MAXS. MATHESON AND JOSEPH RABANI
1324
acid and the -SO3- group on the resin surface. Fejtelson also made the point that the use of hydrocarbon surfaces of ion-exchange resins in model systems for studying hydrophobic bonding may not be valid because of differences in the structure of water between
the resin and the bulk phases. This point has already been discussed above. AcknowEedgment. Thanks are due to Mrs. E. R. Stimson for technical assistance in determining the adsorption isotherms.
Pulse Radiolysis of Aqueous Hydrogen Solutions. I. Reaction of eaq-with Itself and Other Transients.
Rate Constants for
11.
The Interconvertibility of eaq-and H’
by Max S. Matheson and Joseph Rabani Chemistry Division, Argonne National Laboratory, Argonm, Illimia
(Received October $8,1984)
The technique of pulse radiolysis has been applied to aqueous solutions with (up to 100 atm.) and without dissolved H2 in the pH range 7-14. This work demonstrated the reaction H OH.,-+ e,,-, by direct observation of the absorption of the e,,- formed as a product. The rate constants for this and several other reactions in this system have been determined. The values obtained are: k(H OH,,-) = (1.8 f 0.6) X 10’; 2k(e,,ea,-) = (1.1 f 0.15) X 1O’O a t pH 13.3; k(eaq- Oaq-) = (2.2 f 0.6) X 1O1O at pH 13; k(0.qHz) = (8 f 4) X IO’; k(e,,H) = (2.5 f 0.6) X lolo; k(esqOH) = (3.0 f 0.7) X 1Olo M-l sec.-I.
+
+ +
+
Introduction This work was originally undertaken to measure the rate constants for reaction of the hydrated electron, e,,-, with the other transient radiolysis products of process 1. A number of OH scavengers such as methHzO
-+
H2, HzO2, H, OH, ea,-, &O,q+, 0H.q-
(1)
anol, ethanol, and ferrocyanide2had been used in our previous work to simplify the system. However, H2 seemed preferable as a scavenger for OH since the product of reaction 2 does not introduce a new species
OH
+ Hz
+HzO
+H
(2)
to the system but an H atom, a species already present. Another reaction which appeared likely to be useful The Journal of Physical Chemistry
+
+
+
in simplifying the system further was reaction 3, H
+ OH.,-
+ea,-
(3)
which had been suggested by Baxendale and Hughes3 and by Friedman and Zeltmann.4 This suggestion is supported by work in which externally produced H atoms were introduced into alkaline aqueous chloroacetate solutions6 (since chloroacetate ion can be used (1) Based on work performed under the auspices of the U. S. Atomic Energy Commission. (2) 9. Gordon, E. J. Hart, M. S. Matheson, J. Rabani. and J. K. Thomas, Discussions FcZTadUy Soc., 36, 193 (1963). (3) J. H.Baxendale and G. Hughes, Z . physik. Chem. (Frankfurt), 14, 323 (1958). (4)
H.L. Friedman and A. H. Zeltmann, J . Chem. Phya., 28, 878
(1958).
PULSE RADIOLYSIS OF AQUEOUS HYDROGEN SOLUTIONS
1325
to detect hydrated electrons6) or in which H atoms were produced by radiation in aqueous solutions of varying alkalinity.' That some H atoms are produced in the irradiation of water was proposed by Allan and Scholes.8 Supporting evidence for this proposal was obtained by showing that for added solutes reactive toward H atoms the relative rate constants of H atom reaction are the same in neutral solution containing an electron scavengerg as they are in acid solution where reaction 4 converts all hydrated electrons to H atoms.e*s By now it is ea, -
+ H3Oaq++H + H20
(4)
-
generally accepted that H atoms are produced with a yield of GH 0.6 atom/100 e.v. absorbed in irradiated neutral water. However, it has been proposed that excited water molecules are the precursors of these H atoms.'O To us the present evidence for excited water molecules is not conclusive and, in any case, such a precursor would not affect the results in this paper if the excited water molecules are converted to H atoms in less than 1 psec. Possible effects of longer lived H20* on the rate constants are discussed later. The development of pulse radiolysis techniques' and the direct observation of the spectrum of the hydrated electron with this method11dv12have enabled us to confirm reaction 3 directly. Also, we have used the technique of pulse radiolysis on aqueous solutions of various pH values saturated with up to 100 atm. of H2 to measure the rate constants of reaction of eaqwith itself and with other transient species. This work is described and discussed in this paper. A few preliminary results have been referred to previ0us1y.l~
Experimental The experimental arrangement, including the linear accelerator and optical arrangement, has been described in earlier papers,l1aSeand the setup used in the present work differs only in the substitution of the pressure cell (Figure 1) for the cell and mirror MI of Figure 2 in ref. l l e . A 0.4-fisec. pulse of 1BRIev. electrons with pulse currents in the range from 10 to 180 ma. was generally used. In a few experiments, 5-psec. pulses a t 120 ma. were used to study the effect of high transient concentrations. The pressure cell of Figure 1 is of welded stainless steel construction. The electron beam was focused to pass maximum current through the 1-cm. aperture, 0, of the brass collimator. The focused beam entered the stainless steel cell through the 1.2-mm. thick, 1cm. diameter aluminum window, W1. Microscope slides placed in front of or behind cell C and darkened by irradiation indicated that cell C was uniformly
-c
-w2
-
0 I 2 3 4 5 c m Figure 1. Pressure cell for pulse radiolysis.
(5) (a) J. Jortner and J. Rabani, J . Phys. Chem., 66, 2081 (1962); (b) J. Jortner and J. Rabani, J. A m . Chem. SOC.,83,4868 (1961). (6) (a) E. Hayon and J. Weiss, Proc. 2nd Intern. Conf. Peaceful Uses At. Energy Geneva, 29, 80 (1958); (b) E. Hayon and A. 0. Allen, J. Phys. Chem., 65, 2181 (1961). (7) (a) J. T . Allan, M . G. Robinson, and G. Scholes, Proc. Chem. SOC.,381 (1962); (b) S. Nehari and J. Rabani, J. Phys. Chem., 67, 1609 (1963); (c) J. T. Allan and C. M.Beck, J . A m . Chem. Sac., 86, 1483 (1964). (8) J. T. Allan and G. Scholes. Nature, 187, 218 (1960). (9) (a) J. Rabani, J . A m . Chem. Soc., 84, 868 (1962); (b) J. Rabani and G . Stein. J . Chem. Phys., 37, 1865 (1962) ; (c) G. Scholes and M. Simic, J. Phys. Chem., 68, 1731, 1738 (1964). (10) (a) E. Hayon, Trans. Faraday SOC.,60, 1059 (1964); (b) M. Anbar and D. Meyerstein, J. Phys. Chem., 68, 1713 (1964). (11) (a) M. S. Matheson and L. M. Dorfman, J . Chem. Phys., 32, 1870 (1960); (b) J.. ' l Keene, Nature, 188, 843 (1960); (c) R . L. McCarthy and A. MacLachlan, Trans. Faraday Sac., 56, 1187 (1960); (d) E. J. Hart and J. W. Boag, J . A m . Chem. soc.. 84, 4090 (1962); (e) L. M .Dorfman, I. A. Taub, and K. E. Buhler. J . Chem. Phys., 36, 3051 (1962). (12) (a) J. P. Keene, Nature, 197, 47 (1963); (b) J. P. Keene, Discussions F a ~ a d a ySac., 36, 304 (1963). (13) (a) M. S. Matheson and J. Rabani. Radiation Res., 19, 180 (1963); (b) M. S. Matheson, Radiation Res. S u p p l . , 4 , 1 (1964).
Volume 63,Number 4
April 1566
MAXS.MATHESON AND JOSEPH RABANI
1326
10
9 X
e 7
6
VD 5
4
?
2
I
0
I
1
I
10
20
30
I 40 t(psec)
I 50
1
Figure 2. Second-order decay of absorption of eaq- a t 5780 A. in 0.2 M NaOH equilibrated with 100 atm. of Hz. Light path = 8 cm.; t = 0 corresponds to middle of pulse. Plot is linear to 85% reaction.
irradiated. The cylindrical cell C was 2 cm. in diameter and 1 or 2 em. in length. An aluminum mirror, PI, protected by a copper coating, was evaporated on the window of the cell through which the electron beam entered. Light entered the pressure cell through the 16-mm. thick high purity silica window, Wz. The analyzing light beam from an HBO 107/1 Osram mercury lamp traversed the length of cell C two or four times. The transient changes in absorption were recorded photoelectrically and photographed on an oscilloscope screen.”e Three stainless steel cells were constructed and so machined that they could be interchanged on the support, s, shown in Figure 1. Minor adjustments were required since the mirror M was not independent but was deposited on the silica cell. The silica cells were filled under an atmosphere of argon in a glove box. The triply distilled water had previously been degassed in syringes14which were then introduced into the glove box. The KaOH solutions (unless otherwise stated) were prepared by weighing freshly cut sodium in the glove box and adding the sodium in small pieces to a beaker containing a fraction T h e Journal of Physical Chemistry
of a milliliter of triply distilled water. After the sodium had all reacted, the appropriate volumes of triply distilled water were added and the resulting alkaline solutions were put in syringes. The glove box technique and the preparation of NaOH from Na were adopted to avoid carbonate in the solutions, since it has been shown that OH (or 0-) reacts with carbonate to give a long-lived transient absorbing also a t 5780 W . l 6 , l 6 By acidification and subsequent gas chromatographic analysis, the CO, in 1 M NaOH prepared in the glove box was found to be 5.3 FLM. The carbonate introduced by S a O H would be proportionately lower in less alkaline solutions. Since three pressure cells were available, three silica cells could be filled for each set of experiments. With the silica cells in place, the pressure cells were closed in the glove box and then removed from it. Next each pressure cell was connected to a cylinder of Matheson Co. Ultrapure hydrogen (impurities 1 p.p.m. of HzO, 5 to 9 p.p.m. of argon) and flushed out briefly with a current of H,. Then the cell was filled to a pressure of about 30 atm. of H2 and shaken as vigorously as possible without spilling solution from the silica cell. (The cell neck was made long (Figure 1) to minimize possible spillage.) Subsequently, the exhaust valve VBwas opened and the cell pressure decreased to just above atmospheric pressure. The pressurizing, shaking, and exhausting were repeated three times to assure removal of traces of air. Finally, the cell was filled to the desired pressure, usually 100 atm. (1500 p.s.i.g.). One of the cells was always filled with -0.2 M NaOH and 100 atm. of H, to serve as a control and relative dosimeter. The initial optical density under presumably similar conditions varied as much as 20% in the three different cells when all were filled with the same solution. Perhaps the slight adjustments necessary for optical alignment changed the alignment with the beam, although the vertical axis of adjustment passed 1 cm. from 0 and the horizontal axis 5 em. below aperture 0, both axes intersecting a t pivot point P. When the hydrogen pressurized cells were pulseirradiated it was found the rate of eiiq- decay decreased from one pulse to the next, attaining a constant rate of decay after about 20 pulses. If a cell with 100 atm. of H, was given 500 consecutive pulses (total dose about lo6 rads) but the shutter of the oscilloscope camera opened only for the 20th and 500th pulse, the two ~~
~~
~
~
(14) C. Senvar and E. J. Hart, Proc. Znd Intern. Conf Peaceful Uses A t . Energy Geneva, 29, 19 (1958). (15) S. Gordon, E. J. Hart, M .S. Matheson, J. Rabani, and J. K. Thomas, J . Am. Chem. Soc., 85, 1375 (1963). (16) G.E.Aaams and J. W. Boag, Proc. Chem. Soc.. 112 (1964).
PULSERADIOLYSIS OF AQUEOUS HYDROGEN SOLUTIONS
decay curves photographed as a single curve. For two reasons it was concluded that this effect was probably due to traces of impurities introduced from the gas phase. First, the effect of the initial pulses in slowing the rate occurred a t all pH values from 7 to 14 and, therefore, these impurities could not have been added with the NaOH. Second, if water plus 100 atm. of Hz was pulse-irradiated until the limiting slow rate was attained and then the H2 was removed and replaced by repeated shaking with 1 atm. of argon, then the rate for the first pulse with argon was the same as that measured for pure deaerated water (of the same pH) with the usual syringe and unpressurized silica cell technique. Experiments without H2 were generally made using deaerated triply distilled water (pH 7 to 14) in the usual syringe technique2 and, of course, using only the decay curve for the first pulse. The attainment of the slower constant rate in the presence of 100 atm. of Hz after repeated pulsing indicated that the inolecular Hz02 produced by each pulse probably disappeared in a chain reaction with the Hz. All experiments were carried out at room temperature, 23". For H2-saturated solutions the cell was given about 20 pulses to obtain a constant rate and then generally for each decay curve the absorption decays for three consecutive pulses were photographed superposed to give a single curve. For computer fitting of plots of l/(optical density) vs. time, an IBM 1620 was used.
1327
1.2 or 2.0 or 3.0 X lolo; kll" = 1.2 X lolo; 2k1218rm 0.8 or 1.2 X 10'"; k1aZ1 = 1.4 X 10" M-' set.-'. Although kg(Hz0)22is only about 1.4 X lo3 set.-', a first-order reaction of about lo4sec. was found in our work. This first-order reaction is probably due to impurities and in most experiments was a very small fraction of the initial decay rate. It is the initial decay rate which is most significant in the determination of rate constants. Other reactions such as 2, 3, 14, 15, and 16 are slower and need be considered only if the = =
+ Hz +HzO + H H + OH,,- +eaqH + HzO2 +H2O + OH OH + H20z +HzO + HO2 OH
H
+ H3Oaq'
Hz, H202, H, OH, e,,-, &Oaq+, OH,,-
(1)
+ H30aqf +H + H2O
(4)
ea, eaq-
+ ea,-
eaqeaq-
Hz0
+Hz
+ 2OH.q-
+ OH +OH,,-
+H
H20
+H2
+ OH,,-
+ HzO2 +OH + OHaqeaq- + HzO +H + OH.,-
eaq-
H+H+H2
+ OH H2O OH + OH +H202 + OHa,- +2H2O H
(14) (15) (16)
+
0.q-
ea,-
+ H2
-
OH,,-
+H
+ Oaq- +20Haq+ +20Haq-
(24
H9O
(64
H20
In water, the following reactions should result from the irradiation pulse --+
+ H20
(3)
stable reactant is present in sufficient concentration. k220" = 4.5 X 107, k14'8 = 9 X lo', klj20a = 4.5 X 107, and kls is much smaller than these.13b H z (100 Atm. at p H -13): Determination of k(e,,eaQ-). In the pulse radiolysis of a solution at high pH (-13) and under 100 atm. of Hz (-0.08 M Hz), reactions 4, 6a, 7, 10, 11, and 12a should be eliminated since reactions 2a, 3,23and 13 will be very fast. The
Results and Discussion
H20
+ HZaq+
(2)
4
(5)
(6) (7)
(8) (9) (10)
(11) (12) (13)
Except for (9), reactions 4 through 13 are generally fast, having rate constants 1O1O M-l sec.-l or greater, e.g., k415117 = 2.2 x 10'0; ks'5'17b = 1.2 X 10'"; 2klo18v19
Oaq-
Oaq-
(124
pK of OH is 11.9,20b so reactions 2a, 6a, and 12a replace reactions 2, 6, and 12. Reactions 8 or 8a and ea,-
+ HO:-
+OHaq-
+
Oaq-
@a)
9 should also not be important, since (3) following (9) and (2a), and (3) following (8) or (8a) will regenerate eaq-. Thus, under these conditions ea,- should disappear only by reaction 5, and a plot of l / D us. t should be linear, since for the second-order disappearance of an absorbing species (17) (a) L. M. Dorfman and I. A. Taub, J . A m . Chem. Sac., 85, 2370 (1963); (b) J. P. Keene, Radiation Ree., 22, 1 (1964). (18) (a) J. K. Thomas, J . Phys. Chem., 67, 2593 (1963); (b) J. P. Sweet and J. K. Thomas, ibid., 68, 1363 (1964). (19) H . A. Schwara, ibid., 67, 2827 (1963). (20) (a) H. A. Schwara, i b i d . , 66,255 (1962) : (h) J. Rahani and M . S. Matheson, J . A m . Chem. Soc., 86, 3175 (1964). (21) M. Eigen and L. DeMaeyer, 2. Elektrochem., 59, 986 (1955). (22) This work indicates previously published values are too high and S. Gordon and E. J. Hart have recently found ko 5 25 M-1 sec. - 1 . (23) ks s 2 X 107 M-' sec.-l can be estimated from ref. 7a,h assuming k(H 0 2 ) = 2 X 1010 M - 1 sec.-1.
+
Volume 69, Number 4
April 1965
MAX S. MATHESON AND JOSEPH RABANI
1328
l/Dt
=
1/Do
+ (k/CBXZ)t
(17)
where Do and D, are the optical densities a t times 0 and t , tsX is extinction coefficient of absorbing species s a t the wave length A, and 1 is the optical path length for the analyzing light. Figure 2 indicates the linearity obtained. In Table I are summarized the results of experiments at 100 atm. of Hz and pH 13.3 with varying electron pulse intensities. The intensity of the pulse can be obtained from the initial optical density a t 5780 A., Do in Table I . In most of these experiments a 0.4-psec. pulse was used, and, since eaq- decays very little during 0.4 psec., extrapolation to the middle of the pulse yields a value of Do very close to that appropriate for a pulse of equivalent total energy and negligible duration. However, with the 5-psec. pulses considerable ea,- decayed during the pulse, and D owas taken arbitrarily a t the end of the pulse.
Table I : "k" for 100 Atm. of Hz and Q.2 M NaOH Using Variable Pulse Intensities
~ ~ 5 7 8 0
2.32"' 1.8OC 1.43 1.25c 1.20 1.07 0.600 0.460 0.370 0.300 0.240 0.200 0.150
% ' linearity
No. of
"k"'l
of plot
expt6.
1.2 1.0 1.05 1.2 1.2 1.05 1.35 1.os 1.3 1.3 1.3 1.25 1.4
75 90 90 85
3 3 3 3 2 3 5 3 6 9
90 85 80 90 85 80 80 80 75
1 1
6
a "k" measured from initial linear portion of plot as [ d ( l / D ) / d t ] x EL. * With a 1-cm. cell, two passes with 0.1 M NaOH. Do multiplied by 4 for comparison with other results where 1 = Using a 5-wec. pulse. Do = D a t the end of pulse; 8 cm. others, 0.4-psec. pulse and DOextrapolated to middle of pulse.
In estimating the per cent linearity in the table there was a small error but we believe the error is not more than 5%. In plots such as Figure 2 the initial slope was drawn though the greatest number of experimental l / D values possible while still excluding points which would deviate from the straight line by more than the experimental error. The deviation from linearity would be consistent with the presence of a small amount of a reaction of ea,- with an impurity present a t a constant concentration, since the deviation, The Journal of Physical Chemistry
which from computer calculations seems to require a first-order reaction, cannot be due to eaqH20 at this pH because this reaction would be immediately OH,,-. The reaction is not with reversed by H HzO2 either, since in the presence of the high concentration of Hz used, HzOz regenerates ea,- as previously discussed. The same ea,- decay a t a given intensity was obtained whether the solution was purified by a series of pulses of higher, or of lower, or of equal intensity. A factor of 8 in pulse intensity was tried. The fact that the deviation could not be eliminated by repeated pulse irradiation in the presence of H2 could be explained by a relatively high concentration of inipurity with low reactivity toward ea,-, or by a metal ion impurity which regenerates after each pulse, perhaps by reaction with water.24 The impurity may have been introduced with the H, in filling the cell, perhaps from connecting tubing. If metallic impurities were introduced with the sodium, the amount should be proportionately less for lower pH values. The principal decay of ea,- is definitely second order over a 16fold range of pulse intensity (Table I ) , and yields (using only the results at Do = 0.460 and higher and correcting for the small first-order reaction) 2k5 = (1.08 f 0.05) X 101OM-lsec.-l. te6780= 10,600 ( flo?&)M-l cm.-l was used throughout this paper.25 When the error in is included 2k6 = (1.1 f 0.15) x 1010 M-' sec.-', in satisfactory agreement with previous values: (0.85 0.15) x 10'" at pH 10.g2 and