Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.issp.ac.ru/lhpp/PapersAntonov/81.pdf Дата изменения: Tue Feb 23 20:37:17 2016 Дата индексирования: Sun Apr 10 03:00:10 2016 Кодировка: Поисковые слова: http astrokuban.info astrokuban

Journal

of the Less-Common

Metals. 17% 174 (1991) 343- 350

343

57Fe Mijssbauer study of the hydrides of Ni - Cu alloys
B. Zhang and H. J. Bauer
S&ion Physik, Ludulig-Maximilians-~n~~lersitiit. W-8000 Miinchen (F.R.G.)

M. Baier and F. E. Wagner
Physics Department. Technical University of Munich. W-#046 Garching (F.R.G.)

V. E. Antonov

and T. E. Antonova
Physics, Acadenzy 142432 (U.S.S.R.) of Sciences of the U.S.S.R.

Institute of Solid State Chernogolovka District

Abstract Hydrides of Ni,,,Cu,,.,,, alloys were prepared electrolytically or under high pressures of molecular hydrogen and studied by MGssbauer spectroscopy using dilute substitutional "7Fe as a probe. At a hydrogen pressure of 7 GPa, a hydrogen-to-metal ratio near x = 0.9 could be reached, while at 1.1 GPa or by cathodic charging hydrogen contents near x = 0.6 were obtained, which are already sufficient to suppress ferromagnetism completely. The hydrogen was found to be repelled by the iron probes, but less effectively than in pure nickel. The temperature dependence of the isomer shift
showed that the samples netic phase had hyperfine hydrogenation. began to lose hydrogen near 200 K. The reappearing magparameters practically the same as those of the alloy before

1. Introduction Nickel and copper form a continuous series of solid solutions. Both the saturation magnetization and the Curie temperature decrease with increasing copper content of these alloys, which no longer become ferromagnetic even at low temperatures when the copper concentration exceeds about 60 at.% [I]. Ni--Cu alloys with copper contents up to about 50 at.% can be hydrogenated electrolytically or under high pressures of H, gas even more easily than pure nickel [2- lo], although high hydrogen concentrations are easier to obtain using the latter method. The hydrides retain the f.c.c. structure of the Ni -Cu alloys. At copper concentrations exceeding about 40 at.%, the separation into the x phase containing only few per cent of hydrogen and the hydrogen-rich /j phase disappears [4,9]. (Alternatively, these two phases have been designated as ;`, and yZ [9].) As in pure nickel [ 111, the ferromagnetism of the NiG Cu alloys is suppressed by hydrogenation [ 2.4, 5,7 lo]. Recent Mossbauer studies using "`Fe as a dilute probe in NiH,X have shown [12,13] that the hydrogen is strongly repelled by the iron, such that hydrogen occupies interstitial sites next to iron atoms only at hydrogen-to-

Elsevier Sequoia/Printed

in The Netherlands


344

metal ratios very close to unity. In this context, Massbauer studies on hydrides such as those of Ni-Cu alloys are of interest since the distribution of hydrogen around the iron probes is expected to be changed by the partial substitution of nickel by another element. Moreover, Mossbauer spectroscopy can yield microscopic information on the magnetic properties of such hydrides as it has done for hydrogen-free Ni-Cu alloys [14,15]. As first results of an investigation of the (Ni, _y Cu,,)H, system by 57Fe Miissbauer spectroscopy, X-ray diffraction, and magnetic measurements, this paper reports on M&sbauer experiments with hydrides of Ni0,70C%,30.

2, Experimental details The M~ssbauer absorbers were made from an arc-melted Ni~,~~Cu~.~~ alloy containing 0.15 at.% of isotopically enriched 57Fe. This was rolled to foils about 15 pm thick which were then annealed in a hydrogen atmosphere at 850 "C for several hours and then cooled to room temperature at a cooling rate of about 10 K min-`. The foils were loaded with hydrogen electrolytically at ambient temperature in 0.5 n H,SO, with 0.2 g 1-l of thiourea as a promoting agent 1163 and a current density of 20 mA cme2, or under high pressures of molecular hydrogen. Hydrogen pressures of up to 1.1 GPa were reached in a piston-type pressure cell [8,17-191, while hydrogenation at 7 GPa was performed in a high pressure cell using MnH, as a hydrogen donor [20]. After hydrogenation, the pressure cell was cooled with liquid N, before the pressure was released and the sample removed from the cell. The samples were stored in liquid N, and later transferred into the Mossbauer cryostat without warming. The hydrogen content was determined by outgassing into a calibrated volume, either after the Mossbauer experiments, or for parts of the samples removed before the transfer into the Mossbauer cryostat. During cathodic charging, the formation of the hydrides was monitored by magnetization measurements. In some cases lattice parameters were measured by X-ray diffraction using Fe Kcr radiation.

3. Results and discussion

Figure 1 shows the MGssbauer spectra of Ni*,~~Cu~.~~ of two hyand drides of this alloy. Table 1 summarizes the loading conditions and experimental results. X-ray diffraction at liquid nitrogen temperature showed the sample loaded at 7 GPa to be pure /I phase hydride with a unit cell volume 18% larger than that of the unloaded alloy. The sample loaded at 1.1 GPa was not X rayed, but with a hydrogen content of x = 0.59 it is also expected to be pure p phase [3,4> 9]* This expectation is confirmed by the Mossbauer spectra taken at 4.2 K which, for both hydride samples, do not contain any magnetic fraction. Since


345

I"-

0.69
GPa

\

99

1.1

p" 2

t

Fig. 1. 57Fe Mossbauer spectra of NiO.MCuo.:re,(Nio.iOCu,:,,)H,,, and (NiO,,,Cu,,,)H,,,. samples OXat.% of the metal atoms were substituted by enriched ?Fe. The spectra hydrides were measured with both the absorber and the source of "7Co in rhodium at 4.2 spectrum of the unloaded sample was measured with the absorber at 323 K and the source at TABLE 1 Conditions hydrides
X

In all of the K. The 300 K.

of hydrogenation,

lattice

parameters,

and Mossbauer

results

for (Ni,,,

Cu,, :lo)H,

%

S
(mm s-`) -0.098 & 0.005 f 0.080 +_0.005 + 0.408 + 0.005 0.41 + 0.01 0.50 +_0.01 0.41 & 0.01

(A) 0.0 0.59 0.89 3.542 3.741

Conditions of hydrogenation

1.1 GPa, 25 ,`C, 117 h 7 GPa, 250, C, 24 h

The lattice constants a, were I is about 0.02. S is the mean and W the full width at half unloaded alloy was determined the linewidth was taken from the alloy.

measured at 100 K. The uncertainty of the hydrogen-to-metal ratio isomer shift with respect to the source of s7Co in rhodium at 4.2 K, maximum of the Mossbauer patterns. The isomer shift for the from the magnetic hyperfine spectrum measured at 4.2 K, while a spectrum measured at 323 K, i.e. above the Curie temperature of

the hydrogen-free alloy is ferromagnetic at 4.2 K, the spectrum of Ni,,,,CuO,,, shown in Fig. 1 for comparison with the hydride spectra was taken above the Curie temperature. The second-order Doppler shift (see, for example, ref. 21) is nevertheless expected to have practically no influence on the peak positions of the spectra shown in Fig. 1 since the temperature difference between


346

source and absorber was small in all cases. The centre shifts are therefore representative of the effect of hydrogenation. As in NiH, [12,13] and PdH, [22], loading with hydrogen causes an increase in the isomer shift, which corresponds to a decrease in the electron density at the iron nuclei. The isomer shift with respect to the unhydrogenated alloy is -i-O.18 mm s-l for the hydride with x = 0.59 and +0,51 mm s-l for that with x = 0.89. Previous work, particularly for 57Fe in NiH, [12,13] and PdH, [22], has shown that in these systems the increase in the isomer shift caused by hydrogenation depends mainly on the number i of hydrogen atoms next to the 57Fe probes, and that the shift is, at least approximately, proportional to this number. For the shift per hydrogen neighbour, values of AS = +0.088 mm s-' and i-O.091 mm s-' have been derived for 57Fe in NiH, [ 131 and PdH, [22] respectively. If the present results for (NiO,,,CuO,,,)H, are interpreted in the same manner, the change in isomer shift of +0.51 mm s-l caused by hydrogenation to x = 0.89 corresponds to iron with five or six hydrogen neighbours, i.e. to a situation where all or nearly all octahedral interstitial sites next to the iron probes are occupied. For the sample with x = 0.59, the shift is only +O.l8mm s-l, corresponding to iron with an average of only two neighbours, much less than the average 3.5 neighbours expected for a random occupation of interstitial sites in the f.c.c. lattice at x = 0.59. This shows that the hydrogen has a tendency to avoid sites next to iron, but in (Ni,,,Cu,,,)H, the influence of the repulsive iron-hydrogen interaction is smaller than in pure NiH, [12,13], where iron with close to six hydrogen neighbours has not yet been observed even at x values very close to unity, and hydrogen has been found noticeably to populate the sites next to iron probes only when x exceeds about 0.95. This interpretation of the isomer shifts is supported by the shape and widths of the Mossbauer lines. The shape of the absorption line of pure Ni,,,, is already broadened, non-lorentzian, and Cu,XJ above the Curie temperature slightly asymmetric (Fig. l), presumably because of the distribution of nickel and copper atoms on the 12 metal sites nearest to the iron probes in the f.c.c. lattice. The Mossbauer pattern of (Ni,,.,,Cu0.30)H0.SB differs in shape from that of the unloaded alloy and is substantially broader, indicating a distribution of hydrogen configurations in the vicinity of the Mossbauer atoms. The spectrum of (Ni,,,,,Cu,,,)H,,, is again narrower than that of (Ni0,,,Ct+,30)Ho_59, but still broader than that of the unloaded alloy. Together with the large isomer shift, this indicates that in this case mainly iron with five and six hydrogen neighbours is present. 3.2. Temperature dependence of the spectra of N~,,,,,CU~,~~~ and its hydride The temperature dependences of the Mossbauer spectra were measured alloy and for a sample of ( Ni,,,,CuO,,,)H,,, for the unloaded Ni,,,,Cu,,, prepared electrochemically. The spectra of the unloaded sample (Fig. 2) exhibit a magnetic hyperfine splitting with lines that are moderately broadened by a distribution of hype&me fields. The mean hyperfine fields and centre shifts are shown in Fig. 3 as a function of temperature. The magnetic


Fig. 2. "`Fe Mkbauer spectra uf the hydrogen&-i-ee N~,,,CU,,~, temperatures with the source of "To in rhodium at 300 K.

al"fr>y measured

at different

hyperfine splitting collapses near 300 K, in good agreement with magnetic measurements [l, 51_The temperature dependence of the centre shift could be fitted by a Debye model curve for the second-order Doppler shift [Zl] with a Debye temperature of 8o = 428 K. Relative to dilute 57Fe in pure nickel, the is shifted by 0.02 mm s-l towards higher velocities. resonance in Ni,,,Cu,,, This result is in good agreement with the shifts observed by Window eE al, f34] for Ni-Cu alloys. The mean hypesfine fields at low temperatures and the width of the distribution of hype&no fields are also in good agreement with the results obtained by these authors for Ni,,,, Cu,,,, containing 1 at.% Fe. The mean centre shift of (Ni0,7(1CUg30)H0.G6 more positive by about is 0.14 mm s-l than that of the unloaded alloy. Up to 200 K its temperature dependence can be explained by the second-order Doppler shift (Fig. 4). Abovr; 200 K, the centre shift begins to decrease more rapidly, showing that the sample begins to lose hydrogen within the timespan of one or a few days required to measure a Moissbauer spectrum. Conszquently, this decrease in the isomer shift is irreversible and the shift remains reduced in spectra measured subsequently at 100 K (Fig. 4). The magnetic 2 phase first reappeared, with an intensity of 18% of the area, in a spectrum measured at 100 K after the sample had been at 24-OK for a day. The isomer shift of the reappearing magnetic pattern is practically the same as before hydrogenation,


348

-O----0ZO-

-0-

-0. .\O '0 O'b
\ \

IO-

0,

;Lo-o
1 I

-0,15~

Temperature

tK)

Fig. 3. Temperature dependence shifts S with respect to the source curve fitted to the centre shifts represent the hyperfine field and the hydride at 240 K for a day.

of the mean magnetic hyperfine fields &r and the mean centre of 57Co in rhodium at 300 K for hydrogen-free Ni,,,,Cu,,,. The corresponds to the Debye model with 0o = 428 K. Triangles the centre shift of the c( phase after its reappearance on keeping

0.1543 .OOO . . . .. `. . .._ .. ,,.. 0 .. ... ,.,. `.' 3 E cn ooo0.05~·..-.... l. .".. . . . .._... -..... "..O 0

00

O.`O

l......

.q

`0 0

150 Temperotute

(K

1

200

2

with an Fig. 4. Centre shifts of (Ni0,7aCu,,,)H, with a source of 57Co in rhodium at 300 K (0); 0 had lost some hydrogen during the preceding points are connected by dashed lines. The fully unloaded alloy (see Fig. 3).

initial hydrogen content of n = 0.66, measured centre shift measured at 100 K after the sample measurement above 200 K; the respective data drawn curve represents the centre shift for the

and the same is true for the magnetic hyperfine field (Fig. 3). The small amount of hydrogen that is expected to be dissolved in the c1phase [4,9] thus has no noticeable influence on the Mossbauer pattern. The remaining isomer shift for the iron in the non-magnetic p phase (+0.02 mm s-' with respect to the unloaded alloy at 100 K) is also very small. Outgassing yielded a mean hydrogen content of 3 = 0.21 for this sample, which means that x should be


349

somewhat larger in the non-magnetic fl phase, the exact value depending on the hydrogen content of the tl phase fraction. The small shift observed for the p phase at this concentration again reflects the repulsive interaction between the iron and the hydrogen, which at this low hydrogen concentration virtually prevents hydrogen from occupying sites next to the iron.

4. Conclusions The results of this work show that iron probes in hydrides of Ni-Cu alloys have a repulsive interaction with the hydrogen, but the iron becomes surrounded with hydrogen more easily than in the hydrides of pure nickel. In interpreting these findings, one should bear in mind that there is a tendency for short-range order in Ni-Cu alloys [14], which results in the formation of nickel-rich microclusters. The iron is expected to behave much like nickel and will therefore preferentially be surrounded by nickel atoms. At low and moderate hydrogen concentrations, the hydrogen is expected to prefer interstitial sites that have predominantly nickel neighbours and avoid sites with many copper neighbours as well as sites next to iron. At higher hydrogen contents, the interstitials appear to prefer sites with an iron neighbour to sites with several copper neighbours, with the result that in hydrides of Ni-Cu alloys the iron probes have more hydrogen neighbours than in pure nickel hydride at equal overall hydrogen-to-metal ratios. Details of this behaviour should also depend on the degree of short-range order in the Ni-Cu alloys, and hence on the cooling rate of these. Mossbauer spectroscopy can thus be used as a sensitive tool to study details of the microstructure of the hydrides of NiiCu alloys. A systematic investigation, in which both the copper concentration and the hydrogen content are varied, is expected to yield a better understanding of the hydrogen distribution in these systems.

Acknowledgments We are grateful to I. Dugandiid and J. Fried1 for help with the experiments and stimulating discussions, and to Professor R. Sizmann for support and for his interest in this work. Funding of this work by the Deutsche Forschungsgemeinschaft is also gratefully acknowledged.

References
1 2 3 4 5 R. H. B. H. H. M. Bozorth, J. Bauer, 2. Baranowski J. Bauer, G. J. Bauer, 2. Ferromagnetism, Van Nostrand, Princeton, NJ, 1951, p. 308 ff. Nuturforsch. A, 22 (1967) 1468. and S. Majchrzak, Roczniki Chemii, 42 (1968) 1137. Berninger and G. Zimmermann, 2. Naturforsch. A, 23 (1968) 2023. Angew. Phys., 26 (1969) 87.


350 6 B. B. 7 H. 8 B. 9 10 11 12 13 14 15 16 17 Baranowski and M. Tkacz, EuFo~~~s. Corzf. Abstr. A, 1 (1975) 108. Baranowski and M. Tkacz, Pal. J. C&em., 54 (1980) 819. J. Bauer, J. &fugn. Mugn. Meter., 15-18 (1980) 1267. Baranowski, in G. Alefeld and J. Volkl (eds.), hydrogen in Metals II, Topics in Applied Physics, Vol. 29, Springer, Berlin, 1978. p. 157. V. E. Antonov, I. T. Betash, V. F. Degtyareva, E. G. Ponyatovsky and V. G. Thiessen. Fizz. ~etall. ~e&a~laued. 53 (1982) 677. W. E. Martin, H. J. Bauer, S. M. Filipek and B. Baranowski, J. Mugn. Magn. Mater., 54-57 (1986) 1505. H. J. Bauer and E. Schmidbauer, Nat~r~~sse~sckufte~, II (1961) 425. H. J. Bauer and E. Schmidbauer, 2. Phys., I64 (1961) 367. M. A. Amer. F. E. Wagner and H. J. Bauer, Hyperfine Interact., 41 (1988) 539. M. Amer, M. Baier, H. J. Barter and F. E. Wagner. 2. Pkys. &hem., N.F., 164 (1989) 773. B. Window, C. E. Johnson and G. Longworth, J. Phys. C, 2 (1970) S218. G. Longworth and B. Window, J. P&s. F, 3 (1973) 832. B. Baranowski and M. ~mialowsk~, J. Phys. Chem. Solids, 12 (1959) 206. B. Baranowski and R. Wi~niewski, Bull. Akud. Polon. Sci., 14 (1966) 273. R. Wi~niewski, Rev. Sci. Instrum., 41 (1970) 464. B. Baranowski and W. Bujnawski, Roczniki Chemii, 44 (1970) 2271. E. G. Ponyatovsky, V. E. Antonov and I. T. Belash, in A. M. Prokhorov and A. S. Prokhorav feds.), Pro~Zems in Solid State Physics, Mir, Moscow, 1984, p. 109. G. K. Shenoy, F. E. Wagner and G. M. Kalvius, in G. K. Shenoy and F. E. Wagner (eds.), ~~s$baueF Ksomer Shifts, North~Holland, Amsterdam, 1978, p. 49. F. Prijbst and F. E. Wagner, J. Phys. F, 17 (1987) 2459.

18 19
20 21 22