CHANNELS OF PRODUCT FORMATION AND EXCITED MOLECULES RELAXATION AT MULTIPHOTON DISSOCIATION OF CHLORODIFLUOROMETHANE

C2F, molecules have been for the first time directly proved to be generated in the reaction of vibrationally-excited molecules CF2HC1. The contribution of this reaction to CF2HC1 multiphoton dissociation (MPD) products formation was shown to become predominant at the initial CF2HC1 pressures about ten Torr. Also fast relaxation of the highly excited molecules CF2HC1 on the product molecules C2F 4 was discovered.

INTRODUCTION Chlorodifluoromethan (CF2HCI) is known to be one of the most convenient gases for carbon isotope's laser separation. -Under CO2-1aser irradiation CF2HC1 molecules can absorb several quanta and then dissociate: CF2HCI + nhv CFzHCI* -CF 2 -+-HC1. (1) CFz-radicals recombination leads to the product formation: CF 2 + CF 2 + M CzF4 + M. (2) It was found that CFzHCI decomposition rate and selectivity remain quite high up to pressures of several tens of torr. 5-On the basis of mass-spectrometer analysis of carbon isotope's distribution data in C2F4 authors of Ref. 9 supposed that in addition to (2) one more reaction takes part in CzF,-production. Most probably this reaction is a direct interaction between two vibrationally excited molecules of CF2HCI* CFzHCI* + CFzHCI* C2F4 -+-2HC1.
(3) It was shown that selectivity of reaction (3) is high enough and its contribution into the product formation increases with CFzHC1 pressure. 9 Nevertheless, all evidences 29 30 V.A. DIMAND et al.
for involvement of reaction (3) were indirect and it was interesting to find direct confirmation of reaction (3) existence. Since the presence of CF2 radicals in the 1 )st channel and their absence in the (3)rd one is an important difference between channels (1) and (3), it is quite reasonable to measure CF2-radicals concentration after CO2-1aser pulse and C2F4 concentration when all reactions have been completed. If the results would show that the amount of C2F, is greater than half of CF2-radicals amount, then one can consider that the reaction (3) is responsible for this exceeding. This paper is devoted to description of the results of such experiments. We managed to prove the reaction (3) existence and evaluate its contribution to the final products formation at different CF2HC1 pressures and also to study the influence of the reaction products--C2F and HClmaccumulation on the value of multiphoton dissociation yield of CF2HC1 molecules. EXPERIMENT CF2-radicals absorb ultraviolet radiation in the 235-258 nm band ((1B1)-R 1A electron transition).1 o Vibrational levels v 0, v 2 of the upper electron state and v' 0, v 0 of the lower one take part in the absorption near 248.8 nm.
Extinction coefficient in the maximum of absorption band (2 249 13 These data allow one to find approximate values of absorption cross-section for another temperatures. In our experiments KrF-excimer laser radiation was used to probe the gas and determine CF2-radicals concentration (2Kr F --248 nm). Absorption cross-section value for CFEHC1 at this wavelength does not exceed tr 10 -26 cm 2 tRef. 14)that is much more lower than this one for CF 2. As laser pulse duration is 15-20 ns, only radicals on the definite vibrational-rotational level can interact with KrF-laser irradiation. Really, the time between two collisions of CF2HC1 molecule can be evaluated to be not less than 80 ns at CF2HC1 pressure of 1 Torr. Since 5-10 collisions must occur the rotational relaxation to be completed one can consider that under the probing laser pulse duration z 20 ns the molecule remains 'free' for the process of UV quanta absorption up to pressures of  Torr. That is why raise of radicals population on this level owing to vibrational-rotational exchange with the neighbouring levels can be neglected. This leads to a little less absorption of UV radiation than in the case when equilibrium population of lower level is maintained (at probing of the gas with continuous wave light, for example), but, on the other hand, this makes our method insensitive to gas pressure changes.
The apparatus scheme for the measurements of CF2-radicals concentration is presented in Figure 1. TEA CO E-laser beam  11, passes along the axis of the cell 8. The stainless steel cell is used, its internal diameter is 3 cm and length is 10 cm; windows are made from BaF2. Diameter of CO2-1aser beam at the entrance and at the exit of the cell is 1.2 cm. Part of the laser beam, passing through the cell, is declined by the beam splitter 6 to the entrance of calorimeter 7--it p/rmits to measure irradiation energy which passes through the cell. To measure the CO2-1aser energy at the entrance of the cell we use one more calorimeter to receive the beam declined by the beam splitter 10. Probing UV-laser 4 beam is directed by mirror 3 along the cell axis towards CO2-1aser beam. Diameter of UV-laser beam at the entrance and at the exit of the cell is 0.3 cm. Part of the beam is directed to the photodetector 2 by plate 6 and mirror 5. After passing the cell part of the beam is reflected by plate 10 and mirror 9 to photodetector 1. The photodiodes PD21-KP were used as the detectors of UV-irradiation. These photodiodes are weakly sensitive to 248 nm wavelength irradiation, therefore the cells with rhodamine 6G solution in isopropanol were placed in the front of each of them. Thickness of these cells was 2 mm and diameter was 8 mm. Solution of the rhodamine was hermetically sealed by polypropylene films of 40 #m thickness. To avoid electromagnetic hindrances from electrical discharge of the lasers, we refused from wide band amplifier of the photodiode signal and used two-staged amplifier with amplification factor 103 with internal time constant Zpo 2.10-5 s. The shape of the signal at the exit of such amplifier is defined by time parameters of the amplifying path, and the amplitude of the signal with accuracy of zl/'PO 10-3 (,1 is duration of UV-laser pulse) is proportional to the integral of the probing beam intensity. 15 This proportionality remains constant while the voltage of electric signal from photodiode is much less than characteristic voltage of its internal electric field; just for the photodiode that operates in photovoltage regime this voltage of p-n transition is about 0.6 V. In our case the maximum photodiode signal amplitude was mV, therefore one can consider that our measurements were 32 V.A. DIMAND et al.
performed in the linear regime. With the aid of pulse generator G5-60 and specially made synchronization of excimer and CO2-1asers scheme delay between UV and IR-pulses could be varied from 0 to 30 #s. The measurements of the amplitudes of amplified signals were carried out with the help of oscilloscope C9-8, which was interfaced by the adapter for general use to IBM PC AT. The program was made so that after each pulse the measured data from both photoreceptors were delivered to PC AT, where they were treated and stored. The relative absorption of the probing beam was determined by the expression: where I, Io are the intensities of the probing and calibrating beams, respectively, U1, U2 are the amplitudes of the signals from the photoreceptors and 2, k is the setting factor which was determined from the condition that I/Io 1 when there are no particles in the cell which can absorb KrF-laser irradiation.
To check laser spectrometer operation we filled optical gas cell with SO2. It was found that up to 30torr transmitted light intensity attenuation followed the Lambert-Beer law, and absorption cross-section of SO2 at 2 248 nm turned out to be (7.5 0.3 ). 10-20 cm2, that is near to data of Ref.
14. Under CO2-1aser influence (simultaneously at lines 9R (32)-9R (38)) on CF2HC1 absorption at 2 248 nm was found. We attributed it to CF2 radicals appearance in the irradiated gas volume. Weakening of UV signal increased with the increase ofa delay between CO2 and KrF-lasers pulses from 0 up to 10 #s, then (up to 30/s it remained constant. Radicals CF2 concentration growth after CO2-1aser pulse finishing is due to after-pulse decomposition of CF2HC1 vibrationally-overexcited molecules. Within this time range one can disregard the influence of recombination (2) on CF2 decreasing (characteristic time of this reaction at the initial CF2 concentration of 1015 cm-3 is about 30 ms).16 The calibration of laser spectrometer ,was done by the following way. The cell was filled with the mixture CF2HCI'Ar 20" at the pressure of 2.1 Torr to be irradiated by 11 pulses of CO2-1aser with energy density of 1.26 J/cm 2. The attenuation of the probing UV pulse, delayed a time 9.6/s with respect to the beginning of CO2-1aser pulse, for the first pulse of a series makes up I/I o 0.97 0.02. With the help of mass-spectrometer analysis of the content of the irradiated gas the yield of CF2HCl-dissociation per pulse was found. The dissociation yield was determined with the expression" 17 m/e 40 of argon (argon was used in our experiments as a bench-mark) before and after irradiation of the mixture, n is a number of pulses (n 11 ), a is the fraction of the cell volume illuminated by the CO2-1aser (a 0.18). While substituting to the expression (5) the measured value Nn/No 0.97 0.02, one can find fl 0.014 + 0.0003, but it also means that CF2-radicals concentration in the irradiated volume is equal nCF fl'nCF2Ho 1015 cm -3. The cross section for absorption of UV laser radiation by CF2-radicals is a= (ln(Io/I))/(ncF2.L) 3.10 -18 cm 2 (here Io/I is the deviation of UV laser intensities at the entrance and at the exit of the optical cell, which length is 10 cm). The obtained value of r is approximately 3 times lower than the value found in Ref. 18 at the MPD of CFzHC1 investigation where the vibrational temperature of radicals was considered to be 1160 K. This discrepancy of the results more obviously is caused by the difference in experimental conditions:in Ref. 18 the data were obtained with probing the cell with the continuous wave light, when the molecules of a few vibrational-rotational levels, which are disposed not far from each other, can interact with the beam, while in our case, when the probing is produced by a short pulse of light, only the molecules of the single vibrational-rotational level may take part in absorption.

RESULTS AND DISCUSSION
In Figure 2 the experimental results on concentration of C2F4 produced in the irradiated volume (squares) and concentration of CFz-radicals in the same volume (crosses) dependences on CFzHCl-pressure are shown. For more convenient comparing of the results the scale of ordinate for CzF4 molecules is made two times greater than the scale for CF/. CFz-radicals concentration was measured according to the technique described in the previous part. CzF concentration after n laser pulses was determined with the help of mass-spectrometer analysis by the diminishing of CFzHC1 concentration: [C2F4](n i" (N O N,).
It can be seen from Figure 2 that in the range of CFzHC1 pressures from 0 to 4 Torr CzF4 concentration is two times lower than CF2 concentration in the limits of the experimental error. Since two radicals CF 2 can form only one molecule C2F4 and there is no other ways of CF2 radicals elimination from the system, this result means that in this range of CF2HCl-pressures CF2-radicals recombination is the single source of CzF, formation. When pressure is higher than 4 Torr the quantity of C2F4 formed becomes more greater than the quantity which may be formed in the process of CF2 recombination. It means that when increasing CFzHC1 pressure one more channel of CzF4 formation appears and this channel is more probably the reaction (3). The contribution of this second channel increases with CFzHC1 pressure increase; so at 20 Torr the contribution of the second reaction over-exceeds the contribution of the first one approximately in 1.5 times.
In the range of CF2HC1 pressures lower than 8 Torr the increase of CF 2 radicals concentration was observed, although at the following increasing of gas pressure in the cell radical concentration practically did not change. If one can suggest, that the value of dissociation yield does not depend upon initial gas pressure, the linear growth of radical concentration vs gas pressure should take place. The observed deviation from linear dependence can be explained if non-uniformity of CO z-laser power density along the cell is taken into account, higher is the gas pressure, more strongly does this non-uniformity reveal itself.
Let us consider that CO2-1aser power density changes along the cell according to the expression: @ @o .exp O'co Ns. 1), (8) where 0 is CO z-laser power density at the entrance of the cell, trco,_ is cross-section for absorption of CO2-1aser beam by chlorodifluoromethane molecules, NM is CFzHC1 concentration, is distance from the entrance to the cell. It was found that within the experimental error this law is fulfilled quite perfectly. For dependence of dissociation yield on CO z-laser power density one can use the expression fl tI)m, 17 where m is a constant. So one can consider that dissociation yield would change along the cell according to the expression" /3(1) flo.exp(-m-O'co.Ns.l), where flo is dissociation yield value at the entrance of the cell at power density tI) o. Since CFz-radicals concentration is proportional to dissociation yield fl, its change is described by the expression analogous to (9): n(1) no.exp(-m.aco.N.l), where n(1) is CF 2 radicals concentration at the distance from the entrance of the cell, and no n(0). While treating the experimental results this dependence was not taken into account, and CF2 radicals concentration was considered to be constant along all the cell. So we determined the CFz concentration averaged along the axis of the cell nefe, which is connected with no by the following expression" nff no" 1 (@L/@o )m 11 m-ln(@o/@L) where (I) L is CO2-1aser power density at the exit of the cell, or in a different way: neff no" exp m .trco N M L)). (12) m" aco2 NM. L Using the above mentioned suggestion about the independence of dissociation yield on gas pressure, i.e. no flo" N, one can see that while enhancing pressure lim neee= (13) m trco L N the value of the measured CF2 concentration should approach a constant, and that was experimentally observed. The experimental dependence may be approximated by the function y A. exp B. X)) and one can find with the help of the least square method the values of the coefficients A and B. Using the expressions (12) and (13) it is easy to get the values m 1.03 and flo 0.024. The curve with these parameters is shown in Figure 2. The main feature of the dependence of the relation of CzF4 concentration to CFz concentration upon CF2HC1 pressure is the growth of this relation in all investigated range of pressures. The growth of this relation at low pressures may be explained, if one takes into account that the reaction of the excited molecules is the bimolecular reaction. Product yield of this reaction is proportional to the second power of excited molecules concentration: Pr(l) n*(1) 2 n2.exp(-2.m.aco.N.l), (14) and if one would make a supposition about proportionality of excited molecules concentration to the initial concentration of chlorodifluoromethane Nu, the expression for Pr(1) would be: Pr(1) 7"NZM'exp(--2"m'aco'NM'I)), (15) where is a coefficient.  Figure 2; one can see that the function describes the experimental results quite well. It is also seen by the character of the dependence that at sufficiently high pressures the relation IANulms/neff(vv) should be described by the function y + 9. Nu. Such type of dependence was really observed in our experiments.
At high pressures of CF2HC1 one can .see the slight decrease of the experimental value of CF2 concentration as compared to the calculated curve. This decrease more probably can be caused by the violation of an assumption about constancy of dissociation yield/30 while CF2HC1 pressure changes.
The plot of the relation of the probing beam energy absorbed in the cell to the energy of the incident beam vs number of CO2-1aser pulse at the initial CF2HC1 pressure of 20 Torr is presented in Figure 3. One can see that the initial exponential fall in (Io I)/Io(n) is replaced after n 140 by the plateau, and when n > 200 the UV beam passes the cell without losses in its intensity. The assumption that all the molecules CF2HC1 had already been dissociated up to this moment is incorrect: the mass-spectrometer analysis of the irradiated mixture revealed that after the ,200th laser pulse CF2HC1 content in the gas is equal to 37% of its initial value. The same character has the dependence of the relation of the CO2-1aser energy absorbed in the cell to the energy at the entrance of the cell (curve 2): when n > 120 the absorption of IR emission does not change. Those dependencies may be caused by the n pulse number Figure 3 Plot of (Io I)//o-fraction of UV-beam energy (2 248 nm) absorbed in the cell (curve and of relation of CO2-1aser energy absorbed in the cell to the energy at the entrance of the cell (curve 2) vs CO2-1aser pulse number. Initial CF2HC1 pressure 20 torr. Curve 3, the dependence of Ilgl (lg(lo/1))ll upon the number of the pulse. Io and I, the energy of UV probing beam respectively at the entrance and at the exit of the cell. accumulation of some substance in the cell as a number of pulses rises, which gives rise to fast relaxation of the excited molecules CF2HC1. As the concentration of this substance increases, the moment is reached when both dissociation of CF2HC1 molecules and reaction of C2F, formation at the collisions of vibrationally excited molecules CF2HC1 are ceased. As a result, firstly, CF2-radicals could not be detected in the gas after CO2-1aser pulse, and, secondly, as CF2HC1 remains unchanged so does the part of IR irradiation, absorbed by the mixture. Besides this, there is one practical conclusion following the experimental results: in order to obtain reliable data gas should be irradiated by the limited number of CO2-1aser pulses. The permissible number of pulses one can find from the condition of conservation of the linearity of the dependence Ilgl (lg(I/Io))II (n). Figure 3 demonstrates that this linearity has been kept up to the 60th pulse.
Only two substances are accumulated in the cell; C2F4 and HC1. To answer the question how does each of them influence the relaxation of the excited molecules CF2HC1 we carried out the experiments with model mixtures. The model mixtures were prepared on the basis of the ratio of gaseous components generated in the cell after its irradiation by 200 pulses of CO2-1aser. The values of CF2HC1, C2F4 and HC1 pressures after the 200th pulse at various initial values of CF2HC1 pressure, determined using mass-spectrometer analysis, are presented in Table 1. When CF2HC1 relaxation on C2F4 was investigated the model mixture was prepared from these two gases, their partial pressures being chosen equal to those which were observed in the cell irradiated by 200 CO2-1aser pulses. Figure 4 shows how does the portion of the absorbed in the cell IR beam energy (I)abs./(I) 0 (where (I)abs. (I) 0 (I)L) change when irradiating the real mixture of gases CF2HC1 CEF4 and HC1, obtained by the exposing of the neat CFEHC1 to 200 CO2-1aser pulses, the same dependence for the model mixture (consisting only from CFEHC1 and C2F4) also being represented. One can consider that in the first approximation the curves coincide. It means that CEF4 is the main relaxant of excited molecules CF2HC1. A little more higher absorption of IR beam by the real mixture as compared to the model mixture can be explained by the additional presence of HC1. It is interesting that the energy absorption in gas does not lead to CFEHC1 dissociation:in the whole range of investigated pressures both for real and model mixtures the absorption at 2 248 nm was not observed.
C2F formation owing to direct interaction of excited molecules CFEHCI was not also observed:mass-spectrometer analyses showed that starting with the 200th pulse CF2HC1 concentration remained unchanged.
The experiments in which the influence of HC1 on CF2HC1 MPD yield was investigated were carried out for the model mixture CF2HCI:HC1 CF2iICl-pressuro Tort Figure 4 Plot of fraction of CO2-1aser energy absorbed in the cell vs initial gas pressure" (1) for the 200th pulse to the cell with neat CF2HC1. (2) For the first pulse to the.cell with model mixture CF2HCI + C2F4, prepared according to Table 1. 7.5 Torr'12.5 Torr. The comparison of data of the dependence of UV-beam attenuation on the number of CO2-1aser pulsesfor this mixture and for the neat CF2HC1 points out their full identity (see Figure 5). This means that in our circumstances HC1 practically does not affect the process of CF2HC1 multiphoton dissociation. But at the same time the adding of hydrogen chloride to CF2HC1 increases the amount of IR radiation energy absorbed by the gas (curves 2 and 3 on Figure 5 ). This increase of the portion of the absorbed energy does not lead, however, to the enhancement of the yield of C2F4 formation reaction at CF2HC1 collisions.
Really, the results of the comparison of mass-spectral peaks heights m/e 51 (molecule CF2HC1) N2oo and No, respectively after the irradiation of the gas with 200 pulses and before the irradiation, for the model mixture CF2HC1 + HC1 and for the neat CF2HC1 show that (N200/N0)cF2HCI/(N200/N0)cF2HCI+HCI-" 0.95_ 0.01, i.e. the dissociation yield in the model mixture is even lower than in the neat CF2HC1.
But it means that the additional IR radiation energy absorbed by the gas due to increase of HC1 content is consumed mainly to gas heating, without essential influence on the processes of dissociation and interaction between vibrationally excited molecules CFEHC1. Figure 6 shows how the total CF2HC1 MPD-yield and CF2-radicals yield after the first CO2-1aser pulse change with increasing of C2F4 content in mixture. One can see that the yield of CF2-radicals diminishes in two times at C2F4 pressure near 1.5 Torr. Those   to be equal to (10-6/18) s. While comparing those values one can get an estimation of the rate constant of CFzHC1 relaxation on CzF4"kre 5" 10-lo cm 3 s-1. Thus, the relaxation of vibrationally excited CFzHC1 molecule occurs almost at every collision with CzF4 molecule. This estimate is made under the supposition that CFzHC1 molecule is excited over the dissociation threshold only by 1-2 quanta of CO2-1aser, where the time of the dissociation is less than 10 -7 S. 17 In our experiments the dissociation yield does not exceed a few percent, that is why one can consider that at C2F, pressures higher than 4 Torr each excited CFzHC1 molecule is surrounded by approximately a hundred of "cold" CzF4 molecules. The first stage of CFzHC1 vibrational energy relaxation is the resonance transfer of energy to CzF4 molecules promoted by the nearness of the frequencies of the bands vs (1127 cm -1), v4 + v6 (1214 cm -1) and others of CFzHC1 and of the bands (1186cm -1) C2F4; and also of the bands v2 (1313cm -1) and v7 (1351 cm -1) CFzHC1 and the band Vlo (1337 cm -1) C2F4 .19'z0 Then the vibrationally excited molecules loose their energy due to V-T relaxation. The same relaxation processes, where the presence of the frequency resonances between the excited and quenching molecules plays an important role, were observed for SF 6 multiphoton dissociation in the presence of NH 3 and C 2 H4,17 or for CF 2 HC1 MPD in the presence of CF 3 C1.2 Within the frames of the described mechanism one can also explain a weak action of HC1 on relaxation of excited molecules CFzHCI" the energy of the bands (3023 cm-1 and v + v6 (3435 cm-1 appears to be the most close to the energy of the level v HC1 (2900 cm-1); but because of the significant energy defect vibrational-vibrational exchange in the system CFzHC1 + HC1 would not be fast.

CONCLUSION
For the first time the existence of C2F4 formation at the binary collisions of vibrationally excited molecules CFEHC1 was directly confirmed. The input of this reaction to the products formation at CFEHCI pressures of tens Torr was shown to become predominant. Discovering the fast relaxation of vibrationally excited CF2HC1 molecules on the product's molecules CEF4 due to the resonance energy exchange between those molecules seems to be tte second significant result of this work; the rate constant of this process was estimated to be 5.10-lo cm 3 s-1. The presence of the process of "self-relaxation" restricts the degree of the initial matter conversion to the products at the irradiation ofgas in the limited volume.