Radiation Chemistry

---

10 Radiolysis and Photolysis of the Aqueous Nitrate System

Downloaded by YORK UNIV on November 9, 2012 | http://pubs.acs.org Publication Date: January 1, 1968 | doi: 10.1021/ba-1968-0081.ch010

M A L C O L M DANIELS Chemistry Department and Radiation Center, Oregon State University, Corvallis, Ore. 97331

Radiolysis of the aqueous nitrate system is discussed in terms of (a) indirect effect in dilute solution, and (b) con­ current indirect and direct effects in concentrated solution. Analysis of energy fractionation breaks down (b), gives G(NO -) = 4.0, and demonstrates stoichiometry for di­ rect effect according to 2

NO3¯

2 NO --> 2 NO - + O 3¯

2

2

Scavening with H and n-propyl alcohol quantitatively sup­ ports this. Pulse radiolysis shows NO and NO as inter­ mediates, NO being characteristic of concentrated solutions. Formation and decay of NO is discussed. Primary process of direct radiolysis is formulated as 2

3

2

3

3

NO3-—M M—> (NO2 + O + e-) cage, spur G(NO2-) varies with scavenger and may attain ~17. From photochemical studies participation of lowest excited state is unlikely, but the second state at 6 e.v. may be involved. NO3¯

' T ' h i s paper reviews some of the experimental evidence available concerning the radiolysis of the aqueous nitrate system both by continuous and pulsed beams and correlates it with the available evidence on the ultraviolet photolysis of the same system. Since the early experiments demonstrating that ionizing radiation could indeed cause reduction to nitrite, a considerable amount of information has accumulated, much of it originating in the U.S.S.R. during the last decade. This work has been ably summarized by Pikaev (14) and Sharpatyi (16). Despite this continuing effort, it was apparent that no satisfactory (i.e., quantitatively 153 In Radiation Chemistry; Hart, E.; Advances in Chemistry; American Chemical Society: Washington, DC, 1968.

154

RADIATION CHEMISTRY

1

and qualitatively self-consistent ) account could be given of the radiolytic behavior of this system. This is perhaps not surprising considering the many and varied problems involved, which cover indeed almost the whole gamut of experimental and theoretical activity in radiation chemistry. Thus, to begin to understand this system, it was necessary to start with dilute solutions ( which we shall operationally define, from the concentra­ tion dependence (5) of G ( N 0 " ) , as less than about 0.1M) and elucidate a complete mechanism of radiolysis which, for extension to other situa­ tions of high concentration and high intensity, also had to be quantitative. Since the over-all effect of radiation is reduction, the role of the oxidizing O H radicals and the possibility of their reaction with the nitrate ion is critical. Coincidentally, this work is also relevant to the continuing prob­ lem of stoichiometry in water radiolysis.

Downloaded by YORK UNIV on November 9, 2012 | http://pubs.acs.org Publication Date: January 1, 1968 | doi: 10.1021/ba-1968-0081.ch010

2

The radiolysis of concentrated nitrate solution is characterized by a yield of nitrite increasing continuously with nitrate concentration up to the solubility limit and the occurrence of 0 as a major product. This behavior has prompted various suggestions as to its origin. Sworski (18), observing the effect indirectly as net reduction of eerie ion over concen­ trations up to 0.5M N 0 ~ , suggested that it is caused by the "direct effect" —i.e., the chemical effect of deposition of energy in the nitrate ion, as contrasted with the "indirect effect" manifested in dilute solution. Other suggestions have invoked ( a ) "sub-excitation elections," ( b ) excited states of nitrate ion formed either directly in the energy dissipation processes or indirectly by energy transfer from (hypothetical) excited states of water (16), (c) collective dissociation processes (4). Although attractive and stimulating, such suggestions have generally not been evaluated quantitatively. To do this, the approach must be separated into two stages. First, work must be directed to understanding the chemical proc­ esses involved, through stoichiometry, over-all kinetics, and transient kinetics, so that the radiolysis can be described in terms of a particular "primary process." W e can then appropriately ask how such a primary process can occur. 2

3

W e have investigated in detail the parameters affecting the continu­ ous γ-radiolysis of concentrated solutions (intensity, p H , 0 , scavengers, etc.). Transients were investigated by pulsed electron beam radiolysis and kinetic spectroscopy, and the reactions of the optically accessible excited states of nitrate were investigated by conventional photolysis. This paper represents a survey of our recent results which, taken in con­ junction with the work of others, allows the construction of a model whereby the main features of this system may be understood and may even be predicted. Literature review is necessarily selective for the present purpose (because of doubtful relevance to liquid state processes, low temperature radiolysis, and hence E S R work has been omitted from 2

In Radiation Chemistry; Hart, E.; Advances in Chemistry; American Chemical Society: Washington, DC, 1968.

10.

DANIELS

155

Aqueous Nitrate System

consideration). However, we take this opportunity to emphasize some unanswered problems of general interest which arise from the specific system. y-Radiolysis of Dilute Solutions (5) The low magnitude of G ( N 0 ~ ) (.—0.5) in dilute solutions imme­ diately suggests that N 0 is formed as the result of mutually opposing reduction and oxidation reactions. Before the discovery of the hydrated electron, Bakh (2) suggested that the reducing species (then called H atom) could react readily with N 0 ~ , and this has been confirmed by direct measurement of the rate constants for reaction of the electron (3, 19 ) with ΝΟ-Γ and indirectly for the H atom ( 1 ). It is clear that except in the most dilute solution, e~ w i l l be scavenged rapidly and completely by N 0 2

2

Downloaded by YORK UNIV on November 9, 2012 | http://pubs.acs.org Publication Date: January 1, 1968 | doi: 10.1021/ba-1968-0081.ch010

3

8

e~ 4- N 0 - -> Ν (IV) ( N 0 % H N 0 " , N 0 ) 3

3

2

3

2

(1)

The question then concerns the nature of the oxidizing reaction—i.e., what is the fate of the O H radicals? To answer this, it was necessary to distinguish between the following reactions: (a) reaction with nitrate O H + N 0 - -> Ν(VI) ( N 0 , H N 0 " ) 3

3

4

(2)

followed by N(IV) + N(VI) - * 2 N 0 3

(b) reoxidization of the intermediate Ν ( I V ) O H + Ν(IV) - > N 0 -

(3)

3

or ( c ) oxidation of N 0 ~ 2

O H + N 0 - - » N(IV) ( N 0 , H N 0 ) 2

2

3

(4)

(In these reactions the unidentified species is symbolized by its formal oxidation number—i.e., N 0 by Ν ( V I ) , N 0 by Ν ( I V ) , etc.) The clue to the behavior of O H radicals came from the observation that concen­ tration vs. dose curves for N 0 " formation were initially nonlinear until concentrations of NO » reached about 2 μΜ and were thereafter inde­ pendent of N 0 " . This behavior was observed over a wide range of nitrate concentrations from 10~ to 6.0M, but the nonlinearity was elimi­ nated by adding small quantities of N 0 " before irradiation. These facts clearly indicate that Reaction 4 is the fate of the O H radicals, and this was confirmed by demonstrating competition kinetics (5) between H> and N 0 " for the O H species. The main features of the radioly­ sis i n dilute solution can thus be accounted for by Reactions 1 and 4, followed by a hydrolytic dismutation process 3

2

2

L

2

3

2

2

2 N 0 + H 0 -> 2 H + N O , " + N 0 " 2

2

+

2

In Radiation Chemistry; Hart, E.; Advances in Chemistry; American Chemical Society: Washington, DC, 1968.

(5)

156

RADIATION CHEMISTRY

1

The pulse radiolysis work of Pikaev et al. (15) further verifies this scheme. They measured N 0 " and H 0 formation as a function of inten­ sity at very high intensities. For dilute solutions both G ( N 0 ~ ) and G ( H 0 ) increased considerably at higher intensities, owing to the inci­ dence of Reaction 6 2

2

2

2

2

2

OH + O H ^ H 0 (6) relative to Reaction 4. The extensive investigation of the γ-radiolysis has yielded inde­ pendent relations, allowing the evaluation of the yields of primary species (5). In this way the following results were obtained [where g ( H ) is taken from the measurements of Mahlman ( I I ) ] : 2

2

Downloaded by YORK UNIV on November 9, 2012 | http://pubs.acs.org Publication Date: January 1, 1968 | doi: 10.1021/ba-1968-0081.ch010

2

g(e

+ H) = 3.41

g(OH) = 2.53

g(H )=0.44 2

g(H O )=0.75 2

2

These values are based on experiments for this one system, and stoichiometry is not assumed. Indeed these values can be regarded as a test of stoichiometry for conventional species. A small, but unfortunately inescapable difference of about 0.3 exists between the sum (in equiva­ lents ) of the reducing species and the oxidizing species. This discrepancy can be resolved if a small amount of oxygen ( g .— 0.1) is a product of the radiolysis of water. Evidence to support this suggestion has been presented (6), based on (a) determinations of G ( 0 ) as a function of nitrate concentration, (b) isotopic composition of the evolved oxygen when H O was used (12). 2

2

l s

The understanding which we now have of the mechanism of radioly­ sis in dilute solution provides an adequate quantitative base for pro­ gressing to consideration of the effects in concentrated solution. H o w ­ ever, some current gaps exist in our knowledge. As indicated in Reactions 1 and 4, we are still ignorant of the molecular structures of the species denoted by Ν ( I V ) ; interesting ionic relations and differences of reactivity may exist between their various forms. The mechanism of Reaction 5 is somewhat obscure; in fact, actual evidence for it is sparse. These are of course suitable topics for pulse radiolysis experiments, and we expect to be able to clarify them in the near future. Such experiments may also evaluate the extent of Reaction 3 at high intensities and its rate constant. y-Radiolysis of Concentrated Solutions The immediate problem in considering the radiolysis of concentrated solutions is how to account for the water radiolysis process which still undoubtedly accounts for the major part of the energy dissipation. The simplest approach is to assume that the phenomena characteristic of concentrated solution originate independently of water radiolysis; hence,

In Radiation Chemistry; Hart, E.; Advances in Chemistry; American Chemical Society: Washington, DC, 1968.

10.

DANIELS

Aqueous Nitrate System

157

for a product P, formed both i n dilute and concentrated solutions, we write: G(P)=G(P) H 0 2

f

H20

+ G(P) NO?," * /χυ.·Γ

(A)

where G ( P ) is calculated from total energy absorption (based on elec­ tron density), G ( P ) o is a constant characteristic of dilute solution, and G ( P ) N O - is a constant characteristic of concentrated solutions. F o r the weighing coefficients / o and / . we have taken the fractions of energy deposition in the water and nitrate components based on electron frac­ tions. There are, of course, considerable theoretical uncertainties i n such a procedure, so we can only regard it as semiempirical and judge its utility by its success. One way to test this formulation is in the rearranged form: ^ = G(P) o + ^ 2 î l - G ( P ) , (Β) H 2

3

Downloaded by YORK UNIV on November 9, 2012 | http://pubs.acs.org Publication Date: January 1, 1968 | doi: 10.1021/ba-1968-0081.ch010

H2

N()

r

H 2

/H O

N O

/H O

2

2

Figure 1 shows the yield of nitrite from 0.1 to 6.0M N 0 " treated i n this way (5, 7). Considering the difficulties experienced previously i n fitting such results, we seem to have found a significant functional relationship. Figure 1 also shows the results of Mahlman (11) for G ( 0 ) , which respond even better to this treatment. F r o m the slopes of Figure 1, we obtain G ( N 0 - ) - = 4.0, G ( 0 ) ΝΟ3- = 2.0, and thus we have the first demonstration of stoichiometry for the processes occurring in concen­ trated solution, according to 3

2

2

N 0 3

2

2NO3- - » 2 N 0 - + 0 2

(7)

2

Further verification of the simple assumptions involved i n this treat­ ment has been obtained by using scavengers. It is known that H is a specific scavenger for O H radicals in dilute solution ( 5 ) ; if this holds for concentrated solutions, then the increase in G ( N 0 " ) caused by adding H should be a measure of g ( O H ) . Results for 4.0M N 0 " are shown in Figure 2. G ( N 0 ~ ) for H -saturated solution is indicated by the hori­ zontal dashed line. Owing to uncertainty i n the hydrogen concentration, it was necessary to ensure that scavenging was complete by using η-propyl alcohol at various concentrations ( in He-swept solution ) ( Curve A, Figure 2 ). η-Propyl alcohol and H react similarly with O H and give identical stoichiometry (5). In this way, we find g ( O H ) = 1.9 for 4.0M NO3", which is practically identical with g ( O H ) o * /Η Ο· Clearly this indicates that O H is not an intermediate i n the concentrated solution processes (at low intensities) and that it is produced solely from water according to the electron fraction. 2

2

2

3

2

2

2

H 2

2

The close fit between our model and the data would seem to pre­ clude consideration of energy transfer processes, particularly i n the

In Radiation Chemistry; Hart, E.; Advances in Chemistry; American Chemical Society: Washington, DC, 1968.

158

RADIATION CHEMISTRY

1

higher concentration range above 2 M . Accordingly, we shall henceforth refer to the characteristic effects in concentrated solutions as "direct effect." Other scavenging work (7) has been done with Γ. In this case, yields were increased considerably beyond those anticipated from g ( O H ) o and g(e~) o, suggesting that Γ was interfering in the "direct effect." According to the mechanism found in dilute solution, Γ can react efficiently with N 0 H 2

H2

2

NO.

• NO," + I

(8)

This behavior then suggests that N 0 is an intermediate in the "direct effect" (the present behavior could be and probably is caused partly by increased formation of e~ from the nitrate, the electron then reacting via Reaction 1, but pulse radiolysis work show this is not a sufficient explana­ tion ). Further discussion of Γ scavenging is deferred until a presentation of the results of pulse radiolysis experiments.

Downloaded by YORK UNIV on November 9, 2012 | http://pubs.acs.org Publication Date: January 1, 1968 | doi: 10.1021/ba-1968-0081.ch010

2

Figure 1.

Dependence of G(N0 ~), G(0 ), and G(H 0 ) tions 2

2

2

2

on electron frac­

• : G(NO ~), corrected for spur scavenging of Hz by nitrate Ο : 0[O ), data of Mahlman (11) z

t

In Radiation Chemistry; Hart, E.; Advances in Chemistry; American Chemical Society: Washington, DC, 1968.

10.

DANIELS

Aqueous Nitrate System

159

[SCAVENGER] ,/AM (pH 12) 50

π

100

s

150

200

1

250

1

500

1— 1 /A

525

Γ"

Downloaded by YORK UNIV on November 9, 2012 | http://pubs.acs.org Publication Date: January 1, 1968 | doi: 10.1021/ba-1968-0081.ch010

5h

2h

n-Propanol

Figure 2.

, m M (neutral)

Effect of scavengers on G(NO,~) for He-swept 4.0M NaN0 3

A: Β: C: Δ;

τι-Propyl alcohol at neutral pH HzOz at pH 12 (upper concentration scale) η-Propyl alcohol at pH 12 (upper concentration scale) G(NOf) at pH 12 for a solution 100 μΜ in H Oz and 1 X 10~ M in τι-propyl alcohol 2

2

Pulse Radiolysis of Concentrated Solutions Investigating the pulse radiolysis of aqueous nitrate system by kinetic absorption spectroscopy led to the discovery (8, 9) of the char­ acteristic absorption spectrum of the NO,* radical i n neutral solutions—

In Radiation Chemistry; Hart, E.; Advances in Chemistry; American Chemical Society: Washington, DC, 1968.

160

RADIATION CHEMISTRY

1

> 0.5M. The amount formed was nonlinear in nitrate concentration but agreed well with Equation B. In itself, this strongly indicated that it was not produced by an O H radical reaction (Reaction 2), and this was confirmed by scavenging studies. However, these studies did show that it has a precursor—i.e., it was not itself a primary species of the "direct effect," formed by a reaction such as N 0 - -> N 0 + e~ 3

3

A clue to the possible precursor was found from the fact that in the region 380-530 κημ, another transient was found, decaying more slowly than the N 0 and by second-order kinetics. B y comparison with gasphase spectrum it is believed that this species is N 0 . The amount of N 0 formed is independent of nitrate concentration (whereas N 0 in­ creases), implying that as the amount caused by the "indirect effect" de­ creases, this is balanced by increased formation in the "direct effect." These results agree qualitatively with the deductions from the Γ scaveng­ ing experiments that N 0 is a transient of the "direct effect." Accordingly, it is suggested that N 0 is formed by Reaction 9,

Downloaded by YORK UNIV on November 9, 2012 | http://pubs.acs.org Publication Date: January 1, 1968 | doi: 10.1021/ba-1968-0081.ch010

3

2

2

3

2

3

N0 + O ^ N 0 2

(9)

3

where the precursors originate from a primary process

NO3-— —> ( N 0 + Ο + e~) MM

(10)

2

The iodide scavenging agrees quantitatively with Reaction 10, and further evidence for Reaction 9 is that the formation of N 0 depends strongly on intensity. 3

The fate of the N 0 species is of some interest, particularly since it is characteristic of the "direct effect." Decay kinetics are first order and independent of acidity in acid and neutral solutions. The decay rate increases rapidly in slightly alkaline solutions ( such that the species cannot be observed above p H 11 ) and seems to depend linearly on O H " concentration. Above p H 9, in the presence of 0 , the characteristic absorption spectrum of 0 ~ is seen. A l l this is consistent with the decay scheme's being 3

2

3

N 0 + O H " -> NO3- + O H 3

( 11 )

followed by OH-^H+O-

(12)

Ο" + 0 - » O3- etc.

(13)

2

In neutral and acid solution, the relevant process is believed to be N 0 + H 0 -> N 0 - + H + O H 3

If this is true, N 0

3

2

3

+

(14)

formation leads to a decrease in net formation of

In Radiation Chemistry; Hart, E.; Advances in Chemistry; American Chemical Society: Washington, DC, 1968.

10.

DANIELS

Aqueous Nitrate System

161

nitrite. Here we have an explanation of the results of Pikaev (15). who found the over-all G ( N 0 ~ ) to be independent of intensity in concentrated solution. Insofar as the nitrite yield arising from the water fraction was shown to increase at high intensities, this implies that the nitrite yield caused by the "direct effect" decreases at high intensities; this is clearly consistent with our findings that N O * increases with intensity, and its decay leads to decreased nitrite yields. In summary, then we have a reasonably satisfactory working model for the description of the radiolysis of concentrated solutions and the interrelationship of "indirect" and "direct" effects. However, we must consider the remaining problems. First, the energy absorption caused by the cation ( N a ) has been largely neglected. A t the moment we can say little of the role of the cation, which may indeed have specific effects if it is associated [ C a (17)] or reactive [ T l (4), P u ( 1 3 ) ] . W e have begun to investigate this topic. Secondly, the model presented here suggests that H 0 originates solely from the water as "molecular product." As such, its formation as a function of nitrate concentration should be represented by Equation Β by a straight line of zero slope. Although this appears to be the case at higher nitrate concentrations (Figure 1), anomalous behavior is found between 0.1 and 2.0M N 0 ~ . This may well be correlated with the unusual isotopic results for G ( 0 ) obtained by Mahlman (12), the deviation of nitrite values from strict linearity i n this same region ( Figure 1), and the effects observed by Mahlman (11) in acid solutions of nitrate containing Ce ( I V ) . Further work is needed in this area.

Downloaded by YORK UNIV on November 9, 2012 | http://pubs.acs.org Publication Date: January 1, 1968 | doi: 10.1021/ba-1968-0081.ch010

2

2

2

3

2

Photolysis of Nitrate Solutions Although most of the known facts about nitrate radiolysis can be accounted for on the basis of "primary" formation of reactive species suggested to be N 0 , O, and e~, much less is known about the origin of such species. Also, the possible occurrence of "molecular" processes cannot yet be evaluated. Such a process could, for example, be repre­ sented as 2

N0 --» N0 -+ Ο 3

(15)

2

followed by Ο + N 0 - -> 0 + N 0 3

2

2

(16)

in a solvent cage, where the nitrate of Reaction 16 could be the nearest neighbor of the Ο atom. The photolysis of nitrate seemed to be a promis­ ing way of investigating the occurrence of such processes. In addition, nitrate ion is characterized by a well-defined, relatively forbidden (/ — 10" ) absorption at about 4 e.v. and might provide interesting results 4

In Radiation Chemistry; Hart, E.; Advances in Chemistry; American Chemical Society: Washington, DC, 1968.

162

RADIATION CHEMISTRY

1

pertinent to consideration of energy transfer effects and the excitation of forbidden levels by subexcitation electrons. Accordingly, we have begun an extensive investigation of the ultra­ violet photolysis of nitrate, under conditions comparable to radiolysis of concentrated solutions. Photolyzing i n the first absorption band, it is found (10) that nitrite is formed by two processes in neutral solution, an initial rapid rate being self-inhibited and leading to a lower "residual rate" which is independent of most parameters investigated. The quan­ tum efficiency of the initial process is 4 Χ 10" and that of the "residual" is an order of magnitude less. Both processes can be described on the basis of Reactions 15 and 16 together with Downloaded by YORK UNIV on November 9, 2012 | http://pubs.acs.org Publication Date: January 1, 1968 | doi: 10.1021/ba-1968-0081.ch010

2

0 + N0 ->N0 2

(17)

3

The initial process shows homogeneous competition of Reactions 16 and 17, and the residual process may be described as the occurrence of Reaction 16 as a "solute-cage" reaction. From the magnitude of the quantum efficiency and the difference in behavior between the photolysis and radiolysis, it seems unlikely that the lowest excited state plays a major role in the radiolytic processes. However, nitrate has another, much stronger (/ ^ 0.3) absorption at 6 e.v., and we have recently obtained results which suggest that this level may be involved i n the radiolysis processes (discussed below). Magnitude of Direct Effect Radiolysis,

G(N0 ~)NO 2

S

The model presented for the direct effect radiolysis is essentially that of self-scavenging by nitrate of its own primary products, suggested in Reaction 10. However, the net observable yield, G ( N 0 ' ) ( N O - ) = 4, w i l l be determined i n great part by the extent of primary recombination —i.e., the reverse of Reaction 10. Such recombination may be interfered by using scavengers which w i l l lead to an increased yield of N O - " . Specifically, we have found (7) that moderate concentrations of iodide ion and small changes i n p H both increase G ( N 0 ~ ) o - to 8.0. However, a more remarkable effect is found in alkaline solution at p H 12. Measure­ ments of "molecular yield" H 0 show it to be consumed, most probably by O H radicals (or O") as in dilute solution. Hence, we have used it as a scavenger. Results for 4 . 0 M N O f are shown i n Figure 2, Curve B. A typical scavenging curve is not obtained. T o investigate this further, we used another scavenger (η-propyl alcohol), which we found func­ tioned satisfactorily i n neutral solution. A scavenging curve (Curve C , Figure 2) was obtained, whose limit was significantly less than the maxi­ mum obtained from H 0 . The difference was real as shown by irradiating 4.0M N 0 " at p H 12 i n the presence of η-propyl alcohol and H 0 simul2

2

2

N

3

3

2

:

2

3

2

2

In Radiation Chemistry; Hart, E.; Advances in Chemistry; American Chemical Society: Washington, DC, 1968.

2

10.

DANIELS

163

Aqueous Nitrate System

taneously, each at concentrations corresponding to their maximum effect. This increased G ( N 0 ~ ) by 50% of the value for η-propyl alcohol alone ( Figure 2 ). Our present thinking is that the H 0 ( as H 0 " ) is acting as scavenger for the Ο atom, 2

2

Ο + H 0 - -> Ο" + 0 - + H 2

2

2

2

+

(18)

and we evaluate G ( N 0 ~ ) - to be about 17 under these conditions. This corresponds to 6 e.v. per primary event, which also is the energy of the second ultraviolet absorption band of nitrate. Work in progress on the photolysis may show if this is more than coincidental and may indicate participation of this higher excited state in the radiolysis. Downloaded by YORK UNIV on November 9, 2012 | http://pubs.acs.org Publication Date: January 1, 1968 | doi: 10.1021/ba-1968-0081.ch010

2

N O s

Literature Cited (1) Appleby, Α., Scholes, G., Simic, M . , J. Am. Chem. Soc. 3891 (1963). (2) Bakh, Ν. Α., Medvedovskii, V. I., Revina, Α. Α., Bitiukov, B. D., Proc. All Union Conf. Radiation Chem., 1st, Moscow, 1957, p. 45. (3) Baxendale, J. H . , Fielden, Ε. M . , Keene, J. P., Proc. Roy. Soc. (London) A286, 320 (1965). (4) Bednar, J., Proc. Tihany Symp. 325 (1964). (5) Daniels, M . , Wigg, Ε. E., J. Phys. Chem. 71, 1024 (1967). (6) Daniels, M . , Wigg, Ε. E., Science 153, 1533 (1966). (7) Daniels, M . , Wigg, Ε. E., unpublished results. (8) Daniels, M . , J. Phys. Chem. 70, 3022 (1966). (9) Daniels, M . , unpublished results. (10) Daniels, M . , Meyers, R. V., Belardo, Ε. V., J. Phys. Chem. 72, 389 (1968). (11) Mahlman, Η. Α., J. Chem. Phys. 35, 936 (1961). (12) Mahlman, Η. Α., J. Phys. Chem. 67, 1466 (1963). (13) Miner, F. J., Seed, J. R., Chem. Rev. 67, 299 (1967). (14) Pikaev, A. K., Russ. Chem. Rev. 29, 235 (1960). (15) Pikaev, A. K., Gluzanov, P. Y., Yukubovich, Α. Α., Kinetics Catalysis (USSR) 4, 735 (1963). (16) Sharpatyi, V. Α., Russ. Chem. Rev. 30, 279 (1961). (17) Sowden, R. G., Trans. Faraday Soc. 55, 2084 (1959). (18) Sworski, T. J., J. Am. Chem. Soc. 77, 4684 (1955). (19) Thomas, J. K., Gordon, S., Hart, E. J., J. Phys. Chem. 68, 1262 (1964). RECEIVED January 9, 1968. Paper prepared with the support of the U . S. Atomic Energy Commission, A E C Document No. RLO-2014-3.

In Radiation Chemistry; Hart, E.; Advances in Chemistry; American Chemical Society: Washington, DC, 1968.