EPR Spectra in Irradiated Potassium Perchlorate

A. A. Vasiliev, L. V. Serikov, R. N. Isaev, Yu. A. Zakharov

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

Paramagnetic products in irradiated chlorates and perchlorates of alkali metals have been observed by many authors [1–5, etc.]; the data obtained by them indicate a significant role of radicals in the mechanism of the radiation-chemical decomposition of these salts. To elucidate the poorly studied initial and intermediate stages of the radiolysis process, much can be gained from studies at low temperatures, when the mutual transformations of unstable products are substantially slowed and the latter can serve as objects of direct investigation. In the present work, for this purpose, the EPR spectra of irradiated potassium perchlorate were studied at temperatures from \(-155^\circ\mathrm{C}\) to \(+130^\circ\mathrm{C}\).

Experiment

The experiments were carried out on single-crystal samples grown by slow cooling of a saturated aqueous solution of a “pure” grade \( \mathrm{KClO_4} \) reagent, recrystallized three times in distilled water. The crystals had the shape shown in Fig. 1, with dimensions of about \(5 \times 5 \times 2\) mm.

The samples were irradiated with electrons on an ESG-2.5 accelerator.

An electron beam with an energy of 1.6 MeV was introduced through a cylindrical aperture in the pole piece of the electromagnet directly into the microwave resonator of a modified EPA-2 spectrometer, where the crystal being irradiated was placed.

EPR spectra were recorded either immediately after irradiation or after heating the sample. The required temperature of the sample was maintained by blowing a stream of evaporating nitrogen through the resonator and could be regulated within the range from \(-165^\circ\mathrm{C}\) to \(+20^\circ\mathrm{C}\) with an accuracy of \(\pm 2^\circ\mathrm{C}\).

Fig. 1. Single crystal \( \mathrm{KClO_4} \)

The temperature was measured with a copper–constantan thermocouple. The thermocouple junction was pressed between the sample and the inner surface of a thin-walled ampoule made of “Luch-2” glass, which served to introduce the sample into the resonator. At room temperature, EPR spectra were recorded on an RE-1301 spectrometer. Analysis of gaseous

products released upon heating the irradiated samples was carried out on an MX-1302 mass spectrometer.

Results

1. Irradiation at −155°C.

After irradiation of KClO₄ single crystals, they acquire a cherry-red color and a very intense EPR spectrum appears from paramagnetic centers of at least two types, having different concentrations and different orientational properties (Fig. 2a, b).

Analysis of the spectrum shows that it consists of no fewer than 24 lines, which are divided into three groups, denoted by the numbers I (I′); II (II′) and III (III′). The intensity of the lines denoted by primed numbers is three times smaller than the intensity of the corresponding unprimed lines, and their splittings are in the ratio 0.83 : 1. These ratios, and the fact that each of the identically denoted groups includes 4 equidistant lines of equal intensity, make it possible to assign each group of lines in the spectrum to a radical containing one nucleus of the isotope Cl³⁵ (I, II, III) or Cl³⁷ (I′, II′, III′).

Observations of the change in the spectrum upon rotation of the crystal in a magnetic field show that the lines I (I′) and II (II′) have a strongly anisotropic \(g\)-factor and hyperfine-structure constants, possessing symmetry no higher than axial.

In the case where the magnetic field \(\vec H\) is parallel to the crystallographic axis \(\vec a\) (Fig. 1b), the groups of lines I and II, I′ and II′ merge and give one spectrum of doubled intensity. These groups of lines are similar to the spectrum given in [1]. The lines of group III have a smaller width and smaller hyperfine splitting. Upon rotation in a magnetic field, no anisotropy of them was observed. Holding the samples for 10 min at any temperature in the interval −155–−95°C, followed by cooling and recording of the EPR spectrum at −155°C, showed that the intensity of the spectrum does not change appreciably; at a temperature of −75°C, however, within 10 min complete disappearance of the observed spectrum occurs, and no new lines of comparable intensity appear.

2. Irradiation at +20°C.

At room temperature, an EPR spectrum is observed in irradiated single crystals, produced by at least four types of paramagnetic centers (Fig. 2c). The corresponding lines are denoted by the numbers 1; 2; 3(3′); 4(4′). The lines 3(3′) are completely analogous to the spectrum presented in [2], where it is assigned to the ClO₂ radical. The lines 4(4′) correspond to the data of [3] for the ClO₃ radical. The number 1 denotes a very weakly anisotropic singlet, assigned in [4] to the ion-radical \(\overline{\mathrm O}_3\). There are no literature data concerning the doublet under the number 2. Each of the lines 2, 3(3′), and 4(4′) is split into two lines for all orientations of the crystal except \(\vec H \perp \vec b\) and \(\vec H \perp \vec a\). After 8-hour irradiation with unfiltered light from a DRSh-500 mercury lamp, complete destruction of ClO₃ occurs and an approximately 5-fold weakening of the ClO₂ spectrum; with more prolonged illumination (for 30 hours), the ClO₂ spectrum also disappears completely (Fig. 2e). At the same time, the appearance of any lines that could be assigned to the ClO radical was not detected, although its formation could have been expected in the photolysis of ClO₂. Upon heating irradiated samples that had not been exposed to ultraviolet light to a temperature of +130°C, cracking occurs

single crystals. After this, only a single line remains in the EPR spectrum (Fig. 2d). Mass-spectrometric analysis of the gaseous products released upon heating irradiated KClO$_4$ showed the presence of oxygen.

Thus, the EPR spectrum observed in irradiated KClO$_4$ at $-155^\circ$C indicates the presence of two types of radicals. One of them

Fig. 2. EPR spectra of irradiated KClO$_4$:
a — $T=-155^\circ$C; angle $\vec H$ with $c = 90^\circ$; angle $\vec H$ with $a = 15^\circ$;
b — $T=-155^\circ$C; angle $\vec H$ with $c = 90^\circ$; angle $\vec H$ with $a = 0^\circ$;
c — $T=+20^\circ$C; $\vec H \parallel \vec a$;
d — EPR spectrum of irradiated KClO$_4$ after 30-hour illumination with ultraviolet light;
e — spectrum after heating to $+130^\circ$C.

can be correlated with the spectrum given in [1], attributed to the ClO$_4$ radical, whose formation may be assumed on the basis of the ionization mechanism of the primary stage of radiolysis:

\[ \mathrm{ClO}_4^- \longrightarrow \mathrm{ClO}_4 + e^- . \tag{1} \]

The other may correspond to a paramagnetic center of tetrahedral structure and has no analogs among the known spectra of chlorine–oxygen radicals.

Apparently, the only possible center of this kind in KClO$_4$ may be the radical $\mathrm{ClO}_4^{=}$, formed by the reaction:

\[ \mathrm{ClO}_4^- + e^- \to \mathrm{ClO}_4^{=} . \tag{2} \]

This interpretation of the spectrum is supported by the fact that the resulting concentration of the proposed $\mathrm{ClO}_4^{=}$ particles is approximately 3 times less-

... the concentration of $\mathrm{ClO_4}$, which may be explained by the capture, by various structural and impurity defects of the crystal lattice, of some of the free electrons formed in reaction (1).

Verification of this assumption, however, requires additional experiments.

The parallel disappearance of the spectra of both radicals upon heating the samples indicates the possibility of reverse reactions:

\[ \mathrm{ClO_4^- \to ClO_4 + e,} \tag{3} \]

\[ \mathrm{ClO_4 + e \to ClO_4^- .} \tag{4} \]

The fact that lines 2 disappear in the EPR spectra (Fig. 2c) at $T = 130^\circ\mathrm{C}$, the evolution of gaseous oxygen, and the cracking of the crystals upon heating make it possible to assign these lines to oxygen retained in the lattice, the formation of which under radiolysis has been established by many authors. The paramagnetic state of $\mathrm{O_2}$ may be the ${}^{3}\Sigma_g^{-}$ state. With regard to an analogous doublet in sodium chlorate, such an assumption was made in [5].

REFERENCES

1. J. R. Morton. J. Chem. Phys., 45 (5), 1800, 1966.
2. T. Cole. Proc. Nat. Acad. Sci. U. S., 46, 506, 1960.
3. P. W. Atkins et al. J. Chem. Soc., 4785, 1962.
4. A. V. Dubovitskii, G. B. Manelis. Kinetics and Catalysis, 8, 828, 1965.
5. S. B. S. Sastry, C. Ramasastry. J. Phys. Soc. Japan, 19, 770, 1964.