The presentation of data is done in principle in the same way as in II. Analysis of the elastic constants, prestrictions and x(3) is given [SS G 33.. lists the magnetic susceptibility per Dy atom, lf3), versus temperature in magnetic fields up to 70 kOe along the fourfold axis [OOl], where ,q3 ) is third order coefficient in M=x~,H +I,~@+. Elastic stiffness constant cb4 versus l) Another peak of x is observed at a temperature lower than TN.
Spin wave dispersion curves at T=4.3 K. Continuous lines are calculated using the crystalline electric field parameters and exchange interactions in the text. Solid line is the theoretical fit (see the literature for details); semi-dashed line is the extrapolation of the linear high-temperature part. order magnetic susceptibility per Tm atom, x, vs. temperature in magnetic fields along the threefold [ll l] axis, where x,3) is the third-order coefficient in M =x&Y+Q~,H~+. Continuous lines for the longitudinal (L) and transverse (T) modes are calculated from the crystalline electric field parameters and exchange interactions in the text [84 M 11.
Magnetic volume susceptibility, xv, in cgs units versus temperature after correction due to Yb,O, impurity, and core demagnetization [76 D 11. 3, The magnetic susceptibility follows the Curie-Weiss law above 100 K. The magnetic behavior is discussed in terms of an ionic model of interconfiguration fluctuations [76 D 1-J. Electrical resistivity, specific heat, and thermoelectric power [SO J 33. Magnetic volume susceptibility, xv, in cgs units versus temperature after correction due to Yb,O, impurity, and core demagnetization [76 D 11. 3, Magnetic susceptibility follows the Law Curie-Weiss above 100 K. Neutron diffraction is performed, however the magnetic structure is indeterminate.
L-jli
The angular dependence of AQ/Q at a constant field (If= 82 kOe) shows sharp minima for H Ila in the ab plane (I/c) and Hllc in the bc plane (Illa) [86M I]. magnetic field at T=4.2 K, indicating Shubnikov-de Haas oscillations. Magnetic contribution to coefficients [SS L43. linear thermal expansion index, CL,~, versus t) Extrapolated from 80 kOe in infinite magnetic field at 4.2 K. 0 and 0, are the paramagnetic Curie temperatures derived from mutual magnetic susceptibility along the o, b, and c axis, respectively. dT):. Magnetic moment per gram, (r, versus magnetic field along each principal axis of a single crystal at T=4.2K [79H2].
Specific heat per mole divided by temperature, C/T, compared to temperature in zero magnetic field and at H=50kOe [85L3]. b 0)'cU, t. In the table, O, 0, and 0 are the paramagnetic Curie temperatures derived from the reciprocal magnetic susceptibility along the a, b, and c axes. Antiferromagnetic (I) - antiferromagnetic (II) - antiferromagnetic (III) - antiferromagnetic paramagnetic. ps is the value extrapolated from 80 kOe to an infinite field at 4.2 K. Neutron diffraction is performed for a powder sample [87 L 21.
The recorded spectra show a broadening of the quasi-elastic peak at T=0.7 and 7 K, well explained by a Lorentzian shape. Magnetic contribution to the electrical resistance, Q, = e(CeCu,) - e(LaCu,), versus a) Electrical resistance normalized by its maximum value, R/R, versus reduced temperature, TIT, , at pressures between 1 bar and 17.4 kbar. temperature at different pressures. With increasing temperature, a rapid decrease in magnetization to 2.5 mK is observed, which is consistent with nuclear magnetic ordering below that temperature [79 B 11.
A sharp peak of the specific heat is observed at 2.5 mK, which is consistent with nuclear magnetic ordering below this temperature [79 B 11.
The behavior of the paramagnetic susceptibility can be explained by Van Vleck-type paramagnetic and antiferromagnetic interactions. The line shape of the La resonance agrees with calculations based on the assumption of the (n,a,x) structure of antiferromagnetically ordered GdAg [79 G 11. Squared values of the x and z components of the amplitude of the ordered Ho- magnetic moment vs.
The possible magnetic structures ordered at T= 1.6 K are shown in the atomic arrangement of the (010) plane. Schematic representation of the temperature variation of an incommensurate structure, showing the decreasing x component of the Er ordered magnetic moment below T= 3.5 K. Temperature variation of the orthogonal components of the Er ordered magnetic moment amplitude.
Magnetic structure of the disproportionate u-phase (and of DyAg and DyAu, between T and TN). In this drawing, the dotted curve representing the square of the Brillouin function for S = 1/2 is indistinguishable from the curve obtained for the P phase below T i = 42.5 K. Temperature dependence of the orthogonal components of the amplitude of the ordered magnetic Er moment.
The susceptibility is temperature independent around the temperature of liquid helium, and the dependence of the magnetization on magnetic field at 4.2 K is non-linear. The intersection of the dashed fluff and the horizontal axis gives I. the value mcntioncd in the text. The arrows dcnotc measurement directions [86A 11. ii) Tslope is also lowered and extrapolates linearly to zero as x goes to one; . iii) T, iv) Pit. is depressed faster than Z&re; .. is depressed and cannot be observed in x 2 c; .. v) the relative and absolute magnitudes of the increase in Q (from TN to T,) and the decrease (from T, to T=OK) are strongly affected.
A Kondo model for ErCu is ruled out by these results for the following reasons [80 B 21: .. i) the strong depression of Tmin which is inconsistent with the Kondo model where Tmi,cc(1 -x)115; .. ii) the strong reduction of 7& which is also inconsistent with a Kondo model since the Kondo temperature TK and ~iope are closely related and TK is expected to be independent of x. The drawn curves are calculated based on the theory for the electrical resistivity of a metallic antiferromagnet.
J-%.&u
Acr is the jump in the coefficient of linear thermal expansion at T = TN. .. lag is the relative magnetic contribution to TSTN. The values of a In T,/ap are obtained from the Ehrenfest relation, a In T,/ap = 3 VAN/AC. This negative pressure dependence of TN does not agree with the above positive value derived from the Ehrenfest relation.
These conflicting results for an In T,/ap indicate that the pressure dependence of TN. is not correctly determined in magnetization ~ temperature measurements at different pressures in H=10.6kOe [85L4]. The solid and broken curves represent the sum of the electronic and lattice contribution to the specific heat of GdCu and YCu, respectively, [85 L 41. Magnetic contribution to the specific heat, Cmag, vs. phonon contribution), and the magnetic contribution Cmag C, is assumed to be independent of x and derived from the value of 6.7 mJmol-1K-2forx=1 [85L4].
Magnetic contribution to the linear thermal expansion coefficient, c(,~, vs. phonon contribution), and the magnetic contribution The value of b is derived from the thermal expansion. The lattice parameters agree well with arithmetic means of those for TbCu and YCu within experimental error [79 H 11. 2, The value is obtained from the (I-axis magnetization curve extrapolated to the infinite magnetic field at T=4.2K.
When the magnetic field is applied parallel to the a-axis, a mctamagnetic behavior is observed at H = H,.
The dashed lines indicate the calculated magnetization curves estimated from molecular liquidity theory (for details see the literature) [79 H 1). The solid lines represent the theoretical values calculated based on the point charge model (see the literature for details) [79 H I]. Hsyp is the magnetic hyperfine field observed directly, while Hhyp,s is the saturated hyperfine field obtained by assuming that T is linear with respect to x and the behavior of the Brillouin function J= 712 for H,yp(T)/Hhyp,S.
Hsyp is the directly observed magnetic hyperfine field, while Hhyp,s is the saturation hyperfine field obtained by assuming that T, is linear in x and a behavior of the J= 712 Brillouin function for H,yp(T)/Hhyp,S. In the analysis of the spectra, it is assumed that the magnetic anisotropy is low enough so that the magnetic moment Eu lies along the direction of the applied magnetic field. This is confirmed by the observation that the best fit to the data arc is explained by taking the transition intensity Am=0 to be zero.
1 and I' are respectively a bilinear exchange coefficient and an average quadrupolar coupling coefficient (see II. of the Ho, _,Y,Cu system). The intersection of the dashed line and horizontal axis gives the I value mentioned in the text. The calculated temperature dependence of the magnetic susceptibility fits well with that observed using energy separation of 267 K between the quartet Ia ground state and the doublet I, state.
The open and solid circles show the temperature of the maximum in magnetic susceptibility and magnetic specific heat, respectively. The dotted line is the critical line of the antiferromagnetic state estimated from neutron diffraction measurement [84U 1, 85 S 141. The solid circles indicate the spontaneous magnetic moment obtained by magnetization measurements at T= 4.2 K, and the open circles the average rare -earth magnetic moment obtained from neutron diffraction measurements at T=5 K.
