Nonequilibrium concentrations of the vibrationally excited OH radical in a methane flame measured by cavity ring-down spectroscopy

Article abstract of DOI:10.1016/S0009-2614(97)01443-7

Strong deviation from the Maxwell-Boltzmann equilibrium for the populations of the v″=0 and v″=1 vibrational levels of OH(X ²Π_i) radicals was observed in the preflame zone of the methane/oxygen/nitrogen flame. The OH concentrations for the v″=0 and v″=1 levels were measured using cavity ring-down spectroscopy in the 30 Torr flat flame. In that flame the population ratio for v″=1/v″=0 is ~20 times the equilibrium value at locations ~6 mm below the luminescent zone. Above the flame front this ratio was found to be close to the equilibrium value.

Full text of DOI:10.1016/S0009-2614(97)01443-7

6
March 1998
Chemical Physics Letters 284 Ž1998. 407–411
Nonequilibrium concentrations of the vibrationally excited OH
radical in a methane flame measured by cavity ring-down
spectroscopy
a,b
a
a,)
V.A. Lozovsky , I. Derzy , S. Cheskis
a
School of Chemistry, Sackler Faculty of Exact Sciences, Tel AÕiÕ UniÕersity, Ramat AÕiÕ, Tel AÕiÕ 69978, Israel
b
SemenoÕ Institute of Chemical Physics, Russian Academy of Science, 4 Kosygin str., Moscow 117977, Russian Federation
Received 28 November 1997
Abstract
Y
Y
Strong deviation from the Maxwell–Boltzmann equilibrium for the populations of the Õ s0 and Õ s1 vibrational
2
levels of OHŽX P . radicals was observed in the preflame zone of the methaneroxygenrnitrogen flame. The OH
i
Y
Y
concentrations for the Õ s0 and Õ s1 levels were measured using cavity ring-down spectroscopy in the 30 Torr flat
Y
Y
flame. In that flame the population ratio for Õ s1rÕ s0 is ;20 times the equilibrium value at locations ;6 mm below
the luminescent zone. Above the flame front this ratio was found to be close to the equilibrium value. q 1998 Elsevier
Science B.V.
1
. Introduction
higher than that of many other molecules. This radi-
cal plays the central role in combustion w1,2x. Meijer
et al. w3x using the new method of cavity ring-down
spectroscopy ŽCRDS. observed that the Maxwell–
Boltzmann equilibrium for the vibrational levels Õs
0, 1 and 2 is fulfilled in a Bunsen flame at least as a
good approximation. They used a methanerair flame
at atmospheric pressure and measured the tempera-
ture to be ;1900 K. The results obtained w3x are
In principle, all chemical reactions disturb the
Maxwell–Boltzmann equilibrium. However, fast re-
laxation processes in many cases return the system to
the equilibrium. Therefore, most of the widespread
chemical combustion mechanisms are based on the
assumption of energy equilibrium for most of the
participated species. Nevertheless, especially in case
of the vibrational distribution, some deviations from
the equilibrium are possible. The rate of vibrational
relaxation is usually slower if the quantum of vibra-
tional energy is larger. The OH radical has a quan-
tum of vibrational energy ;3600 cmy1, which is
related probably to after-flame gases, since the flame
front is thin at atmospheric pressure. On the other
side one can expect the strongest deviation from
equilibrium in the flame front and preflame zone,
where the reaction rate is the highest and the relax-
ation processes have not enough time to occur. In the
2
present Letter we measure OHŽX P . concentration
i
)
Y
Y
Corresponding author. E-mail: cheskis@ccsgtau.ac.il
in the Õ s0 and Õ s1 vibrational states in a
0
009-2614r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved.
PII S0009-2614Ž97.01443-7

4
08
V.A. LozoÕsky et al.rChemical Physics Letters 284 (1998) 407–411
low-pressure methaneroxygenrnitrogen flame with
high spatial resolution using CRDS. We found a
strong deviation from equilibrium in the preflame
zone.
measured rate coefficient k and baseline k sŽTq
0
L.rt :
0
1
1
s
Ž
l
.
nlskyk s y t
.
2
Ž .
0
t
0
The schematic of the experimental apparatus is
shown in Fig. 1. The flat flame burner has a porous
2
. Experimental
stainless-steel head ŽMcKenna Products. with a 6 cm
Recently, we used CRDS for the quantitative
Y
diameter flame zone surrounded by a nitrogen shroud
flow. This burner is placed inside a low-pressure
apparatus and operated with a premixed CH rO rN
determination of the OH ŽÕ s0. concentration in a
low-pressure methane flame w4x. CRDS is a highly
4
2
2
sensitive absorption method based on the measure-
ment of the intensity decay rate of a laser pulse
injected into an optical cavity built with highly re-
flective mirrors. CRDS may be considered as a
multipass absorption technique, with an equivalent
absorption path length given by the number of pas-
sages inside the cavity before the injected light inten-
sity decays by 1re. This path length may be of the
order of tens of km in the visible range, where the
best mirrors are available w5x. In contrast to standard
gas flow at a pressure of 30"0.3 Torr. Two long
arms were connected to the low-pressure chamber by
stainless-steel bellows. These arms were closed by
two concave mirrors Žthe radii of curvatures were 0.5
m. forming the ring-down cavity. Both mirrors ŽVirgo
Optics Corp.. had a specified reflectivity R)99.9%.
The observed ring-down time in the empty cavity
was ;0.8 ms, which corresponds to a reflection
Rf99.75%. Slow flows of nitrogen were added
through the thin tubes near the mirrors in order to
decrease contamination of the mirrors by combustion
products. The length of the ring-down cavity was
multipass absorption methods, in CRDS the probe
volume is confined within the fundamental cavity
mode. This provides a spatial resolution of a fraction
of a millimeter, sufficient for measurements in low-
pressure flat flames. Since the first demonstration by
O’Keefe and Deacon w6x, this method has been used
in many different applications Žsee, e.g., the review
of Scherer et al. w7x and refs. therein.. CRDS was
used to detect several important radicals: OH w3,4,8x,
CH in UV w9x and IR w10x, HCO w11,12x, C H w13x,
0
.75 m. The laser beam from the second harmonic of
a dye laser ŽContinuum Corp. Nd-60. was injected
into the cavity through one of the mirrors. The laser
y1
linewidth is narrow, 0.08 cm
as measured by a
Burleigh WA-4500 wavemeter. Two lenses and a
pinhole Ž50 mm diameter. were used to match the
laser beam to the TEM00 modes of the cavity. This
mode matching allows us to obtain the best spatial
resolution because the TEM00 mode has a minimal
diameter in the center of the cavity. In order to
measure the real diameter of the beam inside the
cavity we placed different diaphragms into the cavity
and measured the changes in the ring-down time
dependence on the size of the diaphragms. Using
such a method we estimated the beam diameter to be
;300 mm ŽFWHM..
3
6
5
including measurements in flames w3,4,10,11x.
The absorption signal in CRDS is obtained by
measuring the decay rate coefficient k or inverse
decay time as
1
Ž
TqLqs
Ž
l
.
nl. ,
ks t
s
1
Ž .
t0
where t sLrc is the time of light propagation in
0
the cavity, TqL are the cavity losses due to the
mirror transmission ŽT., absorption and scattering
ŽL.; s Žl. is the absorption cross-section at wave-
We observed the ring-down signal through the
second mirror using a photomultiplier ŽHamamatsu
R-166 or 1P28.. The signal was digitized by a
length l of the sample placed inside the cavity
which has concentration n and length l Žequal in our
case to the burner diameter D.; c is the velocity of
light and L is the cavity length. The absorbance per
pass s Žl.nl can be obtained using Eq. Ž1. from the
LeCroy 9310 oscilloscope and was transferred to the
PC computer after each laser pulse for processing.
The ring-down decay rate was determined for each
curve and obtained values were averaged through
40–60 consecutive laser pulses. Special software

V.A. LozoÕsky et al.rChemical Physics Letters 284 (1998) 407–411
409
Fig. 1. Experimental apparatus for CRDS in a flat flame. SHGssecond harmonic generator ŽKDP crystal.; PMsphotomultiplier;
Msring-down cavity mirrors; Lslenses; PHs50 mm pinhole.
was written to provide real time processing and
averaging of the signal.
burner. Fig. 2 shows fragments of OH CRDS spectra
obtained at different locations above the burner.
These fragments contain both lines of the vibra-
tionally excited radicals Ž1–1 transition. and the
ground state of this radical Ž0–0 transition.. The
3
. Results and discussion
deviation from equilibrium in the spectrum at 5 mm
above the burner can be seen at a glance. The
temperature in this location was measured to be
1050 K. At that temperature the Boltzmann factor for
the vibrational level Õ s1 must be 7=10 . Thus
the ratio in the intensity of the Õs0 and Õs1 lines
must be so high that the Õs1 lines should be
practically invisible in Fig. 2.
We measured by CRDS, the concentration of the
Y
Y
OH radical in the Õ s0 and Õ s1 levels and
compared the ratio obtained with that expected as-
suming a Boltzmann equilibrium. CRDS is useful for
such measurements because the absorption spec-
troscopy measurements are not affected by quench-
ing and predissociation. In order to calculate the
concentration based on the measured absorption, we
used rotational transitions probabilities listed by
Chidsey and Crosley w14x.
Y
y3
Table 1
Flame parameters
In most of our experiments we used a stoichio-
metric methaneroxygenrnitrogen flame Žsee Table
Reactant gas
Flow rate
Mol%
Žsccm.
1
.. This flame has a luminescence zone at ;10 cm.
CH4
450
900
2990
10.37
20.74
68.89
A temperature profile for this flame was measured
using the CRDS of the OH radical in the 306.7–307.0
nm wavelength range w4x. The maximum of the
O2
N2
temperature profile is located at 12 mm above the
Total pressure: 30.0"0.2 Torr.

4
10
V.A. LozoÕsky et al.rChemical Physics Letters 284 (1998) 407–411
Fig. 3. Boltzmann plot of the level populations nj measured by
CRDS at 5 mm above the burner in the stoichiometric methane
flame Žsee Table 1.. The populations n were calculated using Eq.
j
Ž2. and appropriate line strengths w14x.
In order to make a quantitative estimation of
deviation from equilibrium we calculated the
nonequilibrium factor u. We define its value as
Fig. 2. Fragment of the OH CRDS spectrum measured at two
wns1xrwns0x
Y
locations in the flame. The lines originated from the Õ s1 level
us
,
3
Ž .
Ž
wns1xrwns0x
.
equil
are shown in the lower spectrum Ž5 mm above the burner. by
asterisks.
where wns0x and wns1x are the concentration of
OH in the ground and first excited vibrational states,
respectively. We measured the ratio wÕs0xrwÕs1x
using the relative line intensities listed by Chidsey
More accurate evidence of the nonequilibrium can
be seen in Fig. 3. This figure represents Boltzmann
Y
plots of the OH lines from both the Õ s0 and
Y
and Crosley w14x. Fig. 4 shows the dependence of the
Õ s1 levels. It can be seen that the slope of plots
Y
Y
from the approximated Õ s0 and Õ s1 points are
Y
roughly equal, whereas the plot for Õ s1 is shifted
up. It shows that rotational temperature measured
Y
Y
using the Õ s1 and Õ s0 rotational line is approx-
Y
imately the same, but the intensities of all Õ s1
lines are higher. It demonstrates in its turn that the
effect could not be caused by artifacts, for example
by the non-ideal flatness of the flame. Generally
speaking, CRDS being a line-on-site technique, mea-
sures the concentration averaged through all the opti-
cal pass. Therefore, if the laser beam passes through
a small zone of high temperature caused by the
curvature of the flame front, it can increase the
Y
observed absorption of the ‘hot’ Õ s1 line. How-
Fig. 4. Dependence of the nonequilibrium factor u on the distance
from the burner. The temperature profile shown was measured
ever, this effect must in the same fashion be seen for
the ‘hot’ rotational lines with high J numbers. Good
linearity of the Boltzmann plot shows that non-
flatness does not substantially affect the results.
Y
using the rotational level of the Õ s0 state of OH. The solid line
in the temperature profile is the smooth approximation of the
measured experimental points shown by solid circles. Flame con-
ditions are as in Table 1.

V.A. LozoÕsky et al.rChemical Physics Letters 284 (1998) 407–411
411
u factor on the distance above the burner. The figure
shows that at distances above 10 mm, the concentra-
tion of vibrationally excited OH radicals is close to
equilibrium Žuf1.. It also shows that the ratio of
the band intensity for the Ž0, 0. and Ž1, 1. transitions
measurements with the simulation results. It will
produce an additional picture of the pathways of
formation and destruction of these radicals.
used in this work is correct. However, at low dis-
Acknowledgements
Y
tances, the concentration of OHŽÕ s1. can be more
than 20 times higher than the equilibrium value.
This research was supported by the Israel Science
Foundation administrated by the Israel Academy of
Science and Humanities and by the James Franck,
German–Israeli Binational Program in Laser Matter
Interaction.
Y
Y
Note, that even in this case the ratio of wÕ s1xrwÕ
s0x is only 0.1. However, the equilibrium value at
this temperature is much lower, which gives uf20.
We have performed additional experiments in two
other flames of different equivalent ratios: ws0.8
and ws1.2. In both these flames we observed prac-
tically the same behavior: the nonequlibrium factor
had approximately the same value as in the stoichio-
metric flame Žws1. at the same distances below the
flame front Žluminescence zone..
References
w1x J. Warnatz, in: W.C. Gardiner ŽEd.., Combustion Chemistry,
Springer, New York, 1984, pp. 197–360.
We can speculate about the origin of the vibra-
tionally excited OH. Several exothermic chemical
reactions produce OH in the flame. The reactions
w x
D.R. Crosley, Combust. Flame 78 1989 153.
Ž .
2
w3x G. Meijer, M.G.H. Boogaarts, R.T. Jongma, D.H. Parker,
A.M. Wodtke, Chem. Phys. Lett. 217 Ž1994. 112.
w4x S. Cheskis, I. Derzi, V.A. Lozovsky, A. Kachanov, D.
Romanini, Appl. Phys. B Ž1997, in press..
HqHO ™ OHqOH , D Hsy36 kcalrmol,
2
w5x D. Romanini, A.A. Kachanov, N. Sadeghi, F. Stoeckel,
OqH CO ™ OHqHCO ,
2
Chem. Phys. Lett. 264 Ž1997. 316.
D Hsy13.9 kcalrmol
could be considered. It is possible also that some
contribution of the diffusion of OHŽÕs1. from the
upper hot zone must be also taken into account.
w6x A. O’Keefe, D.A.G. Deacon, Rev. Sci. Instrum. 59 Ž1988.
2
544.
w7x J.J. Scherer, J.B. Paul, A. O’Keefe, R.J. Sayakally, Chem.
Rev. 97 1997 25.
Ž
.
w8x J.J.L. Spaanjaars, J.J. ter Meulen, G. Meijer, J. Chem. Phys.
Y
1
07 Ž1997. 2242.
The small concentration of OHŽÕ s1. in the
w9x P. Zalicki, Y. Ma, R.N. Zare, E.H. Wahl, J.R. Dadamio, T.G.
Owano, C.H. Kruger, Chem. Phys. Lett. 234 Ž1995. 269.
w10x J.J. Scherer, K.W. Aniolek, N.P. Cernanski, D.J. Rakestraw,
nonequilibrium zone could not probably substantially
affect the total kinetics of the combustion, because
most of the reactions could not probably be acceler-
ated, essentially with the vibrational excitation of the
radicals. Nevertheless, in other cases such an influ-
ence can be more important. Moreover, the explicit
including of the vibrationally excited OH radical in
the combustion mechanism could give an additional
test to that mechanism when comparing the results of
J. Chem. Phys. 107 1997 6196.
Ž
.
w11x J.J. Scherer, D.J. Rakestraw, Chem. Phys. Lett. 265 Ž1997.
1
69.
w12x L. Zhu, D. Kellis, C.-F. Ding, Chem. Phys. Lett. 257 Ž1996.
87.
w13x T. Yu, M.C. Lin, J. Phys. Chem. 99 Ž1995. 8599.
4
w
1
4 I.L. Chidsey, D.R. Crosley, J. Quant. Spectrosc. Radiat.
x
Transfer 23 Ž1980. 187.