Vibrational Relaxation of O 3 (ν 2 ) by O( 3 P)

Vibrational Relaxation of O3(ν2) by O(3P)Karen J. Castle,*,† Labe A. Black,‡ and Tara J. Pedersen†

†Department of Chemistry, Bucknell University, 203 Rooke Chemistry Building, Lewisburg, Pennsylvania 17837, United States‡Department of Chemistry and Biochemistry, University of Montana, 32 Campus Drive, Missoula, Montana 59812, United States

ABSTRACT: Laboratory measurements of the rate coefficient for quenching of O3(ν2)by ground-state atomic oxygen, kO(ν2), at room temperature are presented. kO(ν2) iscurrently not well known and is necessary for appropriate nonlocal thermodynamicequilibrium modeling of the upper mesosphere and lower thermosphere. In this work,a 266 nm laser pulse photolyzes a small amount of O3 in a slow-flowing gas mixture of O3, Xe,and Ar. This process simultaneously produces atomic oxygen and increases the temperature ofthe gas mixture slightly, thereby increasing the population in the O3(ν2) state. Transient diodelaser absorption spectroscopy is used to monitor the populations of the O3(ν2) and groundvibrational states as the system re-equilibrates. Relaxation rates are measured over a range ofquencher concentrations to extract the rate coefficient of interest. The value of kO(ν2) wasdetermined to be (2.2 ± 0.5) × 10−12 cm3 s−1.

■ INTRODUCTION

Ozone is a significant contributor to the thermal and radiativestructure of Earth’s mesosphere/lower thermosphere (MLT)region. O3 molecules absorb solar radiation and are photolyzedto heat the atmosphere both directly through the formation oftranslationally excited products and indirectly through thecreation of electronically or vibrationally excited products thatwill go on to heat the atmosphere via collisional energy exchangeprocesses.1 In the lower atmosphere, collisional energy exchangedominates the energy balance when compared with nonlocalradiative terms associated with atmospheric and solar radiation.In this case, the vibrational level populations of ozone follow aBoltzmann distribution with a local temperature. This situation iscalled local thermodynamic equilibrium (LTE). At the altitudesabove ∼65 km, the collision frequency is too low to maintain thepopulations of ozone vibrational levels in thermal equilibriumwith the surrounding gas. In this case, the collisional, photo-chemical, and radiative transfer processes related to populationand depopulation of the vibrational levels must be taken intoaccount in describing the vibrational state populations. Theseconditions are known as nonlocal thermodynamic equilibrium(non-LTE). Knowledge of ozone mixing ratios in the MLT and adetailed understanding of the kinetics and dynamics with whichozone exchanges energy with collocated atoms andmolecules arenecessary for accurate non-LTE modeling of the temperatureand density structure of the region. Because the upperatmosphere is especially sensitive to chemical and dynamicalprocesses occurring at lower altitudes, early evidence of globalatmospheric change may be found by monitoring species such asozone in the MLT. Experiments that help reduce uncertainty inkey kinetic parameters in O3−M molecular energy-transferinteractions (whereM is any one of several ambient species) maybe of fundamental importance in tracking long-term variations inclimate.

Ozone is a bent molecule with three normal modes ofvibration: a symmetric stretching mode (ν1) at 1103 cm−1, abending mode (ν2) at 701 cm−1, and an asymmetric stretchingmode (ν3) at 1042 cm

−1. The notation (ν1ν2ν3) will be used toidentify a particular vibrational state in this manuscript. Figure 1

Received: January 8, 2014Revised: May 23, 2014Published: June 3, 2014

Figure 1. Lowest vibrational energy levels of O3, labeled by thesymmetric stretch (ν1), bend (ν2), and asymmetric stretch (ν3) as(ν1ν2ν3). The solid lines indicate the V−V and V−T processes that areincluded in the kinetic model; the V−T process being measured in thiswork is shown with a thick line. The dashed lines indicate the spec-troscopic transitions used for detection.

Article

pubs.acs.org/JPCA

© 2014 American Chemical Society 4548 dx.doi.org/10.1021/jp500224j | J. Phys. Chem. A 2014, 118, 4548−4553

pubs.acs.org/JPCA

gives an energy level diagram for the lowest energy vibrationalstates. It should be noted that the (100) and (001) states arecoupled through near-resonant V−V energy exchange and canbe approximated reasonably well as a single, doubly degeneratelevel.The ground and three lowest excited vibrational states of O3 in

the lower mesosphere have been measured through microwaveobservation,2 but O3 mixing ratios at higher altitudes in the MLTare often inferred from 9.6 μm emission profiles measuredby limb-scanning sounding instruments on orbiting satellites.Interpretation of these emission profiles is complicated by theoccurrence of non-LTE conditions, sparking the development ofnumerous non-LTE models.3−9 Accurate models requireknowledge not only of the kinetic parameters for O3(ν3) butalso of the other low-lying vibrational states. The O atom den-sity in theMLT peaks at 1012 cm−3 near 100 km.10 Because of thepotential efficiency with which O3 and O exchange energy,O3−O energy exchange could be significant in non-LTE modelsfor nighttime conditions when the higher-altitude O3 is not beingphotodepleted by the sun.In general, vibrationally excited ozone could be deactivated

through the vibrational−vibrational (V−V) process

+ ⎯ →⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯ +−O (001, 100) M O (010) Mk v

3( )

3M,V V 1,3

(1)

and the vibrational−translational (V−T) processes

+ ⎯ →⎯⎯⎯⎯⎯ +O (001, 100) M O Mk v

3( )

3M 1,3

(2)

and

+ ⎯ →⎯⎯⎯⎯ +O (010) M O Mk v

3( )

3M 2

(3)

where M is an atmospherically relevant species such as N2, O2, orO.11 Processes 1 and 3, which are relevant for the current work,are indicated in Figure 1. When M = O, there is also a reactivechannel given by

+ ⎯ →⎯⎯vO ( ) O( P) 2Ok v

33 ( )

2r

(4)

The goal of the present work is to experimentally measure thevalue of kO(ν2) with greater precision than the current literaturevalue.Rosen and Cool12,13 were the first to perform vibrational

relaxation measurements of O3(ν) by various collider species.They excited ground-state O3 to the (001) level with a CO2 laserand then monitored the (101) fluorescence using a filtered InSbdetector at 4.8 μm. The (101) level was populated by vibrational“ladder-climbing” from the lower energy level populations.The authors developed a kinetic model to derive the “systemrelaxation rate” and estimated that stretch manifold populationsrelax a factor of three times faster than those for the bendmanifold. They also concluded that stretch manifold populationsare quenched exclusively through the bend manifold before theenergy is converted into translation, a finding that has beenconfirmed by two more recent studies.14,15 Thus, the processdescribed in eq 2 will not be considered in this work.Other groups have used a variety of excitation and probe

techniques to measure O3 vibrational relaxation in collisions withO3, O2, N2, and other species. All of the experiments used thesame technique to excite the O3: a rovibrational line in the O3(001) manifold was pumped by a 9.6 μm CO2 laser line near-resonant with an O3 (001)−(000) rovibrational transition. Manyof the experiments have employed somewhat indirect detection,

for example, emission from the (101) band at 4.8 μm orspectroscopic techniques with relatively low spectral resolution,for example, broadband IR emission or UV absorption. Elegantbut complex kinetic arguments were used to infer the excited-state population evolutions following excitation of the (001)level. Zeninari et al.15 used such a model to derive relaxation ratesof vibrationally excited O3 by O2 and N2, among other species,from the phase information in the photoacoustic response toCO2 laser excitation of O3.Two groups have used higher-resolution techniques to more

specifically detect individual O3 rovibrational levels involvedin relaxation processes. In one set of papers, Flannery andSteinfeld16,17 used high-resolution diode laser absorptiondetection of individual rovibrational states following CO2 laserexcitation. However, they were mainly interested in the rate ofrotational energy transfer. Menard-Bourcin and cow-orkers14,18−20 performed experiments in which the O3 vibra-tional populations were selectively detected by a CW CO2 laser,again tuned to a near-resonant O3 line, but in this case to detectthe excited vibrational levels via hot and combination band ab-sorption. They used various lines to selectively detect the vibra-tional levels of interest, and measured, among other quantities,kM(ν2) for M =O2 and N2. Both the Menard-Bourcin et al.14 andZeninari et al.15 publications place the values of kO2

(ν2) and

kN2(ν2) in the (2 to 3) × 10−14 cm3 s−1 range.In the only published measurement of the quenching rate of

vibrationally excited O3 by O atoms, performed byWest et al.,21,22

a pulsed CO2 laser was used to excite O3 to the (001) level. Oatoms were created by using a visible laser to dissociate a smallamount of O3. The O-atom population lifetime was monitoredusing 130 nm resonance fluorescence, with and without excitingthe O3 vibration. Assuming nonreactive quenching, the valuekO(ν2) = 3 × 10−12 cm3 s−1 was determined with factor-of-twouncertainty, larger values being inferred if the quenching wasassumed to be reactive according to the process in eq 4. Theauthors later estimated that at least 70% of the quenchingproceeds via vibrational relaxation.22 However, this result hasbeen disputed by Rawlins and coworkers,23 whose model of IRemission data in microwave-excited O2/Ar mixtures implied thatchemical reaction dominates with a rate coefficient exceeding10−11 cm3 s−1. West et al.21,22 also determined kM,V−V(v1,3) = 9 ×10−12 cm3 s−1 with the same factor of two uncertainty. Thereaction of O with vibrational ground-state O3 is relatively slowwith a rate coefficient of 8 × 10−15 cm3 s−1.24

In this work, we present a room-temperature laboratorymeasurement of kO(ν2). The experimental approach involvesimposing a temperature perturbation on the system to tem-porarily disturb the equilibrium conditions and monitoring thevibrational state populations as a function of time as the systemreturns to equilibrium.25 Measuring the characteristic decay ratesof the vibrational state populations as functions of O atom con-centration allows for determination of kO(ν2), as will beexplained in the Results and Discussion section.

■ EXPERIMENTAL METHODS

The temperature-jump/transient diode laser absorption ap-proach used in this work was similar to that used in our previouswork on the CO2−O system.26,27 A slowly flowing gas mixture ofO3 in Xe and Ar was passed through a 1 m path length reactioncell. The O3 was generated by flowing O2 through an ozonator(Ozomax) and stored on a cold silica gel column before beingflowed into the reaction cell through a needle valve. Ar and Xe

The Journal of Physical Chemistry A Article

dx.doi.org/10.1021/jp500224j | J. Phys. Chem. A 2014, 118, 4548−45534549

were flowed via mass flow controllers (MKS Instruments) at flowrates between 10 and 100 sccm. Total pressure in the flow cellwas maintained at a steady pressure between 10 and 50 Torrusing a roughing pump. The gas mixture was subjected to a266 nm fourth harmonic pulse from a Nd:YAG laser(Continuum), dissociating a small fraction of the O3 to formatomic oxygen. O3 partial pressures were generally <1 Torr, andvariable amounts up to 100 mTorr were dissociated with eachlaser pulse. The Nd:YAG laser was typically run at a repetitionrate of 5.4 Hz with a fluence of ∼18 mJ per pulse.The initial 266 nm dissociation of O3 produces O(1D) +

O2(1Δg) and O(

3P) + O2(3Σ−

g) at 85 and 15% branching ratios,respectively.24 Excited O(1D) is quickly quenched to O(3P) bythe efficient quenchers Xe (7.5 × 10−11 cm3 s−1)28 and O3 (2.4 ×10−10 cm3 s−1).24 The O(1D) product is not energetic enough fordirect O3 vibrational excitation. Vibrationally excited O2(

1Δg,v)and O2(

3Σ−,v) are efficiently quenched by the O2 (from de-composition of O3) flowing through the cell together with theO3.

29,30 The O atom density is accurately determined knowingthe initial O3 concentration (obtained by monitoring absorptionof the 254 or 297 nm line of a Hg arc lamp by the O3 in thereaction cell) and the photolysis yield (determined through theknown absorption cross-section of O3 at 266 nm and the mea-sured Nd:YAG laser fluence over the reaction volume).In addition to producing O atoms, the absorption of UV

radiation by O3 also leads to a temperature-jump in the system. Adetailed account of the thermalization of the absorbed laserenergy that leads to the temperature-jump can be found in Castleet al.26 The magnitude of the temperature increase in the presentwork is variable between about 1 and 22 K, depending on theexperimental conditions. Following the photolysis pulse, the O3vibrational level populations will redistribute according to thenew temperature, the ground-state population shifting upward.To illustrate the magnitude of the population shift, consider aBoltzmann distribution among the lowest four energy states ofozone at 298 K. The relative populations in the (000), (010),(001), and (100) states will be 95.7, 3.2, 0.6, and 0.5%,respectively. A Boltzmann distribution at 320 K (assuming thelargest T-jump of 22 K) gives the populations in these same statesas 94.4, 4.0, 0.9, and 0.7%, respectively. Thus, ∼1.3% of the totalpopulation is shifted upward out of the ground state by the smalltemperature jump, and of that shifted population, two-thirds willgo to the (010) state, while one-third goes to the (001) and (100)states.Vibrational energy transfer in the O3−O system was

monitored using a rovibrational-state-selective detection techni-que, transient diode laser absorption spectroscopy. The 10 MHzbandwidth of the diode laser (Laser Components) allowsselective detection of individual rovibrational states, even formolecules such as O3 with highly congested rovibrationalmanifolds. Unlike using a CO2 laser as the probe, which relieson fortuitous line coincidences, the diode laser can be set to anarbitrary wavelength over the O3(ν3) spectral region. Previouswork primarily used more indirect means of detection;complicated kinetic models were often used to infer the rate ofν2 deactivation relative to deactivation of the stretching modes.In this work, each excited vibrational state can be probed directly,thus giving a more direct means for monitoring the vibrationalrelaxation. Because rotational populations are interconverted ona much shorter time scale than the vibrational relaxation, any ofthe rotational states can be used to determine the vibrational-level population evolution following laser excitation.

Figure 2 shows a representative experimental absorptionspectrum of 100 mTorr of O3 in Ar bath gas. Included are the(001)−(000) band asymmetric-top rotational line assignmentsindicated with the quantum number notation JKa,Kc. The spec-trum gives an indication of the signal-to-noise ratio associatedwith absorption of the diode laser by O3 in this spectral regionand indicates the spectral transition used to acquire most of thekinetic data.Dichroic mirrors were used to overlap the counter-propagating

pump (Nd:YAG) and probe (diode) laser beams. Transientabsorption signals, which represent the time-evolving populationof the rovibrational state being probed, were obtained by fixing thefrequency of the diode laser to a peak of anO3(ν3) transition in the1035−1060 cm−1 range. Absorption was monitored as a functionof time with a liquid-nitrogen-cooled HgCdTe (MCT) detector(Infared Associates). A pair of 8−12 μm antireflection-coated Gewindows was used to reduce the amount of scattered light reachingthe detector and laser diode. Two irises were inserted in the diodelaser path to restrict the diameter of the detection beam to 1.5 mm,significantly smaller than the 5.8 mm diameter photolysis beam. ALeCroy Wavesurfer oscilloscope was triggered by the photolysispulse and was used to record the transient absorption curves. Therate coefficient was determined by measuring the quenching rateas a function of O atom concentration, which was controlled byvarying the amount of dissociated O3.

■ RESULTS AND DISCUSSIONA representative set of transient absorption curves used to derivekO(ν2) is given in Figure 3. Each plot represents the transientabsorption signal obtained for a different quencher concen-tration, with [O] values varying between 7.6 × 1013 and 4.2 ×1014 cm−3 across the data set. The total pressure in the reactioncell was maintained at 32 Torr and the bath gas flow rateswere Xe = 50 sccm and Ar = 10 sccm. The photolysis laser wasoperating at a fluence of 18 mJ per pulse. These particular datawere obtained by fixing the diode laser frequency to coincidewith an absorption peak corresponding to the (001)216,15 ←(000)206,14 transition of O3 (see Figure 2) and monitoring thepopulation of the lower state as a function of time after thephotolysis pulse. Note that the sign of the transient absorptionsignal is arbitrary and determined by the settings on thepreamplifier used in the data acquisition process.

Figure 2. Experimental absorption spectrum of 100 mTorr O3 in the1055 cm−1 region. The lines were assigned using the HITRAN 2004database and are indicated with the notation JKa,Kc.



The shape of each transient absorption curve in Figure 3 canbe understood as follows. At t = 0 there is an offset in intensitydue to initial O3 loss following photolysis by the dissocia-tion laser. The rise side of a given transient absorption curverepresents the evolution of the vibrational state population to anew equilibrium at the post-temperature-jump temperature. Theslower decay side of the curve corresponds to thermal re-equilibration of the gas mixture back to the original temper-ature.31 The solid lines shown in Figure 3 were obtained by fittingthe data to a difference of two exponential functions using amodified version of the home-written Levenburg-Marquartnonlinear least-squares fitting algorithm that was used in ourCO2−O work.27 The longer term decay constant related tothermal re-equilibration is fairly constant at∼200 s−1 for all of thedata acquired in this experiment, which is reasonable given theexperimental conditions.The rise portion of each transient absorption curve contains

the kinetic information on interest here as it represents thesystem reaching equilibrium at the post-temperature-jumptemperature. To extract the rate coefficient of interested, we fitthe data using Kintecus v. 5.20 chemical modeling software.32

The user programs in the desired kinetic model and initialconcentrations of each species and the software identify the bestfit of the data to the model. The model used in this analysis wassimilar to model 3 of Manuilova et al.33 The (000), (010), (001),

and (100) states were included along with all relevant V−V andV−T processes. The value of kM,V−V(ν1,3) from eq 1 was fixed atWest et al.’s value of 9 × 10−12 cm3 s−1, and chemical reaction ofexcited O3 molecules with atomic oxygen was assumed to be theminimum, as suggested by West et al. No energy levels higherthan (100) were included in the model, which is appropriategiven the small temperature jumps (<10 K) imposed on thesystem.The rise portion of each transient absorption curve was fit

within the context of the model with all parameters fixed exceptfor collisional quenching of the (010) state. This rate parameterwas then extracted from each transient absorption curve usingthe fitting feature in Kintecus. A sample fit is shown in Figure 4;

the (010) quenching rate in this measurement was determinedto be 1255 s−1, while the O atom density was 1.0 × 1014 cm−3.This rate parameter was determined for each of 129 transientabsorption curves taken over the course of 15 data sets covering arange of O atom concentrations. Each transient absorption curverise is represented by∼2000 data points averaged over 3000 lasershots. For each data set, the (010) quenching rate was plottedversus O atom density to generate a pseudo-Stern−Volmer plot,such as the example shown in Figure 5. The slope of the best-fit

Figure 3. Raw transient absorption data for five different O atomdensities. The same scale is used on the Y axis for each curve. Each solidblack line represents the predicted population time evolution from aglobal nonlinear least-squares fit to a double-exponential function.

Figure 4. Sample Kintecus fit of a transient absorption curve taken withan O atom concentration of 1.0 × 1014 cm−3 and a total cell pressure of32 Torr.

Figure 5. Pseudo-Stern−Volmer plot for two combined data sets takenat a total cell pressure of 51 Torr. The slope of the best-fit line representsthe rate coefficient kO(ν2).



line of the pseudo-Stern−Volmer plot gives the desired ratecoefficient, kO(ν2). A Stern−Volmer analysis was performed oneach data set individually. This was necessary because the totalcell pressure varied between 11 and 51 Torr for the various datasets and the intercept of the pseudo-Stern−Volmer plot changeswith total pressure as it originates from quenching of the excitedozone by the bath gases. The resulting kO(ν2) values and un-certainties from this Stern−Volmer analysis are given in Table 1.

The final rate coefficient was taken to be the weighted meanof the individual data sets, or (2.2 ± 0.5) × 10−12 cm3 s−1. Thereported uncertainty represents the standard deviation of themean, which is expected to be larger than the systematic errorthat would arise primarily from uncertainties in the measuredO-atom density. It should be noted that this rate coefficientrepresents collisional quenching at the temperature in thereaction cell after the temperature jump. Because this temper-ature is actually changing over a roughly 10 K range over the datasets (because it is a function of ozone concentration), werecommend designating the kO(ν2) at the median post-temperature jump reaction cell temperature of 301 K.While attempts were made to fit the data with kO,V−V(ν1,3)

and kO,V−T(ν2) both as unknowns, the results were statisticallyambiguous. Further study is needed before an individualmeasured value for kO,V−V(ν1,3) can be reported. The accuracyof kO(ν2) is obviously limited by the accuracy of the chosenkinetic model as well as the rate parameters entered as knownvalues in the Kintecus model.The measurements presented herein are consistent with the

lower precision measurements of West et al.21,22 The resultssuggest that the measured rate coefficient is nearly two orders ofmagnitude larger than vibrational relaxation of O3(ν2) by thedominant species N2 and O2.

5 This works supports the idea thatquenching of vibrationally excited ozone byO(3P) is efficient andmay be important for proper non-LTE modeling of the MLTregion.It should be noted that because temperatures in the MLT can

reach as low as 110 K, it will ultimately be important to knowthe temperature dependence of this rate coefficient. For example,our previous measurements of the rate coefficient for vibrationalrelaxation of CO2(ν2) by O(3P) showed a slight negativetemperature dependence.27 In that case, the rate coefficient for

CO2(ν2)−O(3P) relaxation at the coldest temperatures in theMLT was found to be roughly 50% larger than the roomtemperature value. Until more work is done, some assumptionswill have to be made to scale the room-temperature ratecoefficient kO(ν2) for O3−O to colder temperatures.

■ CONCLUSIONSUsing contemporary experimental techniques, the relaxation ratecoefficient for quenching of the lowest excited vibrational levelby ground-state O atoms, kO(v2), has been measured. Theexperimental approach was similar to our successful method usedto study CO2 (ν2)−O energy transfer: a temperature-jumpperturbation/relaxation technique followed by transient diodelaser absorption detection of vibrational state populations. Thediode laser absorption method permits the facile detection ofO3, which is a key IR-active species in MLT photochemistry andis difficult to detect using standard pulsed UV laser techniques.Overall, the recommended value of kO(ν2) = (2.2 ± 0.5) ×10−12 cm3 s−1 represents a significant improvement on the con-fidence level in the literature for this rate coefficient, which islarge enough to be important in non-LTE models of the MLT.Future work will focus on measuring kO,V−V(ν1,3), determiningthe loss rate through the reactive pathway, further reductionof the uncertainty in kO(ν2), and experimentally measuring thetemperature dependence of both rate coefficients.

■ AUTHOR INFORMATIONNotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSThis material is based onwork supported by theNational ScienceFoundation’s CEDAR Program, grant number 0640063, andbuilds on previous work supported by the NASA GeospaceSciences Program. James Dodd’s assistance with previous work,including the original global fitting algorithm, contributed greatlyto the success of this project. We also thank students TriciaClyde, Alex Kowler, and Michael Brignone for their assistancewith this manuscript.

■ REFERENCES(1) Feofilov, A. G.; Kutepov, A. A. Infrared Radiation in theMesosphere and Lower Thermosphere: Energetic Effects and RemoteSensing. Surv. Geophys. 2012, 33, 1231−1280.(2) Sandor, B. J.; Clancy, R. T. Mesospheric Chemistry of VibrationallyExcited O3 from Diurnal Microwave Measurements of O3(ν1), O3(ν2),O3(ν3), and O3(Ground State). J. Geophys. Res. 2010, 115, D21302/1−D21302/10.(3) Adler-Golden, S. M.; Gruninger, J.; Smith, D. R. Derivation ofAtmospheric Atomic Oxygen and Hydrogen Profiles from Ozone ν3Band Emission. J. Geophys. Res. 1992, 97, 19509−19518.(4) Kaufmann,M.; Gusev, O. A.; Grossman, K. U.; Martin-Torres, F. J.;Marsh, D. R.; Kutepov, A. A. Satellite Observations of Daytime andNighttime Ozone in the Mesosphere and Lower Thermosphere. J.Geophys. Res. 2003, 108, ACH 9/1−ACH 9/14.(5) Manuilova, R. O.; Shved, G. M. The 4.8 and 9.6 μm O3 BandEmissions in the Middle Atmosphere. J. Atmos. Sol.-Terr. Phys. 1992, 54,1149−1168.(6) Mlynczak, M.; Zhou, D. K. Kinetic and Spectroscopic Require-ments for the Measurement of Mesospheric Ozone at 9.6 μm UnderNon-LTE Conditions. Geophys. Res. Lett. 1998, 25, 639−642.(7) Mlynczak, M. G.; Drayson, S. R. Calculation of Infrared LimbEmission by Ozone in the Terrestrial Middle Atmosphere. I. SourceFunctions. J. Geophys. Res. 1990, 95, 16497−16511.

Table 1. Summary of Pseudo-Stern−Volmer Fits for Each ofthe 15 Data Sets Included in This Worka

data set no. pressure/torr kO (ν2)/(10−12 cm3 s−1)

1 11.0 2.6 ± 0.42 11.5 2.9 ± 0.43 13.7 2.7 ± 0.44 30.0 0.1 ± 0.35 30.8 0.8 ± 0.26 32.0 2.0 ± 0.67 32.0 2.4 ± 0.48 32.3 1.4 ± 0.49 32.5 2.1 ± 0.310 49.7 2.4 ± 0.611 51.0 2.2 ± 0.212 51.0 2.1 ± 0.213 57.0 2.2 ± 0.314 57.0 1.4 ± 0.5

aWeighted average and standard deviation are kO(ν2) = (2.2 ± 0.5) ×10−12 cm3 s−1.



(8) Mlynczak, M. G.; Drayson, S. R. Calculation of Infrared LimbEmission by Ozone in the Terrestrial Middle Atmosphere. II. EmissionCalculations. J. Geophys. Res. 1990, 95, 16513−16521.(9) Rawlins, W. T. Chemistry of Vibrationally Excited Ozone in theUpper Atmosphere. J. Geophys. Res. 1985, 90, 12283−12292.(10) Smith, A. K.; Marsh, D. R.; Mlynczak, M. G.; Mast, J. C. TemporalVariations of Atomic Oxygen in the Upper Mesosphere from SABER. J.Geophys. Res. 2010, 115, D18309/1−D18309/18.(11) Lopez-Puertas, M.; Taylor, F. W. Non-LTE Radiative Transfer inthe Atmosphere; World Scientific Publishing Co. Pte. Ltd.: Singapore,2001; pp 200−212.(12) Rosen, D. I.; Cool, T. A. Vibrational Deactivation of O3(101)Molecules in Gas Mixtures. J. Chem. Phys. 1973, 59, 6097−6103.(13) Rosen, D. I.; Cool, T. A. Vibrational Deactivation of O3Moleculesin Gas Mixtures. II. J. Chem. Phys. 1975, 62, 466−476.(14) Menard-Bourcin, F.; Menard, J.; Doyennette, L. VibrationalRelaxation of Ozone in O3-O2 and O3-N2 Gas Mixtures from InfraredDouble-ResonanceMeasurements. J. Chem. Phys. 1991, 94, 1875−1881.(15) Zeninari, V.; Tikhomirov, B. A.; Ponomarev, Y. N.; Courtois, D.Photoacoustic Measurements of the Vibrational Relaxation of theSelectively Excited Ozone (ν3) Molecule in Pure Ozone and its BinaryMixtures with O2, N2, and Noble Gases. J. Chem. Phys. 2000, 112, 1835−1843.(16) Flannery, C. C.; Steinfeld, J. I. Rotational Relaxation Measure-ments in Ozone: Temperature and Collision Partner Effects. J. Chem.Phys. 1992, 92, 8157−8163.(17) Flannery, C. C.; Steinfeld, J. I. State-to-State Ozone RelaxationRates for O3-N2 Collisions. J. Chem. Phys. 1993, 99, 6495−6505.(18) Doyennette, L.; Boursier, C.; Menard, J.; Menard-Bourcin, F. ν1-ν3 Coriolis-Assisted Intermode Transfers in O3-M Gas Mixtures (MO2 and N2) in the Temperature Range 200−300 K from IR Double-Resonance Measurements. Chem. Phys. Lett. 1992, 197, 157−160.(19) Menard, J.; Doyennette, L.; Menard-Bourcin, F. VibrationalRelaxation of Ozone in O3-O2 and O3-N2 Gas Mixtures from InfraredDouble-Resonance Measurements in the 200−300 K TemperatureRange. J. Chem. Phys. 1992, 96, 5773−5780.(20) Menard-Bourcin, F.; Doyennette, L.; Menard, J. VibrationalEnergy Transfers in Ozone from Infrared Double-Resonance Measure-ments. J. Chem. Phys. 1990, 92, 4212−4221.(21) West, G. A.; Weston, R. E., Jr.; Flynn, G. W. Deactivation ofVibrationally Excited Ozone by O(3P) Atoms. Chem. Phys. Lett. 1976,42, 488−493.(22) West, G. A.; Weston, R. E., Jr.; Flynn, G. W. The Influence ofReactant Vibrational Excitation on the O(3P) + O3 BimolecularReaction Rate. Chem. Phys. Lett. 1978, 56, 429−433.(23) Rawlins, W. T.; Caledonia, G. E.; Armstrong, R. A. Dynamics ofVibrationally Excited Ozone Formed by Three-Body Recombination. II.Kinetics and mechanism. J. Chem. Phys. 1987, 87, 5209−5221.(24) Sander, S. P.; Abbatt, J.; Barker, J. R.; Burkholder, J. B.; Friedl, R.R.; Golden, D. M.; Huie, R. E.; Kolb, C. E.; Kurylo, M. J.; Moortgat, G.K.; et al. Chemical Kinetics and Photochemical Data for Use in AtmosphericStudies; Eval. No. 17; JPL Pub: Pasadena, CA, 2011; 1-5-4A-7.(25) Bernasconi, C. F. Relaxation Kinetics; Academic Press: New York,1976.(26) Castle, K. J.; Black, L. A.; Simione, M. W.; Dodd, J. A. VibrationalRelaxation of CO2(ν2) by O(

3P) in the 142−490 K Temperature Range.J. Geophys. Res. 2012, 117, A04310.(27) Castle, K. J.; Kleissas, K. M.; Rhinehart, J. M.; Hwang, E. S.; Dodd,J. A. Vibrational Relaxation of CO2(ν2) by Atomic Oxygen. J. Geophys.Res. 2006, 111, A09303.(28) Schofield, K. Rate Constants for the Gaseous Interactions of O(2

1D2) and O(21S0): A Critical Evaluation. J. Photochem. 1978, 9, 55−68.

(29) Yankovsky, V. A.; Manuilova, R. O. Model of Daytime Emissionsof Electronically-Vibrationally Excited Products of O3 and O2

Photolysis: Application to Ozone Retrieval. Ann. Geophys. 2006, 24,2823−2839.(30) Slanger, T. E.; Copeland, R. A. Energetic Oxygen in the UpperAtmosphere and the Laboratory. Chem. Rev. 2003, 103, 4731−4765.

(31) Pollock, D. S.; Scott, G. B. I.; Phillips, L. F. Rate Constant forQuenching of CO2(010) by Atomic Oxygen. Geophys. Res. Lett. 1993,20, 727−729.(32) Ianni, J. C. Kintecus, Windows Version 5.01, 2014. www.kintecus.com.(33) Manuilova, R. O.; Gusev, O. A.; Kutepov, A. A.; Von Clarmann,T.; Oelhaf, H.; Stiller, G. P.; Wegner, A.; Lopez-Puertas, M.; Martin-Torres, J.; Zaragoza, G.; et al. Modelling of Non-LTE Limb Spectra of i.r.Ozone Bands for the MIPAS Space Experiment. J. Quant. Spectrosc.Radiat. Transfer 1998, 59, 405−422.



www.kintecus.com

www.kintecus.com

Documents

Vibrational Relaxation of O 3 (ν 2 ) by O( 3 P)