COLLISION ENERGY DEPENDENCE OF THE OVERALL RATE CONSTANT FOR THE REACTION NH ( aA ) + HN

The overall rate constant for the reaction NH(aA) + HN has been determined by the laser photolysis of hy_drogen azide (HN3) at 266 nm and 193nm. The visible emission from vibronically excited NHz(ZA) was dispersed and its time-dependent profiles were measured at several wavelengths. The rate constants are dependent not only on the photolysis wavelengths but also on the vibrational levels of the NH2(/ZA) produced in the reaction. The intermolecular potential between NH(aA) and HN was determined to be the form V(R) -C/R (2 < < 4, C: constant) from the analysis with a long-range potential approximation. The interaction between NH(aA) and HN3 is mainly governed by the dipoledipole interaction in the initial stage of the reaction.


INTRODUCTION
Chemiluminescence is due to the reactions which produce emissive excited states.Such highly reactive species as atoms and radicals are frequently related with the production of the electronically excited states.Chemiluminescence was observed in many reaction systems; nevertheless, most of the mechanisms have not completely been resolved yet.The goal of the present study is not only to make clear the reaction mechanism but also to investigate the effect of a collision energy on a chemilumines- cent reaction.Spectroscopic information reveals a source of the emission and a kinetic study gives the reaction mechanism of the chemiluminescence.The visible emission in the HN3 photolysis system was studied in the present work.
There have been a few reports on the chemiluminescence in the photodissociation of hydrogen azide HN3.Okabe first observed the emission over the range of 430-800 nm in the photolysis of HN3 with rare gas lamps (Xe and Kr).He assigned the source of the emission to the transition of N first-positive band (B31-lg-A3u+).
McDonald and co-workers, 2-4 who performed pioneering works on the photolysis of HN3, observed visible emission in the 266 nm photolysis.They proposed that the emission was due to the transition of NH2(/2AI-X2BI) only by comparing with its absorption wavelength.They also proposed a reaction NH(alA, v 0) + HN3 NH2(,2A) + N for the mechanism on the basis of the heat of reaction.Piper et al. observed temporal profiles not only of the visible emission but also of the laser-induced fluorescence (LIF) of NH(cH-aA).They reported that both profiles were consistent with the mechanism proposed by McDonald et al., although their rate constant was about two times larger than that measured by McDonald et al.The mechanism had been accepted with little doubt since then and no study showed the direct information on the origin of the emission by more reliable spectro- scopic technique.The overall rate constant for the reaction NH(aA, v 0) + HN3 has also been reported using various techniques; 6-1 however, there have been few studies on the reaction products, particularly on the electronically excited states.
In our prev_ious study, we reported a direct evidence for the source of the emission [NH2(,2A-X2B)] and for how the emissive species are produced in the HN3 photolysis at 266 nm.We observed dispersed fluorescence spectra that were not necessarily consistent with the mechanism accepted before and gave a modified reac- tion mechanism including internally excited NH(aA).
In the present study, the overall rate constant for the reaction NH(aA) + HN3 has been determined using different photolysis wavelengths (266 nm and 193 nm) and given information on the reaction dynamics.The rate constants were found to be dependent not only on the photolysis wavelength but also on the vibrational levels of the NH2(,2A) produced in the reaction.The rate constants reduce with an increase in the energy of photodissociation and with a decrease in the vibrational energy of the product NH2(/2A).The results are rationalized by the generation of centrifugal potential barrier due to the large translational energy of NH(aA).The relation be- tween the rate constant and collision energy will be discussed.
Both laser beams were trimmed with an aperture to 6 mm in diameter.The cell made of Pyrex tube has two ports for observation with quartz window of 22 mm in diameter and was evacuated with a rotary pump (ULVAC D-950).The pressure of the sample gas was monitored with a capacitance manometer (Baratron 122AA).
Emission from the observation region was collected through a long-pass filter (Toshiba Y-43) and was focused with a lens (f 80 mm) on the entrance slit of a monochromator (Nikon P-250).Spectral resolution was 3 nm (FWHM).The signal from a photomultiplier (Hamamatsu R928) mounted on the exit slit was sent to a home-made amplifier (xl0) and then fed into a gated integrator (Stanford Research Systems SR-250).The signals of 30 pulses were averaged.In order to correct for the shot-to-shot laser fluctuations, undispersed emission was focused with a lens (f 60 mm) on another photomultiplier (Hamamatsu 1P28) through a filter (Toshiba 0-54) at the opposite side of the monochromator.The signal from the 1P28 was amplified (10) and inverted with a home-made amplifier and then averaged with a home-made gated integrator.Both dispersed and undispersed emission signals were fed into an A/D interface and sent to a computer.
When the temporal behavior of the emission intensity was observed, the signal from the photomultiplier was amplified (x300) and digitized with a home-made tran- sient memory whose time resolution was 40 ns.Signals of 15000-30000 pulses were averaged to obtain a good S/N ratio.The relative laser fluence was monitored with a photodiode and averaged with a gated integrator.Data acquisition system was the same as that described in the spectroscopic study.
Hydrogen azide, HN3, was synthesized in vacuo by the reaction between sodium azide and stearic acid according to the procedure in ref. 12.

Spectroscopic Study
Figure 1 shows a dispersed emission spectrum that was observed in the 266 nm photolysis of HN3 (ref. 11).Significant progression appears along with a broad back- ground.Since the overall sensitivity of the detection system was not corrected, the decrease in the intensity at longer wavelengths does not show the correct intensity distribution but reflects the sensitivity of the detection system.Moreover, the filter used (Toshiba Y-43) was transparent only at longer wavelength than 420 nm; the v2'= 15 14 13 ; ( , .Wavelenh / nm Figure 1 Dispersed emission spectrum observed in the HN3/266nm photolysis.Pressure of HN was 150 mTorr with no buffer gas; spe_ctral resolution was 3 nm (FWHM).The assignments are appropriate to the linear configuration for the A2A state (ref. 13).The onset about 420 nm is due to the filter used.
onset of the spectrum at short wavelength is due to the transmission of the filter.
Most of the peaks were assigned to the transition NH2(A(0, v2', 0)-B(0, 0, 0)).4][15] The spectrum shows the excitation of the v2-bending vibrational levels up to at least v2' 15. Figure 2 shows a dispersed emission spectrum in the HN3/193 nm photolysis recorded under the same detection system.The envelope of the spectrum is very similar to that of 266 nm photolysis and the peaks can also be assigned to the transition of NH2(2AI-2B1).
Since the rotational motion of nascent NH(aA) produced in the uv photolysis is not highly excited at any wavele_ngth, 25 there are two possible sources to produce the high vibrational levels of NHz(AZA) in the reaction.One is a vibrational energy and the other is a translation energy of NH(aA).There have been many works on the energy partitioning in the NH(aA) produced in the HN3 photolysis, , 26-31 whereas the vibrational distribution of nascent NH(aIA, v) is still a matter of controversy.Nelson and McDonald 26 have reported the energy partition in the HN photolysis at 266 nm.According to their summary, the vibrational distribution of nascent NH(aA, v) is 1.0/(1.1 + 0.3)/(0.8+ 0.3)/(0.9_+ 0.5) for v 0/1/2/3.Hawley et al.1, however, have recently reported a totally different distribution: 1.0/0.3/0.02 for v 0/1/2.These data were obtained in the same laboratory; the reason for the discrepancy is unclear.The 193 nm photolysis has not been concluded, either.Misra and Dagdigian 29 reported the vibrational distribution to be 1.0/(0.66+ 0.05)/(0.09+ 0.05) for v 0/1/2.Bohn and Stuhl, 3 on the other hand, reported more highly excitated distribu- tion: 1.0/(0.48+ 0.11)/(0.21+ 0.07)/(0.13+ 0.05) for v 0/1/2/3.
A translational energy of NH(aA), on the other hand, has been determined by Nelson and McDonald 26 to be 6000-7000cmfrom Doppler-broadening in the 266 nm photolysis.Rohrer and Stuhl 2v have reported that the nascent NH(aA) has a translational energy of 26000 cm in the 193 nm photolysis.
The emission spectra observed in the 266 nm and 193 nm photolysis are almost the same in the present study.If translational energy is effective to produce highly excited NHz(/ZA), there must be a difference in the spectra owing to the large difference in the translational energy of NH(aA) colliding with HN.Thus we ten- tatively conclude that the vibrational energy in the NH (aA) is more effective to produce highly excited NHz(/ZA) than the translational energy.
Time resolved emission intensity should be observed to obtain information on the difference in kinetics.

Kinetic Study
In the photolysis at 266 nrn, a laser fluence dependence and temporal profiles of dispersed emission were measured at 520, 600 and 660 nm, where the emissions were due mainly to the transitions from NHz(/ZA) V2 t-" 12, 9, and 7 to 2B V2 0, respectively.The emission intensities are linearly dependent on the laser fluence irrespective of the observation wavelengths.Linear dependence does not necessarily correspond to a one-photon process because two-photon process shows the first-order dependence if the first excitation is not effective and there is no process competing with the second absorption.This is not the case because HN3 on the upper potential undergoes fast predissociation. 2 Consequently, the dependence observed shows that an only photon is associated with the production of single NHz(/ZA) molecule.
In the photolysis at 193 nm, the emission of NH2(/2A-2B) was observed at 520, 600 and 630 nm, where the emission due to the transitions from the v2' 12, 9, and 8 levels of NH2(,2A) state.Typical profiles observed are shown in Figure 3.The emission intensity also varied linearly with laser fluence as the same as that observed in the 266 nm photolysis.
The reaction mechanism established in our previous work is as follows: I HN3 + hv(uv) --NH(a) + N2 (1) NH(aA) + HN ---) NH2(A2A1) + N (2a) NH2(A_2A) + HN3 produ_cts (4) NH2(AA) ---) NH2(X2B) + hv' (visible) (5) where the "products" denotes all the possible species produced in quenching and/or reactions.It should be noted that a faster process always governs the rise and a slower one relates to the fall (Fig. 3).In the present experiment, the rise must be related to deactivation of NH2(A2A) because of its extraordinary large quenching rate constant, 32'33 for example, 1.45 x 10cm molecule -1 s -even by helium and (a) Time / us Figure 3 Time resolved emission of NH2(/2A-(2Bt) in the HN3/193 nm photolysis observed at 520 nm (a) and at 630 nm (b).Pressure of HN was 30 mTorr with no buffer gas.The dots denote the observed data and the solid line is the results of simulation.
The absorption cross-sections of HN3 at 266 nm and at 193 nm are 7.4 10 -20 cm and 1.7 10 TM cm2, respectively. 2From the laser fluences and the sizes of the laser beams, the densities of photons at 266 nm and 193 nm were estimated to be 3.6 10 5 and 1.3 10 5 photons cm-2.The ratios of the initial concentration of NH(aA) to that of HN3 are about to be 2.0 10 -4 at 266 nm photolysis and 2.1 10 -2 in the 193 nm photolysis with the quantum yield for the formation of NH(aA). 35herefore, the pseudo-first-order reaction conditions are always satisfied in both the reactions (3) and ( 5), thus there is little change in the concentration of HN3 during the reaction.
Table 1 The overall rate constants for the re_action NH(aA) + HN determined by the analysis of temporal profiles of the emission from NH2(A2A).The rate constants are in units of 10 cm molecule --.The errors denote 20-.

Photolysis wavelength
Observed wavelength The dependence of the first-order decay rate on the pressure of buffer gas (He and Xe) was measured in the study on the 266 nm photolysis.The temporal profiles of the emission were observed through a cut-off filter (Toshiba R-61) which is transparent for the light longer than 610 nm.Thus, the emission observed was mainly due to the transition from NH(/2A, v'<8).Since Xe is known to be a fast quencher of NH(aA) (1.28 x 10cm molecule -S-I), 29'37-39 the apparent first-order decay rate should increase under the presence of Xe.On the other hand, He has much less effective quencher (<1 10 -5 cm molecules-), and can hardly affect the decay rate.Figure 4 shows the dependence of the first-order decay rate on the pressure of the buffer gas.The first-order decay rate is affected by Xe but little by He.  the dependence was measured with low time resolution (1 us) under the static con- ditions, the data shows a qualitative result.Nevertheless, the difference between Xe and He also supports the mechanism that the NH2(A2A) is produced by the reaction between NH(alA) and HN.
The overall rate constants obtained for the reaction NH(alA) + HN3 apparently depend on the photolysis wavelength.The rate constants in the 193 nm photolysis are smaller than those in the 266 nm photolysis.Since no buffer gas was used in the present study, the translational energy of NH(alA) is not relaxed until the first collision with HN3.Therefore, collision energies for the NH(aA) + HN reaction are very different between the 266 nm and the 193 nm photolysis: translational energy of the nascent NH(alA) is 6000-7000 cm -1 26 and 26000 cm -1 27 in the 266 nm and 193 nm photolysis, respectively.The overall rate constant of the reaction NH(aZX) + HN determined by NH(cll-I-alA) LIF was independent of temperature, 4 hence the reaction proceeds with no barrier.However, the apparent decrease in the rate con- stants was shown in the present study with an increase in the translational energy.Note that buffer gas of about 10-20 Torr was added in the LIF experiment and that no buffer gas was used in the present experiment.Since the NH(aA) has a large translational energy without buffer gas, the centrifugal potential barrier on the reac- tion coordinate must be taken into account.
Relation between reaction rate constant and collision energy is expressed by the next equation [Eq.(A-5) in Appendix]: 4 k ET (s-4)/2s, where ET is a collision energy of the system and s is a parameter in the function of the potential (V(R) 1/Rs).The overall rate constants in the present study decrease with an increase in the translational energy (Table 1); as a result, the parameter s for the potential energy at long distance must be less than 4. Potential energies whose radial distributions are consistent with our result (2 < s < 4) are those for charge- quadrupole (s 3) and dipole-dipole interactions (s 3) in a specific intermolecular orientation.No electrostatic interactions have the dependence, 2 < s < 4, when being averaged over possible orientation under Boltzmann distribution.Since there is no charged reactants, the NH(aA) + HN3 reaction system is governed mainly by dipole- dipole interaction at a specific orientation of the molecules.Hydrogen azide HN3 has relatively large dipole moment: , 0.84 and b 1.48 Debey, where u, and '/b denote the dipole moments along principle axes of inertia.Although the dipole mo- ment of NH(aA) has not been reported, the ground state NH(X3Z-) has u 0.98 Debye.Accordingly, it can be concluded that the interaction potential between NH(alA) and HN3 is mainly governed by the dipole-dipole interaction.
The overall rate constants for the reaction NH(aA) + HN3 depends not only on the photolysis wavelength but also on the observation wavelength.As seen in Table 1, the smaller rate constants are obtained at longer observation wavelengths.The difference is more apparent in the 193 nm photolysis than that in the 266 nm photolysis.In other words, a large translational energy suppresses the production of lower vibrational states of NH2(/ZAI).There has been no information as to whether the reaction goes through a complex mechanism or a direct mechanism in terms of the reaction dynamics, it is reasonable that the period of interaction is short in the collisions with high translational energy.If the structure of H-N-H cannot relax to quasi-linear configuration during the interaction time, NHz(/ZA) with low vibrational energy is not likely to be produced.(The equilibrium structure is quasi-linear (ZHNH 144 in the AZA state and bent in the XZB state.) 4: Further experiments are needed into the problem, for example, the decay of NH(aA) produced by the photolysis without buffer gas must be directly measured.

CONCLUSION
The visible emission folio_wing the ultraviolet photolysis of HN3 o_riginates from the vibronically excited NHz(A2A).All the vibrational levels of NHz(A2A) are produced in the reaction of NH(aA) with HN3, and highly vibrationally excited NHz(AZA) is produced from vibrationally excited NH(aA).From the measurement of the photolysis wavelength dependence of the overall rate constants for the reaction NH(aA) + HN3, a long-range potential between the reactants was determined to be the form: (2 < s < 4, C" constant) The value of s indicates that the potential between reactants at long distance is mainly governed by the dipole-dipole interaction.The overall rate constants for the reaction

Figure 2
Figure 2 Dispersed emission spectrum observed in the HN3/193 nm photolysis.The pressure of HN Since

Figure 4
Figure4Dependence of the first-order decay rate (kdecay) on the buffer gas pressure (PM)' Pressure of HN was 20 mTorr.The time resolution was 1/s.Closed circle denotes the results of He and open circle O denotes the effect of Xe.