCN seems
1
to be very unlikely. Insertion of the O͑ D͒ atom into the
H–CN bond can only explain the experimental observations
when the HOCN intermediate lives long enough to ensure
sufficient internal energy transfer into the CN bond.
Another explanation of the reaction dynamics may be an
1
0
,
,
LЈрLmЈ ax
P͑J,JOH ,JCN͒ϭ͑2Jϩ1͒
.
͚ ͚
ͭ
LЈϾLЈmax
JЈ LЈ
1
insertion of the O͑ D͒ atom into the CN bond leading to an
The constraint LЈ р LЈmax is related to the dynamics of the
23
oxazirine intermediate
separating fragments described by Pechukas and is given as
follows:
.
2
/3
1
2
2
1/3
LЈ͑LЈϩ1͒ប р6␮ЈC6
ͩ
ET
ͪ
.
͑19͒
Subsequent hydrogen migration to the N and O atom will
form HNCO and HOCN, respectively. The dissociation of
these intermediates leads to the fragments NHϩCO ͓reaction
E is the translational energy of the products OH and CN, ␮
is the reduced mass and C is the coefficient of the asymp-
totic expression for the attractive part ͑r ͒ of the potential
T
6
Ϫ6
͑5͔͒,
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
J. Chem. Phys., Vol. 103, No. 20, 22 November 1995
130.113.86.233 On: Wed, 10 Dec 2014 04:48:23

1
Kreher, Theinl, and Gericke: Reaction dynamics of O( D) with HCN
8907
between the products. Using a Lennard-Jones potential as in
Ϫ78
6
Ref. 27 we estimate C6 to be 2.2ϫ10
Jm . Setting
2
LЈ(LЈϩ1)ϷLЈ we obtain the upper limit of the orbital an-
gular momentum LЈ͑E in kJ/mol͒
T
3
LmЈ axϷ34ប•
ͱ
E .
͑20͒
T
The probability of forming a reaction complex from the
initial state J , J
, L is given by
O
HCN
1
,
͑
2J ϩ1͒͑2JHCNϩ1͒͑2Lϩ1͒
O
if
otherwise it is zero.
For given values of J , J
͉
JЉϪL
͉
рJр(JЉϩL) and ͉J ϪJ
͉рJЉр͑J ϩJHCN͒,
HCN O
O
24
and L, the probability to
HCN
O
find the products in the rotational states JOH and JCN , is now
obtained by summing over all J and JЉ
FIG. 8. CN͑vϭ0–3͒ product rotational distributions from PST ͑lines͒ and
from experimental results ͑symbols͒. The agreement between both data sets
is very good. Units of the Y axis are relative. The axis is broken ͑‘‘Ϸ’’͒ for
a better comparison of the rotational distribution.
͑
2Lϩ1͚͒ ͚ P͑J,J ,JCN
͒
J J OH
Љ
P͑J ,J
,L,JOH ,JCN͒ϰ
HCN
.
O
͑
2J ϩ1͒͑2JHCNϩ1͒͑2Lϩ1͒
O
͑21͒
The next step is the summation over all possible J , J
and L
O
HCN
distribution up to high rotational energies can well be de-
scribed by phase space theory ͑PST͒. The deviation from the
experimental results are less than 10%. Figure 9 shows the
vibrational distribution of the products OH and CN together
with the state distribution from PST. The vibrational state
population of CN can also be well described by phase space
theory. Using PST one can estimate the upper limit of CN
population in vϭ4 to ϳ2% and in vϭ5 to Ͻ0.1% which is
in agreement with the experimental observations.
The statistical product state distribution of CN seems to
suggests the presence of a long living reaction complex
where complete internal energy randomization explains the
reaction mechanism. However, the OH state distribution can
not be described by phase space theory: In our experiment a
distinct lower population in the rotational and vibrational
states is observed ͑Figs. 9 and 10͒. While the OH rotational
excitation is only slightly lower than statistically expected
P͑JOH ,JCN͒ϰ
P͑J ϭ2,J
,L,JOH ,JCN
͒
HCN
͚ ͚
O
J
L
HCN
ϫf͑JHCN͒f͑L͒.
͑22͒
A summation of the rotational angular momenta of the oxy-
gen atom is not necessary, because of J ϭ2. f͑JHCN͒ is the
O
fraction of HCN in the quantum state J which is given by the
Boltzmann distribution
f͑JHCN͒ϭ͑2JHCNϩ1͒eϪ͓E͑J
͔͒/͓kT
͔
300 K
.
͑23͒
HCN
The orbital momentum L is constrained by a maximum
angular momentum according to
1
0
,
,
LрLmax
LϾLmax
f͑L͒ϭ
ͭ
.
P͑JOH ,JCN͒ in Eq. ͑22͒ is the joint reaction probability of
formation of OH in a specific state JOH , while the partner
product CN is formed in the JCN state. The product state
distribution which is observed in the present experiment of
one product is obtained by summation over all states of the
partner product where conservation of energy has to be con-
sidered. The rovibrational state distribution of CN is given
by
P͑␯CN ,JCN ,E ͒ϰ
P͑JOH ,JCN͒,
͑24͒
av ͚ ͚
␯
J
OH OH
where the vibrational quantum number v is explicitly used.
Equations ͑19͒–͑24͒ were developed into a suitable form
for computer calculations. We neglected the rotational angu-
lar momentum of oxygen by setting JϷLϩJHCN in order to
decrease the computing time. The calculations were carried
out with E ϭ127 kJ/mol, Lр90ប and JHCNр30ប. The or-
av
bital angular momentum LЈmax was calculated by Eq. ͑20͒.
The rotational state distribution of CN in vϭ0–3 to-
gether with the statistical distribution arising from PST are
shown in Fig. 8. For each vibrational level the rotational
FIG. 9. Vibrational distributions of the products OH and CN. The dashed
lines show the calculated using PST and the symbols represent the experi-
mental data. While for CN the agreement is very good the behavior of the
OH fragments can not be described with the statistical theory.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
J. Chem. Phys., Vol. 103, No. 20, 22 November 1995
130.113.86.233 On: Wed, 10 Dec 2014 04:48:23

1
8908
Kreher, Theinl, and Gericke: Reaction dynamics of O( D) with HCN
2
FIG. 10. Boltzmann plot of the rotational distribution of OH͑X ⌸͒. The
1
FIG. 11. Correlation diagram for the reaction of O͑ D͒ϩHCN. The heats of
solid line represents the corresponding statistical distribution from PST.
formation of the reaction intermediates are taken from Poppinger et al. ͑Ref.
28͒. The state energies of HNCO͑AЈ͒ϭϪ24.9 kJ/mol and HNCO͑AЉ͒ϭ475
kJ/mol are taken from Ref. 29.
͑Table II͒, the deviation of the vibrational excitation is much
stronger. A statistical population of the OH fragments allows
vibrational excitation up to vϭ3. Reaction ͑7͒ generates
2
lent, the OH distribution implies that PST cannot rigorously
be used to calculate the complete product state distribution
and anisotropic forces in the exit channel of the reaction
complex have to be considered to explain the reaction
mechanism.
OH͑X ⌸͒ only in the two lowest vibrational states. This
1
analysis suggests that the reaction O͑ D͒ϩHCN
2
2 ϩ
→
OH͑X ⌸͒ϩCN͑X ⌺ ͒ cannot completely be described
by a statistical process. Specific dynamical effects have to be
considered to characterize the state distribution of the OH
products. Strong anisotropic forces in the exit channel of the
reaction complex, for example, would transfer the available
energy nonstatistically to the rotational and vibrational states
2
of the OH͑X ⌸͒ product, as observed in the current study.
ACKNOWLEDGMENTS
In order to explain the reaction geometry we examined
the ⌳-state distribution of OH, which reflects the symmetry
conservation in the reactive process. Reactants which are
This work was supported by the Deutsche Forschungs-
gemeinschaft. We thank Professor Dr. F. J. Comes for useful
discussions and material support. C.K. thanks the Fonds der
Chemischen Industrie for fellowship support.
͑
anti͒symmetric with respect to the point group of the colli-
sion geometry will be transformed to products which are also
anti͒symmetric. By assuming the maintenance of planarity
͑
of the reaction path, the point group of the reaction complex
will be C . Whether the reaction proceeds via the ground
s
state surface AЈ or via the AЉ surface of HOCN can in prin-
ciple be answered by analyzing the ⌳-state distribution of
1
R. J. Cicerone and R. Zellner, J. Geophys. Res. 88, 10689 ͑1983͒.
J. N. Crowley and J. R. Sodeau, J. Phys. Chem. 93, 3100 ͑1989͒.
R. A. Perry and C. F. Melius, Proc. Symp. Combust. Int. 20, 639 ͑1984͒.
A. Szekely, R. K. Hansons, and C. T. Bowman, Proc. Symp. Combust. Int.
2
1
˜ 1
ϩ
OH. The symmetric reactants O͑ D͒ and HCN͑X ⌺ ͒ yield
a symmetric collision complex, HOCN͑AЈ͒, that dissociates
3
4
2
2 ϩ
to symmetric products, OH͓ ⌸͑AЈ͔͒ and CN͑X ⌺ ͒. How-
ever, reaction ͑18͒ can proceed via both surfaces AЈ and AЉ
of HOCN, which leads to OH populations observed in our
experiment ͑Fig. 11͒. We could not find any significant pref-
erence for one of the two ⌳ states ͑AЈ and AЉ͒ of OH. Other
reasons for that effect of low ⌳ selectivity could be nonadia-
batic transitions in the exit channel and the transformation of
electron orbital angular momentum into rotational angular
momentum. Also a strong energy transfer into the degrees of
freedom of the long lived reaction complex, especially into
out-of-plane vibrational modes, causes equal preference of
the ⌳ states.
20, 647 ͑1984͒.
5
B. K. Carpenter, N. Goldstein, A. Kam, and R. J. Wiesenfeld, J. Chem.
Phys. 81, 1785 ͑1984͒.
J. L. Kinsey, Ann. Rev. Phys. Chem. 28, 349 ͑1978͒.
W. Demtr o¨ der, Laser Spectroscopy ͑Springer, Berlin, 1982͒, p. 417.
H. Sasada, J. Chem. Phys. 88, 767 ͑1988͒.
K. R. German and W. S. Gornall, J. Opt. Soc. Am. 71, 1452 ͑1981͒.
A. M. Smith, S. L. Coy, and W. Klemperer, J. Mol. Spectrosc. 134, 134
6
7
8
9
0
1
͑
1989͒.
11
D. H. Rank and G. Skrinko, J. Opt. Soc. Am. 50, 421 ͑1960͒.
A. M. Smith and U. G. Jo”rgensen, J. Chem. Phys. 87, 5649 ͑1987͒.
A. E. Douglas and D. Sharma, J. Opt. Soc. Am. 21, 448 ͑1953͒.
K. H. Slotta, Ber. Dtsch. Chem. Ges. 357, 318 ͑1907͒.
L. Gattermann, Liebigs Ann. Chem. 357, 318 ͑1907͒.
J. R. Wiesenfeld, Acc. Chem. Res. 15, 110 ͑1982͒.
J. C. Brock and R. T. Watsen, Chem. Phys. Lett. 71, 371 ͑1980͒.
K. Mikulecky and K.-H. Gericke, J. Chem. Phys. 96, 7490 ͑1992͒.
P. N. Clough and B. A. Thrush, Chem. Ind. 1971 ͑1966͒.
R. K. Sparks, L. R. Carlson, K. Shobutake, M. L. Kowalczyk, and Y. T.
Lee, J. Chem. Phys. 72, 1401 ͑1980͒.
1
1
1
1
2
3
4
5
16
1
1
1
2
7
8
9
0
In summary, a direct and fast reaction mechanism for the
generation of OH and CN can be excluded. A long living
1
reaction complex is formed by insertion of O͑ D͒ into either
the H–C or the CN bond of HCN. Although the agreement of
the statistical with the experimental CN distribution is excel-
21
A. C. Luntz, J. Chem. Phys. 73, 1143 ͑1980͒.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
J. Chem. Phys., Vol. 103, No. 20, 22 November 1995
130.113.86.233 On: Wed, 10 Dec 2014 04:48:23

1
Kreher, Theinl, and Gericke: Reaction dynamics of O( D) with HCN
8909
2
2
2
2
2
3
4
5
27
K. Mikulecky and K.-H. Gericke, Chem. Phys. 175, 13 ͑1993͒.
P. Pechukas and J. C. Light, J. Chem. Phys. 42, 3281 ͑1965͒.
P. Pechukas and J. C. Light, J. Chem. Phys. 44, 794 ͑1966͒.
J. R. Beresford and G. Hancock, Faraday Discuss. Chem. Soc. 75, 211
W. Forst, Theory of Unimolekular Reactions ͑Academic, New York,
1972͒, p. 134.
D. Poppinger, L. Radom, and J. Pople, J. Am. Chem. Soc. 99, 7806
͑1977͒.
J. W. Rabalais, J. R. McDonald, and S. P. McGlynn, J. Chem. Phys. 51,
5103 ͑1969͒.
28
29
͑1983͒.
2
6
A. Ben-Saul, Chem. Phys. 22, 341 ͑1977͒.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
J. Chem. Phys., Vol. 103, No. 20, 22 November 1995
130.113.86.233 On: Wed, 10 Dec 2014 04:48:23