Academia.eduAcademia.edu

Transient Mössbauer spectroscopy with57Fe

1986, Hyperfine Interactions

Transient MSssbauer experiments with 5rFe are described. A time resolution of 0.7 ns is achieved with a BaF2 scintillator. Polar polyvinylidene fluoride film is shown to be well suited for producing fast mechanical modulation. Pulses of recoilless gamma radiation, of width ~30 ns and heigth 3.5 times the MSssbauer absorption, are generated with stepwise source motion. The efficiency of transient methods in extracting MSssbaner data is demonstrated with sinusoidal motion.

Hyperfine Interactions 29 (1986) 1563-1566 1563 TRANSIENT MOSSBAUER SPECTROSCOPY WITH S7Fe P. HELISTO, E. IKONEN and T. KATILA Department of Technical Physics, Helsinki University of Technology, SF-02150 Espoo, Finland H. S1PILA, Outokumpu Electronics, P.O. Box 27, SF-02201 Espoo, Finland Transient MSssbauer experiments with 5rFe are described. A time resolution of 0.7 ns is achieved with a BaF2 scintillator. Polar polyvinylidene fluoride film is shown to be well suited for producing fast mechanical modulation. Pulses of recoilless gamma radiation, of width ~30 ns and heigth 3.5 times the MSssbauer absorption, are generated with stepwise source motion. The efficiency of transient methods in extracting MSssbaner data is demonstrated with sinusoidal motion. 1. I N T R O D U C T I O N Transient methods are extensively used in NMR and laser spectroscopy. In the if-energy region coherent transient effects are a relatively new d i s c o v e r y / 1 / . In simple terms, transient effects are caused by the delay of the absorber response when reacting to fast changes in the source field. One way of producing such changes is to sweep through the resonance line within the lifetime of the MSssbauer state. The duration of transient phenomena is determined by the width of the resonance. Although saturation of the samples in the sense of NMR and laser spectroscopy is not possible, prominent nonlinear effects due to thickness saturation axe characteristic of MSssbauer transients. Several applications associated with the transients have been developed./2-4/So fax most transient measurements have been performed with the MSssbauer resonance of 67Zn./1-4/ Results of 57Fe measurements have also been reported./2,4-6/ Same type of phenomena have been obtained with rapid magnetization c h a n g e s / 7 / and with synchrotron radiation p u l s e s / 8 , 9 / . We discuss here recent improvements in transient MSssbauer spectroscopy with the isotope 57Fe. Iron has several advantages over zinc. The measurements can be performed at room temperature. Large thickness saturation is easily achieved to enhance the transients. The high coefficient of internal conversion facilitates scattering m e a s u r e m e n t s / 5 , 6 / . On the other hand, the half life of the MSssbauer state, 98 ns, is only one hundredth of that of BTZn. This sets special requirements both on the production of the fast mechanical motion and on the detection of radiation in transient studies. 2. EXPERIMENTAL SETUP The experimental setup was the following. A 5rCo:Rh source, of activity 0.1-0.2 mCi and ~1 mm 2 in area, is attached to a piezoelectric transducer driven by a function generator. The absorber is a 6-t~m-thick 57Feo.1Rh0.g foil. G a m m a radiation transmitted through the absorber is detected in phase with the motion of the source. Because of the high modulation frequencies needed in transient experiments, ordinary multiscaling techniques cannot be utilized. Instead, the counts are registered with a time-to-amplitude converter and the energy discrimination is accomplished by gating the ADC of a multichannel analyzer. 9 J.C. Baltzer A.G., Scientific Publishing Company 1564 P. HelistO, et al., Transient MOssbauer spectroscopy with 5ZFe Scintillation detectors are the most common choice when accurate timing is required. We have obtained a time resolution of 5 ns at 14 keV with a 0.1-mm-thick NaI(T1) scintillator. This is satisfactory for most transient measurements and in agreement with results of, e.g., Ref. / 5 / . Experiments were run as well on a BaF2 scintillator coupled to a Hamamatsu R2076 photomultiplier. A time resolution of 0.7 ns was reached. The resonant absorption in a conventional M6ssbauer measurement was 60 % of that attained with the NaI detector. This decrease is partly due to the larger thickness (0.3 mm) of the BaF2 detector. Several piezoelectric transducer constructions were tested to achieve controlled modulation. One of them was a piece of 28-/~m-thick polar polyvinylidene fluoride (PVDF) film coupled to a plexiglas support. Results obtained with an 80-#m-thick X-cut quartz plate are also presented. 3. RESULTS OF T R A N S I E N T EXPERIMENTS In Fig. 1, transient MSssbauer curves with sinusoidal source motion are shown, when the drive was the PVDF film. Solid curves are least squares fits by a theoretical expression / 2 / to the data. The following (preliminary) results were obtained from the fit to curve a): Fa = 0.162(11) m m / s , r a = 0.0976(15) m m / s , T~ = 6.94(17) and a = 28(3) ns, and from the fit to curve b): s = 0.199(11) m m / s , Fa = 0.0993(9) m m / s , Ta = 7.78(22) and a = 14(2) ns. Here F8 is the Lorentzian width of the source line including broadenings due to nonideal motion, s is the width of the absorber line, Ta is the MSssbauer thickness of the absorber and a is an additional Gaussian time distribution of the modulation; statistical uncertainties of the fits are given in the parentheses. The conversion factor of the drive, derived from the fits, is ~30 p m / V . 100 I ' i I ~ ~ l 8O l o o~ g ~ 80 100 80 500 I 1000 TrME (ns) I 1500 J 2000 Fig. 1. Transient M6ssbauer curves with sinusoidal voltage at 510 kHz applied to a PVDF drive. Voltage amplitudes were 7.5, 15 and 30 V. A negative slope in the data is caused by the dead time of the time-to-amplitude converter. Total time resolution of the measurement was 6.6 ns. P. HelistO, et al., Transient MOssbauer spectroscopy with 57Fe 1565 The transmission curves of Fig. 2 were produced by applying square-wave voltage to the quartz transducer. The source attached to the quartz plate thus moved in a stepwise fashion. The rise- and falltimes of the voltage steps were 22 ns. Curve 2a) was measured with the Nal detector whereas for curve 2b) BaF2 was used. The peak-to-peak effect of curve 2a) is remarkable, ~125 % relative to the off-resonance background. This is 3.5 times the m a x i m u m resonance absorption obtained in a conventional transmission measurement. Also the risetime, 17 ns, and halfwidth, 32 ns, of the curve are notable. Excitation of mechanical resonances of the system causes a slowly varying component to the data. The risetime of curve 2b) is comparable to that of 2a), but the peak-to-peak effect is considerably smaller and the slow component appears to be enhanced. i I I 200 a ) Source Araldd Quartz ,". ~ 9 150 9 % z o " NaI loo ~E AI lmm 32 ns F W H M 9 % ,..,," Z <[ Ix I-- 9 " BQF2 QI 10C .j . _ _ 80 9 I 0 , I , I 100 2o0 TIME (ns) ; I LJ 30o F i g . 2. Transients due to a rapid voltage change of 32.5 V applied to an X-cut quartz plate. Frequency of the square-wave voltage was ~200 kHz. Dashed lines indicate the off-resonance background level. Curve a) is measured with a NaI and curve b) with a BaF2 scintillator. Only the transients corresponding to movement of the source towards the absorber are shown. 4. DISCUSSION Experimental results of Fig. 1 for sinusoidal modulation show the evolution of the resonance lineshape to a complicated transient pattern as the resonance curve is scanned more and more rapidly. The results are in good accordance with theoretical predictions for a thick absorber. The conversion factor of the PVDF drive is more than ten times larger than that of quartz. A simultaneous determination of four parameters affecting the experimental linewidth turned out to be possible from curves la) and lb) where the maximum velocities of the source motion, 0.8 m m / s and 1.6 m m / s , are somewhat larger than the total linewidth. Such amplitudes should be most favourable for the separation of the linewidth parameters./4/ Part of the excess broadening, manifested in the values of Fe and a, can be attributed to temperature drifts in the measuring system, but the motion of the source still seems to include an incoherent component. Despite these nonidealities, the width of the absorber line Fa is very close to 1566 P. HelistO, et al., Transient MOssbauer spectroscopy with 5ZFe the natural linewidth 0.0970 rnm/s of 5ZFe, indicating a negligible distortion of the absorber matrix. The values given by the fits for the MSssbauer thickness of the absorber also show a reasonable reproducibility considering the factor of 2 difference in the modulation amplitudes and the remaining distortions of the modulation. In comparison, an attempt to fit a transmission integral formula to a conventional linewidth measurement, varying ra, ra and Ta, gave anomalously large statistical uncertainties because of the high correlation of the linewidth parameters. Besides a better ability to resolve details of the lineshape, transient methods offer another advantage: the parameter values, although converted here to velocity units, are originally given in terms of the modulation frequency, which can be measured to a much higher precision than the Doppler velocity. Applicability of a BaF2 scintillator to time resolved M6ssbauer spectroscopy is demonstrated in Fig. 2b). The difference between the shapes of curves 2a) and b) mainly stems from ~.he small amount of the 'fast' 220 nm photons emitted by the BaF~ crystal. The average total yield of photoelectrons per a 14-keV quantum was estimated to be 7, which amounts to 25 % of the number given in Ref. / 1 0 / f o r BaF2 coupled to an XP-2020Q photomultiplier. As only 10-20 % of the photoelectrons are caused by the 220 nm photons and the rest originate from the slowly decaying 310 nm photon states, the timing signal may sometimes be erroneously derived from the latter component. The pulses of recoilless "7 radiation of Fig. 2 exhibit a well-defined, narrow time pattern obtained with an inexpensive and simple equipment. As short pulses of resonant "7 radiation have been accomplished only with synchrotron radiation /8/. The resonant counting rate of the pulses is about 5 s -1. An increase by two orders of magnitude to this is readily achieved with stronger sources. The pulse width can be varied by adjusting the risetime and amplitude of the voltage applied to the drive. In conclusion, we have further developed techniques for transient and high frequency experiments with 5ZFe, featuring subnanosecond time resolution with a BaF2 scintillator and high-frequency motion of good quality with polar PVDF. Intense MSssbauer gamma pulses of ~30 ns halfwidth are generated with a quartz transducer and the power of transient methods in extracting MSssbauer linewidth data is demonstrated for sinusoidal modulation. The authors thank Mr. J. Hietaniemi for help in numerical analysis and the Academy of Finland and Emil Aaltonen Foundation for financial support. REFERENCES /1/ /2/ /3/ /4/ /5/ /6/ /7/ /8/ P. HelistS, T. Katila, W. Potzel and K. Riski, Phys. Lett. 85A (1981) 177. P. HelistS, E. Ikonen, T. Katila and K. Riski, Phys. Rev. Lett. 49 (1982) 1209. P. HelistS, E. Ikonen, T. Katila, W. Potzel and K. Riski, Phys. Rev. B 30 (1984) 2345. E. l:konen, P. HelistS, T. Katila and K. Riski, Phys. Rev. A (in press). E.H. Realo, K.K. Rebane, M.A. Haas and J.J. JSgi, JETP Lett. 40 (1984) 1309. P. HelistS, E. Ikonen, T. Katila, J. JSgi, E. Realo and S. Reiman, Helsinki University of Technology Report TKK-F-A571 (1985). G.V. Smirnov, Yu.V. Shvyd'ko, O.S. Kolotov and V.A. Pogozhev, Soy. Phys. JETP 59 (1984) 875. A.I. Chechin, N.V. Andronova, M.V. Zelepukhin, A.N. Artem'ev and E.P. Stepanov, JETP Lett. 37 (1983) 633. /9/ E. Gerdau, R. Rfffer, H. Winkler, W. Tolksdorf, C.P. Klages and J.P. Hannon, Phys. Rev. Lett. 54 (1985) 835. /10/ M. Laval, M. Moszyhski, R. Allemand, E. Cormoreche, P. Guinet, R. Odru and J. Vacher, Nucl. Instr. and Meth. 206 (1983) 169.