HAL Id: jpa-00216972
https://hal.archives-ouvertes.fr/jpa-00216972
Submitted on 1 Jan 1977HAL is a multi-disciplinary open access
archive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.
TIME EFECTS IN 57Fe NMR IN YFeO3
S. Nadolski, W. Zbieranowski
To cite this version:
JOURNAL DE PHYSIQUE Collogue CI, supplement au n° 4, Tome 38, Avril 1977, page Cl-65
TIME EFECTS IN
57Fe NMR IN YFe0
3S. NADOLSKI and W. ZBIERANOWSKI
Institute of Physics Polish Academy of Sciences Warsaw Al. Lotnikow 32/46 Poland
Résumé. — La résonance magnétique nucléaire du "Fe dans des plaquettes minces
monocrys-tallines de YFeCh a été étudiée par écho de spin et par technique stationnaire, dans la gamme de température 1,8 à 450 K. Une modification des signaux est observée après application d'un faible champ magnétique continu. Ce nouvel état est instable ; à 77 K son temps de récupération est d'environ 0,5 heure ; il décroît quand la température augmente. En dessous de 77 K cet effet disparaît. Dans l'état instable le facteur d'amplification devient deux fois plus grand qu'avant l'application du champ. Le signal de spin écho n'augmente pas mais on observe une diminution de l'amplitude des impulsions nécessaire pour obtenir le signal spin écho. Ce fait laisse penser que le volume des parois décroît. On note à 1,8 K quelques particularités caractéristiques d'une transi-tion de phase.
Abstract. — The NMR of 57Fe has been studied in a thin single crystal plates of YFeCh by
spin echo and steady-state methods at the temperature range of 1.8-450 K. The observed signals undergo a change after the application of a weak magnetic DC field. The new state is unstable. The recovery time at 77 K is about 0.5 h and decreases with increasing temperature. Below 77 K these effects disappear. In the unstable state enhancement factor t) is twice as large as before the field application. Spin echo signal does not increase but the amplitude of radiopulses required for the forming of spin echo signal decreases. This fact suggests that the volume of domain walls decreases.
At 1.8 K some features characteristic for the phase transition were noticed.
1. Introduction. — Yttrium orthoferrite Y F e 03 is a
canted antiferromagnet with the magnetization vector directed along the « c » axis [1]. NMR in Y F e 03 was
previously observed by a spin echo [2] and by the steady-state methods [3-5]. In the spectrum corres-ponding to the domain walls the resonance line is split and has two components [4, 6]. The splitting of the resonance line in connected with the dependence of local NMR frequency on the value of an angle 9 inside the domain wall. The 8 is an angle between the easy axis and magnetization in the domain wall. For the uniaxial ferromagnets the following depen-dence is presumed [7, 8] :
v(0) = Vo(l - a sin2 9) . (1)
Where the v0 is the NMR frequency inside the domain.
The theories presented in ref. [7, 8] predict the exis-tence of two resonance lines at frequencies vmax = v0
and vmin = v0(l — a). The enhancement factor r\ for
the vmin line is about one order of magnitude larger
than n for vmax [8]. The enhancement factor depends on
the position inside the domain wall in a following way : n ~ sin 9 [6-9]. However, the orthoferrites show another dependence of the local NMR frequency on the position inside the domain wall [5, 6]. In the approximation of two magnetic sublattices this depen-dence is as follows :
v(0) = v„(l - a sin2 d + {0 sin 2 0) (2)
where £ = 1 for the one sublattice and ^ = — 1 for
the other one. In these conditions, the spectrum corresponding to the domain walls consists also of the two resonance lines at frequencies vmax and vmin.
These lines are associated with the two groups of nuclei, magnetization of which makes with the « c » axis an angle 0max and 0min respectively.
Nuclei, responsible for vmin line are situated closer to
the center of a domain wall than the nuclei giving vmax line. With increasing temperature the cc/j3 ratio
increases, 0max is going closer to 0° or 180° and 9min to
90° (Ref. [5] and Table I). TABLE I
T(K) n(fmin) X lO"6 77(/max) X 10 « —^ r r
?(/m»s) Sin B (/max) 1.8 4.1 — — — 4.2 0.9 0.6 1.5 1.3 60 1.7 1.1 1.5 — 77 2.9 2 1.4 2.4 [5] 300 6.3 1.7 3.7 4.5 [5]
2. Apparatus. — The oscillator used in the experi-ments was a modified nuvistor version of the cricuit designed by Pound and Knight [10]. Gain control provides the steady value of the driving field H^. The oscillator operates at low rf levels and is sensitive only to power absorbed. When the superregenerative mode of operation is required, the quenching pulses must be supplied to the oscillator.
The double modulation of the frequency was applied
C1-66 S. NADOLSKI AND W. ZBIERANOWSKI
[Ill. The low-frequency modulation fm (32 Hz or 320 Hz) enables the tracing of the first derivative of NMR signal. The high-frequency modulation fa (1
+
300 kHz) enables the measurement of the enhacement factor q by a rotary saturation effect [I I]. The coil with a .sample is placed inside a cryostat. In the case of a spin echo spectrometer, a cryostat with a sample is placed inside a coaxial cavity tuned at the NMR frequency [12]. In order to obtain the precise tuning and smaller cavity dimensions, a variable capa- city was applied. The generator of radiopulses as well as the low-noise preamplifier are coupled with the cavity by means of a loop. In order to measure preci- sely the spin echo amplitude, the pulse gated integrator was applied.3. Experiment. - YFeO, samples, used in the experiments were single crystal plates 0.3 mm thick having a mass of 0.1 g. The abundance of 57Fe was natural. Some chromium impurities were found in these at examining them in the 707 A FIG. 1.
- Rotary saturation in YFe03. Driving field
type arrangement. cc c >> axis in these plates was H I = 0.6 mOe.
perpendicular to the sample surface. In the steady state technique, the NMR signal at 77 K was so
large, that one single plate was enough to carry out performed, but at the temperatures 4.2, 60 and 450 K. the experiment. For spin echo investigations 20 such The measured frequencies which correspond to these plates were used. In order to observe NMR in the temperatures were as follows :
domain walls, YFeO, crystals were located in that way, that cc c N axis was parallel to the rf magnetic
field H I .
In these conditions, H I field causes the oscillations of the domain walls [4]. The investigations were carried out by a steady-state technique in the temperature range of 1.8 - 450 K and by spin echo method at 77 K.
In the cw experiments the width of the NMR signals was proportional to the driving field H I . This fact is the evidence for the high saturation of NMR signal. Dispersive part of NMR signal is only observed. The measurements of enhancement factor q for both lines were carried out by the rotary saturation method [ l l , 131. RS effect is based on the additional resonance, that appears in the rotating frame. This resonance occurs at H1/2. y field under the influence of frequency modulation of rf signal [ll]. RS effect causes the decrease of nuclear magnetization vector precessing around the effective field. As a result the observed dispersive signal decreases. The smallest signal is obtained when the oscillator is modulated a t fre- quency : 4.2 K : v,,, = 75.75 MHz v,,, = 76.27 MHz 60 K : v,,, = 75.57 MHz v,,, = 76.13 MHz 450 K : vmin = 55.23 M H z . 77K f,, = 75.40 MHz
fmax
= 75.9 7 MHz f, = ( ~ 1 2 7t).(H,/2.r) (Fig. 1).
Below 4.2 K the rapid increase of y factor appears (Table I), simultanously the second line disappears. The lines obtained by a superregenerative technique were much narrower allowing the precise frequency measurements. The NMR in the YFeO, was investi- 400 K [4, 51. In this paper similar investigations were
gated until now at the temperature range of 77- F,. 2. - The variations of NMR line shape. I) stable state ;
TIME EFFECTS IN 57Fe NMR IN YFeOs C1-67
The observed signals undergo a violent change after the application of a small magnetic DC field along the (( c )) axis (Fig. 2). The new state is not stable,
and after some period of time the situation is the same as before the field application. Now, if the field is switched off after the signal amplitudes are established, once more the unstable state is achieved. By a spin echo method in 77 K a weak, broad line was observed. The measurements of spin echo amplitude were carried out as a function of the driving field H I for both states (Fig. 3a). Taking into account the diffe- rences in the relaxation times T, it appears, that the spin echo amplitude does not change practically (Fig. 3b). These effects do not depend on the magni- tude and sign of the applied field. Apart from NMR investigations, the influence of external field on the imaginary part of susceptibility
X"
wzs examined. The measurements were performed by observing the changes of rf signal amplitude on the output of the oscillator. In this experiment the rf signal amplitude was not stabilized. In the unstable state the suscepti-FIG. 3. - a) Spin echo amplitude vs driving field H I x-stable state, '0-unstable state, 1) 2 2 = 32.5 ps ; 2) 2 2 = 65 ps ; r-separation between the rf pulses. b) Spin echo amplitude vs the
separation between rf pulses.
bility
X"
increases. All the effects described above, disappear with the decrease of temperature. At 60 K these effects were not observable. The recovery time decreases with the increase of temperature. At 77 K the NMR signal amplitude as well as theX"
is stabilized after the time of about 0.5 h.4. Discussion. - Regarding the fact, that time
effects do not depend on the magnitude of the applied DC field, one can assume, that they are caused only by the changes of domain wall positions. Immediately after the change of domain wall position the change of its dynamics properties occurs. Since the intensity of the first line increases more than 4 times and its width about two times, one can associate this with the fact, that y factor is twice as large as before. For the second line, y increases only 1.4 times (Fig. 2).
On the other hand in spin echo experiments in the unstable state the optimum value of driving field H , is
3 dB lower (1.4 times). The spin echo signal can be only connected with the second line. A similar situa- tion occurs for CrBr, [8]. A. Hiari et al. [9] obtained, that the oscillating nuclear magnetization is enhanced by the factor A.6/ V less than the driving field. ( A
being the wall surface, 6 wall thickness, V sample volume). No increase of the spin echo signal indicates, that in the unstable state the volume of domain walls responsible for NMR absorption decreases.
The reasons for such a behaviour of domain walls have not been clarified so far. It seems, that this beha- viour described above of domain walls can be caused by phenomenon similar to the static conversion [14].
The enhancement factors (Table I) change slowly in the range of 4.2-300 K. This is connected with the change of anisotropy field. The reasons for the rapid increase of NMR signal below 4.2 K are not clear. The similar effects occur in orthoferrites in the neigh- bourhood of phase transitions [15]. It is probable, that chromium impurities contained in our sample can cause an effect of phase transition at very low tempe- rature.
References
[I] WHITE, R. L., J . Appl. Phys. 40 (1969) 1061.
[2] KOVTUN, N. M., SOLOVIEV, E. E., SHEMYAKOV, A. A., Zh. Eksp. Teor. Fiz. Pisma 14 105 (1971).
[3] ZALESSKII, A. V., Zh. Eksp. Teou. Fiz. Pisma 12 (1970) 468. [4] ZALESSKII, A. N., LOBACZEV, A. N., DEMYANTES, L. N.,
IVASHCHENKO, A. N., MELNIKOV, 0 . K., LEBEDEVA, T. G., Zh. Eksp. Teor. Fiz. 61 (1971) 2337.
[5] ZALESSKII, A. V., ZVEZDIN, A. K., ZHELUDEV, I. S., SAW- NOV, A. M., LEBEDEV, A. F., Phys. Stat. Sol. (b) 73 (1976) 317.
[6] ZVEZDIN, A. K., Zh. Eksp. Teor. Fiz. 68 (1975) 1434. [7] Tu~ov, Y. A., TANKEYEV, A. P., KURKJN, M. I., Fiz. Met.
Metalloved. 28 (1969) 385,29 (1970) 747.
[8] COBB, C. H., SACCARINO, V., BUTLER, M. A., REMEIKA, F. P., YASUOKA, H., Phys. Rev. B 7 (1973) 307.
[9] HIARI, A., EATON, J. A., SEARLE, C. W., Phys. Rev. B 3
(1971) 68.
[lo] POUND, R. V., KNIGHT, W. D., Rev. Sci. Instrum. 21 (1950)
219.
[I 11 MENDIS, E. F., ANDERSON, L. W., Phys. Rev. B 2 (1970) 569.
[12] LE DANG KHOI, VEILLET, P., Rev. Sci. Instrum. 45 (1974) 759.
[13] REDFIELD, A. G., Phys. Rev. 98 (1955) 1787.
[14] HAISMA, J., VAN MIERLOO, K. L. L., DRUYVESTEYN, W. F.,
ENZ, U., Appl. Phys. Lett. 27 (1975) 459.
[15] DOROSHEV, V. D., KHARNACHEV, A. S., KOVTUN, N. M., SOLOVIEV, E. E., CHERVONENKIS, A. Ya., SHEMYA-
