Direct kinetic measurements on reactions of atomic carbon, C(3P), with O2 and NO at temperatures down to 15 K

Article abstract of DOI:10.1063/1.481448

A CRESU (Cinetique de Reaction en Ecoulement Supersonique Uniforme) apparatus has been used to measure rate coefficients for the reactions of C (³P) with O₂ and NO at temperatures from 295 K down to 15 K. C (³P) atoms were, for the first time in a kinetic study, monitored directly by vacuum ultraviolet laser-induced fluorescence. The rate coefficients for both reactions increase as the temperature is lowered, matching the expressions K(C+O₂) = (4.9±0.8)×10^-11 (T/298 K)-^(0.32±0.08) cm³ molecule^-1s^-1 and k(C+NO)=(1.5±0.4)×10^-10 (T/298K)^{-(0.16±0.14)} cm³ molecule^-1s^-1.

Full text of DOI:10.1063/1.481448

Direct kinetic measurements on reactions of atomic carbon, C ( 3 P), with O 2 and NO at
temperatures down to 15 K
Delphine Chastaing, Sébastien D. Le Picard, and Ian R. Sims
Citation: The Journal of Chemical Physics 112, 8466 (2000); doi: 10.1063/1.481448
View online: http://dx.doi.org/10.1063/1.481448
View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/112/19?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Theoretical investigation of exchange and recombination reactions in O ( P 3 ) + N O ( Π 2 ) collisions
J. Chem. Phys. 126, 054304 (2007); 10.1063/1.2430715
Quantum real wave-packet dynamics of the N ( S 4 ) + N O ( X ̃ Π 2 ) → N 2 ( X ̃ Σ g + 1 ) + O ( P 3 ) reaction on
the ground and first excited triplet potential energy surfaces: Rate constants, cross sections, and product
distributions
J. Chem. Phys. 124, 174303 (2006); 10.1063/1.2186643
Ab initio ground potential energy surface ( 3 A ″ ) for the O ( 3 P)+ N 2 O reaction and kinetics study
J. Chem. Phys. 115, 2540 (2001); 10.1063/1.1381010
Temperature dependent rate constants for the reactions of gas phase lanthanides with N 2 O
J. Chem. Phys. 111, 562 (1999); 10.1063/1.479336
H+N 2 O collision dynamics: Branching ratios for T–V energy transfer and the effect of N 2 O vibrational
excitation on the reaction kinetics
J. Chem. Phys. 107, 4537 (1997); 10.1063/1.474815
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:
146.189.194.69 On: Thu, 18 Dec 2014 22:17:02

JOURNAL OF CHEMICAL PHYSICS
VOLUME 112, NUMBER 19
15 MAY 2000
3
Direct kinetic measurements on reactions of atomic carbon, C„ P…,
with O2 and NO at temperatures down to 15 K
´
Delphine Chastaing, Sebastien D. Le Picard, and Ian R. Sims
School of Chemistry, The University of Birmingham, Edgbaston, Birmingham, B15 2TT, United Kingdom
͑Received 22 February 2000; accepted 25 February 2000͒
´
´
A CRESU ͑Cinetique de Reaction en Ecoulement Supersonique Uniforme͒ apparatus has been used
to measure rate coefficients for the reactions of C (3P) with O2 and NO at temperatures from
295 K down to 15 K. C (3P) atoms were, for the first time in a kinetic study, monitored directly by
vacuum ultraviolet laser-induced fluorescence. The rate coefficients for both reactions increase
as the temperature is lowered, matching the expressions k(CϩO2)ϭ(4.9Ϯ0.8)ϫ10Ϫ11
Ϫ͑0.32Ϯ0.08͒
(T/298 K͒
cm3 moleculeϪ1 sϪ1 and k͑CϩNO͒ϭ͑1.5Ϯ0.4͒ϫ10Ϫ10 (T/298 K)Ϫ(0.16Ϯ0.14)
cm3 moleculeϪ1 sϪ1
. © 2000 American Institute of Physics. ͓S0021-9606͑00͒01719-0͔
INTRODUCTION
ing crossed molecular beams. References to these earlier
studies may be found in our previous paper.1
Ground state atomic carbon, C(2p2 3PJ) has been de-
tected in a wide range of astronomical environments, and is
particularly abundant in dense interstellar clouds. Reactions
of C(3P) are thought to be important in the synthesis of
long-chain carbon-containing radicals observed there, and so
there is considerable interest in the measurement of the rate
coefficients for C(3P) reactions down to the temperatures
prevailing in these dense clouds ͑10–50 K͒.
EXPERIMENTAL TECHNIQUE
In order to measure the rate of the prototypical atom–
diatomic radical reaction between C and NO, as well as to
confirm our earlier results, we sought to devise a scheme for
directly detecting the relative concentration of C(3P) formed
in the CRESU flow, using one-photon vacuum ultraviolet
͑VUV͒ laser-induced fluorescence ͑LIF͒. We initially at-
tempted to use a two-photon scheme first demonstrated by
Bersohn and co-workers,2 whereby a two-photon UV excita-
tion is followed by VUV fluorescence. However, this proved
impractical for kinetic measurements, as the precursor mol-
ecule, C3O2, was very effectively photolyzed by the probe
laser, presumably via multiphoton absorption, resulting in a
high and constant background signal. We therefore turned to
one-photon VUV LIF employing the (3s 3P–2p 3p) transi-
tions around ␭ϭ166 nm. One-photon VUV LIF detection of
C(3P) has been demonstrated previously,3 but never in a
kinetic study. Generation of the required 166 nm radiation is
not possible using standard frequency-tripling techniques in
rare gas mixtures, and so we chose instead to employ two-
photon resonant frequency mixing in xenon.4
The CRESU method, as applied to neutral–neutral reac-
tions, has already been described in detail by Sims et al.5
Full specifications of the Birmingham CRESU apparatus
have also been given recently by James et al.6 Here, we give
a brief description of the CRESU apparatus, emphasizing the
aspects peculiar to this study.
The heart of the CRESU apparatus is an axisymmetric
Laval nozzle, mounted on a moveable reservoir within a
vacuum chamber. All the temperatures ͑apart from 295 K͒ in
the gas flows were achieved by the isentropic expansion of
the gas mixture prepared in the reservoir through the nozzle
and into the main chamber. This expansion generates a su-
personic flow of gas in which the Mach number, the tem-
perature, the density of the gas, the mole fraction of the
co-reagent and the velocity of the gas stream are constant
along the axis of the flow. A range of nozzles was employed
We recently performed the first ever measurements on
the kinetics of C(3P) reactions below room temperature, us-
ing an indirect chemiluminescence technique to follow the
concentration of C(3P) with time in a CRESU ͑reaction ki-
netics in uniform supersonic flow͒ apparatus.1 Reaction rates
were determined by observing the chemiluminescence from
NO (B 2⌸) which is generated in the reaction between
C(3P) atoms and NO2. As C(3P) was not observed directly,
it was not possible to confirm experimentally that any ex-
cited spin–orbit population formed in the photolysis was re-
laxed. It was argued that, as the spin–orbit splittings in
C(3P) are rather small, the 3P1 and 3P2 states lying 16.4 and
43.4 cmϪ1, respectively, above the ground state P0, relax-
3
ation would be very rapid. Furthermore, it was necessary to
consider the possibility that other photolysis products of
C3O2 were responsible for the observed signal. It was con-
cluded that they were not. However, experimental confirma-
tion of these points awaited the current work employing di-
rect detection of atomic carbon. We had also hoped to
measure the rate of reaction of C(3P) with NO in our previ-
ous study, but this proved impossible owing to the reaction
of NO with NO2 present in excess as part of the chemilumi-
nescence detection scheme.
Apart from the results published in our earlier paper,
kinetic measurements on reactions of carbon atoms have
been confined to room temperature or above. Husain and
co-workers have performed kinetic measurements on the
widest variety of reactions using atomic resonance absorp-
tion to detect C(3P). Becker and co-workers are the only
group to have used LIF detection, via a two-photon transi-
tion. Some dynamical studies have also been carried out us-
0021-9606/2000/112(19)/8466/4/$17.00
8466
© 2000 American Institute of Physics
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:
146.189.194.69 On: Thu, 18 Dec 2014 22:17:02

Reactions of C(3P) with O2 and NO
8467
J. Chem. Phys., Vol. 112, No. 19, 15 May 2000
TABLE I. Rate coefficients k, for the reaction between C(3P) atoms and O2
͑in carrier gas M͒ measured over the temperature range 15–295 K.
͓O2]/1013
͓M͔ / 1016
molecule cmϪ3
molecule
No. of
points
k/10Ϫ11cm3
T/K
M
cmϪ3
moleculeϪ1 sϪ1
15
27
54
83
207
295
He
He
Ar
N2
N2
Ar
5.05
4.65
5.36
4.88
5.83
17
0–9.2
0–7.2
0–6.0
0–0.8
0–5.6
0–23
11
10
11
11
11
11
11.2Ϯ0.7a
12.2Ϯ1.2
7.74Ϯ0.84
7.27Ϯ0.79
6.52Ϯ0.33
4.09Ϯ0.16
aErrors quoted as Ϯt␴ statistical error, where t is the appropriate value of
Student’s t distribution for the 95% point.
resonance was ensured by monitoring the Xeϩ multiphoton
ionization signal in a separate cell. Another, visible dye laser
beam ͑pulse energy ϳ20 mJ͒ at a wavelength of around 560
nm ͑corresponding to frequency ␯vis) was generated in a sec-
ond identical dye laser pumped by the same Nd:YAG laser
͑Spectra Physics, GCR 170͒, but at 532 nm, and combined
and co-propagated into the Xe cell. VUV radiation resulted
at a frequency VUVϭ2␯UVϪ␯vis . Tuning the visible dye
v
laser in the range ␭visϭ560–565 nm, corresponded to
␭
VUVϭ165.44–165.88 nm. The resultant VUV beam inter-
sected the CRESU flow and the photolysis beam at the focus
of an optically fast CaF2 condenser lens pair, which imaged
any resulting fluorescence onto the photocathode of a VUV
solar blind photomultiplier tube ͑Electron Tubes, type 9403͒
after passing through a VUV interference filter ͑Acton Re-
search͒ centered at 159.8 nm ͑23.4 nm bandwidth͒.
Examples of VUV LIF spectra are shown in Fig. 1 at
room temperature and at 15 K. Relaxation of the nascent
spin–orbit distribution was seen at the lowest temperatures
studied, and was observed to take place very rapidly ͑in
ϳ1–2 ␮s͒. Kinetic measurements were only started after this
relaxation had taken place. Measurements of the actual rates
of spin–orbit relaxation will be the subject of a future pub-
lication.
FIG. 1. VUV LIF spectra of the 3s 3P–2p 3P transition in atomic carbon,
observed in the CRESU flow at 295 K ͑upper spectrum͒ and 15 K ͑lower
spectrum͒. The transitions are labeled according to the upper and lower
͓
͔
spin–orbit levels involved (J , J ).
Ј
Љ
in the present work, each providing a particular temperature
and density for the selected carrier gas. Experiments were
carried out in He, Ar, and N2.
C(3P) atoms were created by the 10 Hz pulsed laser
photolysis of carbon suboxide, C3O2, using the 193 nm ra-
diation from an excimer laser ͑Lambda-Physik, Compex
102͒, which propagated through the throat of the Laval
nozzle and along the axis of the flow. The fluence of the
beam was about 50 mJ cmϪ2 in the CRESU chamber. C3O2
was synthesized and used as before.1
For each co-reagent concentration, a C(3P) kinetic de-
cay trace was recorded by systematically varying the time
delay between the photolysis and probe laser pulses. Each
decay consisted of 100 points, and was averaged for 6 laser
shots per point. They were fitted to yield values of k1st for
each gas mixture on which experiments were carried out at a
particular temperature. The flows of co-reagent ͑O2, or NO;
Air Products͒, and the carrier gas ͑He, Ar, or N2; Air Prod-
ucts͒ were taken directly from the cylinders and regulated by
means of mass flow controllers ͑MKS͒. Knowledge of the
total gas density from Pitot-tube measurements and of the
individual gas flows enabled the concentrations of the co-
reagent, and hence the corresponding second-order rate co-
efficient, to be calculated.
Detection of C(3P) atoms was achieved using VUV LIF
at wavelengths around 166 nm, corresponding to the
3s 3P–2p 3P electronic transition. 166 nm radiation was
͓
͔
generated by two-photon resonant four-wave frequency
mixing4 in xenon. A gas cell, with a path length of around 20
cm, was equipped with a quartz input lens ͑10 cm focal
length͒ and a CaF2 recollimation lens ͑10 cm focal length͒ at
the output, leading directly into the CRESU chamber. Two
tuneable dye laser beams were introduced into the cell,
which contained 35 mbar of Xe. The first, at a wavelength
255.94 nm was the frequency-doubled output of a 355 nm
Nd:YAG-pumped dye laser ͑LAS, LDL 205͒. The output
energy was ϳ2 mJ. The UV frequency (␯UV) was set to the
Xe 5p–6p ͓212, 2͔ two-photon resonance. Excitation of this
RESULTS AND DISCUSSION
Experiments to determine the kinetic behavior of carbon
atom reactions with a reagent R ͑RϭO2 or NO͒ were per-
formed both at room temperature in a subsonic flow, and at
low temperatures ranging from 15 K to 204 K in He, Ar, or
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:
146.189.194.69 On: Thu, 18 Dec 2014 22:17:02

8468
J. Chem. Phys., Vol. 112, No. 19, 15 May 2000
Chastaing, Le Picard, and Sims
TABLE II. Rate coefficients k, for the reaction between C(3P) atoms and
NO ͑in carrier gas M͒ measured over the temperature range 15–295 K.
͓M͔ / 1016
͓NO͔/1013
molecule
cmϪ3
No. of
points
k/10Ϫ10 cm3
T/K
M
molecule cmϪ3
moleculeϪ1 sϪ1
15
27
54
83
207
295
He
He
Ar
N2
N2
Ar
5.05
4.65
5.36
4.88
5.83
18
0–4.6
0–1.8
0–1.2
0–0.8
0–4.9
0–9.1
11
14
11
9
10
11
2.28Ϯ0.17a
2.01Ϯ0.11
2.50Ϯ0.08
1.73Ϯ0.26
1.69Ϯ0.15
1.2Ϯ0.09
aErrors quoted as Ϯt␴ statistical error, where t is the appropriate value of
Student’s t distribution for the 95% point.
FIG. 3. Rate coefficients k for the reaction of C(3P) atoms with NO plotted
on a log–log scale against temperature. The filled symbols show the results
of this work: circles ͑᭹͒ denoting Ar carrier gas; squares ͑᭿͒ N2; and
triangles ͑᭡͒ He. The solid line shows the result of a nonlinear least-squares
fit to this new data giving kϭ(1.5Ϯ0.4)ϫ10Ϫ10 (T/298 K)Ϫ(0.16Ϯ0.14)
cm3 moleculeϪ1 sϪ1. The room temperature result of Braun ͑Ref. 10͒ is
shown as an open circle ͑᭺͒, that of Husain and Young ͑Ref. 12͒ as an open
diamond ͑छ͒, that of Becker et al. ͑Ref. 13͒ as an open square ͑ᮀ͒, and that
of Bergeat et al. ͑Ref. 15͒ as an open inverted triangle ͑᭞͒.
N2 buffer gases, using a number of Laval nozzles. Results for
RϭO2 and NO are summarized in Tables I and II, and the
temperature dependences of the rate coefficients for these
reactions of C(3P) are displayed in Figs. 2 and 3. Table III
compares previously measured values with the rate coeffi-
cients that have been determined in the present work at room
temperature.
The new results for the reaction
(T/298 K)Ϫ0.34 cm3 moleculeϪ1 sϪ1). Possible explanations
for the observed negative temperature dependence were dis-
cussed in our previous paper.1
3
C P͒ϩO →COϩO
͑1͒
͑
2
display a remarkably good agreement with our previous
measurements using the indirect chemiluminescence detec-
tion technique. This lends confidence to both sets of mea-
surements. An unweighted nonlinear least-squares fit to the
combined data set yields the following temperature depen-
dence: k1ϭ(4.9Ϯ0.8)ϫ10Ϫ11 (T/298 K)Ϫ(0.32Ϯ0.08) cm3
moleculeϪ1 sϪ1, with errors quoted as Ϯ2␴, where ␴ is the
standard error. This compared favorably with the tempera-
ture dependence determined previously (kϭ4.7ϫ10Ϫ11
Our results for the reaction
3
C P͒ϩNO→CNϩO
͑2a͒
͑2b͒
͑
→COϩN
are in only fair agreement with the rather scattered results of
previous studies at room temperature, as can be seen in Table
III. An unweighted nonlinear least-squares fit over the tem-
perature range 15–295 K yields the following temperature
dependence for the overall rate coefficient: k2ϭ(1.5Ϯ0.4)
ϫ10Ϫ10 (T/298 K)Ϫ(0.16Ϯ0.14) cm3 moleculeϪ1 sϪ1, with er-
rors quoted as before. The rapid rate at room temperature,
combined with the mild negative temperature dependence, is
characteristic of radical–radical or atom–radical reactions
occurring on attractive potential surfaces.1 In such cases,
TABLE III. Comparison of rate coefficients for the reactions of C(3Pj) with
O2 and NO at room temperature from the present work with those from
previous measurements.
kϳ300 K(C(3P)ϩR→products)
/10Ϫ10 cm3 moleculeϪ1 sϪ1
RϭO2
Methoda
Reference
RϭNO
0.33
1.1
FP–VUVAb
FP–VUVAb
FP–VUVAb
PLP–LIF
FT–CL
FT–RF
10
11
12
13
14
15
1
0.33Ϯ0.15
0.26Ϯ0.03
0.47Ϯ0.03
0.16Ϯ0.02
0.25Ϯ0.04
0.437Ϯ0.015
0.409Á0.016
0.73Ϯ0.22
0.48Ϯ0.08
0.16Ϯ0.02
0.27Ϯ0.02
0.54Ϯ0.08
—
FIG. 2. Rate coefficients k for the reaction of C(3P) atoms with O2 plotted
on a log–log scale against temperature. The filled symbols show the results
of this work; circles ͑᭹͒ denoting Ar carrier gas; squares ͑᭿͒ N2; and
triangles ͑᭡͒ He. The open circles show the results of our previous mea-
surements over the same temperature range with chemiluminescence detec-
tion ͑Ref. 1͒ and the solid line shows the result of a nonlinear least-squares
fit to the combined data set, giving kϭ(4.9Ϯ0.8)ϫ10Ϫ11 (T/298
K)Ϫ(0.32Ϯ0.08) cm3 moleculeϪ1 sϪ1. The room temperature results of Husain
and Young ͑Ref. 12͒ and Becker et al. ͑Ref. 13͒ are shown as an open
diamond ͑छ͒ and an open square ͑ᮀ͒, respectively.
PLP–CL
PLP–VUV LIF
1.20Á0.09
This work
aFP, flash photolysis; VUVAb, vacuum ultraviolet absorption; PLP, pulsed
laser photolysis; LIF, laser-induced fluorescence; FT, flow tube; CL, chemi-
luminescence; RF, resonance fluorescence; VUV LIF, vacuum ultraviolet
laser-induced fluorescence.
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:
146.189.194.69 On: Thu, 18 Dec 2014 22:17:02

Reactions of C(3P) with O2 and NO
8469
J. Chem. Phys., Vol. 112, No. 19, 15 May 2000
some collisions may occur over surfaces that do not lead to
thermodynamically accessible products. The observed mild
negative temperature dependence may then result from an
increase in the proportion of collisions which occur on sur-
faces that do lead to products. The P1 and P2 spin–orbit
states of C(3P) lie at energies equivalent to 23 K and 62 K,
respectively, above the ground state 3P0. The populations in
the individual spin–orbit states change markedly through the
range of temperature covered in the present experiments.
Therefore, if reaction is more probable from lower spin–
orbit states this alone could lead to a negative temperature
dependence, independent of the dynamics of the reactive
collisions.7
In contrast to the reaction between C and O2, recent ab
initio potential surface and rate coefficient calculations have
been performed on the CϩNO system. Beghin et al.8 em-
ployed the ACCSA method on a long range electrostatic/
dispersion potential, giving a rate coefficient at room tem-
perature of 3.43ϫ10Ϫ11 cm3 moleculeϪ1 sϪ1, with a marked
negative temperature dependence. Andersson et al.,9 in a
very recent study, present quasiclassical trajectory calcula-
tions on a new ab initio potential surface. They obtain a rate
coefficient of ϳ8ϫ10Ϫ11 cm3 moleculeϪ1 sϪ1 at room tem-
perature, and almost no temperature dependence. However,
they took a constant correction factor of 1/9 to allow for the
nonreactive collisions as explained above. At room tempera-
ture, and certainly below, this factor is temperature depen-
dent, and fuller treatment could bring their results into agree-
ment with our experimental data. Further comparisons with
theoretical calculations, as well as with new crossed-beam
measurements at very low collision energies being per-
formed in Bordeaux by Costes and co-workers, will be the
subject of a future paper.
͑under its TMR program͒ for a studentship ͑DC͒ and post-
doctoral fellowship ͑SDLP, as part of the Research Network
on Astrophysical Chemistry͒. The EPSRC Laser Loan Pool
at the Rutherford–Appleton Laboratory provided an
Nd:YAG-pumped dye laser system, for which we express
our thanks. We are also very grateful to Professor Ian W. M.
Smith for valuable discussions, and to Stuart Arkless, Tony
Rothin and Steve West for their skilled technical assistance.
3
3
1 D. Chastaing, P. L. James, I. R. Sims, and I. W. M. Smith, Phys. Chem.
Chem. Phys. 1, 2247 ͑1999͒.
2 P. Das, G. Ondrey, N. Vanveen, and R. Bersohn, J. Chem. Phys. 79, 724
͑1983͒.
3 M. Rowekamp, A. Goehlich, and H. F. Dobele, Appl. Phys. A: Solids
Surf. 54, 61 ͑1992͒.
4 R. Hilbig and R. Wallenstein, IEEE J. Quantum Electron. 19, 194 ͑1983͒.
5 I. R. Sims, J. L. Queffelec, A. Defrance, C. Rebrion-Rowe, D. Travers, P.
Bocherel, B. R. Rowe, and I. W. M. Smith, J. Chem. Phys. 100, 4229
͑1994͒.
6 P. L. James, I. R. Sims, I. W. M. Smith, M. H. Alexander, and M. B.
Yang, J. Chem. Phys. 109, 3882 ͑1998͒.
7 R. A. Brownsword, S. D. Gatenby, L. B. Herbert, I. W. M. Smith, D. W.
A. Stewart, and A. C. Symonds, J. Chem. Soc., Faraday Trans. 92, 723
͑1996͒.
8 A. Beghin, T. Stoecklin, and J. C. Rayez, Chem. Phys. 195, 259 ͑1995͒.
9 S. Andersson, N. Markovic, and G. Nyman, Phys. Chem. Chem. Phys. ͑in
press͒.
10 W. Braun, A. M. Bass, D. D. Davis, and J. D. Simmons, Proc. R. Soc.
London, Ser. A 312, 417 ͑1969͒.
11 D. Husain and L. J. Kirsch, Trans. Faraday Soc. 67, 2025 ͑1971͒.
12 D. Husain and A. N. Young, J. Chem. Soc., Faraday Trans. 2 71, 525
͑1975͒.
13 K. H. Becker, K. J. Brockmann, and P. Wiesen, J. Chem. Soc., Faraday
Trans. 2 84, 455 ͑1988͒.
14 G. Dorthe, P. Caubet,. T. Vias, B. Barrere, and J. Marchais, J. Phys.
Chem. 95, 5109 ͑1991͒.
ACKNOWLEDGMENTS
15 A. Bergeat, T. Calvo, G. Dorthe, and J. C. Loison, Chem. Phys. Lett. 308,
7 ͑1999͒.
We are grateful to EPSRC for a substantial research
grant to construct the CRESU apparatus. We thank the EU
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:
146.189.194.69 On: Thu, 18 Dec 2014 22:17:02