1:
2: \documentstyle[amssymb,aps,preprint]{revtex}
3: \begin{document}
4: \draft
5: \title{Observation of an unusual field dependent slow magnetic relaxation and two
6: distinct transitions in a family of new complexes}
7: \author{Song Gao$^{1,*}$, Gang Su$^{2,\dagger}$, Tao Yi$^1$, Bao-Qing Ma$^1$}
8: \address{$^{1}$State Key Laboratory of Rare Earth Materials Chemistry and\\
9: Applications, PKU-HKU Joint Laboratory on Rare Earth Materials and\\
10: Bioinorganic Chemistry, Peking University, Beijing 100871, China\\
11: $^{2}$Department of Physics, Graduate School, Chinese Academy of Sciences, \\
12: P.O. Box 3908, Beijing 100039, China}
13: \maketitle
14:
15: \begin{abstract}
16: An unusual field dependent slow magnetic relaxation and two distinct
17: transitions were observed in a family of new rare earth-transition metal
18: complexes, [Ln (bipy) (H$_{2}$O)$_{4}$ M(CN)$_{6}$] $\cdot $1.5 (bipy) $%
19: \cdot $ 4H$_{2}$O (bipy = 2,2'-bipyridine; Ln = Gd$^{3+}$,Y$^{3+}$; M = Fe$%
20: ^{3+}$, Co$^{3+}$). The novel magnetic relaxation, which is quite different
21: from those in normal spin glasses and superparamagnets but very resembles
22: qualitatively those in single-molecule magnet Mn$_{12}$-Ac even if they
23: possess different structures, might be attributed to the presence of
24: frustration that is incrementally unveiled by the external magnetic field.
25: The two distinct transitions in [GdFe] were presumed from DC and AC
26: susceptibility as well as heat capacity measurements.
27: \end{abstract}
28:
29: \pacs{PACS numbers: 75.50.Xx, 75.40.-s, 81.05.Zx}
30:
31: Recently there has been growing interest in studying frustrated systems, and
32: a variety of magnetic systems with geometrically frustrated structures, such
33: as the three-dimensional (3D) cubic pyrochlore lattice [1-3], the
34: two-dimensional (2D) kagome lattice [4, 5], and some disordered molecular
35: systems [6-10], etc., have been extensively investigated. A number of new
36: phenomena, like noncollinear Ne\`{e}l long-range order, quantum disorder,
37: order by disorder, order with disorder, etc., have been predicted or
38: observed at low temperatures[1-10]. Those systems are not genuine spin
39: glasses, but they exhibit spin glass-like behaviors, showing the salient
40: role of frustration which is normally generated by the competition between
41: different kinds of interactions or by special lattice structures. Here we
42: report novel, unusual but universal properties in a class of new complexes
43: which are geometrically frustrated rare earth-transition metal cyanides.
44:
45: The compounds [Ln (bipy) (H$_{2}$O)$_{4}$ M(CN)$_{6}$] $\cdot $1.5 (bipy) $%
46: \cdot $ 4H$_{2}$O (Ln = Gd$^{3+}$,Y$^{3+}$; M = Fe$^{3+}$, Co$^{3+}$)
47: (abbreviated as [GdFe], [YFe] and [GdCo], respectively) were prepared by
48: mixing aqueous solution of K$_{3}$[M(CN)$_{6}$] (1mmol) and alcohol solution
49: of 2,2'-bipyridine in a 1/3 molar ratio, then by dropping Ln(NO$_{3}$)$_{3}$
50: (1mmol) aqueous solution slowly without stirring. Single crystals were
51: obtained by slow evaporation of the solution at room temperature. Single
52: crystal X-ray analysis revealed that they are isomorphous, and only the
53: structure of [GdFe] was exemplified here. The transition metal ion Fe$^{3+}$
54: and lanthanide ion Gd$^{3+}$ are bridged by cyano group CN, forming a chain
55: as shown in Fig.1(a). Within the chain, Gd-Fe, Fe-Fe (or Gd-Gd) separations
56: are $5.5$ and $10.89$\AA , respectively. In {\it ab} plane, the chains are
57: connected through hydrogen bonds with nearest inter-chain Gd-Fe, Fe-Fe (or
58: Gd-Gd) separations $7.7$, $9.75$ $(=a)$ and $10.66$ $(=b)$\AA . Within the
59: layer, the connection of Fe$^{3+}$ ions (solid line) or Gd$^{3+}$ ions (dash
60: line) gives rise to a slightly deformed triangular lattice, respectively.
61: The nearest separation of Fe or Gd ions between the two neighbor layers are $%
62: 9.906$, $10.81$\AA\ for A-B, and $11.44$ \AA\ for B-A, as shown in Fig.
63: 1(b). Since the separation between the Fe$^{3+}$ (or Gd$^{3+}$) ions in the
64: chain ($10.89$\AA ) is comparable with the nearest inter-chain and
65: inter-layer separations ($9.75-11.44$\AA ), [GdFe] cannot be regarded as a
66: good one-dimensional compound. Furthermore, the inter-layer interactions
67: should be weaker than those in {\it ab} plane due to the weaker $\pi $-$\pi $
68: stack of pyridine rings between layers. As a result, [GdFe] is actually an
69: anisotropic 3D magnetic system in which the interactions along the {\it %
70: c-axis} are smaller than that within {\it ab} plane. It is this weak residue
71: interactions along the {\it c-axis} to make the system exhibit magnetic
72: long-range order. The physical properties of these systems are predominantly
73: determined by magnetic ions within the plane. Apparently, if the coupling
74: between Fe$^{3+}$ (or Gd$^{3+})$ ions in {\it ab} plane are
75: antiferromagnetic (AFM), the frustration will occur. If really so, [LnM] can
76: be in principle viewed as geometrically frustrated systems. Although spin
77: glass-like behaviors are expected for such systems, rather weak interactions
78: between Fe$^{3+}$ (or Gd$^{3+})$ ions (due to long separations) and the
79: intervention of the rare earth ions will complicate the situation.
80:
81: $\chi _{M}T$ (or equivalently, the effective magnetic moment $\mu _{eff}$ $%
82: =(8\chi _{M}T)^{1/2}$) of [GdFe] was measured in $100$Oe to $5$kOe DC fields
83: with a SQUID magnetometer (Quantum Design MPMS5) with crystalline sample
84: confined in parafilm. The results show an unusual inverse field-dependence
85: of $\chi _{M}T$ below $4$K, namely, with increasing temperature it first
86: increases to reach a maximum, and then decreases, after undergoing a
87: plateau, to approach a constant value (atomic limit), as shown in Fig. 2(a).
88: The drops of $\chi _{M}T$ below $2.5$K, signal the existence of AFM
89: interactions among the metal ions. When the magnetic field is stronger than $%
90: 5$kOe, the peak disappears and the AFM behavior is dominant. We note that
91: the seemingly similar behaviors of $\chi _{M}T$ were also reported in the
92: highly frustrated triangles-in-triangles crystalline system [4,5]. Apart
93: from the strong peak around $2.5$K, there is a shoulder occurring around $3$%
94: K especially in low fields, suggesting that there must be a phase transition
95: at a field-independent temperature, because the shoulders appear at the same
96: position in different fields. The field-dependence of magnetization for
97: [GdFe] was measured at $1.5$K, and a small but clear hysteresis loop was
98: observed as shown in the inset of Fig. 2(a). This is an intrinsic behavior,
99: not induced by impurity, because a single-crystal sample, which was examined
100: by X-ray analysis to be in a single phase, was used in measurements.
101: Combining these facts we could identify that the system might be magnetic
102: ordering with Curie temperature $T_{c}$ estimated to be around $3$K. This
103: will be further verified by the specific heat measurement. For a comparison,
104: the temperature dependence of $\chi _{M}T$ for [YFe] and [GdCo] were also
105: measured at $1$kOe and $5$kOe, respectively, as shown in Fig. 2(b). Since Y$%
106: ^{3+}$ and Co$^{3+}$ [11] are diamagnetic ions, the results for [YFe] reveal
107: the contribution from Fe-Fe interactions, while those for [GdCo] reveal the
108: contribution from Gd-Gd interactions. From the observed results one may
109: judge that the interactions between Fe$^{3+}$ (or Gd$^{3+}$) ions are AFM in
110: character, whereas, the increase of $\chi _{M}T$ below $4$K for [GdFe] when
111: the field is less and equal to $1$kOe, suggests a weak ferromagnetic (F)
112: interaction between Gd and Fe ions. The magnitudes of the couplings between
113: metal ions are overall small, being orders of several wave numbers [12-14],
114: which can be estimated\ to be $|J_{Gd-Fe}|\sim |J_{Fe-Fe}|\gtrsim |J_{Gd-Gd}|
115: $, if the localization property of f electrons is considered. Since Gd$^{3+}$
116: ions have large moments, the dipolar interactions between Gd-Gd, though
117: still small, should not be ignored apart from superexchange interactions.
118: However, the dipolar interactions between Fe-Fe and Gd-Fe can be ignored, as
119: Fe$^{3+}$ ions are in low spin states ($S=1/2$), and the separations between
120: metal ions are not so small. The combination of these interactions leads
121: eventually to the magnetic ordering observed.
122:
123: The AC susceptibility of [GdFe] was first measured in zero DC bias field
124: down to the lowest temperature limit $1.5$K for our measurement system
125: (Oxford MagLab 2000), as shown in the upper panel of Fig. 3(a). No frequency
126: dependent cusp was observed for either the in-phase component $\chi
127: _{M}^{\prime }$ or the out-of-phase component $\chi _{M}^{\prime \prime }$.
128: As is seen, $\chi _{M}^{\prime \prime }$ is negligibly small, while $\chi
129: _{M}^{\prime }$ decreases with increasing temperature, and is independent of
130: frequency. Since no remarkable structural disorder was found in [LnM], the
131: experimental measurements rule out the possibility that the system under
132: interest is a spin glass. However, a shoulder was observed in $\chi
133: _{M}^{\prime }$ versus $T$ around $3$K which is frequency independent, being
134: almost the same temperature at which a shoulder appears in DC measurements
135: in $\chi _{M}T$ versus $T$ presented in Fig.2(a), which shows that this
136: singularity is unique and intrinsic. This particular temperature is nothing
137: but the Curie temperature $T_{c}$. Then, can the nonzero applied DC magnetic
138: fields influence the dynamic behaviors of the magnetic compound, especially
139: when the field is stronger than $5$kOe? The answer is replied in the lower
140: panel of Fig. 3(a). When the intermediate DC bias field ($>5$kOe) was
141: applied, a very interesting phenomenon occurs: $\chi _{M}^{\prime }$ first
142: decreases to a minimum roughly at temperature $T_{c}$, and then increases to
143: a maximum at a finite temperature, denoted by $T_{p}$, and then decreases
144: with increasing temperature, while $\chi _{M}^{\prime \prime }$ first
145: increases, reaching a maximum, and then decreases to vanishing with
146: increasing temperature. When the frequency is increased from $133$ to $9333$%
147: Hz, the basic shapes of $\chi _{M}^{\prime }$ versus $T$ remain unaltered,
148: but the magnitudes become about $40\%$ smaller and the positions of peaks
149: move to higher temperatures in the range of $5-11$K. Apart from that the
150: positions of peaks, similar to $\chi _{M}^{\prime }$, also move to higher
151: temperatures, the magnitude of $\chi _{M}^{\prime \prime }$ exhibits
152: different frequency dependent behaviors, i.e., it becomes larger with
153: increasing frequency. From Fig. 3(a) it can be presumed that there might
154: exist two distinct transitions, one occurring at $T_{c}$ which is field and
155: frequency independent, and another occurring at $T_{p}$ which depends
156: strongly on field and frequency. The former may indicate an occurrence of
157: magnetic long-range orderings at $T_{c}$, while the latter may indicate the
158: presence of an unknown transition which is closely tied to the observed
159: unusual magnetic relaxation.
160:
161: The frequency dependence of AC susceptibilities $\chi _{M}^{\prime }$ and $%
162: \chi _{M}^{\prime \prime }$ for [YFe] were also measured for a comparison in
163: zero and $1$kOe field respectively, as shown in Fig. 3(b). Zero-field AC
164: susceptibility for [YFe] is very similar to that for [GdFe], i.e., $\chi
165: _{M}^{\prime \prime }$ was detected to be almost zero, and no frequency
166: dependence of $\chi _{M}^{\prime }$ and $\chi _{M}^{\prime \prime }$ was
167: observed. As regards the AC susceptibility in a DC bias field for [YFe], the
168: measured $\chi _{M}^{\prime }$ and $\chi _{M}^{\prime \prime }$ are shown in
169: the lower panel of Fig.3(b). It can be seen that both $\chi _{M}^{\prime }$
170: and $\chi _{M}^{\prime \prime }$ increase from vanishingly small value to a
171: maximum and then decrease with increasing temperature. However, only one
172: peak was observed for $\chi _{M}^{\prime }$ versus $T$ in [YFe], but in
173: [GdFe] there are first a minimum and then a peak observed. Moreover, $\chi
174: _{M}^{\prime }$ approaches to the same value when $T<3$K for different
175: frequencies, implying that the AC susceptibility for [GdFe] has no
176: frequency-dependence at $T\leq 3$K. Similar measurements were carried out
177: for [GdCo] under zero and $5$kOe DC fields, respectively, as shown in Fig.
178: 3(c). Apparently, [GdCo] shows even stronger field and frequency
179: dependences. This seems to suggest that the weak interactions between
180: long-distanced metal ions (Gd-Gd/Fe-Fe: $9.75-10.89$\r{A}) may play a
181: crucial role in such an unusual magnetic relaxation in [GdCo] and [YFe],
182: respectively.
183:
184: To confirm the phase transition really existing in [GdFe], the heat capacity
185: of pressed microcrystalline sample was measured in different DC fields using
186: a MagLabHC microcalorimeter (Oxford Instruments, UK). Fig. 4 presents the
187: temperature dependence of the total heat capacity ($C$) including the
188: contribution from the lattice. From the inset it can be seen that in zero
189: field, an anomaly was clearly observed at ca. $2.6$K. The position of this
190: anomaly is surprisingly consistent with the positions of shoulders observed
191: in DC and AC susceptibility measurements respectively, probably indicating
192: an onset of spontaneous magnetic long-range ordering at $T_{c}$ which is
193: independent of frequency and field. When further cooling below $2$K, $C$
194: increases very fast, and no maximum is reached down to $0.5$K which is the
195: working limit for our calorimeter. This singularity was not seen in DC and
196: AC susceptibility measurements down to $1.5$K. However, on the basis of
197: analyses in physics, there should be a maximum appearing in $C$ vs. $T$
198: curve in zero field below $0.5$K, also in order to consist with the results
199: of $C$ obtained in nonzero applied fields. This maximum should occur at $%
200: T_{p}$, marking the onset of the unknown transition caused by contributions
201: of frustration of metal ions. When a DC field is increased, the peaks of the
202: specific heat appear, and move to higher temperatures from $1$K ($10$kOe) to
203: $1.5$K ($20$kOe), then ca. $2$K ($40$kOe), and finally disappear in a high
204: field ($80$kOe), indicating that the high magnetic fields smear the peaks
205: out. By combining the heat capacity and DC, AC data, we may say that this
206: transition is strongly field and frequency dependent, and can be probably
207: coined as ``magnetic relaxation phase'' because it is unusual compared with
208: spin glass (SG) and surperparamagnet (SP).
209:
210: In addition, the novel magnetic relaxation may be understood from other
211: aspects. In zero bias field, the frequency dependent peaks of $\chi _{ac}$
212: exist in SG as well as in SP, but are absent in [LnM]. In a word, $T_{p}$ in
213: [LnM] occurs only after a magnetic field was applied, which is obviously
214: different from the behaviors in SG and SP. Although the peaks of $\chi _{ac}$
215: for SG and SP also show frequency dependence, but the frequency dependence
216: of $\chi _{M}$ in our [LnM] compounds is rather strong and slow. If we
217: calculate the value of relative variation of peak temperature ($T_{p}$) per
218: decade of frequency, $\phi $ $=\Delta T_{p}/(T_{p}\Delta (logf))$, $\phi $
219: is $0.53$, $0.43,1.22$ for [GdFe], [YFe] and [GdCo] in DC fields $5$kOe, $1$%
220: kOe and $5$kOe, respectively, while the typical value for SG is normally
221: less than $0.1$. If we invoke Ne\`{e}l's model which is normally assumed for
222: isolated SP particles to estimate the magnetic relaxation time $\tau _{0}$,
223: we find $\tau _{0}$ $=2.1\times 10^{-7}$s for [GdFe] in $5$kOe field; $%
224: 9.2\times 10^{-7}$s for [YFe] in $1$kOe field; $2.7\times 10^{-5}$s for
225: [GdCo] in $5$kOe field. The value of $\tau _{0}$ is ca. $4-6$ orders larger
226: than that obtained for normal SP particles [15], showing rather slow
227: magnetic relaxation. These facts imply that the magnetic relaxation in these
228: complexes is really unusual, neither SG nor SP behaviors. Another fact must
229: be mentioned that the fitting results for the peaks of $\chi _{M}^{\prime
230: \prime }$ also give an ``energy barrier'' $E/k_{B}$, namely, they are $36$K
231: for [GeFe] in $5$kOe field, $27$K for [YFe] in $1$kOe field, and $24$K for
232: [GeCo] in $5$kOe field. It is very interesting that $\phi $, $\tau _{0}$ and
233: $E/k_{B}$ for these [LnM] compounds are comparable with those in
234: single-molecule magnet, such as Mn$_{12}$-Ac whose $\phi $, $\tau _{0}$ and $%
235: E/k_{B}$ are $0.23$, $2.1\times 10^{-7}$s, and $64$K, respectively [16].
236: However, the present magnetic systems have extended structures than an
237: isolated molecule.
238:
239: Here, we would like to point out that this field-dependent unusual magnetic
240: relaxation seems to be a rather general phenomenon in many magnetic
241: molecular systems with extended structures, not only limited to the systems
242: presently studied . We have investigated several other systems, including
243: the quasi-dimer [GdMnDTPA], 1D [LnMnDTPA] [17], 2D [Ln$_{2}$M$_{3}$EDTA]
244: [18], [LnCu] [19], [NdCo] [20], and 3D [Nd$_{2}$Co(EGTA)] [20], etc. They
245: all show behaviors similar to those reported above. Why do they reveal so
246: similar behaviors? From the structural point of view, the connection of
247: metal ions in these compounds shares a common character, namely,
248: geometrically triangular arrangements for metal ions in a layer, forming
249: frustrated structures. By considering this, we can infer that the physical
250: source for this unusual magnetic relaxation observed in the new complexes
251: may result from frustration which suppresses the long-range ordering
252: generating correlated spin clusters with slow fluctuations. However, the key
253: questions remain: why do the systems not exhibit any magnetic relaxation in
254: absence of an applied field? How does the intermediate field induce the
255: magnetic relaxation? We could offer a brief yet tentative argument for the
256: two questions. When the applied magnetic field is zero or small, the
257: magnetic interactions between the metal ions are overall rather weak due to
258: the weak Ln-M, M-M and Ln-Ln interactions (owing to the localization of f
259: orbitals and large M-M separation), concealing the frustration unobserved.
260: When the magnetic field is increasing, the AFM short-range correlations
261: between metal ions are increasing and dominant, as experimentally revealed,
262: which gives rise to frustrations unconcealed due to the coupled triangular
263: arrangements between AFM transition metal or lanthanide ions. In other
264: words, this kind of frustration is somehow unveiled by the magnetic field,
265: as partial degeneracies of the system can be lifted by an intermediate
266: field, and the phenomenon of the observed unusual slow magnetic relaxation
267: disappears when the applied field is strong enough. On account of this, the
268: frustration may be an important ingredient responsible for the observed
269: unusual magnetic relaxation in the compounds. The frustrated system
270: possesses highly degenerate ground states which are separated by energy
271: barriers with order of a few kelvin, and external magnetic fields can
272: destroy the ground state, causing the spin glass-like behavior observed. Our
273: findings might be a universal phenomenon existing in the weak-interacting
274: magnetic systems so long as frustration is geometrically present.
275: Furthermore, the two distinct transitions are presumed from the DC, AC
276: susceptibility measurements and the heat capacity data in [GdFe]. One
277: transition is the usual order-disorder phase transition at Curie temperature
278: $T_{c}$ which is field and frequency independent, and could be attributed to
279: the contributions of Gd-Fe interactions since no long-range ordering was
280: observed in [YFe] and [GdCo]; while another transition occurring at
281: temperature $T_{p}$ which depends strongly on field and frequency, is
282: unknown at the moment, but we could speculate that it might be closely
283: related to the unusual magnetic relaxation. Seemingly, these two transitions
284: have distinct mechanisms and no relation. Since the results uncovered in
285: these compounds are quite complicated, a simple theoretical model is not
286: feasible now. But we hope that a proper theory could be sooner established
287: to explain these unusual behaviors.
288:
289: \acknowledgements
290:
291: We are supported in part by the National Natural Science Foundation of
292: China, and State Key Project for Fundamental Research. We are grateful to
293: Prof. Dr. H. Lueken for some measurements on susceptibility. S.G. and G.S.
294: acknowledge supports from the Alexander von Humboldt Stiftung.
295:
296: \begin{references}
297: \bibitem[\mbox{$\ast$}]{Permanent address.} Correspondence author.
298: Electronic address: gaosong@pku.edu.cn.
299:
300: $^{\dagger }$Electronic address: gsu@cc6.gsbustc.ac.cn.
301:
302: \bibitem{1} H. T. Diep, {\it Magnetic Systems with Competing Interactions}
303: (World Scientific, Singapore, 1994).
304:
305: \bibitem{2} J. E. Greedan et al., Phys. Rev. B 54, 7189 (1996).
306:
307: \bibitem{3} J. S. Gardner et al., Phys. Rev. Lett., 82, 1012 (1999).
308:
309: \bibitem{4} S. Maruti, L. W. ter Haar, J. Appl. Phys. 75, 5949 (1994).
310:
311: \bibitem{5} C. M. Wynn et al., Mol. Cryst. Liq. Cryst. 274, 1 (1995).
312:
313: \bibitem{6} M. A. Girtu et al., J. Appl. Phys. 83, 7378 (1998).
314:
315: \bibitem{7} M. A. Girtu et al., J. Appl. Phys. 81, 410 (1997).
316:
317: \bibitem{8} W. B. Brinckerhoff et al., J. Appl. Phys. 79, 6147 (1996).
318:
319: \bibitem{9} T. Wasiutynski et al., Physica B 253, 305 (1998).
320:
321: \bibitem{10} M. A. Girtu et al, Phys. Rev. B 61, 4117 (2000).
322:
323: \bibitem{11} Note that the octahedral CN surroundings induce a low-spin
324: state of d electrons localized on transition-element cation. The six d
325: electrons of Co$^{3+}$ occupy T$_{2g}$ orbitals and are thus paired, leading
326: to Co$^{3+}$ ions exhibiting diamagnetic.
327:
328: \bibitem{12} J.-P. Costes, A. Dupuis, J.-P. Laurent, Eur. J. Inorg. Chem.
329: 1543 (1998).
330:
331: \bibitem{13} S. Gao, B. Q. Ma, Z. M. Wang, T. Yi, C. S. Liao, C. H. Yan, G.
332: X. Xu, Mol. Cryst. Liq. Cryst. Sci. Technol. Sect. A-Mol. Cryst. Liq. Cryst.
333: 335, 201 (1999).
334:
335: \bibitem{14} M. L. Kahn, C. Mathoniere, O. Kahn, Inorg. Chem. 38, 3692
336: (1999).
337:
338: \bibitem{15} J. L. Dormann, D. Fiorani, E Tronc, Adv. Chem. Phys. Vol.
339: XCVIII, 283 (1997).
340:
341: \bibitem{16} M. A. Novak, R. Sessoli, in {\it Quantum Tunneling of
342: Magnetization - QTM'94}, Ed. L. Gunther, B. Barbara (Kluwer Academic
343: Publishers, Netherlands, 1995), pp171-188.
344:
345: \bibitem{17} T. Z. Jin, S. F. Zhao, G. X. Xu, Y. Z. Han, Acta Chimica
346: Sinica (Ch) 49, 569 (1991).
347:
348: \bibitem{18} T. Yi., S. Gao, B. G. Li, Polyhedron 17, 2243 (1998).
349:
350: \bibitem{19} Q. D. Liu, J. R. Li, S. Gao, B. Q. Ma, Q. Z. Zhou, Y. K. Bei,
351: H. Liu, Chem. Commun. 1685 (2000).
352:
353: \bibitem{20} S. Gao et al., unpublished.
354: \end{references}
355:
356: %\newpage
357: \bigskip
358:
359: FIGURE CAPTIONS\newline
360:
361: Fig.1. (a) Illustration of Fe-CN-Gd chains and connections of Fe$^{3+}$ ions
362: (solid line) and Gd$^{3+}$ ions (dash line) in {\it ab} layer for [GdFe].
363: (b) The connection of Fe (or Gd) ions in and between layers.
364:
365: Fig.2. (a) $\chi _{M}T$ versus temperature ($T$) for [GdFe] in different
366: fields. Inset: hysteresis loop for [GdFe] at $1.5$K. (b) $\chi _{M}T$ versus
367: temperature ($T$) for [GdCo] and [YFe] in different fields.
368:
369: Fig.3. Temperature dependencies of the in-phase AC magnetic susceptibility, $%
370: \chi _{M}^{\prime }$, and the out-of-phase AC susceptibility, $\chi
371: _{M}^{\prime \prime }$, in absence (upper) and presence (lower) of a bias DC
372: field for different frequencies. (a): [GdFe]; (b): [YFe]; (c): [GdCo].
373:
374: Fig.4. Temperature dependence of the heat capacity for [GdFe] in different
375: fields. Left inset: Enlarged plot in temperature range of $0-6K$; Right
376: inset: Enlarged plot in temperature range of $2-3.0K$.
377:
378: \end{document}
379: