Vibrational Relaxation of O3 (ν2) by O (3P)

Jun 3, 2014 - Karen J. Castle*†, Labe A. Black‡, and Tara J. Pedersen†. † Department of Chemistry, Bucknell University, 203 Rooke Chemistry Bu...
0 downloads 0 Views 2MB Size
Article pubs.acs.org/JPCA

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) is currently not well known and is necessary for appropriate nonlocal thermodynamic equilibrium 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 of the gas mixture slightly, thereby increasing the population in the O3(ν2) state. Transient diode laser absorption spectroscopy is used to monitor the populations of the O3(ν2) and ground vibrational states as the system re-equilibrates. Relaxation rates are measured over a range of quencher concentrations to extract the rate coefficient of interest. The value of kO(ν2) was determined to be (2.2 ± 0.5) × 10−12 cm3 s−1.



INTRODUCTION Ozone is a significant contributor to the thermal and radiative structure of Earth’s mesosphere/lower thermosphere (MLT) region. O3 molecules absorb solar radiation and are photolyzed to heat the atmosphere both directly through the formation of translationally excited products and indirectly through the creation of electronically or vibrationally excited products that will go on to heat the atmosphere via collisional energy exchange processes.1 In the lower atmosphere, collisional energy exchange dominates the energy balance when compared with nonlocal radiative terms associated with atmospheric and solar radiation. In this case, the vibrational level populations of ozone follow a Boltzmann distribution with a local temperature. This situation is called local thermodynamic equilibrium (LTE). At the altitudes above ∼65 km, the collision frequency is too low to maintain the populations of ozone vibrational levels in thermal equilibrium with the surrounding gas. In this case, the collisional, photochemical, and radiative transfer processes related to population and depopulation of the vibrational levels must be taken into account in describing the vibrational state populations. These conditions are known as nonlocal thermodynamic equilibrium (non-LTE). Knowledge of ozone mixing ratios in the MLT and a detailed understanding of the kinetics and dynamics with which ozone exchanges energy with collocated atoms and molecules are necessary for accurate non-LTE modeling of the temperature and density structure of the region. Because the upper atmosphere is especially sensitive to chemical and dynamical processes occurring at lower altitudes, early evidence of global atmospheric change may be found by monitoring species such as ozone in the MLT. Experiments that help reduce uncertainty in key kinetic parameters in O3−M molecular energy-transfer interactions (where M is any one of several ambient species) may be of fundamental importance in tracking long-term variations in climate. © 2014 American Chemical Society

Ozone is a bent molecule with three normal modes of vibration: a symmetric stretching mode (ν1) at 1103 cm−1, a bending mode (ν2) at 701 cm−1, and an asymmetric stretching mode (ν3) at 1042 cm−1. The notation (ν1ν2ν3) will be used to identify a particular vibrational state in this manuscript. Figure 1

Figure 1. Lowest vibrational energy levels of O3, labeled by the symmetric stretch (ν1), bend (ν2), and asymmetric stretch (ν3) as (ν1ν2ν3). The solid lines indicate the V−V and V−T processes that are included in the kinetic model; the V−T process being measured in this work is shown with a thick line. The dashed lines indicate the spectroscopic transitions used for detection. Received: January 8, 2014 Revised: May 23, 2014 Published: June 3, 2014 4548

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

The Journal of Physical Chemistry A

Article

for example, emission from the (101) band at 4.8 μm or spectroscopic techniques with relatively low spectral resolution, for example, broadband IR emission or UV absorption. Elegant but complex kinetic arguments were used to infer the excitedstate population evolutions following excitation of the (001) level. Zeninari et al.15 used such a model to derive relaxation rates of vibrationally excited O3 by O2 and N2, among other species, from the phase information in the photoacoustic response to CO2 laser excitation of O3. Two groups have used higher-resolution techniques to more specifically detect individual O3 rovibrational levels involved in relaxation processes. In one set of papers, Flannery and Steinfeld16,17 used high-resolution diode laser absorption detection of individual rovibrational states following CO2 laser excitation. However, they were mainly interested in the rate of rotational energy transfer. Menard-Bourcin and coworkers14,18−20 performed experiments in which the O3 vibrational populations were selectively detected by a CW CO2 laser, again tuned to a near-resonant O3 line, but in this case to detect the excited vibrational levels via hot and combination band absorption. They used various lines to selectively detect the vibrational levels of interest, and measured, among other quantities, kM(ν2) for M = O2 and N2. Both the Menard-Bourcin et al.14 and Zeninari 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 by West et al.,21,22 a pulsed CO2 laser was used to excite O3 to the (001) level. O atoms were created by using a visible laser to dissociate a small amount of O3. The O-atom population lifetime was monitored using 130 nm resonance fluorescence, with and without exciting the O3 vibration. Assuming nonreactive quenching, the value kO(ν2) = 3 × 10−12 cm3 s−1 was determined with factor-of-two uncertainty, larger values being inferred if the quenching was assumed to be reactive according to the process in eq 4. The authors later estimated that at least 70% of the quenching proceeds via vibrational relaxation.22 However, this result has been disputed by Rawlins and coworkers,23 whose model of IR emission data in microwave-excited O2/Ar mixtures implied that chemical reaction dominates with a rate coefficient exceeding 10−11 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. The reaction of O with vibrational ground-state O3 is relatively slow with a rate coefficient of 8 × 10−15 cm3 s−1.24 In this work, we present a room-temperature laboratory measurement of kO(ν2). The experimental approach involves imposing a temperature perturbation on the system to temporarily disturb the equilibrium conditions and monitoring the vibrational state populations as a function of time as the system returns to equilibrium.25 Measuring the characteristic decay rates of the vibrational state populations as functions of O atom concentration allows for determination of kO(ν2), as will be explained in the Results and Discussion section.

gives an energy level diagram for the lowest energy vibrational states. It should be noted that the (100) and (001) states are coupled through near-resonant V−V energy exchange and can be approximated reasonably well as a single, doubly degenerate level. The ground and three lowest excited vibrational states of O3 in the lower mesosphere have been measured through microwave observation,2 but O3 mixing ratios at higher altitudes in the MLT are often inferred from 9.6 μm emission profiles measured by limb-scanning sounding instruments on orbiting satellites. Interpretation of these emission profiles is complicated by the occurrence of non-LTE conditions, sparking the development of numerous non-LTE models.3−9 Accurate models require knowledge not only of the kinetic parameters for O3(ν3) but also of the other low-lying vibrational states. The O atom density in the MLT peaks at 1012 cm−3 near 100 km.10 Because of the potential efficiency with which O3 and O exchange energy, O3−O energy exchange could be significant in non-LTE models for nighttime conditions when the higher-altitude O3 is not being photodepleted by the sun. In general, vibrationally excited ozone could be deactivated through the vibrational−vibrational (V−V) process kM,V − V(v1,3)

O3(001, 100) + M ⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯⎯→ O3(010) + M

(1)

and the vibrational−translational (V−T) processes kM(v1,3)

O3(001, 100) + M ⎯⎯⎯⎯⎯⎯→ O3 + M

(2)

and kM(v2)

O3(010) + M ⎯⎯⎯⎯⎯→ O3 + M

(3)

where M is an atmospherically relevant species such as N2, O2, or O.11 Processes 1 and 3, which are relevant for the current work, are indicated in Figure 1. When M = O, there is also a reactive channel given by k r(v)

O3(v) + O(3P) ⎯⎯⎯→ 2O2

(4)

The goal of the present work is to experimentally measure the value of kO(ν2) with greater precision than the current literature value. 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 laser and then monitored the (101) fluorescence using a filtered InSb detector 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 “system relaxation rate” and estimated that stretch manifold populations relax a factor of three times faster than those for the bend manifold. They also concluded that stretch manifold populations are quenched exclusively through the bend manifold before the energy is converted into translation, a finding that has been confirmed by two more recent studies.14,15 Thus, the process described 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 with O3, O2, N2, and other species. All of the experiments used the same technique to excite the O3: a rovibrational line in the O3 (001) manifold was pumped by a 9.6 μm CO2 laser line nearresonant with an O3 (001)−(000) rovibrational transition. Many of the experiments have employed somewhat indirect detection,



EXPERIMENTAL METHODS The temperature-jump/transient diode laser absorption approach used in this work was similar to that used in our previous work on the CO2−O system.26,27 A slowly flowing gas mixture of O3 in Xe and Ar was passed through a 1 m path length reaction cell. The O3 was generated by flowing O2 through an ozonator (Ozomax) and stored on a cold silica gel column before being flowed into the reaction cell through a needle valve. Ar and Xe 4549

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

The Journal of Physical Chemistry A

Article

were flowed via mass flow controllers (MKS Instruments) at flow rates between 10 and 100 sccm. Total pressure in the flow cell was maintained at a steady pressure between 10 and 50 Torr using a roughing pump. The gas mixture was subjected to a 266 nm fourth harmonic pulse from a Nd:YAG laser (Continuum), dissociating a small fraction of the O3 to form atomic oxygen. O3 partial pressures were generally