NOTES
4398 NzO scavenges electrons to yield Nz and O-lO
+ NzO -+ NzO- -+ NZ+ 0(4) G(N2) is reduced with increasing DzS concentration but at the same time G(H2 + H D + Dz + Nz) remains e
constant. A reaction of an electron with DzS yielding Dz and H D must be superceding a similar one with XzO. The results of Sagert and Blair'l show that a t the 320 concentration we employ, almost all the Nz arises from reaction 4 and very little, if any, from the secondary reaction
0-
+ NzO +Nz +
02-
(5)
which increases G(Nz) in gas phase radiolysis of saturated hydrocarbons by 55% above G(electrons).12 It is to be noted that at 8 mol % DzS, G(XZ)is less than one half its value in the absence of DZS. A Stern-Volmer type plot using the yields shown in the figure gives ICe+N20/JCe+D2S = 6. The same value is obtained with cyclohexane as solvent. A possible mechanism for the production of Dz via electron capture is that given by von Meissner and Hengleinl as the mechanism operating in irradiated aqueous DzS solutions
DzS
+ e -+
be accounted for by an inefficient electron scavenging mechanism without any positive charge transfer. Although we have shown that electron capture by DzS occurs we cannot say whether positive charge scavenging occurs at the same time. The results of Bone and Futrel15 suggest that both charge exchange and proton transfer might occur. However a limiting G(D2) of -3 a t high DzS concentration as found by von Meissner and Hengleinl is consistent with electron scavenging being the sole process for the production of Dz.
Acknowledgment. This work mas supported by the National Research Council of Canada and Atomic Energy of Canada Ltd, Commercial Products. (10) G. R. A. Johnson and J. M. Warman, Trans. Faraday Xoc., 61, 1709 (1965). (11) N. H. Sagert and A. S. Blair, Can. J . Chem., 45, 1351 (1967). (12) J. M.Warman, J . Phys. Chem., 71,4066 (1967). (13) K. Jager and A. Henglein, 2.Naturforsch., 21a, 1251 (1966). (14) F. Williams, J . Amer. Chem. Soc., 86,3954 (1964). (15) J. W. Buchanan and F. Williams, J . Chem. Phys., 44, 4377 (1966). (16) J. R. Nash and W. H. Hamill, J . Phys. Chem., 66, 1097 (1962). (17) P. J. Dyne, Can. J . Chem., 43,1080 (1965). (18) G. R. Freeman, J . Chem. Phys., 46,2822 (1967).
fD + sZ
D2S-
I + HSD
The D atom will be scavenged by D2Sto give Dz. Von hleissner and Hengleinl rejected the idea of electron capture by DzS in saturated hydrocarbon solvents with low dielectric constants since the effects of solvation are not great enough to overcome the activation energy of 1.56 eV observed for the formation of SH- in the gas phase.la If this argument holds, then the extrapolation of the appearance potential of negative species from gas to liquid phase is not valid for the system under discussion. Electron capture by DzS explains two inconsistencies found in the radiolysis of this solute in cyclohexane and n-hexane. (a) Whereas the addition of the efficient proton acceptors ethanol-d and ammonia-& to liquid cyclohexane does not lead to any increase in the total hydrogen yield14p15, the addition of DzS causes an increase which at 5 mol yo is -1 G unit. This increase can now be attributed to the reactions of DzS with electrons which do not normally participate in hydrogen production. Such increased yields are found with HI16 and HCl" as electron scavengers. (b) Freeman's has made theoretical calculations of positive ion scavenging in liquid hydrocarbons. He finds that agreement between theory and experiment can be obtained for the DzS-n-hexane system only if a reduced mobility is assumed for the negative entity. We have found that with the same basic theory the DZyields can The Journal of Physical Chemistry
Field Independence of Photoinjection into Hydrocarbon Solution
by A. Prock and AI. Djibelian Department of Chemistry, Boston University, Boston, Massachusetts 08816 (Received April 81, 1989)
Photoinjection from rhodium electrode into hydrocarbon solutions has been reported previously,' where a simple thermodynamic analysis based on energy requirements led to an estimate of solvation energies of anionic species. Objections were later raised in private communication that the effect of the large applied electric field was not taken into account. That is, under high enough electric field strength there would be sufficient potential drop across the Helmholtz double layer for a nonnegligible amount of energy to be coupled into the electrode reaction. The solvation energy calculated without accounting for this energy would be too high. The question is, are we missing an energy term under the conditions of the experiment? This note describes experiments on one of our previous systems which carries photoinjection measurements down to 1% of the original applied voltage, or (1) A. Proolr, M.Djibelian, and S. Sullivan, J . Phys. Chem., 71,3378 (1967).
4399
NOTES
presented but shows similar behavior. The original conclusion concerning this work, which ignored the effect of the electric field, is strongly supported by this result. s The lowest applied voltage used in these experiments ‘H x is judged to be low enough to make negligible the elec0 s tric field contribution to the energetics of the reaction 0 0.40 on the following basis. A simple calculation given d below shows that even for 10 V applied, the diffuse z 0 double layer thickness is around 2500 A. Taking the 0 0 Helmholtz double layer to be around 5 8, we find that IO r of the 5 V applied between bulk solution and an eleca. trode, then a t most only the fraction, 5/2500 of the I I \d voltage, or 10 mV, appears across the Helmholtz double V 110 0 IO 20 30 layer to aid energetically in the transfer of charge. VOLTS APPLIED For such low conductivity media, applied voltages of Figure 1. Apparent photoconductivity (AZph/ v,,,)us. applied even tens of volts would not be very effective in energy voltage for 0.10 M anthracene in benzene under white light. transfer at the electrode, but since the amount of A 200-W super pressure mercury lamp plus water and energy coupled rises with the square of applied voltage, saturated NaNOt filter supplies light of k > 404 nm. then at some hundreds of volts applied the field should supply considerable energy to the discharge process. The calculation of double-layer thickness under applied field is based on the differential equation of forced diffusi~n.~JWe retain all terms in the equation, and assume further that in the dark the positive current equals the negative current. This is reasonable since it is almost certain that the dark current is caused by traces of impurity. The usual definitions are D , diffusion coefficient; n+,n-, ionic concentrations, dark; p, mobility, assumed the same for both ions; and ‘p, potential. The forced diffusion equations are for the one-dimensional problem 0.80
-
1
I
I
A I , 10 20 30 V I
0
I IO
dni. J+ = -De dx
VOLTS APPLIED
Figure 2. Apparent photoconductivity (Az,,h/ V,,,) 08. applied voltage under same conditions as in Figure 1, with the addition of a Schott 546 nm filter with HBW of 27 nm to produce incident green light.
about 4% of the original applied field strength. The system studied was anthracene in bemene, the system which is best frorn the aspect of showing very low photocurrent for reversed polarity, Le., the rhodium electrode positive. The system has been described previously; the electrode spacing was reduced to 0.010 cm, and applied. voltage was in the range 1.14-112 V. White light (X > 404 nm) and green light (X = 546 nm, HW = 27 nm) was obtained using a 250-W super pressure mercury lamp with filters, and red light (A = 660 nm, HW = 60 nm) was supplied by a 450-W xenon lamp, with filters. Figures 1 and 2 present data of apparent conductivity change (photoconductivity) as a function of applied voltage for the white light and for the green light. It appears that the photoinjection effect shows no tendency to vanish as the applied voltage decreases towards zero. The graph for red light is not
J-
=
De-dn-
pe
- pe
dx
dP nt dx
dv n-
dx
For steady state the divergence of the sum vanishes. Along with the relation D / p = k T / e , the result is div JT = 0 = -De-
d2 (n+ dx
+ n-)
-
De2 d2‘p _.LT dx2 (n+
e2dp d + n-) - D- (n++ n-1 LT dx dx
(3)
Equations 1 and 2 are subtracted and the result set equal to zero to obtain an evaluation of the third term in eq 3. This latter equation can then be put into the form d2q
+
z2
AD
q = 0, where X =
[l
+ (x, & g)zl” (4)
(2) A. Prock and G. McConkey, “Topics in Chemical Physics,” Elaevier, Amsterdam, 1962, p 242. (3) J. Gavis, J.Chern. P h w , 41,3787 (1964).
Volume 78,Number 12 December 1969
NOTES
4400 Charge density, p, is e(%+ - n-). The length, Xo, is the double layer thickness under no externally applied field.
I n our system, where conductivity is so low (
