PROCEEDINGS
OF THE TOMSK ORDER OF THE RED BANNER OF LABOR POLYTECHNIC
INSTITUTE named after S. M. KIROV

Vol. 176
1970

RADIOLYSIS OF SOLID INORGANIC SALTS

WITH OXYGEN-CONTAINING ANIONS

Yu. A. Zakharov, V. A. Nevostruev

(Presented by the scientific seminar of the Department of Radiation Chemistry)

The level, pace, and scope of research in the field of the radiation chemistry of solid inorganic compounds have been increasing steadily in recent years. In connection with this, the amount of information contained in the specialized literature on these questions is now very large. For publications issued up to 1962, this information was systematized by V. V. Sviridov [1]; more recent reviews in this field are lacking.

At the same time, it is precisely in recent years that research has characteristically become more detailed, with the use of spectroscopy in various ranges and of refined analytical methods for studying radiolysis products; this has led to the possibility, previously absent, of discussing details of the mechanism of radiolysis of certain groups of salts. This applies to the greatest extent to the radiolysis of oxygen-containing compounds—a complex, as a rule multistage, process that includes the formation of a large number of intermediate paramagnetic particles.

Among oxygen-containing inorganic salts, the group of radiation-unstable compounds attracts the closest attention; the investigation of their radiolysis is convenient from the methodological point of view (simple structure and the possibility of obtaining single crystals; the presence among the products of radicals and molecules that can be calculated and determined spectroscopically, etc.). The regularities of radiolysis of these salts have been studied in relatively the greatest detail, and the approaches, methods, results obtained, and conclusions used in doing so, besides being of particular interest, may be used (in methodological and problem-setting respects) in studying the radiolysis of other compounds.

In accordance with the foregoing, in the present review data on radiation transformations in the group of salts most thoroughly studied in this respect are systematized and discussed—mainly nitrates, nitrites, chlorates, bromates, and perchlorates of alkali metals. Special attention is given to the most recent investigations and to the views formulated on their basis.

II. Radiolysis of Nitrates

Of all inorganic salts with an oxygen-containing anion, nitrates are studied most intensively. Within this group of salts, the nitrates of alkali metals have been studied in the greatest detail [2–10],

in less detail—the nitrates of silver [2, 11], alkaline-earth metals [2, 12—13], and lead [2, 14—16].

1. PRODUCTS OF RADIOLYSIS

a) Results of chemical analysis

The generally accepted method for determining the final products of nitrate radiolysis is the determination of nitrite by the diazotization reaction after dissolving the irradiated nitrate in water. In the overwhelming majority of works, both those that appeared in the literature after the publication of Sviridov’s monograph [1] and those considered in that book, the solutions of irradiated nitrates were analyzed only for nitrite. In those works in which the gases released upon dissolution were analyzed (see, for example, [2, 7, 16]), it was shown that oxygen is the gaseous product of radiolysis. The ratio \(G(\mathrm{NO}_2^-)/G(\mathrm{O}_2) \simeq 2\) is maintained in all cases, with the exception of the initial stage of radiolysis of lead nitrate [16].

However, as was shown by Cunningham [7], radical products such as \(\mathrm{O}^-\), \(\mathrm{O}_2^-\), \(\mathrm{NO}_2\), and NO may also contribute to the yield of the radiolysis product determined as nitrite by the diazotization reaction. Indeed, by comparing the results of two analytical methods (cerium and diazotization), it was established in that work that, upon irradiation of \(\mathrm{KNO}_3\) and \(\mathrm{NaNO}_3\) at 77 K, the radiation-chemical yield of nitrite proper amounts to at most no more than half of the value obtained in the analysis for nitrite by the diazotization method. This deviation is smaller when potassium nitrate is irradiated at room temperature, but remains significant in the case of \(\mathrm{NaNO}_3\). Apparently it may be considered proven that the principal products of radiolysis at room temperature, at least for potassium and sodium nitrates, are \(\mathrm{NO}\), \(\mathrm{NO}_2^- + \mathrm{NO}_2\), and oxygen. The presence of nitrogen in the gaseous products of radiolysis noted in [17], as shown by Cunningham [7], is the result of desorption of air during irradiation of nitrates. It should be noted that the oxygen determined after dissolution of the irradiated nitrate in water is present in the crystal lattice at low doses, apparently not only in the form of molecules or atoms, but also in the form of \(\mathrm{O}^-\), \(\mathrm{O}_2^-\), and \(\mathrm{O}_2^{2-}\) ions; this explains to some extent its retention in the lattice during irradiation. In this connection, the assumptions [18] concerning the possibility of metal formation during irradiation of nitrates appear unacceptable.

Since chemical analysis of irradiated nitrates is carried out after dissolving them in water—a process during which the chemical state of the radiolysis products may change—it is difficult by this method to detect either the primary radiolysis products or the final products in the nitrate lattice. The most reliable information about them can be obtained by studying the optical absorption spectra and EPR spectra of irradiated nitrates, since these methods do not require destruction of the crystal lattice.

b) Optical absorption in irradiated nitrates

One of the first studies of optical absorption in irradiated \(\mathrm{NaNO}_3\) was that of Pringsheim [19]. At the temperature of liquid nitrogen an absorption band was found with a maximum at

335 mµ, and at room temperature, in addition to this band, another appears with a maximum at 355 mµ.

The first of the bands was attributed by the author to an electron-excess center, the second to \(\mathrm{NO_2^-}\) ions, since sodium nitrate, to whose melt nitrite had been added, also has an absorption band with a maximum at about 355 mµ. Absorption maxima at \(\sim 350\) mµ were also found by Johnson [16] in studying reflection spectra in irradiated \(\mathrm{KNO_3}\) and \(\mathrm{Pb(NO_3)_2}\). The most detailed study of this question in the case of \(\mathrm{KNO_3}\) is the work of Cunningham [4]; it therefore seems of interest to consider this work in somewhat greater detail. In Fig. 1, taken from [4], the absorption spectrum of a crystal of

Fig. 1. Optical absorption spectrum in \(\mathrm{KNO_3}\) irradiated with \(\gamma\)-rays at 4.2°K [3]. The points represent experimental values. Curve 1 was obtained by summing curves 2, 3, 4, 5, 6, and 7.

\(\mathrm{KNO_3}\) irradiated at 4.2°K is presented. In the dose range below \(10^{20}\) eV/g, six absorption bands were found, both at 4.2°K and at 77°K, with maxima at wavelengths 520, 355, 325, 290, 268, and 250 mµ. Study of the behavior of these bands under irradiation with light of a definite wavelength and with changes in temperature, as well as comparison with the behavior of EPR signals under this type of treatment, enabled the author to identify these absorption bands.

Thus, for example, the bands at 268 and 250 mµ decrease slightly when a crystal irradiated at 4.2°K is stored at the same temperature for 30 min; further storage for another 1 hour does not lead to changes in absorption at these wavelengths. Annealing for 16 hours at 77°K of a crystal irradiated at the same temperature decreases the absorption at 268 and 520 mµ by a factor of two, and at 250 mµ by a factor of three. On the basis of the analogous behavior of the EPR signals [3, 4] attributed to \(\mathrm{NO_2^{-2}}\) ion-radicals and \(\mathrm{NO_3}\) radicals, and also from experiments involving bleaching of these bands and the EPR signals, the absorption at

268 and 520 mμ was attributed to ion-radicals \(\mathrm{NO_2^{-2}}\), and the absorption at 250 mμ to \(\mathrm{NO_3}\) radicals.

Irradiation with light of wavelength 520–680 mμ decreases the absorption at 520, 290, 268, and 250 mμ. Since light of this wavelength should be absorbed by the ion-radicals \(\mathrm{NO_2^{-2}}\), the weakening of absorption at 520 and 268 mμ was attributed to a decrease in \(\mathrm{NO_2^{-2}}\) due to ionization. The weakening of the bands at 290 and 250 mμ was explained by a decrease in the concentration of hole centers due to capture of the electron released upon ionization of \(\mathrm{NO_2^{-2}}\). Thus, these experiments also confirm the above identification of the absorption bands at 520, 268, and 250 mμ. It should be noted that the absorption bands at 325 and 355 mμ remain unchanged under such treatment, which was explained by the electron-excess nature of the centers responsible for these bands. During storage of the crystals at 77°K, the absorption at 325 mμ also decreases; moreover, if crystals with bands previously bleached by light and attributed to \(\mathrm{NO_2^{-2}}\) are annealed, the decrease in absorption at 325 mμ is accompanied by the appearance and growth of absorption at 270 and 520 mμ.

On the basis of these data and of data concerning the behavior of the EPR signal attributed to \(\mathrm{NO_3^{-2}}\) [3], the absorption band at 325 mμ was assigned to the ion-radicals \(\mathrm{NO_3^{-2}}\).

The center responsible for the band at 290 mμ, which is stable at liquid-helium and liquid-nitrogen temperatures but is annealed upon heating to room temperature, was not identified. As for the absorption band at 355 mμ, which Pringsheim [19] attributed to nitrite ions and which is stable upon heating the crystal up to room temperature, it was likewise not identified in this work. The assignment of this band to \(\mathrm{NO_2^-}\) ions in [4] is doubtful, since, because of the very small extinction coefficient of nitrite at 355 mμ, the absorption caused by it at the concentrations obtained in this work and determined chemically lies beyond the experimental error. In a later work [20], this absorption band (C-center) was identified by Cunningham as the oxyhyponitrite ion \(\mathrm{N_2O_3^{-2}}\).

Upon irradiation of \(\mathrm{KNO_3}\) with a dose greater than \(10^{20}\) eV/g, an additional absorption band appearing at 410 mμ was found. Since the EPR signal attributed to \(\mathrm{NO_2}\) [3] also appears at doses exceeding \(10^{20}\) eV/g, this absorption band was assigned to \(\mathrm{NO_2}\). Bleaching of the \(\mathrm{NO_2^{-2}}\) band with light of wavelength greater than 520 mμ also leads to a decrease of the 410 mμ band, which is apparently due to capture of an electron by an \(\mathrm{NO_2}\) molecule. It is appropriate to note here that in [21, 22] the absence of a signal attributed to \(\mathrm{NO_2}\) in \(\mathrm{KNO_3}\) irradiated at liquid-nitrogen temperature was established if the initial nitrate contained no nitrite or if the pure nitrate was not heated between irradiations to room temperature. However, this result apparently does not contradict Cunningham’s data, since the doses used in [21] were less than \(10^{20}\) eV/g, although in a later work by these authors [5] \(\mathrm{NO_2}\) was found in \(\mathrm{KNO_3}\) irradiated at 77°K with a dose of \(2 \cdot 10^{19}\) eV/g.

In sodium nitrate [20] irradiated at 77°K, one absorption band with a maximum at 335 mμ was found (Fig. 2). However, analysis of the curve shape and experiments with UV-light bleaching allowed the author to identify two particles responsible for this band.

($\mathrm{NO_2^{-2}}$ and $\mathrm{NO_3^{-2}}$) and to detect absorption at $\sim 410\,m\mu$, attributed to $\mathrm{NO_2}$. It is interesting to note that in this work [20] an inhibiting effect was observed of $\mathrm{Ag^+}$ ions introduced into the $\mathrm{KNO_3}$ lattice

Fig. 2. Light absorption in $\mathrm{NaNO_3}$. 1—irradiation with X-rays at 77°K for 75 minutes (right-hand scale); 2—irradiation with $\gamma$ rays at 77°K for 30 minutes (left-hand scale); 3—same as 2, after irradiation with light of $\lambda > 510\,m\mu$ at 77°K for 40 minutes

and $\mathrm{NaNO_3}$, on the growth of absorption bands assigned to nitrogen–oxygen centers.

c) EPR spectra in irradiated nitrates

The signals of electron paramagnetic resonance have been studied in considerable detail in irradiated sodium [24—29] and potassium nitrates [3, 5, 7, 21, 22]. There is also information on EPR spectra in irradiated silver [11], strontium [12], and lead [15] nitrate. In all the nitrates studied, the presence of $\mathrm{NO_2}$ has been established. The $\mathrm{NO_3}$ radical has been detected in $\mathrm{KNO_3}$ [3, 5], $\mathrm{NaNO_3}$ [26], and $\mathrm{Pb(NO_3)_2}$ [15], and $\mathrm{NO_3^{-2}}$ in $\mathrm{KNO_3}$ [3, 5, 21, 22] and in $\mathrm{Sr(NO_3)_2}$ [12]. A similar paramagnetic species has also been found in $\mathrm{AgNO_3}$; however, in this case the unpaired electron is not localized only on the anion, but also has splitting on the silver nucleus [11]. The presence of $\mathrm{NO}$, $\mathrm{NO_2^{-2}}$, and $\mathrm{O_2^-}$ in irradiated $\mathrm{KNO_3}$ is indicated only in work [3]. Thus, in one and the same nitrate and under identical or similar conditions, different authors have detected different paramagnetic species. Thus, for example, upon irradiation of $\mathrm{KNO_3}$ with a dose of less than $10^{20}$ eV/g, which Cunningham considers to be the threshold dose for the formation of $\mathrm{NO_2}$, at 77°K Zeldes [5] identified the presence of $\mathrm{NO_3^{-2}}$, $\mathrm{NO_3}$, and $\mathrm{NO_2}$, whereas Cunningham found only $\mathrm{NO_2^{-2}}$ [3]. Although the doses in these works differed by approximately an order of magnitude, the differences in the results cited by the author cannot be explained solely by the difference in irradiation dose. These discrepancies are more clearly seen from the characteristics of the spectra assigned to paramagnetic species in irradiated $\mathrm{KNO_3}$, given in Table 1.

Despite the fact that the EPR spectra in irradiated $\mathrm{NaNO_3}$ have been studied in many works, only the spectrum attributed to nitrogen dioxide has been reliably interpreted. In two works [26, 29], in addition to $\mathrm{NO_2}$, $\mathrm{NO_3}$ radicals were identified. As can be seen from Table 2, which gives the charac-

Table 1

Characteristics of the EPR spectra of paramagnetic particles in irradiated KNO$_3$

Presumed radical	Values of the $g$ factor	HFS constants (gauss)	Literature	Note
NO	2.008 2.010 2.005	3.5 4.0 31	[3]	Primary product at 77°K and 4°K.
NO$_2$	1.9996 1.9932 2.0055 1.999	62.9 50.3 48.7 113	[21] [3]	Heating to room temperature between two irradiations at 77°K or the presence of KNO$_2$ in the nitrate. KNO$_3$ powder, dose $>3 \cdot 10^{20}$ eV/g.
NO$_3^{2-}$	2.025 2.025 2.008	9 9 9	[3]	4°K, dose $3 \cdot 10^{18}$ eV/g. At 77°K it is absent; upon annealing at 77°K for 2 h it decreases by a factor of two.
NO$_3^{2-}$	2.0057 2.0057 2.0015	63.4 31.8 31.8	[21]	77°K, dose $4 \cdot 10^{19}$ eV/g, $H \parallel a$.
NO$_2^{2-}$	2.006 2.006 2.002	61 32 32	[3]	Primary product at 4 and 77°K. Crystal axis $c \perp$ the rotation axis and $\parallel H$.
NO$_3$	2.006 2.006 2.001	0 0 0	[3]	4°K.
NO$_3$	2.0232 2.0232 2.0031	4.3 3.5 3.5	[5]	77°K, rapidly decomposes at this temperature.
O$_2^-$	1.958	0	[3]	Singlet. Appears upon heating to 77°K of a crystal irradiated at 4°K, or upon measurement at 4°K of a crystal irradiated at 77°K.

characteristics of the EPR signals in NaNO$_3$ and the irradiation conditions, different spectra are assigned to the NO$_3$ radical. Jesse and Katsumata [26] assign to this radical a spectrum analogous to that found in irradiated KNO$_3$ by Cunningham [3]. Addae, however, identified NO$_3$ in agreement with Zeldes [5] (cf. the spectra of NO$_3$ in Tables 1 and 2).

Thus, the presence of NO$_2$ and NO$_3$ in irradiated NaNO$_3$ may be regarded as established. If the data of Tables 1 and 2 are compared, it is evident that the presence of NO$_2$ also admits of no doubt in KNO$_3$. In [27] it was shown that the matrix has practically no effect on the characteristics of the EPR spectrum due to NO$_2$. Therefore the presence of NO$_2$ in other nitrates also raises no doubt, since in them too a characteristic EPR spectrum is observed, as is seen from Table 3.

Comparing the data presented in Tables 1, 2, and 3, it is not difficult to notice some contradictions in the identification of individual paramagnetic particles. In this connection the following remarks seem appropriate to us.

1. If the spectral characteristics of certain paramagnetic particles remain the same or similar in the lattices of different nitrates (which, apparently, is in fact the case), then the EPR spectrum attributed by Cunningham to \(\mathrm{NO_2^{-2}}\) in \(\mathrm{KNO_3}\) [3], by Zeldes to \(\mathrm{NO_3^{-2}}\) in \(\mathrm{KNO_3}\) [22], by Mosley and Moulton to \(\mathrm{NO_3^{-2}}\) in \(\mathrm{AgNO_3}\) [11], and by Jackard to \(\mathrm{NO_3^{-2}}\) in nitrate-containing KCl [23], must be assigned to one of these

Table 2

Characteristics of EPR spectra in \(\mathrm{NaNO_3}\)

Irradiation conditions	Paramagnetic particle	\(g_{\parallel}\langle 111\rangle\)	\(g_{\perp}\langle 111\rangle\)	\(A_{\parallel}\langle 111\rangle\)	\(A_{\perp}\langle 111\rangle\)	Literature	Note
\(300^\circ\mathrm{K}\) \(10\)—\(250\) Mrad	\(\mathrm{NO_2}\)	2.0059	1.9982	48.0	58.4	[25]
\(77^\circ\mathrm{K}\)	\(\mathrm{NO_2}\)	2.0058	1.9964	50.5	59.1	[27]
\(170^\circ\mathrm{K}\)	\(\mathrm{NO_2}\)	2.0056	1.9966	51.1	58.4	[28]
\(300^\circ\mathrm{K}\)	\(\mathrm{NO_2}\)	2.0050	1.9970	53.0	57.8	[28]
\(300^\circ\mathrm{K}\)	\(\mathrm{NO_2}\)	2.006	1.9989	53	59	[26]	Disappears at \(400^\circ\mathrm{K}\)
\(300^\circ\mathrm{K}\) \(10^7\)—\(10^8 r\)	?	2.006	2.0097	0	0	[26]	—“—
\(300^\circ\mathrm{K}\) —“—	?	2.0064	2.0147	0	0	[26]	Stable above \(530^\circ\mathrm{K}\).
\(300^\circ\mathrm{K}\) —“—	?	2.006	2.0237	0	0	[26]	Disappears at \(400^\circ\mathrm{K}\); at \(300^\circ\mathrm{K}\) irreversibly converts into the three preceding ones.
\(77^\circ\mathrm{K}\) —“—	\(\mathrm{NO_3}\)	2.0004	2.0053	0	0	[26]	Disappears at \(400^\circ\mathrm{K}\); at \(300^\circ\mathrm{K}\) irreversibly converts into the three preceding ones.
\(77^\circ\mathrm{K}\) \(10^{18}\) eV/g	\(\mathrm{NO_3}\)	2.0022	2.0217	4.11	3.51	[29]
\(77^\circ\mathrm{K}\) and \(300^\circ\mathrm{K}\)	?	2.0030	2.0225	0	0	[29]

paramagnetic particles. The same also applies to the identification of spectra attributed by Cunningham in \(\mathrm{KNO_3}\) [3] and by Zhdanskii and Shrobeck in \(\mathrm{Sr(NO_3)_2}\) [12] to the particle \(\mathrm{NO_3^{-2}}\), and by Livingston and Zeldes in \(\mathrm{KNO_3}\) [5] and by Golding and Henchman in \(\mathrm{Pb(NO_3)_2}\) [15] to the particle \(\mathrm{NO_3}\). This question was considered in rather great detail in the work of Atkins and Symons [30]. On the basis of comparison of the characteristics of these spectra with the spectra of isoelectronic radicals, they assigned the first spectrum to the radicals \(\mathrm{NO_3^{-2}}\) and the second to \(\mathrm{NO_3^{0}}\).

1. The spectrum attributed by Cunningham [3] and Jackard [23] to NO radicals, in [30], is assigned to \(\mathrm{NO_2^{-2}}\) particles.

In [7], however, Cunningham showed, by chemical analysis and by the EPR method, the accumulation of NO with dose with acceptable agreement. These data are shown in Fig. 3. In view of this, and also owing to the very strong anisotropy of the hyperfine-splitting constant, Cunningham’s point of view on this question is probably closer to reality.

As for the identification of the radicals \(\mathrm{NO_3}\) and \(\mathrm{NO_3^{-2}}\), for the time being it is difficult to give preference to any one of the opinions on this matter found in the literature. It would also be desirable, as in the case of NO, to have some independent proof of their presence in the crystal. The available data on optical spectra, of course, may serve

Table 3

Characteristics of EPR spectra in irradiated nitrates of nontransition metals

Paramagnetic particle	Irradiated salt	\(g_x\)	\(g_y\)	\(g_z\)	\(A_x\)	\(A_y\)	\(A_z\)	Literature	Note
\(\mathrm{NO_3}\)	\(\mathrm{AgNO_3}\)	2,009	1,9978	2,0039	50,7	48,5	69,6	[11]	Irradiation at 77 K. The spectrum grows in parallel with the spectrum with accumulation of \(\mathrm{Ag^{+2}}\) upon heating the crystals from 77 to 140 K.
—“—	\(\mathrm{Sr(NO_3)_2}\)	1,9975	1,9975	2,0054	59	59	50	[12]	Room temperature. Formed upon irradiation at \(10^5\) r, if the nitrate contains nitrite.
—“—	\(\mathrm{Pb(NO_3)_2}\)	1,995	1,995	2,004	57	57	50	[15]	Room temperature.
\(\mathrm{NO_3}\)	\(\mathrm{Pb(NO_3)_2}\)	2,029	2,029	1,998	\(\sim 1\)	\(\sim 1\)	\(\sim 2\)	[15]	—“—
\(\mathrm{NO_3^{-2}}\)	\(\mathrm{AgNO_3}\)	2,002	2,007	2,000	35,8	37,1	67,4	[11]	Temperature 77 K. The unpaired electron is localized not only on \(\mathrm{NO_3^-}\), but also has hyperfine splitting on Ag.
—“—	\(\mathrm{Sr(NO_3)_2}\)	2,0081	2,0081	2,0027	0	0	1,2	[12]	Room temperature. Dose \(\sim 10^5\) r.

…some confirmation of Kanningham’s point of view, but these data cannot be considered completely independent of EPR, since the absorption bands were identified, to some extent, on the basis of EPR data.

Fig. 3. Formation of NO in KNO$_3$ and NaNO$_3$ irradiated at 77°K. Curves a and c are results of chemical analysis; b and d were obtained by the EPR method.

As we have seen above, the indicated contradictions are also inherent in the identification of EPR spectra in irradiated NaNO$_3$.

2. KINETICS OF RADIOLYSIS

It was already noted above that the products of nitrate radiolysis are various nitrogen and oxygen oxides and oxyanions, and not only nitrite and oxygen, as had been assumed before the publication of Kanningham’s work [7]. However, in this same work it was shown that the concentration of radical products (NO, in particular) rapidly reaches saturation both under irradiation at 77°K (Fig. 3) and at room temperature. On this basis, and also because of the absence of data on the kinetics of accumulation of individual radiolysis products that contribute to the concentration determined by the diazotization reaction, we shall henceforth consider the kinetic regularities of nitrate radiolysis in terms of the yield of nitrite. It should be borne in mind here that the quantity $G(\mathrm{NO}_2^-)$ includes several quantities [7]: $G(\mathrm{NO}_2^-)$, $G(\mathrm{NO}_2)$, $G(\mathrm{NO})$, $G(\mathrm{O}^-)$, etc.

a) Dependence of the radiation-chemical yield on the absorbed dose

The independence of $G(\mathrm{NO}_2^-)$ from dose for individual nitrates has been noted in many works [7, 8, 10, 16, 18, 31–33]. Along with this, however, in some cases a decrease in the radiation-chemical yield of nitrite with increasing dose was found. The most characteristic such course of radiolysis is for potassium nitrate [32, 34, 35], and moreover na-

accumulation of nitrite in this case was described by two intersecting straight lines with different slopes. As was shown in the work of Johnson [2, 32], deviation from linearity in the accumulation of nitrite with increasing dose is, generally speaking, characteristic of any nitrate, and approximation of this dependence by a straight line is possible only in the region of small doses, which differ for different nitrates. This also makes understandable the fact that, in the case of KNO$_3$, nonlinearity in the accumulation of NO$_2^-$ with dose was observed especially often, since deviation from linearity for KNO$_3$ occurs already at a dose of \(3 \cdot 10^{21}\) eV/g, whereas in the case of NaNO$_3$ this dose is \(\sim 2 \cdot 10^{22}\) eV/g and was not reached in many studies.

In the work of Cunningham and Heal [35], in which doses above \(10^{23}\) eV/g were used, a decrease in \(G(\mathrm{NO}_2^-)\) with increasing dose was also observed for a number of nitrates. For KNO$_3$ the kinetics of radiolysis was described by the equation

\[ \alpha = 1 - e^{bD}, \tag{1} \]

where \(\alpha\) is the fraction of nitrate decomposed, and \(D\) is the absorbed dose.

This question was studied in greatest detail for the nitrates of potassium, cesium, sodium, barium, lead, and silver by Johnson [2, 32]. In these works, using the following kinetic scheme for decomposition:

\[ \mathrm{MNO}_3 \xrightarrow{\gamma} \mathrm{MNO}_2 + \mathrm{O}, \tag{2} \]

\[ \mathrm{O} + \mathrm{NO}_2^- \rightarrow \mathrm{NO}_3^-, \tag{3} \]

\[ \mathrm{O} + \mathrm{NO}_3^- \rightarrow \mathrm{NO}_2^- + \mathrm{O}_2 \tag{4} \]

and from the assumption of a stationary concentration of oxygen atoms, a second-order equation was obtained for nitrite accumulation of the form

\[ a[\mathrm{NO}_2^-]^2 + [\mathrm{NO}_2^-] = bD, \tag{5} \]

where \(a\) and \(b\) are constants for the given salt, and \(D\) is the absorbed dose.

Fig. 4. Formation of nitrite in NaNO$_3$ [2]. \(o, x\) — experimental points representing, respectively, \([\mathrm{O}_2]\) and \(1/2[\mathrm{NO}_2^-]\). The solid line was calculated from equation (5).

The applicability of equation (5) proved to be sufficiently good for CsNO$_3$, NaNO$_3$, Ba(NO$_3$)$_2$, and Pb(NO$_3$)$_2$. As an illustration, Fig. 4 shows the experimentally found accumulation of nitrite and oxygen.

in NaNO₃ and the accumulation of nitrite calculated from equation (5). In contrast to the foregoing, in the case of KNO₃ and AgNO₃ it proved impossible to describe the radiolysis by this equation. It should be noted that equation (5) was obtained without taking into account the change in concentration during irradiation. From the above scheme of radiolysis (equations 2—4), taking into account the nonconstancy of \([\mathrm{NO}_3^-]\), the following equation was obtained in [2]:

\[ a \lg (1-a)+\frac{a}{1-a}=bD, \tag{6} \]

where

\[ a=\frac{[\mathrm{NO}_2^-]}{[\mathrm{NO}_3^-]_0}; \]

\(a\) and \(b\) are constants,

\(D\) is the absorbed dose, which proved applicable for describing the accumulation of nitrite in KNO₃.

For the kinetics of radiolysis of AgNO₃ it was not possible to find a suitable equation from the decomposition scheme according to reactions (2—4).

It is not difficult to see that all three kinetic equations (1, 5, 6), at sufficiently low doses, can be simplified with some approximation to an expression of the form

\[ [\mathrm{NO}_2^-]=bD \tag{7} \]

which is what is observed in most works carried out using small absorbed doses.

The applicability of kinetic equations derived from the radiolysis scheme (2—4) can in no way serve, of course, as confirmation of the mechanism of radiolysis according to reactions (2—4), but indicates only that at high irradiation doses it is necessary to take reverse reactions into account.

b) Effect of dose rate and LET

The effect of dose rate on nitrate radiolysis was studied incidentally while solving other problems in works [10, 33, 35]. This question was considered specifically by Johnson [36]. When the dose rate was varied from \(10^{17}\) to \(10^{20}\) eV/g sec, using potassium, sodium, cesium, and lead nitrates as examples, this work showed the absence of its influence on \(G(\mathrm{NO}_2^-)\), as also in works [10, 33, 35] in the range of other dose rates and other types of radiation.

It is interesting to note that the \(G\) values obtained in this work using electrons of energy 1 and 0.5 MeV agree with the values obtained for γ-irradiation. Thus, the conclusion that the dose rate does not affect the radiation-chemical yield of nitrite over a wide range of values (from \(\sim 10^{15}\) to \(10^{20}\) eV/g sec) made it possible to interpret unambiguously the results on the influence of the type of radiation on \(G(\mathrm{NO}_2^-)\). Table 4, borrowed from the work of Hochanadel [10], gives the values of \(G(\mathrm{NO}_2^-)\) obtained in the radiolysis of nitrates by different types of radiation. The closeness of the values of \(G(\mathrm{NO}_2^-)\) obtained under the action of X-rays and γ-rays at 150°C to the values of \(G(\mathrm{NO}_2^-)\) obtained under the action of α-rays allowed the author to conclude that thermal spikes are formed in the tracks of α-particles. Therefore, the increase in the radiation-chemical yield of nitrite with increasing LET was attributed to local heating along the tracks of heavy particles. From the dependence of \(G(\mathrm{NO}_2^-)\) on temperature

upon irradiation with light types of radiation was estimated to have an apparent activation energy of 2.8 kcal/mole.

Table 4

Effect of LET on \(G(\mathrm{NO_2^-})\) in nitrates [10]

Irradiated salt	\(\alpha\)—particles, 3.4 MeV, 25°C	\(\alpha\)—particles, 3.4 MeV, 120°C	\(\gamma\)-rays \(^{60}\mathrm{Co}\), 30°C	\(\gamma\)-rays \(^{60}\mathrm{Co}\), 150°C	X-rays, 44.5 keV, 15°C	X-rays, 44.5 keV, 150°C
\(\mathrm{LiNO_3}\)	0.7		0.02–0.2	0.03–0.4	0.02
\(\mathrm{NaNO_3}\)	1.3	1.1	0.27	1.0	0.4	1.1
\(\mathrm{KNO_3}\)	2.2		1.5	3.0	2.0	3.0
\(\mathrm{CsNO_3}\)	1.4		1.6	2.3	1.4	1.2
\(\mathrm{Ba(NO_3)_2}\)	1.6		1.8	1.6	1.8	1.8

(c) The role of the nature of the cation

From the very first works devoted to the question under consideration, a definite dependence of \(G(\mathrm{NO_2^-})\) on the nature of the cation in nitrates was noted, a dependence that is absent in the case of aqueous solutions. The generally accepted explanation of this dependence at present is allowance for the influence of the free volume of the lattice on the probability of the reverse reaction (3). This question has been considered in sufficient detail in previous reviews [1, 37], and therefore we shall not dwell on it in detail. It is appropriate to note, however, that despite the existing correlation between \(G(\mathrm{NO_2^-})\) and the magnitude of the free volume, one should not seek the causes of the change in \(G(\mathrm{NO_2^-})\), for example, along the series from \(\mathrm{LiNO_3}\) to \(\mathrm{CsNO_3}\), solely in the difference in the magnitude of the free volume in the same series, especially at the initial stage of radiolysis. Apparently, an essential role should be played not so much by the spatial characteristics of the lattice as by the electronic ones, since the radiation interacts with the electron shells of the constituents of the lattice. In all the works considered here, attention is focused only on the interaction of the radiation with the nitrate ion. Therefore one would expect an increase in \(G(\mathrm{NO_2^-})\) with an increase in the electron fraction of the anion in a series of salts of the same type, for example, alkali-metal nitrates. Probably because of the opposite dependence observed experimentally, this question has not received due attention in the literature. In our view, alongside \(\mathrm{NO_3^-}\), the interaction of the cations with the radiation should make a substantial contribution to the yield of radiolysis products. A similar point of view has already been expressed in the literature [3, 4, 38] without detailed development, and was first confirmed experimentally using \(\mathrm{AgNO_3}\) as an example [11].

Indeed, if the formation of radiolysis products in nitrates were caused only by the interaction of \(\mathrm{NO_3^-}\) with the radiation, then \(G(\mathrm{NO_2^-})\), calculated per 100 eV of energy absorbed only by the anion, should be the same in one and the same salt when different radiations are used. The radiation-chemical yields of nitrite calculated by us per 100 eV of energy absorbed only by the cation, \(G^{k}_{(\mathrm{NO_2^-})}\), or by the anion, \(G^{a}_{(\mathrm{NO_2^-})}\), on the basis of literature data on \(G(\mathrm{NO_2^-})\) for \(\mathrm{KNO_3}\) and \(\mathrm{NaNO_3}\), are given in Table 5. Analysis of this table permits the following considerations to be expressed:

Table 5

Effect of cations on \(G(\mathrm{NO}_2^-)\) in nitrates

Salt	Type of radiation	\(G^a(\mathrm{NO}_2^-)\)	\(G^k(\mathrm{NO}_2^-)\)	Literature source for calculating \(G^a\) and \(G^k\)
\(\mathrm{KNO}_3\)	\(\gamma\)—\(\mathrm{Co}^{60}\)	\(2.4 \pm 0.2\)	\(4.0 \pm 0.3\)	[32–34]
\(\mathrm{KNO}_3\)	RL—45 keV	24.5	2.1	[35]
\(\mathrm{KNO}_3\)	RL—200 keV	22.8	2.4	[8]
\(\mathrm{KNO}_3\)	\(\alpha\)—3.4 MeV	3.4	6.1	[10]
\(\mathrm{KNO}_3\)	\(e\)—1.0 and 0.5 MeV	2.3	4.2	[36]
\(\mathrm{NaNO}_3\)	\(\gamma\)—\(\mathrm{Co}^{60}\)	\(0.25 \pm 0.05\)	\(0.7 \pm 0.1\)	[32, 34]
\(\mathrm{NaNO}_3\)	RL—45 keV	0.8	0.7	[35]
\(\mathrm{NaNO}_3\)	RL—200 keV	0.6	0.7	[8]
\(\mathrm{NaNO}_3\)	\(e\)—1.0 and 0.5 MeV	0.3	0.8	[36]
\(\mathrm{NaNO}_3\)	\(\alpha\)—3.4 MeV	1.7	5.5	[10]

1. As should be expected, for \(\gamma\)-rays from \(\mathrm{Co}^{60}\) and for electrons, whose energy absorption is proportional only to the electron density, both \(G^a_{(\mathrm{NO}_2^-)}\) and \(G^k_{(\mathrm{NO}_2^-)}\) coincide in one and the same salt.

2. The radiation-chemical yield of nitrite per anion in the case of X-rays differs by an order of magnitude from \(G^a_{(\mathrm{NO}_2^-)}\) in the case of \(\gamma\)-irradiation in \(\mathrm{KNO}_3\) and by a factor of three in \(\mathrm{NaNO}_3\), while the corresponding values of \(G^k_{(\mathrm{NO}_2^-)}\) differ only by a factor of two for \(\mathrm{KNO}_3\) and coincide for \(\mathrm{NaNO}_3\). Since the regularities of energy absorption for soft X-rays and for \(\gamma\)-rays from \(\mathrm{Co}^{60}\) differ greatly, it is precisely here that the contribution of the anion and cation to the nitrite yield should become manifest. These results therefore make it possible to speak of the predominant contribution to the nitrite yield of the cation, rather than the anion.

3. The very large difference in the values of \(G^k_{(\mathrm{NO}_2^-)}\) for \(\mathrm{KNO}_3\) and \(\mathrm{NaNO}_3\), as well as the noncoincidence of \(G^a_{(\mathrm{NO}_2^-)}\) and \(G^k_{(\mathrm{NO}_2^-)}\) for \(\alpha\)-rays with the corresponding values for \(\gamma\)-rays and electrons, make it possible to conclude that the processes of interaction of the cation with radiation are not the only source of radiolysis. Apparently, spatial factors in the case of light types of radiation and track phenomena in the case of \(\alpha\)-particles also exert a substantial influence.

d) Influence of other factors

The data on the magnitude of \(G_{(\mathrm{NO}_2^-)}\), obtained by different authors, as shown in Sviridov’s monograph, differ very greatly. At present there are already some experimental facts that make it possible to consider the possible causes of the divergence among the results of different authors. One such possible cause may be the unequal dispersity of the irradiated nitrates. In work [2] it was noted, for example, that different \(G(\mathrm{O}_2)\) values are observed in crystals of different dispersity. Since the presence of trapped oxygen promotes the occurrence of reverse reactions, the dis—

dispersity should also affect \(G(\mathrm{NO}_2^-)\). We studied this question in greater detail [8], and it was shown that the influence of the sizes of the irradiated crystals is especially great for finely dispersed powders; moreover, with increasing dispersity not only \(G(\mathrm{O}_2)\) increases, but also \(G(\mathrm{NO}_2^-)\).

In the same work we studied the influence of heterophase impurities of semiconducting oxides on the rate of radiolysis of \(\mathrm{KNO}_3\) and \(\mathrm{NaNO}_3\). It was found that oxides with the smallest electron work function \(\varphi\) accelerate radiolysis most strongly, whereas oxides possessing a large \(\varphi\) retard it. Since the catalytic activity of the oxides we studied correlates with the electron work function, it was assumed that additives—donors of electrons \((I_{\text{add.}} < I_{\text{salt}})\)—accelerate radiolysis, whereas oxides—acceptors with respect to nitrates—retard decomposition.

Similar data were obtained by Cunningham [20, 31] with a homophase addition of \(\mathrm{AgNO}_3\) in the lattice of irradiated sodium nitrate. The decrease in the rate of radiolysis of sodium nitrate in the presence of \(\mathrm{Ag}^+\) ions in the lattice was attributed to the acceptor properties of \(\mathrm{Ag}^+\). On the other hand, the introduction of \(\mathrm{Tl}^+\) ions, which are electron donors, as shown in [39], accelerates the radiolysis of \(\mathrm{KNO}_3\). Although there are still not enough experimental data in this respect to draw unambiguous conclusions about the mechanism of the influence of impurities, it may be stated that the purity of the irradiated salt, as well as the dispersity of the crystals, has a substantial influence on radiation-chemical stability.

3. MECHANISM OF RADIOLYSIS

The observed ratio \(G(\mathrm{NO}_2^-)/G(\mathrm{O}_2) \simeq 2\), as well as the agreement of the kinetics of accumulation of nitrite and oxygen with the equation derived on the basis of reactions (2—4), served as the basis for interpreting the mechanism of nitrate radiolysis according to the equations given above. However, recently, in the works of Cunningham [4, 7, 20, 31] and in studies of EPR spectra in irradiated nitrates discussed above, the complexity of both the primary and secondary radiation-chemical processes during irradiation of nitrates has been convincingly shown. The contradictions noted above in the interpretation of EPR spectra in irradiated nitrates concern mainly the identification, and not the existence, of particular paramagnetic particles. Therefore, at present it is possible to describe individual putative stages of nitrate radiolysis on the basis of the experimental data considered above.

The primary processes should include the processes of ionization and excitation:

\[ \mathrm{NO}_3^- \rightsquigarrow \mathrm{NO}_3^{-*},\quad \mathrm{NO}_3,\quad \bar e, \tag{8a} \]

\[ Me^+ \rightsquigarrow Me^{++} + \bar e. \tag{8б}{}^{1)} \]

Since excitation of the electronic shells of the cation does not lead to chemical reactions, the process \(Me^+ \rightsquigarrow Me^{+*}\) may be disregarded.

If the formation of \(\mathrm{NO}_3\) during irradiation of nitrates was experimentally observed by many authors (Tables 1, 3), then reaction \((8б)\) in experiment

\(^{1)}\) Here and below \(Me^+\) denotes a normal cation (\(\mathrm{Na}^+\), \(\mathrm{K}^+\), \(\mathrm{Sr}^{++}\), \(\mathrm{Ba}^{++}\), etc.), and \(Me^{++}\) denotes an ionized cation, i.e., one that has lost an additional electron (\(\mathrm{Na}^{++}\), \(\mathrm{K}^{++}\), \(\mathrm{Sr}^{+++}\), etc.).

established only in [11]. It should be said that \(\mathrm{NO_3}\) can apparently be formed not only as a result of reaction \((8^{a})\), but also through the following reaction:

\[ Me^{++}+\mathrm{NO_3^-}\to Me^{+}+\mathrm{NO_3}, \tag{9} \]

which confirms the role of the cation considered above. The radicals \(\mathrm{NO_3}\) formed in reactions \((8^{a})\) and (9) are unstable already at \(77^\circ\mathrm{K}\) (at room temperature \(\mathrm{NO_3}\) has been found only in lead nitrate [15]) and may decompose according to the reactions:

\[ \mathrm{NO_3}\to \mathrm{NO_2}+\mathrm{O}, \tag{10^{a}} \]

\[ \mathrm{NO_3}\to \mathrm{NO}+\mathrm{O_2} \tag{10^{b}} \]

or serve as traps for thermalized electrons:

\[ \mathrm{NO_3}+e\to \mathrm{NO_3^{-*}}\rightleftarrows \mathrm{NO_2^-}+\mathrm{O}. \tag{11} \]

Since the steady-state concentrations of \(\mathrm{NO_3}\) and thermalized electrons are small, reaction (11) is apparently unlikely. On the other hand, Cunningham [3, 7] found that \(\mathrm{NO}\) is the primary product, while \(\mathrm{NO_2}\) appears only at large values of the absorbed dose. In [21, 22] it is also noted that \(\mathrm{NO_2}\) is a secondary product. Therefore reactions \((10^{a})\) and (11) appear unlikely, in contrast to \((10^{b})\).

The decay of \(\mathrm{NO_3^{-*}}\), produced by reactions \((8^{a})\) and (11), should apparently make a certain contribution to the radiolysis products. Since the electron affinity of \(\mathrm{NO_3}\) exceeds the energy of the \(\mathrm{N—O}\) bond, the fragment oxygen atom should have an excess kinetic energy of about \(0.2\ \mathrm{eV}\), if one assumes that the electron-affinity energy of the \(\mathrm{NO_3}\) radical is localized on a particular bond. On the other hand, in a solid the cage effect should manifest itself to the maximum extent. As a result of these factors, the excitation energy of \(\mathrm{NO_3^{-*}}\) will partly be dissipated through the crystal lattice, and part of the \(\mathrm{NO_3^{-*}}\) will decompose into nitrite and oxygen atoms.

The secondary electron, becoming thermalized in the course of inelastic and elastic collisions, is stabilized at some trap. Such traps may be defects initially present in the lattice, products of ionization of the cation or anion, radiolysis products formed in secondary (see below) processes (\(\mathrm{O}\), \(\mathrm{O_2}\), \(\mathrm{NO_2}\), etc.), and, finally, normal lattice sites. In the latter case it is natural to expect for nitrates of alkali and alkaline-earth metals that the thermalized electron will be stabilized on the anion rather than on the cation, since the redox potential \(\dfrac{N^{+5}}{N^{+4}}\) should be much higher than for \(Me^{+}/Me\). Consequently, the main primary product, besides \(\mathrm{NO_3}\), will be the result of stabilization of the thermalized electron, namely:

\[ \mathrm{NO_3^-}+\overline{e}\to \mathrm{NO_3^{-2}} \tag{12} \]

or, in the case of \(\mathrm{AgNO_3}\), where the cation is also an effective electron acceptor,

\[ \mathrm{AgNO_3}+\overline{e}\to (\mathrm{AgNO_3})^{-}. \]

Reaction (12) has been confirmed in many works [3, 4, 5, 11, 12, 20, 21, 26] and occurs only at 77 or \(4^\circ\mathrm{K}\). Upon heating above \(77^\circ\mathrm{K}\), \(\mathrm{NO_3^{-2}}\)

disappears. This ion-radical was not observed if irradiation was carried out at 300°K, nor in any nitrates except \(\mathrm{Sr(NO_3)_2}\) [12].

In potassium and sodium nitrates, the absorption band and EPR spectrum attributed to \(\mathrm{NO_3^{-2}}\) disappear completely upon heating the irradiated crystals to 300°K. At this temperature only the optical absorption bands attributed to \(\mathrm{N_2O_3^{-2}}\) and \(\mathrm{NO_2}\) are stable [4, 20]. Consequently, in addition to reaction \(10^{\mathrm{b}}\), particles with an excess of electrons must contribute to the formation of the final radiolysis products:

\[ \mathrm{NO_3^{-2} \to NO_2^- + O^-}, \tag{13} \]

\[ \mathrm{NO_3^{-2} \to NO_2 + O^{-2}}, \tag{14} \]

\[ \mathrm{NO_3^{-2} \to NO_2^{-2} + O}, \tag{15} \]

\[ \mathrm{NO_3^{-2} \to NO + O_2^{-2}}. \tag{16} \]

The relative contribution to the yield of the final radiolysis products from transformations of oxidized radical particles with an electron deficiency (reactions 8–11) and of reduced ones—with an excess of an electron (reactions 12–16)—is still difficult to assess. It is highly probable that these two groups of reactions lead to different products. As noted above, the relative importance of the individual reactions within each group is also not the same.

It should be noted that the rates of the reactions considered above are only a function of dose and of the electron fractions of the cation and anion. Most of them must be independent of the free volume, since they are irreversible processes. Indeed, in [7] it was found that \(G(\mathrm{NO})\) in \(\mathrm{KNO_3}\) and \(\mathrm{NaNO_3}\) is the same upon irradiation at 300°K and differs only very slightly upon irradiation at 77°K. This fact is experimental evidence for the role of the cation considered above and is consistent with the absence of an effect of dose rate on the radiolysis of nitrates.

In our opinion, spatial factors should have a greater effect on secondary processes, including the interaction of the products of reactions (8–16). It is seen from Fig. 3 that \([\mathrm{NO}]\) in \(\mathrm{NaNO_3}\) reaches saturation already at a dose of about \(1 \cdot 10^{21}\) eV/mole, whereas in \(\mathrm{KNO_3}\) the corresponding dose is approximately twice as high; moreover, the stationary concentration of NO in \(\mathrm{NaNO_3}\) is also lower than in \(\mathrm{KNO_3}\). Thus, the reactions for the consumption of NO become dependent on the lattice parameters. According to Cunningham, such reactions may be:

\[ \mathrm{NO + O \to NO_2}, \tag{17} \]

\[ \mathrm{NO + O_2 \rightleftarrows NO_3^{*} \rightleftarrows NO_2 + O}. \tag{18} \]

\[ \mathrm{NO + NO_3 \to N_2O_4^{*} \rightleftarrows 2NO}, \tag{19} \]

\[ \mathrm{NO + NO_2^{-2} \to N_2O_3^{-2}}. \tag{20} \]

In addition,

\[ \mathrm{O + O \to O_2}. \tag{21} \]

Consequently, it cannot be assumed [1] that the radiolysis of nitrates is not accompanied by the formation of vacant lattice sites. However, the formation of molecular oxygen in excess more than compensates for the formation of vacancies and creates internal pressure in the crystals, which leads to destruction of the crystal lattice. Apparently, in the process of disordering of the lattice, ionization of the cation also plays a rather important role. Since reaction (9), for example, is strongly exothermic, the energy released in this process can be converted into thermal energy

cation, which will lead to its displacement from the lattice site. The changes in the density and heat of solution of irradiated NaNO₃ and KNO₃ observed in [32] are consistent with these assumptions.

Since the radiolysis processes of potassium and sodium nitrates have been studied in relatively greater detail, our discussion concerns mainly these processes. These considerations can, apparently without great risk, be extended to the nitrates of alkaline-earth metals. As for the nitrates of heavy metals or metals exhibiting variable valence, caution is needed in interpreting the mechanism, since in this case entirely different electron–ion processes may occur, the importance of which follows from consideration of the effect of impurities on the radiolysis of nitrates [8, 20, 31, 39].

III. Radiolysis of Nitrites

Nitrites of alkali metals are radiation-resistant compounds. According to the most recent results obtained in this field [41], irradiation of NaNO₂ up to doses of ~250 Mrad does not lead to the appearance of an optically noticeable amount of electron color centers, colloidal metal, or gaseous radiolysis products. The only paramagnetic molecules induced by radiation are NO₂ and N₂O₄ (in the interpretation of [40—43]). According to [40, 42], the EPR spectrum of NaNO₂ irradiated at a temperature of 77°K consists of three lines of equal intensity with a total splitting of 136 gauss. This strongly anisotropic triplet was assigned by the authors [40, 42] with sufficient justification to the NO₂ molecule lying in the bc plane of the crystal [44]. The NO₂ molecules produced at low temperature are unstable and disappear completely when the crystals are heated to room temperature.

Irradiation of NaNO₂ crystals at room temperature leads to consequences somewhat different from those occurring at low temperatures. The EPR spectrum of NaNO₂ irradiated with doses of about 100 Mrad consists in this case of an anisotropic triplet and a superimposed quintet. The calculation of the NO₂ molecule and the study of the triplet symmetry carried out by the authors [43] showed that it belongs to NO₂ lying in the ac plane of the crystal, i.e., rotated by 90° relative to the position of NO₂ in low-temperature radiolysis. The five-component spectrum with a component intensity ratio of 1 : 2 : 3 : 2 : 1 is most probably due to the ion-radical N₂O₄⁻ located in the bc plane, whose unpaired electron has equal densities on both nitrogen nuclei. The symmetry of N₂O₄⁻ and the character of the kinetic accumulation curve of N₂O₄⁻ allow one to assume [41] that the ion-radical is formed as a result of displacement of the anion NO₂⁻ (or, more probably, of the NO₂ molecule) into a neighboring cation vacancy (with a rotation by 180°) and the formation of an N—N bond in the bc plane. In this process, vacancies may be used that either already existed in the crystal before irradiation (predominantly) or were induced by radiation. In accordance with this point of view, the absence of N₂O₄⁻ during low-temperature irradiation of NaNO₂ may be associated with hindered rotation of NO₂ and with the small number of vacancies in the nitrite at a temperature of 77°K.

The data listed above agree well with the results available in the literature on optical spectra (in the visible and IR regions) in irradiated NaNO₂ [44—46]. At room temperature, the absorption spectrum of irradiated NaNO₂ has a broad maximum in the region 550—400 mµ, attributed to NO₂. Studies of spectra in polar-

in polarized light confirmed the conclusion that, at room temperature, the molecule is located in the plane of the crystal \(ac\).

IV. Processes upon Irradiation of Chlorates, Bromates, and Iodates

The radiolysis of chlorates, bromates, and iodates of alkali metals has been studied far from uniformly. Whereas for the first two groups of salts the composition of the final radiolysis products, their yields, and the kinetics of accumulation and destruction of molecular products have in recent years been investigated in considerable detail, for iodates such information is lacking. Optical spectra and EPR signals have been described in the literature only for chlorates; for bromates, work of this kind has the character of brief communications, while the behavior of iodates has not been studied at all. This unevenness in the study of the properties of irradiated materials is connected with the greater radiation stability of bromates and iodates and, for the latter, also with the more complex structure and arrangement of the lattice.

1. COMPOSITION OF THE RADIOLYSIS PRODUCTS

After the first studies on the radiolysis of chlorates [17, 62], in which the presence of oxygen, \(\mathrm{Cl^-}\), and \(\mathrm{ClO_2^-}\) in the products was established, the question of the composition of the radiolysis products was first investigated in detail by Hill [58].

By analyzing solutions of irradiated \(\mathrm{KClO_3}\) and from the absorption spectra of the crystals, he determined the following composition of products: \(\mathrm{O_2}\), \(\mathrm{Cl^-}\), \(\mathrm{ClO^-}\), \(\mathrm{ClO_2^-}\), \(\mathrm{ClO_4^-}\), and also the presence, in small amounts, of \(\mathrm{Cl_2O_6}\) and \(\mathrm{ClO_2}\). Here \(\mathrm{O_2}\), \(\mathrm{Cl^-}\), \(\mathrm{ClO^-}\), and \(\mathrm{ClO_2^-}\) were determined directly, while the presence of \(\mathrm{ClO_4^-}\) was inferred from data on the deficit of free \((\mathrm{O_2})\) and bound oxygen in the products and was confirmed by study of the IR spectra of the irradiated salt. The conclusion regarding the presence of \(\mathrm{Cl_2O_6}\) and \(\mathrm{ClO_2}\) was drawn from analysis of the optical spectra. In the same work it was shown that the listed chlorine–oxygen products are in fact formed in the solid, and do not appear upon dissolution of the irradiated crystals; the latter is evidenced by the presence, in the optical absorption spectra of \(\mathrm{KClO_3}\), of bands characteristic of chlorine–oxygen molecules (see below).

In the first studies on the radiation chemistry of bromates [68, 69] it was established that irradiation of them in a nuclear reactor with mixed radiation—\(\gamma\)-rays, fast and slow neutrons—leads to the accumulation of radiolysis products: \(\mathrm{Br_2}\), \(\mathrm{Br^-}\), \(\mathrm{BrO^-}\), \(\mathrm{BrO_2^-}\); here the value \(G(\mathrm{BrO_3})\) is constant (the accumulation of products is linear with irradiation dose). In later works [71—74] it was shown that the forms of oxidized radiolytic bromine products in irradiated bromates are analogous to the chlorine–oxygen products found in the radiolysis of chlorates: \(\mathrm{Br^-}\), \(\mathrm{Br_2}\), \(\mathrm{BrO^-}\), \(\mathrm{BrO_2^-}\). Among the products there is also \(\mathrm{O_2}\), while the presence of \(\mathrm{BrO_4^-}\) has not been definitively established.

2. KINETIC REGULARITIES OF RADIOLYSIS

The \(G\) values of some products of the radiolysis of potassium chlorate, first found in works [17, 62], are, for irradiation of the salt with X-rays at room temperature: 1.57, 1.2, and 0.4 for oxygen, \(\mathrm{ClO_2^-}\), and \(\mathrm{Cl^-}\), respectively. Radiolytic oxygen obtained ...

is retained in the lattice of \( \mathrm{KClO_3} \) with difficulty (up to 8% conversion, after which cracking of the crystals occurs). The accumulation of \( \mathrm{Cl^-} \) and \( \mathrm{Cl_2^-} \) is linear in the range of doses studied in [62] and depends on the temperature of heating after irradiation.

Fig. 5. Accumulation curves for the main radiolysis products in \( \mathrm{KClO_3} \) at room temperature:
\(1\)—\( \mathrm{O_2} \); \(2\)—\( \mathrm{Cl^-} \); \(3\)—\( \mathrm{ClO_2^-} \); \(4\)—\( \mathrm{ClO^-} \).

The accumulation curves for the main products of radiolysis of \( \mathrm{KClO_3} \), first obtained in [58], are shown in Fig. 5. The “initial” radiation yields calculated from these data, as well as from the results of radiolysis at \(-196^\circ\mathrm{C}\) (\(G\) at a dose of \(0.38 \cdot 10^{23}\) eV/g), are:

a) at \(t = 25^\circ\mathrm{C}\)—1.3; 0.5; 2.0; 2.5; 0.2;

b) at \(t = -196^\circ\mathrm{C}\)—1.1; 0.5; 1.5; 2.0; and 0.2 for \( \mathrm{Cl} \), \( \mathrm{ClO^-} \), \( \mathrm{Cl_2^-} \), \( \mathrm{O_2} \), and \( \mathrm{ClO_2^-} \), respectively.

The following features are characteristic of the kinetics of radiolysis of \( \mathrm{KClO_3} \):

1. A constant rate of accumulation of \( \mathrm{Cl^-} \) and \( \mathrm{O_2} \), and saturation of the accumulation curves for \( \mathrm{ClO_2^-} \) and \( \mathrm{ClO^-} \).

2. A slight influence of temperature in the range 77–300°K on the values of \(G\).

3. Probable formation of \( \mathrm{ClO_4^-} \) or \( \mathrm{Cl_2O_5} \) in appreciable amounts.

In the only study [63] on the radiolysis of \( \mathrm{NaClO_3} \) chlorates under the action of heavy radiation (\(\alpha\)-particles), it was shown that the radiation yield \(G_{\mathrm{Cl^-}}\) in this case is 2.12 (at room temperature) and, in contrast to the effects under \(\gamma\)-irradiation, does not change after heating of the irradiated salt. On the basis of this fact, the author [63] suggested a large role of thermal effects (in the track) in the radiolysis of chlorates by \(\alpha\)-radiation.

Closely related to work on the radiolysis of chlorates is a group of studies on the Szilard–Chalmers effect in this class of compounds [64–66]. Among the products in neutron-irradiated chlorates, the presence of \( \mathrm{Cl^-} \), \( \mathrm{ClO_2^-} \), \( \mathrm{ClO^-} \), \( \mathrm{Cl_2} \), and \( \mathrm{ClO_4^-} \) is noted; the isotope \( \mathrm{Cl^{36}} \) formed by the \((n,\gamma)\) reaction is retained almost completely in the form of \( \mathrm{Cl^-} \). In [65], according to data on the distribution of induced activity

(i.e., the content of Cl\(^{36}\)) in the forms \(\mathrm{ClO_3^-}\) and \(\mathrm{ClO_4^-}\), it was concluded that the formation of \(\mathrm{ClO_4^-}\) is more probable upon oxidation of \(\mathrm{ClO_3^-}\) than by the oxidation reaction of \(\mathrm{Cl^-}\).

In the radiolysis of \(\mathrm{KBrO_3}\) under the action of mixed reactor radiation, the fraction of decomposed bromate increases linearly with increasing irradiation dose [68—69]. At the same time, according to [70], \(G_{\mathrm{Br^-}} = 1 \pm 0.1\) for \(\gamma\)-radiation and fast electrons of a broad energy spectrum, and is 23—30 \((G_{\mathrm{Cl^-}})\) for mixed neutrons.

The radiolysis of alkali-metal bromates has been studied in greatest detail in the later works of Boyd and coauthors [71—74]. Data on the yield of the principal radiolysis products under the action of \(\mathrm{Co}^{60}\) \(\gamma\)-radiation [71] are summarized in Table 6.

Table 6

Values of \(G_0\) for the radiolysis of bromates in alkali metals upon irradiation with \(\mathrm{Co}^{60}\) \(\gamma\)-radiation

Salt	\(G(\mathrm{BrO_2^- + BrO^-})\)	\(G(\mathrm{Br^-})\)	\(G(-\mathrm{BrO_3^-})\)
\(\mathrm{LiBrO_3}\)	0.21	0.13	0.23
\(\mathrm{NaBrO_3}\)	0.64	0.88	1.48
\(\mathrm{KBrO_3}\)	0.46	1.07	1.53
\(\mathrm{RbBrO_3}\)	0.82	1.02	1.85
\(\mathrm{CsBrO_3}\)	0.49	2.23	2.73

The accumulation of \(\mathrm{Br^-}\), linear in the initial stage of radiolysis (approximately up to \(1 \cdot 10^{23}\) eV/g), subsequently begins to proceed at a gradually decreasing rate for all bromates except \(\mathrm{LiBrO_3}\); the degree of deviation from linearity increases along the series from \(\mathrm{NaBrO_3}\) to \(\mathrm{CsBrO_3}\). For the latter, the following dependence of the yield \((y)\) on the absorbed dose \((x)\) is valid:

\[ y = \kappa_1 x / 1 + \kappa_2 \cdot \sqrt{x} \]

To an even greater extent in the course of radiolysis, the value \(G(\mathrm{BrO^- + BrO_2^-})\) decreases.

In the temperature interval 77—358°K, the rate of radiolysis depends only very weakly on the irradiation temperature. The ratios

\[ \frac{G(\mathrm{Br^-})}{G(\mathrm{Br^-})} \]

and

\[ \frac{G(-\mathrm{BrO_3^-})}{G(-\mathrm{BrO_3^-})} \]

at 357 and 77°K are smallest for \(\mathrm{CsBrO_3}\) [1, 16] and largest for \(\mathrm{RbBrO_3}\) [1, 47]. For \(\mathrm{CsBrO_3}\), the rate of accumulation of \(\mathrm{Br^-}\) is practically independent of the irradiation dose rate as well. The isotopic composition (for \(\mathrm{LiBrO_3}\)) and the method of preparation of the compounds (for \(\mathrm{LiBrO_3}\) and \(\mathrm{RbBrO_3}\)) have only a weak effect on the rate of radiolysis of the bromates.

One of the central conclusions of work [71] is the establishment by the authors of a correlation between the free-volume value of the bromates and their radiation stability, similar to that found for a number of other inorganic salts [75]. This correlation, however, is valid only for a group of similar compounds (alkali-metal bromates).

In a later work [72], Boyd and coauthors established that the \(G\) values of the radiolysis products of bromates in a neutron-irradiation field—

... are considerably higher than in the case of γ-irradiation. As a consequence of this, and in order to clarify the question of the specific features of radiolysis under the action of neutrons, the same authors studied in detail the radiolysis of bromates in a mixed reactor-radiation field [73]. It was established analytically that, in the total absorbed dose, more than half is due to internal irradiation of the bromates by radiation generated by unstable reaction products \((n,\gamma)\) and by recoil nuclei. In mixed reactor radiation, the fraction of “pure” γ-radiation amounts to more than 50% of the total dose absorbed by the bromates. The radiation yields obtained in radiolysis under such conditions are given in Table 7.

Table 7

Values of \(G\) of the main products of bromate radiolysis in a reactor-radiation field

Salt	\(G(-\mathrm{BrO}_3^-)\)	\(G(\mathrm{Br}^-)\)	\(G(\mathrm{BrO}^-)\)	\(G(\mathrm{BrO}_2^-)\)
\(\mathrm{Li^7BrO_3}\)	0.47
\(\mathrm{NaBrO_3}\)	1.4	0.59	0.22	0.59
\(\mathrm{KBrO_3}\)	1.3	0.63	0.2	0.54
\(\mathrm{RbBrO_3}\)	2.4	0.96	2.23	0.31
\(\mathrm{CsBrO_3}\)	3.4	0.61
\(\mathrm{LiBrO_3}\)	1.4	0.58	0.26	0.47

Comparison of the data in Tables 6 and 7 shows that, for all bromates except \(\mathrm{LiBrO_3}\), the yields of products under γ- and reactor irradiation are very close. The increase in the yield for \(\mathrm{LiBrO_3}\) is associated with the high efficiency of neutron capture by \(\mathrm{Li^6}\) nuclei. In this case, unlike the situation in γ-radiolysis, the isotopic composition of \(\mathrm{LiBrO_3}\) naturally affects the rate of the process.

Among the alkali-metal bromates, radiolysis has been studied in greatest detail for \(\mathrm{CsBrO_3}\) [74].

The kinetic regularities of product accumulation in this case proved to be qualitatively similar to those observed for \(\mathrm{KClO_3}\) (see above). The accumulation curves of the bromoxy products proceed with saturation, so that at doses of the order of \(10^{24}\ \mathrm{eV/g}\) their yield practically becomes equal to zero. In contrast to the bromoxy products, the accumulation of \(\mathrm{Br^-}\) and \(\mathrm{O_2}\) is linear up to high doses. Annealing phenomena are very pronounced; in this connection, the relative content of \(\mathrm{Br^-}\) and bromine oxides in the products depends on the temperature of radiolysis. The yields of all products do not depend on the dose rate of irradiation. The temperature dependence of the radiolysis rate proved to be complex for \(\mathrm{CsBrO_3}\), which is associated with the complexity of the annealing processes in bromates, proceeding both with the formation of final products of the process and with reverse oxidation of the products to \(\mathrm{BrO_3^-}\).

EPR and optical spectra in irradiated salts

At present, the form of the EPR spectrum has been studied in relatively great detail only in potassium and sodium chlorates irradiated with comparatively small doses \((<10^6\ \mathrm{rad})\). Data on signals in \(\mathrm{CsClO_3}\) and \(\mathrm{RbClO_3}\) are extremely limited; as for bromates and iodates, there is only a general indication of the possibility of observing signals in these salts and a brief description of their form in powdered materials [47].

The EPR spectrum in irradiated alkali-metal chlorates has a complex form (Fig. 6) and has not yet been completely and sufficiently correctly interpreted.

As regards the identification of each signal in the spectrum, the literature usually contains contradictory opinions.

After a qualitative description of the signals in irradiated $\mathrm{KClO_3}$ [48], the EPR of this salt was first studied in detail by Hasty and co-workers [48, 50]. According to [48], the spectrum of $\mathrm{KClO_3}$ irradiated at $20^\circ\mathrm{C}$ consists of two groups of lines: a broad central singlet with a $g$-factor greater than 2 and a group of four strongly anisotropic quartets of relatively lower intensity, which the authors assigned to a single singlet. The 16 lines of the spectrum disappear completely when the crystals are heated to $105^\circ\mathrm{C}$; in this process the crystal becomes opaque, milky in color.

The results of the authors’ calculation of the STS value of the quartets gave them grounds to assume that this signal belongs to the grouping $\mathrm{ClO{-}ClO_3^-}$ with a delocalized valence electron. In this case the $\mathrm{ClO}$ radical is obtained as a result of decomposition of $\mathrm{ClO_3^0}$. The four quartets realized in the spectrum are assigned to different combinations of the isotopes $\mathrm{Cl^{35}}$ and $\mathrm{Cl^{37}}$ in the complex $\mathrm{ClO{-}Cl_3^-}$, which have different values of the nuclear magnetic moments. The authors ascribe the central singlet to the $\mathrm{O_2^-}$ radical (without any clear grounds for doing so). The authors of [49], on the contrary, believe that the central singlet in the spectrum of irradiated $\mathrm{NaClO_3}$ should (as in $\mathrm{KClO_3}$) more probably be assigned to the $\mathrm{O_3^-}$ radical ($g = 2.003—2.015$) than to $\mathrm{O_2^-}$, since the latter possesses a large anisotropy [50], which is not characteristic of the singlet in the $\mathrm{NaClO_3}$ spectrum. In contrast to [48], in [49], in addition to the singlet and the complex multiquartet spectrum, the presence of a poorly resolved doublet (Table 8) and two further lines hidden by the quartets was shown. The first doublet was tentatively assigned to the $\mathrm{ClO_2}$ radical ($g = 2.01$ according to 51), and the second to oxygen or to an oxygen species (Table 8).

Fig. 6. EPR spectrum of irradiated $\mathrm{NaClO_3}$ (irradiation and recording at room temperature). $A_1$ and $A_2$ are two quartets ($\mathrm{Cl^{35}}$ and $\mathrm{Cl^{37}}$); $B$ and $C$ are singlets.

In the same work the authors concluded that there is no signal in $\mathrm{NaBrO_3}$ irradiated with $\gamma$-radiation.

Atkins and co-workers, who investigated the analogy between the spectra of chloroxy radicals ($\mathrm{ClO}$, $\mathrm{ClO_2}$, $\mathrm{ClO_3}$) frozen in various matrices and the form of the EPR signals in irradiated chlorates and perchlo-

in chlorates, they believe that the false signal is more probably associated with the formation of \(\mathrm{Cl—ClO_3^-}\) than with the ion-radical \(\mathrm{ClO—ClO_3^-}\). As for the central singlet, the authors [51] definitely assign it to \(\mathrm{O_3^-}\), on the basis of the presence of a similar signal in \(\mathrm{KO_3}\) and of correlation with the optical spectrum of ozonide.

Table 8

Character of EPR signals in irradiated chlorates

Work	\(\mathrm{O_3^-}\) \((\mathrm{O_2^-})\)	\(\mathrm{ClO_2}\)	\(\mathrm{ClO}\)	\(\mathrm{ClO_3^0}\) \(\mathrm{ClO—ClO_3^-}\)
[58]	\(g = 2.00\)
[49]	\(g = 2.00\) \(A = 11\) gauss	\(g = 2.01\quad A = 57\)	\(g = 2.022\)	\(g = 2.0023\)
[51]		\(g = 2.0102\)
[52]	\(g = 2.008\)
[53]	\(g = 2.0338\)			\(g = 2.0124\)
[61]		\(g = 2.0032\)		\(2.003\)

In the most recent work devoted to EPR spectra in irradiated \(\mathrm{NaClO_3}\) [52], it was shown that at irradiation doses up to \(\sim 10^5\) roentgens three paramagnetic centers are observed, while with further irradiation their number increases. The multicomponent spectrum (also found in all preceding works), in contrast to [48] and [51], is assigned by the authors to the \(\mathrm{ClO_3}\) radical, on the grounds that the established form of the center—a pyramid with its axis along the \(\langle 111\rangle\) direction of the crystal—corresponds to the form of \(\mathrm{ClO_3^0}\) located at the normal site of the chlorate anion. The axial symmetry established experimentally for the center does not allow it to be assigned to \(\mathrm{ClO_2}\), which more probably has rhombic symmetry [51]. A similar conclusion regarding the nature of the signal was reached also in [61]; in addition, three more centers were found there in \(\mathrm{KClO_3}\), one of which was assigned to \(\mathrm{ClO_2^0}\). In addition to the complex signal, the authors found two more central singlets, which they did not identify.

The position of the \(g\)-factor of one of the singlets (\(g = 2.008\)) is close to the position of the signals assigned in [48] and [49] to the ion-radical \(\mathrm{O_3^-}\). The center responsible for the second signal (\(g_{\parallel} = 2.007;\ g_{\perp} = 2.02\)) has not been established.

The results obtained in [53] show that the character of the EPR signals in powdered salts irradiated at room temperature is qualitatively similar in all chlorates of the alkali and alkaline-earth metals. The values of the \(g\)-factors of the centers established in that work, however, differ from those found earlier for single crystals. Without introducing new assumptions concerning the identification of the signals, the author adheres in this respect to the point of view of [48].

As follows from the foregoing, there is no unified point of view on the question of the models of the centers responsible for absorption in chlorates in the microwave region.

At present, it appears possible with sufficient certainty to decipher the central singlet by assigning it to \(\mathrm{O^-}\). Of the propose-

of the possible models of the multicomponent spectrum, apparently the more probable one is the radical \(\mathrm{ClO}_3^{0}\).

In contrast to the earlier studies [47, 54], information was obtained for the first time on the ESR spectra of alkali-metal chlorates at a temperature of 77°K during irradiation and recording, as well as data on the kinetics of signal accumulation and the regularities of signal annealing in a number of chlorates and on the relationship of the accumulation rate to the parameters of the crystal lattice. Irradiation of salts at low

Fig. 7. ESR spectrum of irradiated \(\mathrm{KClO}_3\) (\(\vec H\) [111]), \(D\) — multicomponent signal

temperatures with large doses (up to \(10^8\) rad) made it possible for the first time to obtain signals in bromates and iodates and to observe distinctly signals in \(\mathrm{KClO}_3\) that had not previously been detected.

In studies [55, 57], which are mainly in the nature of brief communications, signals of nuclear quadrupole resonance on \(\mathrm{Cl}^{35}\) nuclei in potassium and sodium chlorates are described.

The authors note a broadening of the lines after irradiation of chlorates, which is a consequence of perturbation of the crystal field by radicals and ions—products of radiolysis. Of interest is the conclusion of one of the cited works that the value of \(G\) is larger upon irradiation of \(\mathrm{NaClO}_3\) with neutrons than upon radiolysis by \(\gamma\)-radiation.

Information on optical absorption spectra in irradiated salts of the \(\mathrm{MeHalO}_3\) type is extremely limited.

Hill, who studied the absorption spectra of thin single-crystal plates of \(\mathrm{KClO}_3\) [58] irradiated with X-rays to high doses (\(\sim 10^{23}\) eV/g), established the presence in the spectra of three absorption bands—at 260 m\(\mu\), 310 m\(\mu\), and in the region of 450 m\(\mu\). By analogy with the positions of the absorption bands of hypochlorite anions in aqueous solutions [59], he assigned these bands respectively to \(\mathrm{ClO}^{-}\), \(\mathrm{ClO}_2^{-}\), and \(\mathrm{Cl}_2\mathrm{O}_6\), whose presence he confirmed by analysis of a solution of the irradiated salt. In contrast to Hill’s data, in [48] and [60] only one intense band with a maximum at 461 m\(\mu\), stable at room temperature, was observed in irradiated \(\mathrm{KClO}_3\). Under the same conditions, in \(\mathrm{NaClO}_3\) in [60] two absorption peaks shifted into the short-wavelength region of the spectrum were observed. A more thorough study of the coloration

NaClO\(_3\) at room temperature, carried out in [67], showed that in the spectrum of the irradiated salt three absorption regions are distinguished—near 260 m\(\mu\) and with maxima at 302 and 420 m\(\mu\). These centers are not produced by irradiation of the chlorate with light (region 220–400 m\(\mu\)), by additive coloration, or by heating the crystals in an oxygen atmosphere. The absorption centers are optically stable, but are destroyed when the irradiated crystal is heated to 140°C.

Absorption in the region of wavelengths below 260 m\(\mu\), and at 302 m\(\mu\), naturally, as in KClO\(_3\), should be associated with ClO\(^{-}\) and ClO\(_2^{-}\). If one assumes that in irradiated NaClO\(_3\) all absorption bands are shifted into the short-wavelength region of the spectrum relative to the bands in KClO\(_3\), then it may be supposed that the 420 m\(\mu\) band in NaClO\(_3\) is analogous to the 460 m\(\mu\) band in KClO\(_3\). In [48] the maximum in KClO\(_3\) at 460 m\(\mu\) is assigned to the O\(_2^{-}\) center located in an anion vacancy. And finally, [53] contains one more supposition, according to which the coloration of irradiated chlorates (the author, unfortunately, does not say which absorption band is meant) is due to ClO—ClO\(_3^{-}\) centers.

The discrepancy between the data of [58] and [48, 53, 60] may be connected with the fact that in the latter works irradiation was carried out with comparatively small doses, which evidently did not make it possible to obtain bands from a number of products. As regards the model of the center responsible for the 460 m\(\mu\) band, the most probable species should be considered to be Cl\(_2\)O\(_6\) molecules or O\(_2^{-}\).

On the mechanism of radiolysis

The data presented above on optical-absorption and EPR spectra, as well as on the nature and accumulation of the final products in the radiolysis of alkali-metal chlorates, provide a basis for constructing a tentative mechanism of the radiation decomposition of this group of compounds. To the greatest extent the statements made are applicable only to sodium and potassium chlorates, since the necessary experimental data have been obtained only for these two salts.

In the radiolysis of chlorates the following elementary stages are probable:

\[ \begin{aligned} &\hspace{7.5em}\longrightarrow \mathrm{ClO}^{-} + \mathrm{O}_2, && \text{(a)}\\ &1.\quad \mathrm{ClO}_3^{-} \rightsquigarrow \mathrm{ClO}_3^{-*} \begin{cases} \longrightarrow \mathrm{ClO}_2^{-} + \mathrm{O}, & \text{(б)}\\ \longrightarrow \mathrm{Cl}^{-} + \mathrm{O}_2 + \mathrm{O}, & \text{(в)}\\ \longrightarrow \mathrm{ClO} + \mathrm{O}_2^{-}, & \text{(г)}\\ \longrightarrow \mathrm{ClO}_2 + \mathrm{O}^{-}; & \text{(д)} \end{cases} \end{aligned} \]

\[ 2.\quad \mathrm{ClO}_3^{-} \rightsquigarrow \mathrm{ClO}_3^{0} + e^{*}; \]

\[ 3.\quad e^{*} \to e; \]

\[ \begin{aligned} &\hspace{7.5em}\longrightarrow \mathrm{ClO}_2 + \mathrm{O}, && \text{(a)}\\ &4.\quad \mathrm{ClO}_3^{0} \longrightarrow \begin{cases} \longrightarrow \mathrm{ClO} + \mathrm{O}_2,\\ \longrightarrow \mathrm{Cl} + \mathrm{O}_2 + \mathrm{O}; & \text{(б)} \end{cases} \end{aligned} \]

\[ 5.\quad \mathrm{ClO}_3 + e \longrightarrow \left|\,\text{products } 1,\ \mathrm{a} - 1,\ \mathrm{д};\right. \]

\[ 6.\quad \mathrm{O}_2^{-} + \mathrm{O} \longrightarrow \mathrm{O}_3^{-}, \]

\[ \mathrm{O}_2/\square a + e \longrightarrow \mathrm{O}_2^{-}/\square a; \]

1. \[ \mathrm{O} + \mathrm{O} \longrightarrow \mathrm{O}_2; \qquad (a) \]
   \[ \mathrm{Cl} + \mathrm{Cl} \longrightarrow \mathrm{Cl}_2; \qquad (b) \]

2. \[ \mathrm{ClO}_2^- \longrightarrow \mathrm{Cl}^- + \mathrm{O}_2, \qquad (a) \]
   \[ \mathrm{ClO}^- \longrightarrow \mathrm{Cl}^- + \mathrm{O}; \qquad (b) \]

3. \[ \mathrm{ClO}_2^- + \mathrm{O}_2 \longrightarrow \mathrm{ClO}_3^- + \mathrm{O}, \qquad (a) \]
   \[ \mathrm{ClO}^- + \mathrm{O}_2 \longrightarrow \mathrm{ClO}_3^-; \qquad (b) \]

4. \[ \mathrm{ClO}_3^- + \mathrm{O} \longrightarrow \mathrm{ClO}_4^- . \]

In radiolysis, both excitation (1) and ionization (2) of the anions are probable as primary stages. Upon dissociation of excited \(\mathrm{ClO}_3^{-*}\), a broad range of chlorine–oxygen products may be formed, determined analytically or spectroscopically (reactions 1,a—1,g). The formation of chlorine–oxygen anions (stages 1,a—1,c), however, is more probable than reactions 1,d and 1,e, owing to the higher electron affinity of chlorine–oxygen species [76] than that of oxygen. The products thereby formed have been clearly determined analytically (see above), and the presence of some of them directly in the crystals has also been established in studies of the optical absorption spectra in irradiated crystals.

The following experimental facts confirm ionization in chlorates:

1. the presence of a many-line signal in the spectra of irradiated \(\mathrm{NaClO}_3\) and \(\mathrm{KClO}_3\), attributed most probably to \(\mathrm{ClO}_3^0\);

2) the formation during radiolysis of small amounts of \(\mathrm{Cl}_2\mathrm{O}_6\) [58], produced in the bimolecular interaction of \(\mathrm{ClO}_3^0\).

After the act of ionization, the excited electron loses its energy (the stage of electron thermalization (3)), after which it is captured by some trap.

The radical \(\mathrm{ClO}_3^0\), being an unstable species, may undergo thermal decomposition (especially during radiolysis in the region of elevated temperatures); the radicals \(\mathrm{ClO}\) and \(\mathrm{ClO}_2\) formed in this process may be present in the radiolysis products, but the question of their presence as intermediate or final species has been clearly resolved only for \(\mathrm{ClO}_2^0\).

In view of the high electron affinity of the \(\mathrm{ClO}_3^0\) radical (\(E = 3.96\) eV, according to [76]), their recombination may also lead to cleavage of bonds within the radical. The possibility of this is determined to a considerable extent by the ratio between the electron-affinity energy of \(\mathrm{ClO}_3^0\) and the energy of cleavage of the Cl—O bonds in chlorate. The latter quantity is \(2.38\) eV in \(\mathrm{KClO}_3\) [17]. Consequently, it may be assumed that recombination has the character of dissociative attachment of the electron to \(\mathrm{ClO}\) (reaction 5).

In the EPR spectra of irradiated chlorates, one of the singlets has been assigned to an oxygen radical; this provides grounds for introducing reaction 6, and, judging from later works, stage 6,a is the more probable.

By analogy with bromates it may be considered that annealing of the radiolysis products, which apparently also takes place in the very process of radiolysis at medium and high temperatures, proceeds both by the path of thermal conversion of unstable products (reactions 8,a,b) and with the reverse oxidation of chlorine–oxygen radiolysis products by radiolytic oxygen (reactions 9,a,b). In the radiolysis of chlorates, as primary acts

apparently reactions 1 and 2 proceed in parallel. It is not possible at present to determine definitively the further fate of the \(\mathrm{ClO_3^0}\) radicals formed in stage 2 (decomposition or dissociative recombination with an electron). However, data on the absence of an influence of dose rate on the accumulation of paramagnetic particles in chlorates make it possible to suggest that the more probable process is decomposition of \(\mathrm{ClO_3^0}\) (reaction 4). In this connection, and in accordance with what was set forth above, one may hope that in chlorates the formation of neutral chlorine–oxygen radicals \((\mathrm{ClO}, \mathrm{ClO_2})\) and anions \((\mathrm{ClO^-}, \mathrm{ClO_2^-})\) occurs by different reactions (respectively 4 and 1).

Since in a number of works the presence of the \(\mathrm{ClO_4^-}\) anion in the products of radiolysis of chlorates was shown, and in work [65] there is a statement on the formation of \(\mathrm{ClO_4^-}\) by further oxidation of the chlorate anion, reaction 10 should be introduced as a tentative one.

Similarly to the scheme of radiolysis presented, Baberkin [62] also considers the parallel occurrence of excitation and ionization leading to the formation of radiolysis products to be probable, but regards the role of ionization processes as insignificant.

Hill, in his work [58], considers only excitation of chlorate anions as the primary process; he does not take into account the set of elementary stages of radiolysis considered by him, although the formation of a number of radical products, whose presence was later established by the EPR method, is not considered.

The formation, as primary products in the radiolysis of chlorates, of \(\mathrm{Me^{++}}\) cations by the reaction

\[ \mathrm{Me^+ \to Me^{++} + e} \tag{11} \]

and the subsequent interaction of \(\mathrm{Me^{++}}\) with a neighboring anion

\[ \mathrm{Me^{++} + ClO_3^- \to Me^+ + ClO_3^0} \tag{12} \]

is in principle apparently also possible, just as was proposed above for nitrates. For chlorates, however, it is not yet possible to estimate the contribution of the cation to the overall yield of products.

The formation of radicals and final molecular products in the radiolysis of alkali-metal bromates has been studied in far less detail than in the case of chlorates. In the literature, only the question of the composition and accumulation of bromide and bromine–oxygen anions with a low degree of oxidation of bromine has been covered to a sufficient extent. In this connection, the mechanism of radiolysis of bromates presented in [72–73] appears less definite and justified than the set of elementary stages that can be presented for chlorates. It is reasonable to assume (and comparison of the composition of the molecular products of radiolysis of chlorates and bromates confirms this) that the nature of the elementary stages of bromate radiolysis should, in the main, be similar to those occurring in the radiolysis of chlorates. The differences in the mechanisms of radiolysis that exist for these two groups of salts can be determined only by careful comparative study of the optical and EPR spectra of the salts.

V. Radiolysis of Perchlorates

Of the group of salts considered in the present review, perchlorates are the least stable to the action of radiation. As products of the radiolysis of potassium perchlorate, after dissolution of the irradiated salt in water, Hill [77] identified \(\mathrm{Cl^-}\), \(\mathrm{ClO_3^-}\), and \(\mathrm{O_2}\). However, Baberkin [62] found that chloride appears only after heating

irradiated samples above 370°K. In [77] it was also suggested that hypochlorite and chlorite are formed; their presence in irradiated perchlorates of alkali and alkaline-earth metals was experimentally demonstrated in later works [38, 78, 79]. The accumulation of the main radiolysis products with absorbed dose differs little from linearity. The values of the radiation-chemical yields of the radiolysis products of perchlorates of various metals are given in Table 9.

Table 9

Radiation-chemical yields of the radiolysis products of perchlorates at room temperature

Salt	$G(\mathrm{ClO_3^-})$	$G(\mathrm{ClO_2^-})$	$G(\mathrm{ClO_2})$	$G(\mathrm{ClO^-})$	$G(\mathrm{Cl^-})$	$G(\mathrm{O_2})$	$G(-\mathrm{ClO_4^-})$	Literature
$\mathrm{LiClO_4}$	2,80	0,15	0,59	0,10	0,12	2,15	3,76	78
$\mathrm{NaClO_4}$	3,57	0,17	0,11	0,09	0,42	2,96	4,36	[78]
$\mathrm{NaClO_4}$	2,8	0,75	0,08	0,1	0,43	—	4,2	[38]
$\mathrm{KClO_4}$	2,99	0,18	0,12	0,09	0,45	2,68	3,83	[78]
$\mathrm{KClO_4}$	—	—	—	—	0,94	—	3,46	[79]
$\mathrm{RbClO_4}$	4,06	0,20	0,12	0,14	0,75	3,84	5,27	[78]
$\mathrm{RbClO_4}$	—	—	—	—	1,14	—	4,86	[79]
$\mathrm{CsClO_4}$	5,28	0,22	0,10	0,17	1,07	5,28	6,84	[78]
$\mathrm{CsClO_4}$	—	—	—	—	1,94	—	8,84	[79]
$\mathrm{Mg(ClO_4)_2}$	4,29	0,14	0,07	0,03	0,15	2,62	4,67	[78]
$\mathrm{Ca(ClO_4)_2}$	3,44	0,00	0,51	0,08	0,12	1,99	4,15	[78]
$\mathrm{Sr(ClO_4)_2}$	3,90	0,19	0,14	0,11	0,19	2,61	4,53	[78]
$\mathrm{Ba(ClO_4)_2}$	1,76	0,84	0,42	0,12	0,06	2,18	3,20	[78]

As can be seen from this table, the main radiolysis products of perchlorates are chlorate and oxygen. The other products are formed with relatively small yields. In [38, 78] it was shown that $G(-\mathrm{ClO_4^-})$ in $\mathrm{KClO_4}$ and $\mathrm{NaClO_4}$ is almost independent of temperature over a wide temperature range. From Table 10, taken from the work of Prince and Johnson [78], it can be seen that, with increasing irradiation temperature of $\mathrm{KClO_4}$, $G(\mathrm{ClO_2^-})$ and $G(\mathrm{ClO_2})$ decrease monotonically, while $G(\mathrm{ClO^-})$ passes through a maximum. These facts were attributed by the authors to thermal decomposition of the indicated products.

The instability of $\mathrm{ClO_2}$ molecules above 340°K [80] was also confirmed by the EPR method. At room temperature in irradiated $\mathrm{KClO_4}$, in addition to $\mathrm{ClO_2}$ [80, 81], paramagnetic particles $\mathrm{O_3^-}$ and an unidentified radical were detected. At 77°K, $\mathrm{ClO_3}$ radicals additionally appear [80]. The $\mathrm{ClO_3}$ radical was also found in irradiated $\mathrm{NH_4ClO_4}$ [82—84] at room temperature. However, under these conditions it is very unstable. The presence of other paramagnetic particles formed upon irradiation of $\mathrm{KClO_4}$ was not detected in $\mathrm{NH_4ClO_4}$. In addition to $\mathrm{ClO_3}$, stable ion-radicals $\mathrm{NH_3^+}$ [82—86] are formed in irradiated ammonium perchlorate at 300°K. It should be noted that not only the intermediate radiolysis products of ammonium and potassium perchlorates differ, but also the final products determined by chemical methods.

Thus, in [87, 88] it was shown that the main products of the radiolysis of \(\mathrm{NH_4ClO_4}\) are \(\mathrm{Cl_2}\) and \(\mathrm{Cl^-}\), which is associated with oxidation of the ammonium ion. In the perchlorates of metals of the first and second groups of the periodic system, however, the nature of the cation affects only the magnitude of the radiation-chemical yield. The dependence of \(G(-\mathrm{ClO_4^-})\) on the nature

Table 10

Effect of irradiation temperature on the radiolysis of \(\mathrm{KClO_4}\) [78]

\(t^\circ\mathrm{C}\)	\(G(-\mathrm{ClO_4^-})\)	\(G(\mathrm{ClO_3^-})\)	\(G(\mathrm{ClO_2^-})\)	\(G(\mathrm{ClO_2})\)	\(G(\mathrm{ClO^-})\)	\(G(\mathrm{Cl^-})\)	\(G(\mathrm{O_2})\)
\(-196\)	4.0	2.7	0.58	0.18	0.00	0.5	3.0
\(-80\)	4.0	2.8	0.49	0.17	0.00	0.6	2.7
\(-16\)	3.7	2.8	0.26	0.06	0.06	0.5	2.4
\(-8\)	3.5	2.6	0.26	0.09	0.06	0.5	2.6
\(0\)	3.4	2.6	0.20	0.07	0.06	0.5	2.2
\(20\)	3.7	2.9	0.19	0.09	0.08	0.5	2.5
\(72\)	4.1	3.3	0.19	0.09	0.00	0.5	2.7
\(260\)	4.9	4.2	0.00	0.00	0.00	0.7	2.0
\(295\)	3.8	2.9	0.00	0.00	0.00	0.9	1.0

of the cation, as in other salts, is attributed to a change in the free volume of the crystal lattice [78, 79]. Attention is drawn to the fact that, in the case of perchlorates, the influence of the nature of the cation is much less pronounced than in the case, for example, of nitrates. Apparently, this is connected with the fact that the predominant importance in the radiation-chemical stability of perchlorates in the series of cations of the alkali and alkaline-earth metals belongs not to the free volume, but to the electron affinity of the cation and anion [38]. However, for a confident interpretation of the mechanism of perchlorate radiolysis, the information available in the literature is still insufficient. In this connection, it would be highly desirable to study the intermediate products of perchlorate radiolysis. The accumulation of experimental results, such as the data on EPR spectra in irradiated perchlorates [80—85] and on optical absorption spectra [59], will make it possible to study in greater detail the intermediate stages of perchlorate radiolysis. From this point of view, the schemes of individual intermediate stages of radiolysis proposed in [78, 80] should be regarded as highly tentative.

VI. Annealing of Radiation Damage

A number of investigators, mainly in recent years, have undertaken the study of processes of restoration of the properties and stoichiometric composition of irradiated materials under the action of various kinds of treatments on the irradiated compounds.

The work carried out in this direction comprises two groups of studies with somewhat different objectives.

Most of the work is devoted to the study of annealing phenomena in preparations (mainly inorganic oxygen-containing radiation-stable salts) irradiated with slow neutrons. Cons-

mainly considered questions of isotope separation using the Szilard–Chalmers effect and the change with time in the distribution of isotopes among individual chemical forms in irradiated materials. In view of the indirect relation of this question to the subject of the radiation chemistry of solid compounds, and also because two detailed reviews [89, 90] have recently appeared in this field, in the present work we do not consider in detail the questions of annealing redistribution of recoil nuclei.

A comparatively small number of publications requiring consideration are devoted to questions of changes, during post-radiation annealing, in the concentration of radiolysis products in solid salts.

1. ANNEALING PHENOMENA IN IRRADIATED NITRATES

Most work on the annealing of radiolysis products has been carried out using barium nitrates and nitrates of certain heavy metals as objects. The first qualitative observations of the annealing of radiation products in nitrates were obtained, however, for KNO$_3$ [91]. In this and subsequent works, the change with time in the concentration of nitrite in nitrates irradiated by $\gamma$-radiation or by fast electrons was studied.

The material available in the literature is devoted exclusively to the kinetics of thermal annealing; post-radiation changes in the products under the influence of other factors have been studied extremely little.

In [92] it was established that, when Pb(NO$_3$)$_2$ crystals are heated at $t = 150^\circ$C, the concentration of nitrite in them decreases at a rate that is maximal in the initial period of annealing and acquires a constant value after 30–40 hours of heating. Atmospheric oxygen does not participate in the annealing; thermal decomposition of solid solutions of Pb(NO$_3$)$_2$—Pb(NO$_2$)$_2$ at 150°C is likewise not observed [93]. All this gave the authors grounds to believe that, during annealing of irradiated nitrates, oxidation of nitrate by molecular oxygen—a radiolysis product—takes place. In a series of works it was established that disturbances of the crystal lattice of nitrates play a very important role in annealing. Any treatment of irradiated crystals that leads to an increase in their defectiveness may be the cause of an increase in the annealing rate.

Thus, in [94] it is shown that pressing $\left(40\ \dfrac{T}{cm^2}\right)$ of irradiated Pb(NO$_3$)$_2$ powder leads to a noticeable increase in the annealing rate. A similar (quantitatively) effect is produced by grinding the irradiated lead nitrate, carried out before beginning its heating at 150°C [95]. In both cases the acceleration is characteristic of the initial (rapid) stage of annealing, while the constant rate of the process at later stages remains unchanged. At the same time, grinding the crystals itself causes only very weak annealing, whereas pressing of irradiated Pb(NO$_3$)$_2$, on the contrary, itself leads to a considerable decrease in the nitrite concentration. The fraction annealed in this case is 0.141 for a pressing pressure of $10\ \dfrac{T}{cm^2}$ and 0.181 for $20\ \dfrac{T}{cm^2}$ [14]. The authors tentatively associate the acceleration of annealing after pressing with the formation, under large nonuniform loads, of a significant number of point [13] defects—vacancies and interstitial ions [13], which play an important role in annealing [93, 96].

A study of annealing in irradiated \( \mathrm{Ba(NO_3)_2} \) [13] showed that in this case the process occurs at higher temperatures than for \( \mathrm{Pb(NO_3)_2} \). In contrast to \( \mathrm{Pb(NO_3)_2} \), barium nitrate is characterized by complete cessation of annealing after 20–40 hours of heating and by a smaller depth of the process (i.e., greater irreversibility of radiolysis). Some results on the thermal annealing of nitrites in irradiated \( \mathrm{TlNO_3} \) and alkali-metal nitrates are contained in [39, 97, 98]. The data presented in these works show that the annealing rate in nitrates depends both on the macroproperties of the crystal lattice [98] (free volume) and on microdisruptions and defects of the matrix, the content of which can be regulated by introducing into the nitrates impurity ions of different valence [39].

The dependence of the annealing rate on macroproperties leads to a regular change in the rate and depth of annealing in alkali-metal nitrates in the series from \( \mathrm{NaNO_3} \) to \( \mathrm{CsNO_3} \). At the same time, the temperature coefficient of the annealing rate in nitrates is substantially greater than the analogous value for the rate of radiolysis. As a consequence, the regular dependences of the radiolysis rate of alkali-metal nitrates at elevated temperatures and under irradiation with \(\alpha\)-particles (high temperatures in tracks) are much less pronounced [10]. On the other hand, the important role of microdisruptions is indicated by the change in the annealing rate upon introducing cations of higher valence into the nitrate lattice, accompanied by an increase in the number of cation vacancies in nitrates.

2. ANNEALING PHENOMENA IN OTHER OXYGEN-CONTAINING SALTS

The various oxidation forms of halides formed during the radiolysis of oxyhalogen compounds are thermally unstable products and may undergo various transformations during storage of irradiated salts. The kinetics of transformations of this kind in sodium chlorate are described in [63, 99]. Heating \( \mathrm{NaClO_3} \) in the temperature range 100–210°C leads to an increase in the yield of \( \mathrm{Cl^-} \) owing to thermolysis of chlorine–oxygen radiolysis products (\( \mathrm{ClO_2} \), \( \mathrm{ClO_2^-} \)). Thermostating the irradiated salt at 50°C, however, causes more complex changes with time in \( G_{\mathrm{Cl^-}} \): a rapid initial increase in \( G_{\mathrm{Cl^-}} \) and a subsequent decrease to a value smaller than \( G_{\mathrm{Cl^-}} \) in unannealed chlorate. The complex character of the kinetic annealing curve provides grounds for assuming that, alongside the process of decomposition of oxyhalogen forms, an essential role upon heating is played by competing reactions of recombination of these forms with oxygen trapped by the matrix, which lead to restoration of the radiation-destroyed anion. In [71, 74, 100, 101] experimental results are presented that confirm the consumption, during annealing, of oxidized forms of the halide by these two channels simultaneously and at rates not greatly differing from one another (for \( \mathrm{CsBrO_3} \) and \( \mathrm{KBrO_3} \)).

Annealing, probably proceeding in the form of thermal decomposition of chlorine–oxygen products, is also characteristic of alkali-metal perchlorates [78], although in this case the question of the nature of annealing (oxidation or thermolysis) was not specifically investigated by the authors. The results reported in [78] permit the assumption that in perchlorates, similarly to what occurs in alkali-metal nitrates [98], the annealing rate depends substantially on the free volume of the crystal lattice, and it is greatest for perchlorates having the maximum free volume.

The existence of this kind of correlation for nitrates and perchlorates of alkali metals makes it possible to raise the question of the generality of the relationship between the annealing rate of radiolysis products and the magnitude of the free volume of the lattice for the entire class of inorganic salts. The existence of such a relationship is understandable in the case when annealing proceeds as recombination (but not thermolysis) of radiolytic products. Improvement of the conditions for mutual diffusion of the products in a series of salts of the same type, in parallel with an increase in free volume, should naturally cause an increase in the annealing rate in the same sequence.

Of considerable interest are studies in which the influence of various kinds of treatments on radiation-stable salts (mainly chromates, bichromates, and phosphates) was investigated after their irradiation with slow neutrons [102—108]. In these studies it was established that treatment of irradiated salts with softer radiation that does not cause their radiolysis (light, γ-rays, fast electrons), as well as pressing or grinding them, accelerates the subsequent thermal annealing of the Szilard–Chalmers reaction products. As in the experiments with nitrates described above, the detailed mechanism of such stimulating treatments is unclear.

The efficiency of retention of the products of the Szilard–Chalmers process is also substantially affected by impurity ions introduced into the crystal lattice [109, 110]. The authors believe that this indicates the important role of ionic and electronic elementary processes in the annealing that occurs in neutron-irradiated inorganic salts.

REFERENCES

1. V. V. Sviridov. Photochemistry and Radiation Chemistry of Solid Inorganic Substances. Minsk, 1964.
2. T. H. Chen, E. R. Johnson. J. Phys. Chem., 66, 2249, 1962.
3. J. Cunningham. J. Phys. Chem., 66, 779, 1962.
4. J. Cunningham. J. Phys. Chem. Solids, 23, 843, 1962.
5. R. Livingston, H. Zeldes. J. Chem. Phys., 41, 4011, 1964.
6. Yu. A. Kulyupin, A. F. Yatsenko. FTT, 5, 3334, 1963.
7. J. Cunningham. J. Phys. Chem., 67, 1772, 1963.
8. Yu. A. Zakharov, V. A. Nevostruev. ZhFKh, 1966 (in press).
9. S. R. Logan. Nature, 199, 369, 1963.
10. C. J. Hochanadel. Rad. Res., 16, 286, 1962.
11. W. C. Mosley, W. G. Moulton. J. Chem. Phys., 43, 1207, 1965.
12. K. Zdansky, Z. Sroubek. Czechosl. Phys. J., B14, 121, 1964.
13. S. R. Mohanty, S. P. Upadhyay. J. Inorg. Nucl. Chem., 25, 309, 1963.
14. S. R. Mohanty, S. P. Upadhyay. Nature, 201, 921, 1964.
15. R. M. Golding, M. Henchman. J. Chem. Phys., 40, 1554, 1964.
16. E. R. Johnson. J. Amer. Chem. Soc., 80, 4460, 1958.
17. G. Henig, R. Lees, M. S. Matheson. J. Chem. Phys., 21, 664, 1953.
18. A. S. Baberkin, M. A. Proskurin, V. D. Orekhov. In: The Action of Ionizing Radiations on Inorganic and Organic Systems, p. 186. Moscow, 1958.
19. P. Pringsheim. J. Chem. Phys., 23, 369, 1955.
20. J. Cunningham. J. Chem. Phys., 41, 3522, 1964.
21. H. Zeldes. In: Paramag. Resonance, vol. 2, Acad. Press, p. 764, 1963.
22. H. Zeldes, R. Livingston. J. Chem. Phys., 37, 3017, 1962.
23. C. Jaccard. Phys. Rev., 124, 60, 1961.
24. W. B. Ard. J. Chem. Phys., 23, 1967, 1955.
25. R. Adde, P. Petit. C. r. Acad. sci., 256, 4862, 1963.
26. K. Gesi, V. U. Kazumata. J. Phys. Soc. Japan, 19, 1981, 1964.
27. R. Adde. C. r. Acad. Sci., 260, 504, 1965.
28. R. Adde. C. r. Acad. Sci., 260, 2781, 1965.
29. R. Adde. C. r. Acad. Sci., 261, 685, 1965.
30. P. W. Atkins, M. C. R. Symons. J. Chem. Soc., 4794, 1962.
31. J. Cunningham. J. Amer. Chem. Soc., 85, 3716, 1963.
32. E. R. Johnson, J. Forten. Disc. Faraday Soc., 31, 238, 1961.
33. C. J. Hochanadel, T. W. Dewis. J. Chem. Phys., 27, 333, 1957.

1. J. Cunningham. J. Phys. Chem., 65, 628, 1961.
2. J. Cunningham, H. G. Heal. Trans. Faraday Soc., 54, 1355, 1958.
3. E. R. Johnson. J. Phys. Chem., 66, 755, 1962.
4. A. S. Baberkin. Problems of Physical Chemistry, 2, 163, 1959; ZhFKh, 35, 373, 1961.
5. V. N. Beletsky. Autoreferat diss., Moscow, 1965.
6. V. V. Boldyrev, V. M. Lykhin, A. N. Oblivantsev. Kinetics and Catalysis (in press).
7. H. Zeldes, R. Livingston. J. Chem. Phys., 35, 563, 1961.
8. N. M. Atherton, R. N. Dixon, G. N. Kirby. Nature, 206, 83, 1965.
9. P. W. Atkins, N. Keen, M. C. R. Symons. J. Chem. Soc., 2873, 1962.
10. J. Tateno, K. Gesi. J. Chem. Phys., 40, 1317, 1964.
11. N. M. Atherton, R. N. Dixon, G. N. Kirby. Trans. Farad. Soc., 60, 1688, 1964.
12. K. Gesi, J. Takagi. Japan J. Appl. Phys., 3, 1267, 1964.
13. L. Zisen. Acta Chem. Scand., 12, 578, 1958.
14. L. V. Serykov, Yu. A. Zakharov. Collection of proceedings of a conference of chemists, Tomsk, 1966 (in press).
15. T. E. Hasty. Bull. Amer. Phys. Soc., 3, 299, 1958; T. E. Hasty, W. B. Ard, W. G. Moulton. Phys. Rev., 116, 1459, 1959.
16. C. Ramasastry, S. B. S. Sastry, Y. V. Murthy. J. Phys. Soc. Japan, 19, 770, 1964.
17. T. Hasty. J. Phys. Chem., 123, 167, 1962.
18. P. W. Atkins, J. A. Brivati, N. Keen, M. C. R. Symons. J. Chem. Soc., 4785, 1962.
19. F. T. Gamble. J. Chem. Phys., 42, 3542, 1965.
20. G. I. Subbotin. Optics and Spectroscopy, 18, 175, 1965.
21. Yu. A. Zakharov, L. V. Serykov. Collection of proceedings of a conference of chemists, Tomsk, 1966 (in press).
22. L. Rendall, W. C. Moulton. J. Chem. Phys., 31, 730, 1959.
23. V. S. Grechishkin. Priroda, 8, 85, 1959.
24. J. Depireuk. J. Dushesne Nature, 181, 759, 1958.
25. H. G. Heal. Canad. J. Chem., 37, 979, 1959.
26. V. A. Nevostruev, Yu. A. Zakharov, I. F. Kolesnikova. Present collection.
27. W. E. Webb, W. G. Moulton, W. B. Ard. Bull. Amer. Phys. Soc. Series, 11, 4, 389, 1959.
28. J. Fanart, B. Thieblemant. C. R. Acad. Sci., 261, 1501, 1965.
29. A. S. Baberkin, M. A. Proskurin, V. O. Orekhov. The Action of Ionizing Radiations on Inorganic and Organic Systems, p. 193, Academy of Sciences of the USSR, Moscow, 1958.
30. C. I. Hochanadel. J. Phys. Chem., 67, 229, 1963.
31. K. J. Mc. Callum, O. S. Holmes. Canad. J. of Chem., 29, 691, 1951.
32. L. J. Sharman, K. J. Mc. Callum. J. Chem. Phys., 23, 597, 1955.
33. C. W. Owens, F. S. Rowland. J. Inorg. Nucl. Chem., 24, 133, 1962.
34. S. B. S. Sastry, C. Ramasastry. J. Phys. Soc. of Japan, 18, 1220, 1963.
35. G. Boyd, J. Cobble, S. Wekler. J. Amer. Chem. Soc., 74, 217, 1952.
36. J. Cobble, G. Boud. J. Amer. Chem. Soc., 74, 1272, 1952.
37. G. Boyd, J. Cobble. J. Phys. Chem., 63, 919, 1959.
38. G. Boyd, F. W. Graham, Q. V. Larson. J. Phys. Chem., 66, 300, 1962.
39. G. Boyd. J. Phys. Chem., 68, 3809, 1964.
40. G. Boyd, Q. V. Larson. J. Phys. Chem., 68, 2627, 1964.
41. G. Boyd, Q. V. Larson. J. Phys. Chem., 69, 1413, 1965.
42. A. N. Oblivantsev, V. M. Lykhin, V. V. Boldyrev [et al.]. Present collection.
43. Handbook. Ionization Potentials and Electron Affinities. Publishing House of the Academy of Sciences of the USSR, 1962.
44. H. G. Heal. Can. J. Chem., 31, 91, 1953.
45. L. A. Prince, E. R. Johnson. J. Phys. Chem., 69, 359, 1965.
46. A. N. Oblivantsev, V. M. Lykhin, V. V. Boldyrev. Journal of the All-Union Chemical Society named after D. I. Mendeleev, 10, 598, 1965.
47. A. V. Dubovitskii, G. B. Manelis. Kinetics and Catalysis, 6, 828, 1965.
48. T. Cole. Proc. Natl. Acad. Sci. U. S., 46, 506, 1960.
49. Yu. M. Boyarchuk et al. Kinetics and Catalysis, 5, 823, 1964.
50. T. Cole. J. Chem. Phys., 35, 1169, 1961.
51. A. V. Dubovitskii, N. Ya. Buben, G. B. Manelis. ZhSCh, 5, 40, 1964.
52. J. S. Hyde, E. S. Freeman. J. Phys. Chem., 65, 1636, 1961.
53. M. Fujimoto, J. R. Morton. Can. J. Chem., 43, 1012, 1965.
54. V. A. Nevostruev, G. N. Bazhenova, Yu. A. Zakharov. Proceedings of a conference of young chemist-scientists, Tomsk, 1966, p. [[unclear: page number not visible]].
55. G. Odian, T. Acker, Th. Pletzke. J. Phys. Chem., 69, 2477, 1965.

1. M. A. Murin, R. V. Bogdanov, S. M. Tomilov. Uspekhi khimii, 33, 619, 1964.
2. S. R. Mohanty. J. Scient. Res. Banaras, 2, No. 2, 1961–62, Lindu Univ.
3. M. A. Proskurin, V. D. Orekhov. Proceedings of the Conference on the Peaceful Uses of Atomic Energy. Moscow, 41, 1955.
4. A. G. Maddock, S. R. Mohanty. Nature, 182, 1797, 1958.
5. A. I. Maddock, S. R. Mohanty. Disc. Far. Soc., 31, 193, 1961.
6. S. R. Mohanty, S. R. Upadhyay. Nature, 199, 169, 1963.
7. S. R. Mohanty, S. R. Upadhyay. Indian J. of Chem., 1, 456, 1963.
8. A. Bees. Adv. in Phys., 1, 43, 1952.
9. A. G. Maddock. Radioch. Acta, 1, 85, 1963.
10. Yu. A. Zakharov, E. P. Abakumov. This collection.
11. P. Patrick, K. J. McCallum. Nature, 194, 767, 1962.
12. G. E. Boyd. Naturwissen, 52, No. 11, 1965.
13. T. Anderson. Nature, 200, 1094, 1963.
14. J. H. Green, G. Harbottle. Trans. Farad. Soc., 49, 1413, 1953.
15. R. F. C. Claridge, A. G. Maddock. Nature, 184, 1932, 1959.
16. A. G. Maddock, J. I. Vargas. Nature, 184, 1931, 1959.
17. A. G. Maddock. Trans. Farad. Soc., 59, No. 26, 26, 1963.
18. N. Getoff. Nature, 199, 593, 1963.
19. N. Getoff. Radioch. Acta, 1, 49, 1963.
20. M. Vlatkovic, A. Aten. J. Inorg. Nucl. Chem., 24, 139, 1962.
21. T. Anderson, A. G. Maddock. Nature, 94, 372, 1962.
22. T. Anderson, A. G. Maddock. Trans. Farad. Soc., 59, 2363, 1963.