455
RADIATION CHEMISTRY OF FLUOROCHLOROMETHANES
The ~ a d ~ a Chemistry i ~ ~ o ~ of Aqueous Solutions of CFCI,, CF,Cl,, and GF,Gl1 airgut I. Balkas, J. H. Fendler, and Robert H. Schuler* 1'Zadiat.h Research Laboratories and Department of Chemistry, Mellon Institute of Science, Carnegie-Mellon University, Pittsburgh, Pennsylvania 16213 and Chemistry Department, Middle East Technical Unizersity, Ankara, Turkey (Received October 39, 1970) Publication costs assisted by the Carnegie-Mellon University and the U. 8. Atomic Energy Commission
Radiation chemical studies of aqueous solutions of the three fluorochloromethanesshow that the initial reaction omirs mainly via the attack of hydrated electrons on the solute to produce chloride ions. A rate constant of ,1010 M-1 see-1 is observed for reaction of hydrated electrons with each of the s-olutes. In the absence of competing reactions the halogenated radical produced as the complement of the chloride ion undergoes secondary hydrolysis reactions. In the case of CFC13,one fluoride and three chloride ions are produced for each electron which reacts with the solute. Hydrolysis is complete except for experiments carried out at very high dose rates or in the presence of very reactive radical scavengers. The chloride ion yields from CF2C12and CF3Clare, respectively, twice and equal to the hydrated electron yield, with the yield from CF&I being notably independent of dose, dose rate, or added radical scavenger. In these cases the fluoride ion yields exhibit 8, very pronounced dose rate dependence and complete hydrolysis occurs only at dose rates less eV g-1 see-1. A common dose rate dependence is observed for these two solutes, and this fact is than taken to indicate that the rate-controlling steps involve the second-order reactions of a difiuorinated radical, probably CF,. The secondary hydrolysis reactions were studied on the 10-6 to low2see time scale using conductometric methods. In each case the second step in the hydrolysis occurs with a period of -10 psec. The third and fourth halide ions are produced in reactions which have periods -1 msec or longer. A number of effects which illustrate the minor importance of complicating side reactions are noted and discussed in some detail.
Radiation chemical studies on aqueous solutions of CH&l have shown that this solute reacts selectively and rapidly with hydrated electrons. One equivalent of chloride ion2 is produced for each electron scavenged by the CH3C1and the reaction responsible is presumably ea,l-
+ ECI -3-R.4- ~
1 -
(1)
We wish to report h f m the results of related studies on aqueous solutions of the three fluorochloromethanes in which large yields of halide ion are observed. I n each case addition of ethylme suppresses the production of all but one chloride ion so that reaction 1 appears to be responsible for the initial attack on the solute. This reaction is then followed by secondary reactions which result, if competing reactions are unimportant, in the complete hydrolysis of the halide. I n all three cases the soiute was examined at a concentration sufficiently low that reaction is expected to be predominantly with electrons which have escaped from the spur, and the yield of reaction L clan be expected to be close to the value of 2.15 observed in 144 CH3Cl solution^.^ Complicating nffefeett of competing secondary reactions a t high dose ra,tes have been explored. These studies are made possible Xby the availability of ion-selective electrodes which permit the ready analysis of both fluoride and chloride ions simultaneously produced. Auxiliary conductometric pulse-radiolysis experiments have been carried out to examine the secondary hydrolysis reactions on the 1 -* to 10-2 sec time scale.
Experimental Section For the steady-state experiments %Intheson CF&l (bp -29.8"), and CFC4 (bp 23.8') (bp -81.4"), @F2CI2 were fractionated on a vacuum line at -- 88' and known amounts (pressure-volume measurement) added to previously outgassed aqueous solutions. The solution concentrations were calculated from the solubility coefficients3 and the known liquid and vapor volumes of the sealed sample tubes. I n the case of CFCl, (a liquid) the amount of added solute was less than the liquid phase solubility limit. For the pulse conductivity studies a flow system was employed, and measurements were made on solutions saturated at atmospheric pressure (saturation concentrations ab 25" ; CFC13, 8 X M , CF2C12,2.3 X n/p, CFaC1, 8.6 X 10-4 1 ~ 9 . 3 The analyses for fluoride and chloride ions m r e by means of ion-selective electrodes using methods completely identical with those employed in previous studies of CH&1 and SFe solutions.2j4 A t the doses required for the chloride determinations suEcient hydrogen ion (-10-4 f44)builds up in the irradiation of a neutral unbuffered solution that it competes significantly with the solute present at low concentration and (1) Supported in part by the U. 8 . Atomic Energy Commission. ( 2 ) T. I. Balkas, J. R.Fendler, and R. H. Schuler, J . Phys. Chem.,
in press. (3) "Matheson Gas Data Book," The Matheson Co., Inc., East Rutherford, N. J., 1966, pp 131, 159, 473. (4) X.-D. Asmus and J. H. Fendler, J . Phys. Chem., 72, 4285 (1968).
The Journal of Physical Chemistry, Vol. 76, No. 4, 1971
456 the use of a buffer is necessary. Most solutions were buffered to $3 6.51with sodium phosphate ( 2 X M Na2NP04acidified slightly with H2S04). The buffer interferes with the use of the fluoride electrode at fluoride Concentrations below .LO-5 M , but satisfactory results could be obtained at doses of l0ls eV/g and greater at which the chloride was determined. I n one series of experimenls base as added to increase the initial pH to --.IO so that the hydrogen ion would be neutralized as it was fornied. 'Within experimental error the results of these experiments were identical with those obtained on the buffered solutions. Irradiations were carried out both with 6oCoy rays and with 2.8-VeV Tian de Graaff electrons. The former covered the dose rate range of l O I 3 to eV g-' sec-l. Absorbed dose ratcs were determined in the y-ray experiments by measurements on the Fricke system. Several experiment s a t moderately higher dose rates were carried out with an electron beam (current -0.1 PA? duration 2-1; sec), and the fluoride yield was mea~ yield. Since dose rate efsured relative to t h chloride fects are very appsrent in the CFzClz and CF&l systems, pulse electron irradiations were used to examine the absolute yields for both fluoride and chloride production at vrty high dose rates. Currents up to 50 mA were used t o produce absorbed dose rates up to 2 X 1 0 2 3 eTT g- 4ec-I~ Doses of 3-5 X loi8eV/g were delivered in single pulses which were monitored by integration of -',hebeam current with a pulse integrator. Dosimetry w:ts in terms of chloride production from Jd; G(Cl-> taken as 2.8) with the observed CN3C1 effect3 interrointed by the electrical measurements. TKOseries of' pulse irradiations were carried out at a beam current of 50 mA and a third series at a current two orders of magnitude Power. Because the irradiation geometry changed somewhat from series to series it was necessary, for comparison purposes, to normalize the integral ed currents by factors determined from the relative effects on methyl chloride solutions. The results can, therefore, be compared only within each series but can in turn be compared to the cesuits on CF3Ciwhere a single chloride is also produced. In general, in spite of significant problems involving detail3 of the jrradizition geometuy, the results within a single series rf runs are reproducible to &3%. Together Tvit'n the W o irradiations, these studies covered a ten order of magnitude variation in absorbed dose rate with easentiallg the same dose being delivered in periods which m n g d from 25 ysec to 5 days. A lmrge number of conductometiic pulse-radiolysis experiments ere carried out which were similar to those described by Beck5 and also previously used in these laboratories in studies of CH3C1solutions.2 Pure water was acidified slightly with perchloric acid t o bring the pM to thr desired value (usually -6). Relative conductivity yields were determined in terms of the experimental unit of mhos conductance per coulomb of The JozsrnaE of Physical C'hemiutry, Val. 76,No. Q, 197'1
T. I. BALKAS, J. H. FENDLER, AND R. electron beam current collected from the conductivity cell (see ref 2). Absolute yields of ionized product were determined by reference to measurements on CHICl solutions saturated at atmospheric pressure on the assumption that in this case the conductivity change results from the production of HCL with a yield of 3.14.2 It is demonstrated here that reaction of hydrated electrons with CFC13, CF,C12, and CF&1 first produces €IC1 so that direct comparison of the initial change with that observed in CH&l solutions is possible. dec-l, this Since the rate constants are -IOio initial reaction must occur in the saturated solutions (-lo+ & on the I ) time scale of ~ 1 0 sec ' ~and an abrupt change corresponding to the production of at least one equivalent of HC1 is expected at the time of irradiation. Because the equivalent conductances of MF and KC1 (395 and 425 mhos/equiv, respectively, at 25') differ bg only 7%, the order in which subsequent reactions occur will have little effect on the interpretation of the results (as long as one operates on the acid side of neutral as was done here). The conductivity experiments carried out during the present study covered the dose range of 2 X to 2 X lo1' eV/g with pulses which were usually of -1-psec duration. The results represented in the duai-trace photographs of Figures 2, 5 , and 6 were all at doses eV/g. At these doses product- concentrations of M are produced. These doses are considonly erably less than in the case of the steady-state experiments so that certain secondary reactions may have a somewhat greater importance. I n particular, hydrogen atoms will be removed by reaction with product oxygen at the higher doses of the steady-state experiments but may start to make a small contribution to the yields at the dose levels at which the conductivitj experiments are carried out.
Results and Discussion Summary of Rate Data. For convenience in the following discussions the rate constants for the reactions of ) hydrated electrons (IC,) and hydrogen atoms ( k ~with the fluorochloromethanes and related compounds are given in Table I. The hydrogen atom data were obtained by measuring the concentration dependence of the decrease in the hydrogen atom signal in esr experiments6 and the rates of the electron reactions with CFC13 and C F d X by competitive studies (present work-see below) and with CF&l by pulse radi~lysis.~ Since the conductivity experiments described below were carried out on saturated solutions, the corresponding yields expected for the initial reaction of electrons with each of the solutes (calculated for the appropriate k,[S]from the relation given in ref 2 ) are given in the final column. (5) G. Beck, Int. J . Radiat. Phys. Chem., 1, 381 (1969). (6) P. Neta, R. W. Fessenden, and R. H. Schuler, t o be published. (7) G. Bullock and R. Cooper, Trans. Faraday SOC.,66, 2055 (1970).
RADIATION CHEMIBTRY OF FLUOROCHLOROMIETHANES _______I________
Table I : Rate I M a and Expected Initial Yields Saturationa oonon,
CC14
CFC1, CFzClz CF361 @F4
C&c1
,-------Rate
oonstants----
M
kHb
0.005 0.008 0.002% 0.0009 0 0002 0.1
4.4 X lo7 1 . 5 X 106 < IOw3
8.35 8.25 3.11 2.11 5.28 3.90 2.21 1.54
2.78 2.80 1.04 0.69 1.72 1.31 0.69 0.50
Solute
N20
a
1.0 X
PI CFC13. Dose rate
=
a(Cl-)/
G(F-)
3.01 2.94 2.99 3.06 3.07 2.9s 3 20 3.08
1.3 X 10l6 eV g-'
soc-1.
readily attack the halocarbon to produce halide ion. Similarly, the data for €3' indicate that at high acid concentrations .cT/herrl all the electrons have been converted into h5drogen atoms perhaps 5% of the Iatter may attack the CBC:&. In ihe absence of acid only a negligible contribution to the yield (-0.05 X 0.6 = 0.03) would result. The rate constant for reaction of hydrogen ailoms with CFC18 is given by esr measuremen@ as 1.7 X IO6 sec-I so that they are presumably removed by more rapid reactions with competing solutes or the peroxide and oxygen which build up t o M' during the course of these irradiations. As is shown in 'Table 111. methanol has no effect on either el- cr I?-- production. At very low doses in extremely pure muter the hydrogen atoms can be expected to attack the CFCI, and give a small additional yield of"halide. The Jouinal of Physical Chemistry, Val. 76,N o . 4, 1.971
From the slopes of standard competition plots of the data of Table I11 the relative rate constants for reaction of hydrated electrons with CFb=18,H+ and NzO are 1 : 1.50:0.56. Taking 2.4 and 0.89 X 1Olo M-lsec-1 as the absolute rate constants for the latter two reactions,O~ C F C Iis, ~ in both cases, 1.6 X Polo M-l sec-1. It should be noted in the table that the ratio of 1 : 3 is maintained for G(F-):G(CI-) in the presence of H+ and N20. Results of experiments at the highest dose rates employed are given in Table IV. It is seen that the chloride and fluoride yields are, respectively, 91 and 86% of those observed at low dose rates. Since radicals seem to be important intermediates, it is indeed expected that for prolonged irradiations a t high dose rates radicalradical reactions will compete with the hydrolysis reactions and a resultant decrease in the observed yield of halide ion should be observed. However, in the studies reported here the irradiations are of io& short duration and the radicals build up only to concentrations of M (depending on whether the combination to reactions are fast or slow). At these levels the radical lifetimes will be a t least a few microseconds. The initial hydrolysis reaction seems to compete eff'ectively since only a small reduction is, in fact, observed. It is noted that if the observed chloride yield k corrected for the one equivalent expected from reaction I the residual (7.62 - 2.78 = 4.84) is 87% of that found at low dose rates and twice the fluoride yield. It appears, therefore, that 13y0of the CFCl2 radicals are lost by radicalradical reactions in these experiments. One additional pulse experiment a t a dose rate of 3 X IOz1eV g-l sec-l gave absolute chloride and fluoride yields of 8.50 and 2.92 and a ratio of 2.91. These latter results are, within experimental error, identical with those st lower dose rates (which is as it should be since from the results a t loz1eV g-1 sec-l, one expects a drop of only one-tenth that at the higher dose rate or 1-2'%)~ Examination of the conductivity change of a saturated solution shows that for each electron scavenged by the CFC13 one equivalent of C1 is formed immediately but that additional ionic species grow in over a period of -5 msec. A typical dual trace oscillogram is shown in Figure 2b and can be compared with the stepwise change in conductivity observed with methyl chloride (Figure 2a). Two regions of growth after the initial incremental change are apparent. First a rapid rise is observed in the region of 10 pxec. This gro-r\ithis essentially over in 50 psec. A longer growth is then observed on the millisecond time scale. This latter increase is exponential in character and has a halfperiod of 1.7 msec which was quite reproducible in a large number of experiments. Using pulses of I-psec duration the conductivity change was measured as a
-
(9) S. Gordon, E. J. Hart, M. S. Matheson, J. Rabani, and 3. K. Thomas, Discussions Faradau Soc., 3 6 , 193 (1963).
RADIATION ~~~~~~T~~
OF!FLUOROCHLOROMETNANES
459
Table IV : Resuits of Pulse Experiments" &-Integrated current,
cx
CXsCl
CFC13 CFzClr
CFsCl
108
0.645 0.505 0. 5SOC 0.640" 0.515 0.693 0.625 0.680 0.750 0 . '750" 0.7700 0.520 0.520 0,650
Chloride produaed, M x 104
ICl-I/Q
1.90 1.42 1.65 1.85 4.10 5.50 3.01 3.50 3.70 3.65 3.40 1.57 1.55 1.80
2.95 2.81. 2.85 2.89 7.96 7.92 4.82 5.13 4.94 4.87 4.42 3.02 2.98 2.77
U(Cl-)
(2.75)b 7.63 7.60 4.62 4.92 4.73 4.65 4.25 2.89 2.85 2.66
Q(F -)
I
.
.
2.45 2.35 1.10 1.20 d 1.15 1.07 2.76 2.91 2.87
Q [F-I/G[Cl-I
... 0.323 0.309 0.240 0.243 d 0.246 0.250 0.95 1.02 1.08
Pulse currents -150 mA for 20 psec. Sample volume 2.0 cc. Doses -3 x 1018 eV g-' delivered a t a dose rate ~ 1 0 eV ~ 3g-1 sec-1. Methyl chloride concentration was 10+ M . All other solute concentrations were M . Sample buffered to p R 6.5. Absolute yields determined by comparing the halide ion produced per unit integrated current with the similar quantity for methyl chloride. Bssurned. Abeolute yield of 2.76 is calculated if it is assumed that 2.0 out of the 2.8 MeV per electron is absorbed in the sample. Series I. Integrated currents normalized by factor of 1.40. All other experiments series 11. Methyl chloride results should be compared only within the sepnrale series. Not measured. Q
b C Figure 2. Oscilloscope recordings of Ihe conductivity changes produced by the I-ptec puise irradiation of solutioris saturated with (a) CHzGI, (b) @FCls,and (c) both CFC13 and ethylene (0.0043 M ) at pfI 5.9. Time scale of lower traces is 10 pseclctn and upper traces is I msec/cm. Doses used were ,loL6 eV/g and HX piroduced -10-6 M . The yields corresponding to the pl~ateausof the upper curves are, respeclively, 3.14 (assumed, cf. ref 2), 14.2, and 5.3. Initial spikes on Iower traces are instrumental with recovery requiring ~5 pbec.
function of dose for closes up to 10'' eV/g. Since the extrapolation i o zero time is somewhat problematical, the changes observed at 50 psec (;.e.) on the first plateau on the lower curve ol Figure 2b) were compared with
the changes observed in methyl chloride. The increase in conductivity was a linear function of dose with a dope of 440 units relative to a slope of 197 units for a saturated methyl chloride solution. This change a t 50 psec corresponds to an HX yield of 7.0. Extrapolation of the growth curve back to aero time, as best one can, gives an estimate of one-half this value or 3.5 for the initial yield. For a soIution 0.008 M in CFCIa the yield of reaction 1 is expected, from the results on methyl chloride, to be 3.2 so that it is clear that the second stage in the hydrolysis is esseni ially complete a t 50 psec. The change observed a t 10 msec was 875 units which, taking into account the slightly lower equivalent conductance of F--, corresponds to a final yield of 14.2. The secondary reactions which lea complete hydrolysis of the CFC1, must occur within the 10-msec time scale. The observed yield is -10% higher than the expected value of 22.8, but the difference can be real since, as pointed out above, the doses are very low and in the absence of competing reactants the hydrogen atoms can react with the C:FGId. From the rate constant for the H atom reaction mentioned above such attack should occur, a t 0.008 M CFCL and in the absence of competing reactions, with a halfperiod of -50 psec. The observed difference is, however, somewhat less than the 4(0.6) = 2.4 increase expected if all hydrogen atoms react in this way and so competing reactions for the hydrogen atoms appear to be present and the apparent half-life will be consequently reduced. An attempt to scavenge the hydrogen atoms with 0.1 M 2-propanol resulted in a 30Oj, reduction of both the short- and long-lived components, presumably as a result of abstraction of hydrogen by The Journal of Phy8icaE Chemistry, Val. Y g r No. 4, 1971
T. I. BALKAS, J. H. FENDLER, AND R. H. SCHULER
460
0 0 0
2
4
t i me -m sec 6
a
10
20
40 60 80 IO0 time p s e c . Figure 3. Relative conductance as a function of time calculated from eq I (solid curves) with rate constants as given in the text. Representative data for CFCPa solutions ( 0 )taken from Figure 2b. Similar data for CFCla solutions containing 0.1 M %propanol (0)&ow a :35% reduction in the yield from the secondary reactions and a corresponding decrease in the period of the initial stage. I n this latter case the long-term reactions unexpectedly show a similar acceleration (dotted line) without, however, a corresponding reduction in the yield so that complications seen to be present.
the GFClt. radical. A rate constant of 2 X lo5 M-l see-‘ for the abstraction process is required to account for the observed competition. I n the type of results obtained here the details of hydrogen atom reactions will be completely obscured by the other secondary processes present. Summarizing the conductivity results on CFC13, we can say that it appears that for each hydrated electron which reacts with the GFCla, as expected one molecule of HGP is produced immediately (Le., in times less than see). A second molecule of HX (very probably WC1) i s then produced in a secondary reaction having a haif-period -.I5 psec (k2 = 46,000 sec-l) and a small additional component resulting from hydrogen atom attack may occiir on this same time scale. Two additional molecules of HX are subsequently produced at much longer times jn tertiary and quaternary reactions with rate constants 7 c ~and k4. From the exponential character of the growth, the third and fourth steps must occur essentially simultaneously, which implies that the third ~ ~ t is e prate controlling; i e . , kq >> k3 = 400 sec-l. If we assume that hydrolysis occurs in this stepwise fashion, then, taking into account the growth and decay of the intermediate produced in the second step and the fact that the equivalent conductance of K F is 93% of that of X-ICI, the time dependence of the Conductivity cliange relative to that for the initial reaction is given by relative conductivity =
The Journal of Phyaical Chemistry, Vol. 76,No. 4, 1971
where fs i s the fraction of the radicals produced in reaction 1 which undergo hydrolysis. This expression is plotted in Figure 3 (with k’s taken as given above) along with representative data from Figure 2b (normalized a t 10 msec) and the curves and data for a 0.1 M 2-propanol solution where the secondary reactions are reduced by 35%. Although minor complications may be present in the form of parallel side reactions, the excellent agreement indicates that eq 2 gives a reasonably complete description of the time dependence of the hydrolysis, The conductivity of a solution additionally saturated with ethylene (0.0043 M ) was also examined, and atypical oscillogram is illustrated in Figure 2c. It is seen that the long-term growth has largely disappeared. The yield of HX at 50 psec is 4.6 and at 5 msec is 5.3. The latter value agrees well with the total of 5.2 estimated Erom the results of steady-state experiments (correcting the total of 4.25 given in Table I by 0.5 for the increased yield a t 0.008 M CFC& and by 0.45 for the somewhat lower ethylene concentration). At 50 psec the excess over the initial value of the yield (estimated to be 3.1) is reduced to about 50% of the excess observed in the absence of ethylene, and the halfperiod of the increase in this region is seduced by roughly the same amount. Assuming that the scavenging reaction is
CFClZ.
+ CzH4--% RCHdXz
*
(2)
a crude estimate of lo4 M-l sec-l can be given for k, by equating k,[C2€14] with 0.693/(17 X It is interesting to note that while the extent of the increase on the millisecond time scale is only 10% of that found in the absence of ethylene, the growth period, itself,
RADIA.TION CBEMXBTRY OF FLUOROCHLOROMETHANES does not appear to be reduced significantly. This fact implies that the CF612 * radical hydrolyzes to a product which is responsible for the long-term growth but which does not react readily with ethylene. If this is so, then it would appear that a yield of (4.35 ( i e . , one-half of 0.7) of CFClz. escapes scavenging by the ethylene (in agreement with the fluoride yield in the steady-state experiments) and that some parallel side reaction contributes to the short-term growth. This interpretation leads to an estimate of 4 X 110' M - l sec- for IC,. CF2C12. Tt is Seen in Table I1 that chloride ion is produced in the radiolysis of lW3 M CFzClz solutions with til yield. .;thick is essentially twice the yield of solvated electrons, The simplest explanation is that the initial reaction occurs solely with solvated electrons and that, as with C:FCls, hydrolysis of the secondary radical rapidly follc~wsto produce quantitatively the second chloride. This explanation is borne out by the other observations reported here. The rate constant for reaction of hydrated electrons with CF2Cl2 was determined by examining the reduction in the chloride yield produced by t'he addition of NzO or Rf. Extrapolation of a,pgropriate plots to infinite N20 and H+ concentration showr; that neither H nor OH. attack contributes significa,ntXy to the chloride yield. Interpretation of the observed reduction in terms of simple competitive scavenging of hydrated electrons by the two solutes gives n value of 1.4 X 1Olo M-I sec-I for the rate constant of reaction 1. CF2Gl2is, in many ways, the most complex of the three solutes. First it will be noted in Table I1 that although the yield of chloride ion from the ethylenefree solution is twice the yield of hydrated electrons, the fluoride ion yield is only 73% as great so that hydrolysis is not complete. 4120th yields are, however, linearly dependent on dose over the range of 1-10 X loi8eV/g. Examination of the yields over a wide range of dose rates (but a t :ipproximately the same total dose) shows that a very i n t e m d h g dose rate effect is present. A summary of &.e results obtained at dose rates up to loz3eV g-l 8ec-l is given in Table V. The detailed data from pulse experiments, which were carried out parallel to those reported above for CFCI3,are given in Table IV. 19ie chloride yield is independent of dose rate up to 1W1eV g-' sec-l and drops off by only 10% in the experiments at 102aeV g-I sec-' where combination of the CF264. radicals is starting to become important. I n general, it would appear that, except for experiments a d very high dose rates or in the presence of a radical scavenger, hydrolysis of the chloride content is complete and that one can, as was done in the competitive studies, take one-half the chloride yield as a good measure of reaction 1, From the results of the y-ray experiments this value is 1,/2(5.46)= 2.73 or only slightly less lhan expected from the results on methyl chloride or CIFCI,. At the lowest dose rates the fluoride yield is essen9
461 ~~~~
_____
Table V: Dose Rate Dependence of Yields from CFzClza
2.0 x 1.1 x 1.3 X 7.5 x 2.2 x 6 X 2 x 3 x 2 x
1013 1014
1016 1015
10'6 1017
10'8 1021 1023
5.47 5.43 5.46b 5.536 5.42 (5.46)" (5.46p 5.28 4.86
5.36 5.05 3.986 3.81b 3.41 1.8'3 1.50 1.49
1.13
0.98 0.93 0.73 0.69 0.63 0.345 0.275 0.282 0.245
a At doses of -3 x 1018 eV/g. Yields are independent of dose from 1 to 7 x 1018 eV/g. c Assumed. The results at higher dose rates indicate that there should be no sigiiificant dependence of the chloride yield on dose rate below dose rates of 1 0 2 0 eV g-1 see-1.
tially identical with the chloride yield. However, the yield decreases as the dose rate is increased and a t very high dose rates a very substantial reduction in the F-: (31- ratio occurs. This decrease emonstrates that the secondary reactions proceed, for the inost part, via the production of chloride ion in the second step and fluoride ions in the third and fourth steps. If fluoride ion was produced at the second step, the complementary fragment would be expected to be identical with that formed in the CFC13 system. Since hydrolysis is known to be complete in the latter case, one would not expect a drop in fluoride production here. The fluoride yield does not, however, go to zero at very high dose rates as might be expected from the negligible yield observed in the presence of ethylene so that complications are unquestionably present. If the decrease observed in the region of 10I6eV g-l sec-l involves the competition between first- and second-order processes, then the second-order processes should be essentially complete at dose rates above 1020 eV g-l sec-;. The fluoride yields observed in the region of IO1*to eV g-l sec-l are constant at about 1.5 and indicate that npproximately 30% of the CF2C1- radicals are completely hydrolyzed. The solid curve of Figure 4 i s calculated on the assumption that 70% of the radicals produced in the secondary reaction undergo a second-order process which does not produce fluoride and that this reaction is in competition with a first-order hydrolysis with the rates being equal at a dose rate of 10lGeV g-' sec-l. The remaining 30% of the radicals are assumed to be hydrolyzed in all cases. At dose rates above ,1023 eV g-l sec-I an additional drop in fluoride production is expected because of the importance of CF2C1. combination reactions (as is evidenced by the decreased chloride yield). The approximate effect expected is indicated in Figure 4 by the dashed curve calculated on the assumption that this effect is 50% complete a t a dose rate of 3 X eV The Journal of Physical Chemistry, Vol. 76,NO.4, 1971
T. I. BALKAS, J. H. FENDLER, AND R. E. SCHULER
462
I
0 _li L'_-
10'~
--. %.
-. -. --._ 1 .
--_--__------ b--- -______ - 1______ 1
1d4
1015
10'6
10%'
1ot8
'
lot9
1020
L
I
lo2'
IOZ2
I
I
ioz4
Dose Rote - ev.g-' sec"
Figure 4. EfYect of dose rate on the ratio of yield of fluoride ion produced to that for electrons scavenged for M CFzClz (e), 8 x 10-3 A4 CFzCll (0),and 10-3 M CFaCl (A)solutions a t pH 6.5. The absolute yields observed for the 8 X IOwaiM CFzCll solutions are 1547, greater than in the other cases because significant scavenging of electrons within the spurs occurs. The lower solid curve is calculated on the assumption that for 707, of the cases a second-order process removes the precursor of fluoride ion in competition with the first-order tertiary hydrolysis (with equal rates at 1016 eV g-1 sec-1). The upper solid curve is similarly calculated 011 the assumption that fluoride ion is always produced in the secondary step and that 85% of the tertiary species undergo a competition similar to that for CFzCl2. The dotted curves give the dose rate effects expected if the second-order reactions were 100% effective in removing the fluoride ion produced in the third and fourth steps. The dashed cuyves on the right represent crude estimates of the dose rate effects expected a t high dose rates because of second-order reactions of the radicals produced in reaction 1.
g-l sec-l and &at combination of CFzCl. with .OH produces a limiting fliloride yield of 0.3. At this point the principal question is as to the source of the fluoride observed in the region of 1O2I eV g-l sec-I. It would seem that experimental problems involving regions of space or time where the radical concenirations are low can be ruled out since only a of the radicals can react under trivial fraction (