In conclusion, we also note that the methods considered in the present paper can be successfully employed not only to investigate In:_xGaxP1_zAsz but also other four-component solid solutions. We wish to thank L. A. Titov for providing specimens of solid solutions isoperiodic with InP. LITERATURE CITED i. 2. 3. 4~ 5. 6. 7. 8.

9. i0. ii. 12. 13. 14. 15. 16.

A. P. Bogatov et al., Fiz. Tekh. Poluprovodn., 9, No. i0, 1956-1961 (1975). Zh. I. Alferov et al., Pis'ma, Zh. Tekh. Fiz., ~, No. ii, 481-483 (1976). J. J. Coleman, N. Holonyak Jr., and M. J. Ludowise, IEEE J. Quantum Electron., ii, No. 7, 471-476 (1975). O. N. Ermakov et al., IEEE Trans. Electron Devices, 26, No. 8, 1190-1194 (1979). L. M. Dolginov et al., Pis'ma Zh. Tekh. Fiz., ~, No. 14, 631-633 (1976). L. M. Dolginov, P. G. Eliseev, and I. Ismailov, Itogi Nauki i Tekh., Ser. Radiotekh., 21 3-115 (1980). V. I. Osinskii, V. I. Privalov, and O. Ya. Tikhonenko, Optoelectronic Structures Based on Multicomponent Semiconductors [in Russian], Nauka i Tekhnika, Minsk (1981). O. N. Ermakov, Methods of Obtaining Solid Solutions of In:-xGaxP, In:-xGaxP~-zAsz and Multilayer Structures Based on Them, Reviews of Electronic Techniques [in Russian], TsNII, Moscow, No. 5(864), 5 (1982). N. A. Gul'ko, S. G. Konnikov, T. B. Popova, and E. K. Tropp, Avtometriya, No. 6, 98-110 (1980). O. N. Ermakov, Elektron. Tekh. Ser. 2 (Semiconductor Devices), No. 6(149), 3-10 (1981). O. N. Ermakov, R. S. Ignatkina, V. P. Sushkov, and M. V. Chukichev, Fiz. Tekh. Poluprovodn., ii, No. 6, 1102-1107 (1977). E. Johnson, "Absorption in the region of the edge of the fundamental band," in: Optical Properties of Semiconductors [Russian translation], Mir, Moscow (1970), pp. 166-277. H. H. Caspers and H. H. Wieder, Solid State Commun., 29, No. 4, 403-406 (1979). Yu. A. Golovanov, O. N. Ermakov, and V. P. Sushkov, Elektron. Tekh. Ser. 2 (Semiconductor Devices), No. 8(151), 24-28 (1981). R. L. Moon, G. A. Antypas, and L. W. James, J. Electron. Mater., 33, No. 4, 635-642 (1974). N. A. Bert et al., Fiz. Tekh. Poluprovodn., 16, No. i, 60-67 (1982).

"NONADIABATIC" PASSAGE IN MAGNETIC RESONANCE I. Z. Rutovskii, A. S. Sasim, and G. G. Fedoruk

UDC 537.635

For the "nonadiabatic" passage of magnetic-resonance spectra (dH/dt >> yH~ and dH/dt > y(AH)2), according to the Bloch model, the magnetization vector cannot follow the changes in the effective magnetic field, in view of which a magnetization "loss" effect should occur~ which manifests itself as a reduction in the signal amplitude by a factor of [y(AH)2/dH/dt] I/2, where y is the gyromagnetic ratio, HI is the amplitude of the alternating magnetic field, dH/ dt is the rate of change of the polarizing magnetic field, and AH is the line width [1-3]. On the other hand, in [4-6], from an analysis of the results of an investigation of transient oscillations, free induction, and the echo in ESR, it was concluded that the Bloch model is unsuitable for describing the dynamics of magnetic-resonance signals. It was shown that the rate of reorientation of the unit magnetic moment in the opposite direction is determined not by the amplitude HI of the alternating magnetic field, but by its frequency, a consequence of which is the prediction that the amplitude of the signal will be constant under conditions of nonadiabatic passage of magnetic resonance. In order to check this conclusion, we investigated the behavior of the signal as a function of the rate of passage of the resonance line. We carried out the experiment using ESR apparatus. We investigated the ESR signals of solutions of sodium in ammonia (AH = 0.07 Oe, concentration of paramagnetic particles ~i02~ 1984.

476

Translated from Zhurnal Prikladnoi Spektroskopii, Vol. 40, No. 4, pp. 657-659, April, Original article submitted January 14, 1983.

0021-9037184/4004-0476508.50

9 1984 Plenum Publishing Corporation

F

7/D

b

'i-

b

2,5

o

3,5~ I

C

0,5

I

JO~ Fig. i

_

I

I

10~

;~-~

,

__

107 d_H, Oe/sec dt

Fig. 2

Fig. I. Absorption signals of solutions of sodium in ammonia for different rates of change of the magnetic field: 1.5 x 104 (a), 3 x 104 (b), and 2 x 105 0e/sec (c). Fig. 2. Amplitude (A) of the absorption signal from Solutions of sodium in ammonia as a function of the rate of change of the magnetic field: a) calculated using the Bloch model, b) for the hypothesis suggested in [5]. The points are experimental results. Ao is the maximum signal amplitude. cm -3) and E~-centers in crystalline quartz (AH = 0.15 0e, concentration ~5.1017 cm-3). The investigations were carried out on a 3-cm band superheterodyne spectrometer with a recording band of about 1.5 MHz. After the video amplifier the signal was observed on an oscilloscope or, if the signal/noise ratio was not high enough, it was processed by a single-channel gated recorder. The magnetic field was varied sinusoidally in the 102-106 Hz band, had an amplitude of up to 5 0e, and was obtained by passing a current either through modulation coils, placed outside the resonator (102-104 Hz), or through an internal modulation loop (i04-i0 ~ Hz). The amplitude of the magnetic-field modulation at the point where the specimen was situated was measured using a Hall detector. The value of HI in the resonator was calibrated from the frequency of the transient oscillations, which is equal to yHI [5]. The absorption signals from solutions of sodium in ammonia are shown in Fig. i for different rates of change of the magnetic field. A curve of the amplitude of this signal as a function of the rate of passage is shown in Fig. 2. The value of the stationary microwave field was chosen to be such that the magnetic system was not saturated (yH~ = 140 0e/see). The measurements were made at room temperature. In our case the conditions for nonadiabatic passage began to be satisfied when dH/dt > 2 x 104 0e/sec, which is also confirmed by the appearance of a beat signal often called "wiggles". In this range of variation the Bloch model predicts a reduction in the signal amplitude, which when dH/dt = 105 0e/sec should more than halve. At the same time, the observed signal had a constant amplitude up to dH/dt = 3 x 105 0e/sec, and its reduction when the rate of variation of the magnetic field was increased further was determined by the spectrometer passband. Unlike the predictions of the Bloch model, the position of the maximum of the absorption signal was also independent of the rate of passage of the lin~, and coincided with the center of the resonance curve. Similar results were obtained for E~-centers in quartz. These results agree with the results obtained previously for nonadiabatic establishment of resonance conditions by pulses of a polarizing magnetic field in ESR [6], and in ferromagnetic resonance [7], which show that the time taken for the resonance-absorption signal to grow is much shorter than the time 2/yAH, predicted by theory. The behavior of the signals described above indicates that there is no magnetizationloss effect, and cannot be explained within the framework of the Bloch model of magnetic resonance. Attempts to modernize this model taking into account spectral diffusion, excitation transfer between paramagnetic centers, and other relaxation processes gave no positive result, since nonadiabaticity conditions are determined by the sortest relaxation process. In addition, an important feature of the results obtained is the fact that the behavior of

477

the signals is the same both in solids and in liquids, where the applicability of the Bloch model is unquestionable. Thus, for a liquid solution of sodium in ammonia the time, which we determined from the line width, is 2.5 x 10 -6 see, and agrees with the spin-lattice and phaserelaxation times for this substance [8]. At the same time, the amplitude of the signal remains unchanged for line passage in a time of about 2 x 10 -7 sec. We therefore concluded that the reason why the model is not justified is the inadequate description not of the relaxation processes, but of the resonance (coherent) processes, and is due to the use of the approximation of free magnetic moments, which independently interact with the microwave field. However, this approximation cannot be applied to the description of coherent magnetic-resonance phenomena is condensed systems, since in this case the average distance between magnetic moments is much less than the wavelength of the acting radiation, and collective effects must be taken into account. In conclusion, we note that the agreement of the decay time and the nature of the variation of the frequency of the wiggles with existing calculations [i] is due to the use in describing them of the general properties of the spectral-time signal transformations, which are unconnected with the limitations of the Bloch model. On the other hand, since the wiggles are due to beats between the microwave field and the signal of free precession, the absence of a magnetization-loss effect makes it necessary to consider the nature of the latter in more detail. LITERATURE CITED i. 2. 3. 4. 5. 6. 7.

B . A . Jacobsohn and R. K. Wangsness, Phys. Rev., 73, No. 9, 942-946 (1948). M. Weger, Bell Syst. Tech. J., 39, 1013-1112 (1960). A. ~6she, Nuclear Induction [Russian translation], IL, Moscow (1963). I . Z . Rutkovskii and G. G. Fedoruk, Zh. Eksp. Teor. Fiz., 78, No. 3, 1237-1239 (1980). G . G . Fedoruk, I. Z. Rutkovskii, D. P. Erchak," and V. F. Stel'mach, Zh. Eksp. Teor. Fiz., 80, No. 5, 2004-2009 (1981). G . G . Fedoruk and I. Z. Rutkovskii, Phys. Status Solidi (b), 112, No. 2, 453-456 (1982). Yu. V. Kornev, O. S. Kolotov, and V. Ya. Sysoev, Dokl. Akad. Nauk SSSR, 261, No. 2, 355-

8.

J.C.

356 (1981). Thompson, Electrons in Liquid Ammonia, OxfordUniv. Press (1976).

ESTIMATION OF LIFETIME OF SULFUR DIOXIDE IN OPTICAL CUVETTE UDC 543.42

G. T~ Burova, V. N. Tselikov, and E. A. Chayanova

In connection with the development of remote-control optical methods of measuring the principal atmospheric contaminants, a problem arose of the calibration of instruments by means of standard cuvettes that contain a known content of the given gas. The cuvette with the gas is the necessary element for calibrating optical gas analyzers of many different types, including laser and correlational spectrometers, those based on the differential absorption method, and also various types of nondispersional apparatus in which gas-filled filters are used. The accuracy of the results of the analysis of gases can be judged only when information is available on the reliability and accuracy of the composition -- property dependences, not only for two component, but also for multicomponent gas mixtures, because the presence of additional admixtures can appreciably influence the form of this dependence. At present, the composition -- property dependences are known for most multicomponent gas mixtures with a relatively low accuracy. The form of the equation correlating the required property of a gas mixture with the concentrations of the components included in it can be established either theoretically or experimentally [i].

1984.

478

Translated from Zhurnal Prikladnoi Spektroskopii, Vol. 40, No. 4, pp. 660-664, April, Original article submitted December 12, 1982.

0021-9037/84/4004-0478508~

9 1984 Plenum Publishing Corporation