Gamma Spectroscopy Using Scintillation Counters

Since the discovery of radioactivity and up to the present time, gamma spectroscopy has been one of the most important and precise methods for studying the structure of nuclei. It is known that measuring the energies of γ-quanta emitted by nuclei, as well as studying γ-γ and γ-β coincidences, makes it possible to establish the energy levels of a nucleus, while investigation of the angular correlation between γ-quanta successively emitted by an excited nucleus gives information about the electric moment of the nucleus. Until recently, however, the possibilities of gamma spectroscopy were limited by the insufficient efficiency of Geiger–Müller counters for γ-quanta (the efficiency of a Geiger–Müller counter for γ-quanta lies in the range from tenths of a percent to several percent, depending on the energy of the γ-quanta and the material of the counter cylinder). The development of a method for counting nuclear particles and γ-quanta by means of scintillation counters made of dense crystals that give a large yield of light quanta per absorbed particle or absorbed γ-quantum opens new possibilities for gamma spectroscopy. In this brief review several works will be considered in which the detection and measurement of the energy of γ-quanta emitted by radioactive substances were carried out with the aid of scintillation counters. The advantages of this new and very rapidly developing method of detecting γ-quanta are connected above all with the fact that the efficiency of scintillation counters for γ-radiation exceeds the efficiency of Geiger–Müller counters by tens of times. In addition, it has been shown that a scintillation counter in combination with a photomultiplier is a linear device in which the pulse at the output of the photomultiplier is proportional to the energy of the electron produced by the γ-quantum in the substance of the scintillation counter. It is precisely this property of the scintillation counter that makes it possible to measure the energy of the γ-quanta absorbed in it. A scintillation counter also possesses all the advantages of a fast counter, whose dead time is practically on the order of a time hundreds of times shorter than the dead time of Geiger–Müller counters. In measurements of γ-γ and β-γ coincidences with the aid of a scintillation counter, the resolving time can be considerably reduced in comparison with Geiger–Müller counters, which greatly increases the ratio of the measured effect to the background of random coincidences. The mechanism of operation of a scintillation counter irradiated by γ-quanta consists in the following…

(see UFN 39, issue 3, 419 (1950)), that a secondary electron, produced in a counter by a gamma quantum absorbed or scattered in the counter, brakes in the counter (if the dimensions of the crystal are sufficiently large), spending its energy on tearing electrons out of the crystal lattice and on exciting fluorescence centers. In this process, part of the excitation energy, over a time of the order of \(10^{-8}\)—\(10^{-9}\) sec, is emitted in the form of light quanta, which, striking the photocathode of a photomultiplier placed next to the counter, produce at its output a pulse proportional to the number of light quanta reaching the photocathode, i.e., proportional to the energy of the secondary electron.

Fig. 1.
Labels in the figure: \(A\)—source; \(B\)—crystal; \(C\)—photomultiplier; \(D\)—amplifier; \(E\)—discriminator; \(F\)—counting circuit.

As an example, let us point out that a quantum with an energy of \(1\ \mathrm{MeV}\), being absorbed in a NaJ crystal, releases on average about 20,000 photons of visible light.

Fig. 2.
Labels in the figure: absorption coefficient, \(\mathrm{cm}^{-1}\); energy, \(\mathrm{MeV}\); Compton effect; photoeffect; pairs.

The simplest setup¹ for counting \(\gamma\)-quanta with the aid of a scintillation counter is shown in Fig. 1. Here \(A\) is a radioactive source of \(\gamma\)-radiation, \(B\) is a crystal counter located near the photocathode of the photomultiplier, \(D\) is an amplifier, \(E\) is a discriminating circuit (it is necessary in order to cut off the so-called dark pulses of the photomultiplier, caused by thermionic emission from the electrodes), and \(F\) is a counting and registering circuit. The crystals most commonly used at present are zinc sulfide, anthracene, calcium tungstate or scheelite \((\mathrm{CaWO_4})\), and sodium iodide activated with thallium \((\mathrm{NaJ}(\mathrm{Tl}))\).

Absorption of \(\gamma\)-quanta in a scintillation counter leads to the formation of secondary electrons as a result of the following processes: 1) Compton scattering; 2) the photoeffect; 3) pair formation, if the energy of the \(\gamma\)-quantum is greater than \(1.02\ \mathrm{MeV}\). Fig. 2 shows the calculated cross sections of these three processes for iodine, which is part of the NaJ crystal. To take into account the influence of absorption of \(\gamma\)-quanta in Na, the ordinate of the Compton-effect curve must be increased by 21%, and that of the pair curve by 4%. In accordance with these three types of electrons produced by \(\gamma\)-quanta, one should expect, in the amplitude distribution of pulses measured by the setup shown in Fig. 1, the appearance of three lines corresponding to pair electrons, Compton electrons, and photoelectrons. Since the recoil-electron energy in the Compton effect varies over wide limits, Compton electrons give the broadest...

... distribution of pulses, while the photoelectrons and pair electrons will give almost monochromatic lines corresponding to the electron energy

\[ h\nu - E_K \]

in the case of the photoeffect from the \(K\)-shell, and to the energy

\[ \frac{1}{2}(h\nu - 1.02)\ \text{MeV} \]

in pair formation. The authors of the papers under review\(^2\) indeed observed all three distributions. In Fig. 3 the \(\gamma\)-spectrum of \(Co^{60}\), obtained in a NaJ(Tl) crystal of dimensions \(1.2 \times 1.2 \times 1.2\ \text{cm}^3\) with a beam collimation equal to \(1^\circ\), is shown. It is well known, in particular from very precise measurements performed on a bent-crystal spectrograph\(^3\), that \(Co^{60}\) emits two lines with energies \(h\nu_1 = 1.17\) and \(h\nu_2 = 1.33\ \text{MeV}\). The first large maximum on the left in Fig. 3 corresponds to the Compton electrons from these two lines. The next two maxima correspond to photoelectrons from the same lines. Since the energy of the \(\gamma\)-quanta is not sufficiently large, the presence of pair electrons

Fig. 3.

Fig. 4.

was not manifested in the spectrum in any noticeable way. It is obvious that the circumstance that each line may be represented by three lines in the spectrum makes the interpretation of complex \(\gamma\)-spectra extremely difficult. Therefore, and also in order to increase the resolving power of this method of gamma spectroscopy, the method of investigating \(\gamma\)-spectra by measuring coincidences of light flashes in two crystals was developed. The idea of these measurements of \(\gamma\)-spectra is clear from Fig. 4. A quantum, entering the crystal, is scattered and creates a recoil electron \(\beta\) and a quantum \(h\nu'\) (going backward in Fig. 4). From the theory of the Compton effect it is known that the energy of the primary \(\gamma\)-quantum is determined by the angle at which the secondary quantum is scattered, and by the energy of the recoil electron. Therefore, if the \(\gamma\)-quanta scattered at a certain angle are somehow fixed, then monochromatic recoil electrons will be recorded in crystal \(x\), and each \(\gamma\)-line in the spectrum will correspond to its own group of recoil electrons, whose degree of monochromaticity depends on the accuracy of the angle \(\theta\) determination. The second crystal \(D\) serves to detect quanta scattered at this angle. A radio-engineering circuit measures the coincidence of light flashes in crystals \(x\) and \(D\) and the magnitude of the pulse at the output of the photomultiplier connected with crystal \(x\). If it is necessary to increase the luminous intensity, which may arise when working with low-intensity radioactive substances, crystal \(D\) may be made in the form of a circular ring. In the experiment corresponding to Fig. 4, the angle was chosen equal to \(150^\circ\). The choice of such an angle is explained by the fact that, as is known from the theory of the Compton effect,

the energy of the recoil electrons corresponding to quanta scattered through angles of \(135—180^\circ\) varies less than for the same measurements of the angle \(\theta\) in any other range of values. The spectrum of \(\gamma\)-rays of \(\mathrm{Co}^{60}\) obtained by this method is shown in Fig. 5, where two separate lines are distinctly visible: \(h\nu_1 = 1.17\) MeV and \(h\nu_2 = 1.33\) MeV, formed by Compton electrons. Comparison of this figure with Fig. 3 shows that, in measurements by the coincidence method, the lines associated with photoelectrons have disappeared, while the lines associated with pair electrons, which, generally speaking, can also be detected by the coincidence method, did not appear because of their low intensity, just as they did not appear in the spectrum of Fig. 3.

Another method for studying \(\gamma\)-spectra requires the use, as the recording instrument, of a cathode oscilloscope with a waiting sweep, triggered by the front of an amplified pulse from a photomultiplier.^4 This same amplified pulse is fed to the deflection plates of the oscilloscope. By photographing the screen of the oscillographic tube with a camera with a moving shutter, one can obtain, in a single photograph, the superposition of all the lines. The deciphering of the photographs is carried out by photometry.

Fig. 5.

Fig. 6.

In Fig. 6 is shown one of the \(\gamma\)-spectra of \(\mathrm{Au}^{198}\) obtained^5 in this way. The upper part of the photograph corresponds to monochromatic photoelectrons. The continuous blackening of the photographic plate results from the superposition of pulses of different amplitudes and is caused by Compton electrons. It should be noted^4 that the latter method is especially important in the study of spectra of radioactive substances with lifetimes on the order of seconds. In this case the entire spectrum can be recorded on the oscilloscope screen in a few seconds and then interpreted photometrically.

It is of interest to compare the resolving power of the new scintillation method of gamma spectroscopy with the resolving power of known methods. Let us turn to Fig. 5, from which it is clear that the half-width of both lines \(h\nu_1 = 1.17\) and \(h\nu_2 = 1.33\) MeV is approximately \(5\%\). Meanwhile, these same lines have been measured^3 on a spectrometer with a bent crystal with an accuracy of \(1/10—1/20\%\). The resolving power of the scintillation method is strongly limited by fluctuations in the number of light quanta arising during the slowing down of secondary

particles and reaching the photocathode, as well as by fluctuations in the amplification of the photomultiplier. However, the simplicity and effectiveness of the new method of gamma spectroscopy are ensuring its ever wider application. With the aid of scintillation counters, a large interval of energies can be detected and measured, from soft X-rays to very hard gamma rays. Evidently, the new method of investigation will also find application in medical and biological work involving radioactivity.

A. Weissenberg

REFERENCES CITED

1. Motton and Mitchell, Operation of the 931-A photomultiplier with a scintillation counter. Nucleonics, January 1949, p. 16.
2. Hofstadter and McIntyre, Phys. Rev. 78, 617, 619 (1950); 79, 389 (1950).
3. Lind, Brown and Du Mond, Phys. Rev. 76, 12, 1838 (1949).
4. Campbell and Goodrich, Phys. Rev. 78, 5, 640 (1950).
5. Hofstadter and McIntyre, Nucleonics, September 1950, p. 32.