The Production of IF ( B 3 I [ ) in the 248 nm Laser Photolysis of Fluorine / Alkyl Iodide Mixtures

Sustained visible emission in the region 440–850 nm from the B → X system of IF is observed when a gas phase mixture (~0.5 mbar) of an alkyl iodide with F2 in He is photolysed at 248 nm by a KrF laser. The total intensity and decay rate of the IF(B) emission is a strong function of the identity of the alkyl iodide and can be correlated with the 248 nm photon yields for the production of I*(2p1/2). The half-lives for the IF(B) decays range from 5 μs for t-C4H9I to 770 μs for n-C3F7I. Decay curves for the I*(2p1/2) concentrations are also measured by atomic fluorescence. The mechanism for IF(B) formation in these systems is discussed and it is suggested that IF(B) is produced either by a recombination process involving I* and F atoms or by multistep collisional excitation of ground state IF(X) by I*. The F atoms can be produced following the photolysis pulse by the reaction of the alkyl radical with molecular fluorine and IF(X) can result either from the reaction of F atoms with the alkyl iodide or from a dark reaction between molecular fluorine and the alkyl iodide.


INTRODUCTION
The search for a visible chemical laser system has recently been concentrated on the diatomic interhalogen molecules and, in particu- " Present address: Chemistry Department, Kansas State University, Willard Hall, Manhattan, Kansas 66506, USA.lar IF.Clyne and McDermid 1'2 suggested in 1977 that an electronic transition laser could be made to operate on the B(31"I0+) X(l+) system of IF and Davis et al. 3,4 have demonstrated pulsed and CW laser operation on the B -X system using optical pumping of ground state IF produced by the low pressure gas phase reaction of F2 with 12.It has been suggested that it should be possible to construct a true chemical laser in which excited IF(B) is produced directly by a chemical reaction or by other means provided that a sufficiently high concentration of IF(B) (>1 x 10 -3 mbar) can be achieved. 4Various gas phase chemilu- minescent reactions and energy transfer processes involving chemically-produced excited metastable species have been shown to produce IF(B), but as yet none of these schemes has generated a sufficient density of IF(B) to achieve lasing. 5n this paper, we report the generation of IF(B) by the 248 nm photolysis of a low-pressure gas phase mixture of F2, He and an alkyl iodide (n-C3F7I, C2F5I, CF3I, CH3I, C2H5I or t-C4H9I).A preliminary account of this work for the system CF3I/F2/He has already been published.6 We find that IF(B) is produced in differing yield over a period varying between 5 #s and 770/s following the photolysis laser pulse, depending on the identity of the alkyl iodide.IF(B) has also been observed 7 following the 248 nm laser photolysis of a mixture of 03, 02 and C3F7I.The time dependence of the IF(B) emission following the laser pulse showed a rise time of 20/s and a decay time of 100/s.The IF(B) was found to be formed with a vibrational temperature of 1800 + 200 K and a rotational temperature of 800 + 50 K.

EXPERIMENTAL
A block diagram of the apparatus used in these experiments is shown in Figure 1.The flow cell and chemiluminescence detection system have been described in detail elsewhere. 9Essentially, a mixture of an alkyl iodide and 5% F2 in He was flowed through a blackened aluminium cell.Typically, the experiments were performed at pressures of---0.51mbar with relative contributions of 4% F2, 78% He and 18% alkyl iodide.The 248 nm output of a KrF excimer laser l " i '1 0 / / (Lambda Physik EMG 103 MSC, 60 Hz, 57 mJ cm-2) was passed through the cell and the resulting ultraviolet and visible emission in the spectral region 200-900 nm was observed at fight angles to the laser axis by a 0.5 m monochromator equipped with a cooled photomulti- plier tube.The signal from the photomultiplier was recorded using an electrometer (Keithley 640) or a boxcar integrator (Brookdeal 9415/ 9425).In addition to measuring the chemiluminescence spectrum, the time dependence of the emission following the laser pulse was recorded for selected spectral features using either the boxcar integra- tor or a transient digitiser (Biodata, 1/s resolution).
Electronically excited iodine atoms, I*(2p1/2) were detected by time resolved atomic fluorescence using a microwave-powered atomic iodine lamp. 1 The lamp was a sealed quartz tube filled with argon to --1 mbar into which iodine was sublimed.The 206.2 nm output of the lamp was selected by an interference filter centred at 206.4 nm (width 13 nm, transmission 14%).The resulting fluorescence, principally at 178.3 nm, was detected at right angles to both the axes of the atomic lamp and the excimer laser by a solar blind photomultiplier (EMI 9413B) in a N2 purged housing.The time-resolved signal was recorded by the transient digitiser and averaged over 103-104 laser shots using a microcomputer.
CFaI and CEFsI (Fluorochem) and the 5% F2 in He mixture (B.O.C.Special Gases) were used without further purification.CHaI (Fisons), CEHsI (B.D.H.), t-C4H9I (Fluka) and n-CaF7I (Hoescht)were purified by repeated freeze-pump-thaw cycles and were stored under vacuum in darkened glass ampoules over a mixture of a 5/ xeolite molecular sieve and copper turnings.This procedure proved effective in remov- ing any free iodine contamination.The iodides were run either from an ice or room-temperature bath.Particular attention was paid to ensur- ing that there were no leaks in the connections to the cell or in the vacuum system to totally eliminate the presence of molecular oxygen that could be efficiently converted to O2(1A) in collisions with 1"(2p1/2).11'12 O2(1A) is known 13-16 to effectively enhance the yield of IF(B) in fluorine/iodide systems.

RESULTS
The emission produced by the different photolyses was quite clearly visible by eye, and the colour varied between purple (for t-C4H9I) and yellow-green (for n-C3F71).The recorded spectra were found to result from two emitters.In the ultraviolet region of the spectrum, from 248 to 480 nm, we observe the dispersed fluorescence from CF2(, 1B 1-- 1A1) following excitation of ground state CF2 to its/k state by the 248 nm KrF laser line.The mechanism of CF2 formation in these systems and its spectroscopy have been described elsewhere. 17The intensity of CF2 emission is greatest for t-C4H9I and CH3I and there is no CF2 observed in the case of n-C3FTI.In the visible region of the spectrum, from 440 nm to beyond 800 nm, emission is observed from the B --X system of IF as is shown in Figure 2 for the case of C2FsI/F2/He.The distribution shows population of v' levels up to 7. Using the procedures described previously, 9 we have obtained rotational and vibrational population distributions for the IF(B) product.(In the case of those systems with intense CF2 emission, the high (v',0) bands of the IF B X emission are overlapped by CF2 lines and only limited determination of the IF(B) vibrational state distribution can be achieved.)The rotational temperatures all lie in the range 325 + 50 K, and all the IF(B) vibrational population distributions were Boltzmann in form and could be characterised by temperatures in the range 700-800 K.
Representative vibrational population distributions are shown in Figure 3 for CF31 and C2F5I.
Time resolved decay curves for the IF(B) and I*(2p/2) concentra- tions are shown in Figures 4 and 5, respectively, and the values for the time taken for the concentrations to fall to half of their peak values are given in Table I.It should be emphasised that these values are not logarithmic half-lives as there is evidence that the shorter-lived dis- tributions for IF(B) in particular are not exponential.The IF(B) values range from 5 to 770 #s, which should be compared with the collision- free lifetime for IF(B) of 7 #s. 2 In all cases, the corresponding decay times for the I*(2px/2) concentration are considerably larger.(The radiative lifetime of I* is 125 ms TM and the residence time for I* in our system is about 5 ms.) The total integrated intensities of the IF(B) emission vary over five orders of magnitude (see Table II) and follow the increasing trend in the IF(B) half-life.The dependence of the IF(B) emission intensity upon laser intensity was determined for the CF31 and CH3I systems by monitoring the (0,4) band whilst varying the laser pulse energy.For CF3I, the emission intensity was found to vary as the laser intensity raised to the power 1.6, whilst for CH3I the dependence was 1.7.l*(ZP1/2) to decay to half their peak values, following 248 nm photolysis of the iodide with F2 in He.Insufficient density prevented a measure- ment of I* in the case of t-CaH9I.DISCUSSION IF(B) is observed when all of the iodides are photolysed at 248 nm in the presence of molecular fluorine.However, the total intensity and duration of the emission is strongly dependent on the identity of the iodide involved.A study of the information contained in Tables I and II shows that there is a strong correlation between both the duration and the integrated intensity of the IF(B) emission and the quantum yield, *, for the production of I* (2p1/2) from the 248 nm photolysis of the iodide.A similar but less well defined correlation also exists between the intensity and duration of the IF(B) emission and the rate constant, kq, for quenching of I* (2p/2) by the various iodides.(There is an increase of about three orders of magnitude in the values of kq in going from the fluoroto the hydrogenated alkyl iodides as the excitation energy of I* is better matched by the large C-H vibrational frequencies.)The 248 nm absorption cross sections for the iodides all lie in the same range [(2.60-9.43)x 10 -19 cm2molecule-1] 21,28,29 so that the initial I*(2p1/2) concentrations do not vary significantly in the various mixtures.It is not unreasonable to suggest that I* is a key reagent in the production of IF(B) in these photolyses.A comparison of the time profiles for the I* and IF(B) concentrations (Table I and Figures 4 and 5) shows that the decay of I* is always much longer than the corresponding IF(B) decay.We see that I* is always in excess and thus deduce that some other species in addition to I* is responsible for creating IF(B) and that it is the rate of production of this species that determines the IF(B) decay times.
In our preliminary account of the results for the system CF3I/F2/He,6 we suggested that IF(B) was formed by a process that was first order in both I* and F atoms (possibly a recombination reaction) and had an effective bimolecular rate constant of-8.0 x 10 -12 cm 3 molecule -1 s-1.We presented the results of a kinetic model that fitted both the I* and IF decay curves.This model assumed that the initial photolysis step in which the I* atoms were created RI + hv--R + I* (1) was followed by a reaction of the radical with F2 to yield an F atom R + F2--) RF + F (2) and then the recombination of the I* and F atoms to give IF(B) Despite being highly exoergic, reaction (2) has a high activation energy which can be overcome by the internal and kinetic energy supplied to the radical in the photolysis step (reaction (1)).This reaction is then likely to control the rate at which IF(B) is produced as it determines the supply of F atoms.The major decay channels for I* that were also taken into account included quenching by F2, IF and RI. 6 Fatoms were also removed by the fast reaction F + RI --IF(X) + R (4) and regenerated by the slow reaction I* + F2--) IF(X) + F. ( Whilst there appears to be no kinetic data for reaction (2) involving the radicals of this study, there is data for the corresponding reactions with Br2 in the case of n-C3F7, C2F5 and CF3 30 and with C12 in the case of cn3, C2H5 and t-C4H9. 31If as suggested above, it is the supply of F atoms via reaction (2) that determines the IF(B) decay time, then we would expect the fastest rate to be associated with the shortest period for IF(B) emission.For the family of reactions R + C12, the relative rate constants are in the ratio 1:5:11 for R cn3, C2H5 and t-C4H9, 3 respectively, and for the reactions R + Br2, the rate constants are in the ratio 1:0.26:0.38 for R CF3, C2F5 and n-C3F7, 3 respectively, showing a correlation of faster R + X2 rate with shorter IF(B) half-life.
These trends give general support to the role played by fluorine atoms and their creation via reaction (2).It should also be noted that a small quantity of fluorine atoms will also be created by the 248 nm photolysis of F2 (---101 atoms cm -3 compared with an initial I* concentration of 5 1013 atoms cm-3).The kinetic model 6 which was successful in describing the time decays of IF(B) and I* for the CF3I/F2/He system was not found to adequately describe the systems involving the other iodides.In general, whilst the I* decays could be modelled, the model always predicted longer IF(B) decays than were measured resulting from an oversupply of fluorine atoms.However, it was not possible to justify the introduction of any additional F atom removal pro- cesses.
Another possible method of IF(B) production that must be con- sidered is that of collisional excitation of IF(X) by some energy rich species, possible I*.Two recent studies have been performed in which IF(B) is produced following laser photolysis of various mixtures in which it has been proposed that IF(B) is created by energy transfer reactions with ground state IF.David et al. 8 photolysed a mixture of CF3I/SF6/N/HN3 at 10.6 m.At this wavelength, the SF6 is dissociated to F atoms which rapidly react with the CF3I to produce IF(X).At an equally fast rate, the fluorine atoms react with HN3 to form azide radicals which are converted by the N atoms into metastable N2 (A2E +u).The metastable nitrogen efficiently pumps IF(X) into IF(B). 32e time profile for IF B X emission extends for several milliseconds following the laser pulse and is identified with the F + HN3 rate which is the first step in the production of metastable N2.
Both O2(1A) and I*(2Pm) have similar excitation energies (7603 and 7882 cm-1, respectively) and it is possible that in our system we can create IF(B) from IF(X) by steps ( 10) and (11), with M* I*.A convenient intermediate IF* state is provided by the recently dis- covered A' (31-I(2)) state which lies at an energy of 13,250 cm-. 34The excitation energy of the I* atom requires that the IF(X) in step (10) is already vibrationallyoexcited (v" > 9).In our system, IF(X) can be generated by the fast reaction (4) using fluorine atoms created by reaction (2).However, reactions of the form of (4) are known 3-37 to produce IF(X) in low vibrational levels with a Boltzmann population distribution decreasing from v" 0, so that an initial collisional excitation step must be included IF(X) + I* --IF(X,v" > 9) + I.

(12)
When we model this mechanism for IF(B) formation using near gas kinetic rates for steps (10)-( 12), we cannot get satisfactory agreement for the IF(B) decay curves if we assume that all the IF(X) is produced via reactions (2) and (4) following the laser pulse.The IF(B) decay curves have too long a grow-in time and overestimate the decay times.
If we allow there to be an initial non-zero IF(X) concentration, then we can obtain satisfactory agreement with the experimental results.It is possible to justify having such an initial IF(X) concentration in two ways.Firstly, the concentration of IF(X) will build up following the laser pulse and it is possible that it has not all been removed by the time of the next pulse.The concentration of IF(X) produced this way would vary considerably in magnitude with the repetition rate of the laser, but it is our observation that the IF(B) decay curves are independent of laser repetition rate.A second method of producing IF(X) would be via a dark reaction between molecular fluorine and the iodide F2 + RI IF(X) + RF (13)  which would produce a small steady state concentration of IF(X).It is possible that there is a combination of the dark reaction providing an initial source of IF(X) and reactions ( 2) and (4) producing further IF(X) following the laser pulse.This might explain the non- exponential IF(B) decay curves that are seen.

CONCLUSIONS
It is clear that the mechanism for IF(B) formation in the 248 nm photolysis of alkyl iodide/fluorine mixtures is far from established.In particular, it is uncertain what the precursors of IF(B) are.It would seem clear that I*(2p1/2) plays a major role and is likely to be a direct progenitor.It is not clear whether the other precursor is an F atom or IF(X) and it is necessary to determine the time-resolved concentra- tions for these key species in order to determine a better model for the mechanism.To continue to improve and develop new methods of producing IF(B), it is important that we should be able to understand and model systems as apparently simple as these in which there are only two reagents; an alkyl iodide and molecular fluorine.Of the various systems for producing IF(B) that we have investigated, 5 this one appears to be the most capable of producing high densities of IF(B).It is possible that laser gain could be achieved by stimulating some of the IF B ---X transitions near 600 nm in these systems.The possibility of constructing such a photolysis-initiated chemical laser is worthy of further study.

Figure 3
Figure 3 Logarithmic plot of the vibrational population distributions for IF(B) pro- duced in the 248 nm photolysis of some RI/F2/He mixtures for R CF3 (e) and C2F5 (I).
The straight lines represent Boltzmann distributions.
is then proposed that IF*(B) is produced by successive energy transfer processes of the form IF(X) + M* --) IF* + M (10)IF* + M* --IF(B) + M(11) David etal. 8ave observed emission from IF(B) lasting for a few milliseconds by photolysing SF6 at 10.6/m with a pulsed CO2 laser to produce F atoms in the presence of HN3, CF3I and N atoms.

Table I Table
of the times taken for the concentrations of IF(B) and TableIIThe relative integrated intensities of the IF(B) emission, Ires, the 248 nm quantum yield for I*(2p1/2) production, *, and the rate constant for quenching of I* by the various iodides, kq.The values of * come from Refs. 19-22and of kq from Refs.