Academia.eduAcademia.edu

Weak ferromagnetism in Gd2CuO4 and Gd2−xRxCuO4±δ(R = Ce, Sr)

1994, Physica C: Superconductivity

Gd2CuO4 developes, at TN, a long range AF order with a weak ferromagnetic (WF) component. This component is allowed by in-plane distortions of the tetragonal T' structure and persists above TN, induced by the magnetic field, in a large temperature interval. From de magnetization measurements, we have accurately determined the WF transition, which follows the characteristic law of classical ferromagnets, h = t(' Y+13), with (¥+[~) = 1.8(2). The WF component persists in Gdl.g5Ce0.15CuO4+8, even after a reduction procedure, although the Ce 4+ doping lowers 'IN. A set of Gd2-xSrxCuO4-8 samples, prepared via a sol-gel method at a maximum temperature of 900"C, showed an expanded unit cell without distortions, the absence of a WF component and strongly modified magnetic properties of the Gd sublattice.

Physica C 235-240 (1994) 1569-1570 North-Holland PHYSICA Weak ferromagnetism in Gd2CuO4 and Gd2.xRxCuO4+6(R - Ce, Sr). A. Rouco a, J.L. Garcfa-Mufioz a, F. P6rez a, J.M. Navarro a A. Fuertes a S. Pifiol a, X. Obradors a and M. Tovar b. a Institut de Ci~ncia de Materials de Barcelona-CSIC. Campus UAB, 08193 Bellaterra, Spain. b Centro At6mico Bariloche, C.N.E.A. and I.B., Universidad National de Cuyo, 8400 Bariloche, Argentina. Gd2CuO4 developes, at TN, a long range AF order with a weak ferromagnetic (WF) component. This component is allowed by in-plane distortions of the tetragonal T' structure and persists above TN, induced by the magnetic field, in a large temperature interval. From de magnetization measurements, we have accurately determined the WF transition, which follows the characteristic law of classical ferromagnets, h = t('Y+13),with (¥+[~) = 1.8(2). The WF component persists in Gdl.g5Ce0.15CuO4+8, even after a reduction procedure, although the Ce4+ doping lowers 'IN. A set of Gd2-xSrxCuO4-8 samples, prepared via a sol-gel method at a maximum temperature of 900"C, showed an expanded unit cell without distortions, the absence of a WF component and strongly modified magnetic properties of the Gd sublattice. The symmetry between hole and electron doping of the R2CuO4 compounds [1] has impelled the study* of both kinds of doping in Gd2CuO4, which exhibits intermediate cell parameters between those of the parent compounds of n-type and p-type SC, respectively, The first step in this way is a full understanding of the WF transition of the undoped G d 2 C u O 4 , since it is absent in the parent compounds of SC. We have measured the dc magnetization, Mdc, vs. T, up to 340 K, of Gd2CuO4 single crystals, in fields up to 50 kOe applied parallel to the ab planes. The paramagnetic (PM) contribution of the Gd sublattice has been determined from Mdc data in 50 kOe, measured between 340 and 540 K. Samples of Gdl.85Ce0.15CuO4+8 have been prepared by means of solid state reaction at 1100°C and reduced at 700°C in As atmosphere. We have measured their ac susceptibility. A set of samples of Gd2_xSrxCuO4_ 8, with x = 0, 0.05, 0.1, 0.15 and 0.2, has been prepared via sol-gel at a maximum temperature of 900°C. Two other sets with x = 0.01, 0.02, 0.03 and 0.04 and x = 0, 0.05, 0.1, 0.15 and 0.2 have been obtained by the same procedure, but reaching a temperature of * "~zs work has been supported by the DGICYT (PB920849) and the EEC(CT92-0087, CT92-0785). 950°C. These samples have been characterized by Xray powder diffraction, thermogravimetric analysis (TGA), SEM, EPR, ac susceptibility, de magnetization and resistivity measurements. The WF transition can be determined from the inflection points of the Mdc(T) curves in different fieid~, hi Fig. l(a), wc ,~hc;wsome ef ~hese Mdc(T) curves for G d 2 C u O 4 single crystals, after subtracting the PM contribution from the Gd sublattice. The inflection points are plotted in Fig. l(b), with a fit to the function h = t(7+[3). This function can be deduced from a scaling law equation of state which, in the vicinity of the ferromagnetic critical temperature, has the form m(h,t) = tbF(h/t(?+13)), where m, h=H/Ho and t = (T-TN)/T N are reduced variables. The best fit gives the parameters ~/+~ = 1.8(2), similar to that expected from MF theory and observed in Y2CuO4 [2], and . . . "r~ TN = "~""~ app~uammL~y I~OIIIk..IU~IiI w l•t h .h~. defined at the susceptibility peak [3]. Doping with 7.5% of Ce lowers the N6el temperature down to TN = 165 K, but it does not suppress the WF component [4]. We have observed that a later reduction at 700°C in Ar atmosphere has no effect on the magnetic properties, although this reduction probably removes around 0.04 O atoms/f.u. [5]. This is a clear indication that the 0921-4534/94/$07.00 © 1994 - Elsevier Science B V All r,ghtq reqerved SS/.q 0921-4534(94)01351-9 .4 Rouco et ai /Physica C 235-240 (1994) 1569-1570 1570 electron doping by Ce4+ is not equivalent to that produced by oxygen reduction in this kind of cupmtes. o.12 , ,, (a) _ kOe • o 240 20 15 260 Co 280 300 T (K) -- 320 340 .. 0 _~ I0 ~ 5 0 260 270 280 290 300 T (g) 310 all these measurements indicate an expanded T" structure, without distortions, for type-I samples, instead of the distorted structure of type-II samples, typical of Gd2CuO4. As expected from a non-distorted T structure, the de magnetization and the ac susceptibility measurements of type-I samples do not show the typical WF transition. Moreover, the zd'l(T) of these samples only obeys a C-W law at high fields and the corresponding effective moment is only =4.8 I,tR/Gd atom. In addition, the EPR spectra of Gd ions in these samples are less intense and the resonance g-factors are shifted to lower values than the free ion g=1.99. We interpret these results as indicative of (a) the suppression of the WF component of the Cu sublattice and Co) the presence of strong AF correlations in the Gd sublattice, that persist until the decomposition temperature. 320 Figure 1. (a) Mdc-XGdH vs. T measured, after FC, with different fields applied parallel to the CuO2 planes of Gd2CuO4 single crystals. (b) WF transition of the Cu sublattice of Gd2CuO4 single crystals. The solid line is a fit to h = t('t+l~). On the other hand, the Gd2-xSrxCuO4.181 samples with x>_0.05 synthesized at a maximum temperature of 900°C, which we will call type-I samples, show expanded T' cell parameters (a - 3.94 ,/L c - 12.00-12.05 A). However, the undoped compeund and the samples with 0<x<_0.2 synthesized above 900°C (type-II samples) show the typical parameters of Gd2CuO4 (a --- 3.895 A, c = i 1.88 A). SEM analysis has conf'trmed similar mean grain sizes, although the type-lI samples prepared at higher temperatures have a more homogeneous size distribution than type-I samples. The TGA has shown an oxygen deficiency in all the samples, ranging from 181= 0.03 for x=0 to 181 = 0.17 for x=0.2. When reducing in an Ar-H atmosphere, all the samples undergo two successive decompositions, ~,r,r~ n ; * ;..-, but the d,,,,,,mpos .... n temperatures za-e higher for type-I samples, thus showing that they are mute stable. Finally, the resistivity at room temperature has been found to be between 2 and 5 orders of magnitude larger for type-lI samples. In summary, o 3 I I ° TO,) 0 50 300 T (K) Figure 2. Mdc/H vs. T measured in 50 Oe for the type-II Gdl.96Sr0.04CuO4.181 and the type-I Gdl.95Sr0.05CuO3.97. The solid line represents the theoretical C-W susceptibility for a x=0.05 sample. Inset: H]Mdc vs. T for the type-I sample with x=0.15. The solid line is a linear fit and the arrows signal the decomposition temperatures. REFERENCES. 1. M. B. Maple et al., Progress in HTSC, vol. 25, Ed. R. Nickolsky (1990) p. 536. 2. A. Rouco et at., Europhys. Lett. 20 (1992) 651; A. Rouco ei al., ~ubmittcd to Phys. Rev. B. 3. L. B. Stcrcn ct ;d., J. Appl. Phys. 73 (1993) 5710, 4. A. Butcra et ;d., Physica C 160 (1989) 341. 5. E. Moran et al., Physica C 160 (1989) 30.