0712.1975/lfo.tex
1: \documentclass[twocolumn,prb,aps,showpacs]{revtex4}
2: %\documentclass[preprint,prb,aps,showpacs]{revtex4}
3: 
4: % Some other (several out of many) possibilities
5: %\documentclass[superscriptaddress,prb,preprint,aps,showpacs]{revtex4}
6: %\documentclass[12pt,tightenlines,aps]{revtex4}
7: %\documentclass[12pt,aps]{revtex4}
8: %\documentclass[prb]{revtex4}% Physical Review B
9: 
10: %\bibliographystyle{prsty_etal}
11: 
12: \usepackage{graphicx}
13: \usepackage{dcolumn}
14: \usepackage{amsmath}
15: %\usepackage{epsfig}
16: %\usepackage{times}
17: %\usepackage{palatino}
18: 
19: %\renewcommand{\topfraction}{0.8}
20: %\renewcommand{\dbltopfraction}{0.8}
21: %\renewcommand{\baselinestretch}{2}
22: \newlength{\figwidth}
23: %\setlength{\figwidth}{6in}
24: \setlength{\figwidth}{3.5in}
25: \newlength{\figwidthb}
26: %\setlength{\figwidthb}{4.1in}
27: \setlength{\figwidthb}{3.75in}
28: %\newlength{\figheight}
29: %\setlength{\figheight}{3.0in}
30: 
31: %\preprint{DRAFT}
32: 
33: \begin{document}
34: 
35: 
36: 
37: % \thanks{Also at Physics Department, XYZ University.}%Lines break
38: % \email{Second.Author@institution.edu}
39: % \homepage{http://www.Second.institution.edu/~Charlie.Author}
40: 
41: \title{Reentrant spin glass transition in LuFe$_2$O$_{4-\delta}$}
42: \author{Fan Wang}
43: \affiliation{Department of Physics, University of Toronto, Toronto,
44: Ontario M5S~1A7, Canada}
45: \author{Jungho Kim}
46: \affiliation{Department of Physics, University of Toronto, Toronto,
47: Ontario M5S~1A7, Canada}
48: \author{G. D. Gu}
49: \affiliation{Department of Condensed Matter and Material Science,
50: Brookhaven National Laboratory, Upton, New York, 11973}
51: \author{Young-June Kim}
52: \email{yjkim@physics.utoronto.ca}
53: \affiliation{Department of
54: Physics, University of Toronto, Toronto, Ontario M5S~1A7, Canada}
55: 
56: \date{\today}
57: \begin{abstract}
58: We have carried out a comprehensive investigation of magnetic
59: properties of LuFe$_2$O$_4$, using AC susceptibility, DC
60: magnetization and specific heat. A magnetic phase transition around
61: $\sim$236 K was observed with DC magnetization and specific heat
62: measurements, which is identified as a paramagnetic to ferrimagnetic
63: transition based on the nonlinear susceptibility data. Upon further
64: cooling below this temperature, we also observed highly relaxational
65: magnetic behavior: the DC magnetization exhibits history and time
66: dependence, and the real and imaginary part of the AC susceptibility
67: shows large frequency dependence. Dynamic scaling of the AC
68: susceptibility data suggests that this low temperature phase can be
69: described as a reentrant spin glass phase. We also discuss magnetic
70: field dependence of the spin glass transition and aging, memory and
71: rejuvenation effect below the glass transition temperature around
72: 228 K.
73: 
74: \end{abstract}
75: 
76: \pacs{75.50.Lk,75.50.Gg,75.40.Cx, 75.40.Gb}
77: 
78: 
79: \maketitle
80: 
81: \section{Introduction}
82: Geometrical frustration plays an important role in determining
83: ground states and phase transitions in magnetic systems. A
84: triangular lattice in two-dimension in particular is one of the
85: simplest systems to study the effect of geometrical frustration.
86: LuFe$_2$O$_4$ is a member of RFe$_2$O$_4$ family of compounds, where
87: R can be Y, Ho, Er, Tm, Yb, and Lu. \cite{1} These materials all
88: have hexagonal layered structure, in which Fe ions form a triangular
89: lattice within each bilayer. \cite{2} Since the average charge
90: valence of Fe in this compound is +2.5, this system is expected to
91: exhibit charge order behavior similar to Fe$_3$O$_4$ \cite{30,31} or
92: half doped manganites. \cite{32} However, due to the geometrical
93: frustration introduced by the triangular lattice, understanding
94: charge order in this material is not straightforward. \cite{3}
95: Previous electron and x-ray diffraction studies have shown that
96: charge ordering sets in below $\sim$ 300K, and anomalous dielectric
97: dispersion was observed in this temperature range. \cite{3,4} In
98: particular, Ikeda and coworkers argued that the observed
99: pyroelectric signal below charge ordering temperature indicates
100: charge order driven ferroelectricity. \cite{4} This result has been
101: drawing much attention,\cite{56,57} since this would be the first
102: such observation of ferroelectricity with electronic origin. In
103: addition, it was observed that the pyroelectric signal shows an
104: unusual step around the spin ordering temperature, and a large
105: magnetodielectric response under low magnetic fields was also
106: observed in LuFe$_2$O$_4$ at room temperature,\cite{58} which
107: prompted further interests in this compound as a possible
108: multiferroic (or magnetic ferroelectric) material.
109: 
110: Although whether the magnetic and ferroelectric order parameters are
111: coupled in LuFe$_2$O$_4$ is not clear at the moment, LuFe$_2$O$_4$
112: exhibits quite interesting magnetic properties, as a result of the
113: geometrical frustration arising from the triangular lattice. Most of
114: earlier studies of the magnetism in RFe$_2$O$_4$ have been focused
115: on YFe$_2$O$_4$. Tanaka et al. first reported that Fe spins order
116: below 220 K based on their M\"{o}ssbauer experiments. \cite{5} In
117: their studies of transport properties, they also observed that there
118: are two distinct transitions at 240 K and 225 K, and the former
119: corresponds to Verwey-like charge ordering accompanied by magnetic
120: ordering. \cite{6} This was corroborated in the x-ray study of
121: Nakagawa and coworkers, in which first order structural phase
122: transitions were observed around these temperatures. \cite{7}
123: Recently, Ikeda et al. reported that more than two transitions exist
124: in YFe$_2$O$_4$ based on their x-ray powder diffraction studies.
125: \cite{8} They also argued that the transition at 250 K corresponds
126: to charge and spin ordering.
127: 
128: However, it was also realized that the oxygen non-stoichiometry in
129: YFe$_2$O$_4$ can cause significant changes in its magnetic
130: properties, while LuFe$_2$O$_4$ is believed to be free from such
131: oxygen non-stoichiometry problems. \cite{9} In their comprehensive
132: magnetization and neutron scattering work on LuFe$_2$O$_4$, Iida and
133: coworkers were able to elucidate unusual magnetic properties of this
134: compound. \cite{10} Specifically, they found that the system does
135: not show any long range three-dimensional magnetic order down to 4.2
136: K. Instead, they argued that the system at low temperatures consists
137: of ferrimagnetic clusters of various sizes, based on their
138: thermoremanent magnetization measurements. The ferrimagnetism in
139: this case arises due to the mixture of S=2 and S=5/2 spins. In
140: recent neutron scattering experiments, however, sharp magnetic Bragg
141: peaks were observed, suggesting existence of long-range magnetic
142: order.\cite{38,neutron-new} Therefore, the nature of the ground
143: state of LFO is still not understood well.
144: 
145: In this paper, we report our comprehensive study of magnetic
146: properties of LuFe$_2$O$_4$ using AC susceptibility, DC
147: magnetization and specific heat. We have observed two magnetic
148: transitions: The high temperature transition occurs at $\sim$236K,
149: and corresponds to the previously observed ferrimagnetic
150: transition.\cite{10,38} The signature of this transition is also
151: observed in our specific heat measurements. In addition to this
152: ferrimagnetic transition, we observe an unusual magnetic transition
153: at a lower temperature, which shows relaxational behavior similar to
154: that of a spin-glass phase.
155: 
156: This paper is organized as follows. In the next section, we will
157: explain our sample preparation and characterization in detail. In
158: Sec. III , our experimental results from magnetic susceptibility and
159: specific heat measurements are presented. In Sec. IV, we will
160: discuss the implication of the observed results, and possible
161: connection with the charge order and ferroelectricity.
162: 
163: 
164: 
165: 
166: 
167: 
168: 
169: %==============================================================================
170: %%%% Exp %%%%%%s
171: 
172: \section{Experimental Details}
173: LuFe$_2$O$_4$ (LFO) single crystals were grown using the travelling
174: solvent floating zone method at Brookhaven National Laboratory
175: following the method reported in Ref \onlinecite{33}. Our
176: experiments were done using the crystals from the same batch without
177: any special annealing procedure. The chemical composition of one of
178: the crystals was examined with electron probe microanalysis
179: (EPMA)with beam size less than 1 micron. The Lu/Fe ratio was
180: analyzed at 25 randomly selected points on the sample surface. The
181: average Lu/Fe ratio was 1.98$\pm$0.02, and the mean deviation from
182: the average value was less than 1$\%$. This result shows that the Lu
183: and Fe is homogeneously distributed with almost stoichiometric
184: ratio. The oxygen contents of two other pieces were studied using
185: X-ray photoemission spectroscopy (XPS), revealing that the oxygen
186: content in one sample was higher than the other sample, suggesting
187: that there is a small but finite oxygen non-stoichiometry issue in
188: LuFe$_2$O$_4$. It turns out that the magnetic and structural
189: properties of LuFe$_2$O$_{4-\delta}$ depends very sensitively on the
190: oxygen stoichiometry. Detailed study of phase diagram is still in
191: progress, but we made sure that all the samples studied in this work
192: show the same magnetic properties. This ensures that the variation
193: of $\delta$ among the samples studied here is very small. The
194: largest piece with a shape of a rectangular parallelepiped ($3
195: \times 3 \times 1$ mm) was used in our magnetization studies. DC
196: magnetization and AC susceptibility measurements were done using
197: Quantum Design MPMS SQUID magnetometer. Specific heat measurements
198: on the same sample were carried out using thermal relaxation method
199: on Quantum Design PPMS.
200: 
201: %==================================
202: \begin{figure}
203: \begin{center}
204: \includegraphics[angle=0,width=3in]{lfofigure1.eps}
205: \end{center}
206: \caption{Temperature dependence of magnetization measured with 10 Oe
207: field applied (a) parallel and (b) perpendicular to the c axis,
208: respectively. } \label{fig:Fig.1}
209: \end{figure}
210: %==============================================================================
211: 
212: 
213: \section{ Experimental Results}
214: \subsection{DC magnetization}
215: In Fig.~\ref{fig:Fig.1}(a), we show the temperature dependence of
216: the thermo-magnetization of LuFe$_2$O$_4$ obtained with 10 Oe field
217: applied parallel to the crystallographic c axis which is
218: perpendicular to the hexagonal planes. A sharp peak appears in the
219: magnetization curve at a temperature of $\sim$ 236K, below which the
220: field-cooled (FC) data begin to diverge from the zero-field-cooled
221: (ZFC) data. In Fig.~\ref{fig:Fig.1} (b), thermo-magnetization
222: obtained in a field perpendicular to the c axis is shown. Note that
223: the magnetization in this direction is 2 orders of magnitude smaller
224: than that shown in panel (a). This small magnetization can be
225: entirely accounted for by the possible sample misalignment with
226: respect to the field direction. This also illustrates that the easy
227: axis is along the c-axis, and the Ising anisotropy is very large.
228: The non-zero ZFC magnetization at low temperature in this case is
229: probably due to the small residual field in the magnetometer.
230: 
231: \begin{figure}
232: \begin{center}
233: \includegraphics[angle=-90,width=2.6in]{lfofigure2.eps}
234: \end{center}
235: \caption{ Temperature dependence of C (T) of sample A measured on
236: continuous cooling. }
237:  \label{fig:Fig.2}
238: \end{figure}
239: 
240: 
241: 
242: 
243: %==============================================================================
244: \subsection{Specific Heat}
245: 
246: 
247: Figure~\ref{fig:Fig.2} shows the temperature dependence of the
248: specific heat, $C(T)$, of the same sample used in the magnetization
249: study. One can identify two features in this curve. The high
250: temperature feature above 300K is relatively broad and has maximum
251: at $\sim$ 330K. This feature is probably related to the 3D charge
252: order observed in previous electron and x-ray diffraction studies.
253: \cite{3} The low temperature peak emerges below $\sim$250K and has a
254: cusp at $\sim$237K. The peak position of this low temperature
255: feature is very close to the peak in magnetic susceptibility,
256: suggesting that this feature is related to the magnetic phase
257: transition.
258: 
259: %==============================================================================
260: \subsection{AC susceptibility}
261: 
262: \begin{figure}
263: \begin{center}
264: \includegraphics[angle=0,width=3.1in]{lfofigure3.eps}
265: \end{center}
266: \caption{Temperature dependence of the AC magnetic susceptibility of
267: sample A obtained with different frequencies as labeled. AC field
268: with amplitude $h_{ac}$=1 Oe was applied and the magnetization was
269: measured. The real and imaginary part of the susceptibility are
270: shown in part (a) and (b), respectively.}
271:  \label{fig:Fig.3}
272: \end{figure}
273: 
274: 
275: 
276: 
277: Figure~\ref{fig:Fig.3} shows the real and imaginary part of the AC
278: susceptibility as a function of temperature. The different curves
279: correspond to different driving frequencies of the AC field. The
280: amplitude of AC field was kept constant at $h_{ac}$=1 Oe. A
281: well-defined peak is observed for the real part of the
282: susceptibility $\chi'$ at 236 K and the low-temperature tail of this
283: peak decreases with increasing frequency. The imaginary part of the
284: susceptibility, $\chi''$, appears below $\sim$ 240 K, and consists
285: of two peaks. The high temperature component, appearing as a
286: shoulder, is located at $\sim$ 237 K and grows as frequency
287: increases, while the peak position remains the same. On the other
288: hand, the low temperature peak grows and shifts to higher
289: temperature with increasing frequency. Such a behavior is commonly
290: observed in spin glass systems.
291: 
292: \begin{figure}
293: \begin{center}
294: \includegraphics[angle=-90,width=3in]{lfofigure4.eps}
295: \end{center}
296: \caption{The dynamic scaling of the reduced temperature vs
297: $\tau_{max}(T_f)=2\pi / \omega$ in a log-log scale for sample A. The
298: solid line is Eq.~\ref{equ:n} with $z\nu=6.6$, $\tau_0=1\times
299: 10^{-13}$ s and $T_g=228.5$ K}
300: \label{fig:Fig.4}
301: \end{figure}
302: 
303: For a spin glass system, with decreasing temperature the spin
304: dynamics become sluggish, so that it takes longer time for a spin to
305: relax, and the maximum relaxation time increases accordingly. When
306: an external AC magnetic field with a driving frequency $\omega /
307: 2\pi$ is applied to a spin glass system, if the maximum relaxation
308: time $\tau$$_{max}$ is longer than $\omega / 2\pi$, the system will
309: not be able to keep up with the oscillating field and become out of
310: equilibrium. Therefore, one can define the freezing temperature,
311: $T_f$, as the temperature, at which $\tau_{max}=2\pi /\omega$. As a
312: result, $T_f$ is a function of driving frequency $\omega$.
313: Experimentally $T_f (\omega$) can be determined either from the
314: maximum of $\chi'(\omega$) or from the inflection point of $\chi''
315: (\omega)$.\cite{23,29} Since the maximum of $\chi'$ is difficult to
316: identify due to the second peak located at 236K, we use the
317: inflection point of $\chi'' (\omega)$ to determine $T_f$. The
318: maximum relaxation time and $T_f (\omega)$ can be modeled with
319: conventional critical slowing down \cite{13}
320: \begin{eqnarray}
321: \tau_{max}=\tau_0(T_f/T_g-1)^{-z\nu},\label{equ:n}
322: \end{eqnarray}
323: where $T_g$ is the spin-glass transition temperature, $z$ is the
324: dynamical exponent, $\nu$ is the usual critical exponent for the
325: correlation length and $\tau$$_0$ is the microscopic flipping time
326: of the fluctuating spins. The scaling of the AC susceptibility is
327: shown in Fig.~\ref{fig:Fig.4}, and the best fit to Eq.~(\ref{equ:n})
328: yields $T_g=228.5 \pm 0.5$ K, $z\nu=6.6 \pm 1.1$ and $\tau_0
329: =10^{-13.0 \pm 1.6}$ s. The value of $\tau_0$ is very close to the
330: microscopic spin flip time $\sim 10^{-13}$ seconds in other spin
331: glass systems. \cite{28,29} The value of $z\nu$ is within the range
332: of well-known spin-glasses such as CuMn(4.6 at.\%) ($z\nu$=5.5)
333: \cite{14} and CdCr$_2$(In)S$_4$ ($z\nu=7$). \cite{15} This value of
334: $z\nu$ is also close to the value obtained from numerical
335: simulations in three-dimensional (3D) Ising spin glasses.
336: \cite{16,17,18} This scaling analysis indicates that the low
337: temperature phase is quite possibly a spin-glass phase. Thus, taken
338: together with the heat capacity and the susceptibility data, LFO
339: seems to undergo a continuous phase transition from a Curie
340: paramagnetic phase to ferrimagnetically ordered phase at $\sim$ 236K
341: and then to a reentrant spin-glass phase below 228K.
342: 
343: 
344: 
345: 
346: %==============================================================================
347: 
348: \subsection{Nonlinear susceptibility}
349: \begin{figure}
350: \begin{center}
351: \includegraphics[angle=-90,width=2.8in]{lfofigure5.eps}
352: \end{center}
353: \caption{M vs H curves of LFO at different fixed temperatures
354: measured after cooling under a 200 Oe magnetic field. The solid
355: lines are the fitting results using Eq.~\ref{equ:m}.}
356: \label{fig:Fig.5}
357: \end{figure}
358: 
359: Figure ~\ref{fig:Fig.5} shows the magnetization (M) versus field (H)
360: curves at different temperatures after cooling under a 200 Oe
361: magnetic field. At low temperatures, since the zero-field
362: magnetization is non-zero, there is thermal remanent magnetization
363: (TRM), which almost vanishes above $\sim$232K. At high temperatures,
364: there is no TRM and the slope of the M vs H curve is largely
365: determined by linear susceptibility, which shows maximum at 236 K
366: (see Fig.~\ref{fig:Fig.1}). At low temperatures, the magnetization
367: initially shows negative slope due to the relaxational behavior of
368: spin-glass phase.
369: 
370: 
371: Nonlinear susceptibility is a valuable tool to study the critical
372: behavior of a magnetic phase transition. In particular, the
373: spin-glass susceptibility is believed to be directly proportional to
374: the non-linear susceptibility.\cite{44,45,46} If a system undergoes
375: a SG phase transition, the magnetization M can be expanded by odd
376: order of H.
377: 
378: \begin{eqnarray}
379: M=M_0+a_{1}H+a_{3}H^{3}+....\label{equ:m}
380: \end{eqnarray}
381: 
382: 
383: The data in Fig.~\ref{fig:Fig.5} (above 50 Oe) were fitted using
384: Eq.~(\ref{equ:m}), and the fitting results are shown as solid lines.
385: The constant offset $M_0$, linear susceptibility $a_1$ and the third
386: order nonlinear susceptibility $a_3$ obtained from the fits are
387: shown in Fig.~\ref{fig:Fig.6}. $M_0$ starts to grow below $\sim$235K
388: and then increases rapidly below $\sim$230K. The temperature at
389: which $M_0$ increases rapidly is close to $T_g$, suggesting that
390: $M_0$ behaves like the spin-glass order parameter. When the
391: temperature dependence of $M_0$ below 230K is fitted using $M_0 \sim
392: (T_g-T)^{\beta}$, best fits are obtained with $T_g=229$ K and
393: $\beta=1.0 \pm 0.2$, as shown by the solid line in
394: Fig.~\ref{fig:Fig.6}(a). This critical exponent $\beta$ is
395: consistent with that of canonical spin-glass systems.\cite{52,53}
396: The linear susceptibility peak in $a_1$ at $\sim$236K
397: (Fig.~\ref{fig:Fig.6}(b)) was also observed in DC magnetization and
398: heat capacity, and this indicates the existence of a magnetic phase
399: transition at $T_c \approx 236$K. If we fit the linear
400: susceptibility for $T>T_c$ to $a_1 \sim (T-T_c)^{-\gamma}$,
401: $\gamma=1.4 \pm 0.3$ is obtained, and the fitting result is plotted
402: as the solid line in Fig.~\ref{fig:Fig.6}(b). As temperature is
403: lowered, the third order nonlinear susceptibility $a_3$ shows a
404: negative minimum at 236K, and then it abruptly changes sign at 234K
405: and then show a broad positive peak at $\sim$230K. This behavior is
406: reminiscent of the recent observation that $a_3$ near a
407: ferromagnetic transition diverges negatively and positively as $T_c$
408: is approached from above and below, respectively.\cite{Nair03}
409: Therefore our observation of the zero-crossing of $a_3$ around 234K
410: seems to suggest that this high temperature magnetic transition is
411: described as ferrimagnetic (FM) transition. This assignment is also
412: consistent with earlier studies.\cite{10} The behavior of $a_3$
413: around 230K, however, is quite unusual for a spin-glass system. It
414: was observed that the non-linear susceptibility ($a_3$) is negative
415: with a cusp at the paramagnet to spin glass transition in Ref. 25
416: and 30. We observe broad positive peak around $T_g$, which may be
417: due to the reentrant nature of the spin-glass phase in LFO. In fact,
418: similarly unusual behavior have been observed in other reentrant
419: spin-glass systems.\cite{50,51} Our nonlinear susceptibility data
420: also supports the phase diagram suggested by AC susceptibility data.
421: \begin{figure}
422: \begin{center}
423: \includegraphics[angle=-90,width=3in]{lfofigure6.eps}
424: \end{center}
425: \caption{(a) Constant offset $M_0$. The solid line is the fitting
426: using $(T_g -T)^{\beta}$. (b) Linear susceptibility $a_1$. The data
427: above $T_c=235.6$K is fitted by $(T-T_c)^{-\gamma}$, as shown by the
428: solid line. (c) The third order of nonlinear susceptibility $a_3$.
429: $M_0$, $a_1$ and $a_3$ were obtained by fitting the data in
430: Fig.~\ref{fig:Fig.5} into Eq.~\ref{equ:m}. } \label{fig:Fig.6}
431: \end{figure}
432: 
433: 
434: 
435: 
436: 
437: 
438: %==============================================================================
439: \subsection{Non-equilibrium phenomena}
440: 
441: 
442: Although spins are considered ``frozen'' in the spin glass phase,
443: due to the slow dynamics, the system simply does not  reach the
444: equilibrium state within the experimental time scale. As a result,
445: the spin glass system exhibits non-equilibrium phenomena. One such
446: example is aging. When a spin-glass system is cooled below $T_g$,
447: the spin-glass domain grows. Since this domain growth occurs
448: logarithmically in time, it is customary to define the relaxation
449: rate $S \equiv (1/H){\partial M/\partial
450: \log(t)}$.\cite{Fischer-book} In Fig.~\ref{fig:Fig.7}, we show our
451: data for $S(t)$ as a function of $\log(t)$. Note that the sample was
452: cooled down to 0.87$T_g$ $\sim$ 200 K in the absence of magnetic
453: field. After waiting for a certain time ($t_w$=1000s, 5000s and
454: 10000s) without external field, the magnetization was recorded as a
455: function of time after a 10 Oe magnetic field was applied. As can
456: been seen from the figure, $t$ at which the maximum relaxation rate
457: occurs increases with increasing $t_w$, and in fact it is almost
458: equal to $t_w$. This kind of aging behavior illustrates
459: non-equilibrium dynamics of domain growth, and has been observed in
460: other spin glass systems.\cite{29,35}
461: 
462: \begin{figure}
463: \begin{center}
464: \includegraphics[angle=-90,width=3in]{lfofigure7.eps}
465: \end{center}
466: \caption{Relaxation rate $S$ defined in the text is plotted as a
467: function of $\log(t)$ at $T=0.87T_g$ ($T_g=228.5$ K). Each curve is
468: obtained by measuring at $H=10$ Oe after waiting for $t_w$ following
469: the cool down.}
470:  \label{fig:Fig.7}
471: \end{figure}
472: 
473: 
474: 
475: Another interesting example of non-equilibrium dynamics of
476: spin-glass system is the so-called memory effect. In order to show
477: this effect, we have measured temperature dependence of $M(T)$ in
478: two distinct routes. The first, $M_{ref}$, was obtained by cooling
479: in 10 Oe magnetic field from 300 K down to 50 K at a constant
480: cooling rate of 2 K/min and then heating back continuously at the
481: same rate. In the second route, M was recorded on cooling in 10 Oe
482: at the same rate from 300 K to 50 K with two halts at $T_1=160$ K
483: for 72000 s and at $T_2=200$ K for 48000 s. During the halts, the
484: external field is turned off to let the magnetization relax. After
485: each halt, M shows a clear deviation from the reference as
486: illustrated in Fig.~\ref{fig:Fig.8}, due to aging. After reaching 50
487: K, the sample temperature is increased continuously at 2 K/min rate
488: in $H = 10$ Oe. During the reheating, the system exhibits a
489: step-like feature at both $T_1$ and $T_2$. The jump at $T_1$ is not
490: very pronounced, but clear jump in $M(T)$ around $T_2$ is clearly
491: visible. This suggests that the system somehow remembers the history
492: of halts during cooling. Exceeding the halt points, $M$ recovers to
493: the reference value and the system is called rejuvenated. Such
494: aging, memory and rejuvenation behavior was observed in other spin
495: glass systems as well.\cite{28,42,43} These observations also
496: suggest that the low temperature phase of LFO is consistently
497: described as a spin-glass.
498: 
499: 
500: 
501: \begin{figure}
502: \begin{center}
503: \includegraphics[angle=-90,width=3in]{lfofigure8.eps}
504: \end{center}
505: \caption{ The relative magnetization $M-M_{ref}$ is plotted as a
506: function of temperature. The magnetization measured on continuous
507: cooling in $H=10$ Oe field, is plotted as solid symbols. During the
508: cooling, there were two halts at $T_1=0.7T_g$ and $T_2=0.87T_g$
509: ($T_g=228$ K). The open symbols denote the measurements done on the
510: reheating. The reference $M_{ref}$ was obtained by continuous
511: cooling and reheating in 10 Oe. The cooling and heating rate in both
512: measurements were 2 K/min.}
513:  \label{fig:Fig.8}
514: \end{figure}
515: 
516: %==============================================================================
517: \subsection{Magnetic field dependence}
518: 
519: 
520: 
521: In Fig.~\ref{fig:Fig.9}, AC susceptibility at 10 Hz driving
522: frequency is plotted as a function of temperature for different
523: external static magnetic fields H. As can be seen in the figure,
524: $\chi'$ is suppressed by the magnetic field. As the field increases,
525: the main peak of $\chi'$ decreases and a double-peak feature
526: emerges. The ferrimagnetic phase transition temperature determined
527: from specific heat and nonlinear susceptibility ($\sim$236 K) is
528: quite close to the position of the high temperature peak in $\chi'$,
529: which slightly increases with increasing field. The low temperature
530: peaks in $\chi'$ and $\chi''$ correspond to the spin-glass
531: transition. As the field increases, the low temperature peak in
532: $\chi'$ decreases and finally disappears under $\sim$ 1 T (as shown
533: in the inset). The peak in $\chi''$ also shifts to lower temperature
534: with increasing field. Under very high external field (above 1 T),
535: the spin-glass transition seems to be completely suppressed.
536: \begin{figure}
537: \begin{center}
538: \includegraphics[angle=0,width=2.9in]{lfofigure9.eps}
539: \end{center}
540: \caption{ (a) The real and (b) the imaginary part of AC magnetic
541: susceptibility versus temperature. The driving frequency was fixed
542: at $\omega/2\ pi=10$ Hz and $h_{ac}$=1 Oe. Each curve was obtained
543: under different applied static  magnetic field of $H$. The inset
544: shows the data obtained with $H=1$ T.}
545:  \label{fig:Fig.9}
546: \end{figure}
547: 
548: Figure~\ref{fig:Fig.10} shows the field versus temperature phase
549: diagram. Due to the large uncertainties associated with determining
550: transition temperature of two nearby phase transitions, the error
551: bars at low field is relatively large. However, the field dependence
552: of the PM-FM transition is very weak in this region, while it is
553: clear that the transition temperature gradually increases with
554: increasing field at high field. The spin-glass transition
555: temperature also exhibits substantial field dependence. The SG-FM
556: transition temperature is suppressed rapidly as a small field is
557: applied. In addition, a threshold field $h_0(\omega)\sim 200$ Oe is
558: observed, below which $T_g$ does not show a systematic change with
559: the change of the field. According to the mean field theory, there
560: exists a phase boundary in $H-T$ phase diagram called
561: de~Almeida-Thouless (AT) line,\cite{19} whereby a spin-glass phase
562: can only exist under this boundary (in the low field region). The AT
563: line is given by \cite{19}
564: 
565: \begin{eqnarray}
566: H\propto{(1-\frac{T_{g}(H)}{T_{g}(0)})^{3/2}}, \label{equ:p}
567: \end{eqnarray}
568: 
569: Here $H$ is the external magnetic field and $T_g(H)$ is the
570: field-dependent glass transition temperature.\cite{22} Our data fits
571: this relation very well as shown in Fig.~\ref{fig:Fig.10}, in which
572: a linear relationship between $T_g$ and $H^{2/3}$ is clearly
573: illustrated.
574: 
575: 
576: 
577: %==============================================================================
578: 
579: 
580: \section{Discussion}
581: We have presented a variety of experimental evidences showing that
582: the LFO sample goes through re-entrant spin glass transition around
583: 228 K. However, this observation is quite puzzling in several
584: aspects. The first is the microscopic origin of the spin-glass
585: behavior. Conventionally, disordered spin arrangements or
586: interactions (random site or random bond) are necessary to produce
587: magnetic frustration required for spin-glass behavior. However, this
588: system, LFO, is considered highly stoichiometric, and in addition,
589: there exists charge ordering below 300 K, which implies that the
590: arrangement of Fe$^{2+}$ and Fe$^{3+}$ spins are regular. Therefore,
591: this system seems to possess only geometrical frustration as a
592: necessary ingredient for spin glass behavior. Unless spin-glass
593: behavior can arise from pure geometrical frustration, one must find
594: the missing ``disorder'' in this system to explain the observed
595: spin-glass behavior.
596: \begin{figure}
597: \begin{center}
598: \includegraphics[angle=-90,width=3in]{lfofigure10.eps}
599: \end{center}
600: \caption{Field vs. Temperature phase diagram. In order to show AT
601: line, we plot $H^{2/3}$ versus $T_g (\omega)$. The open and closed
602: symbols denote the low and high temperature transition temperatures
603: as described in the text. The thick solid line is the linear fit to
604: the AT line (Eq.~\ref{equ:p})} \label{fig:Fig.10}
605: \end{figure}
606: 
607: Before discussing the disorder effect in this system, it is useful
608: to examine magnetic interactions present. In LFO, both the Fe$^{2+}$
609: and Fe$^{3+}$ ions are in their high spin configuration, with the
610: spin angular momentum S=2 and S=5/2, respectively. The exchange
611: interactions between the Fe$^{2+}$-Fe$^{2+}$ and Fe$^{3+}$-Fe$^{3+}$
612: are presumably antiferromagnetic through the superexchange path via
613: the oxygen ions. However, the magnetic interaction between the
614: Fe$^{2+}$ and Fe$^{3+}$ ions requires further consideration. Note
615: that the Fe$^{2+}$ is in $d^{6}$ configuration, and the Hund's rule
616: dictates that the extra electron in this ion, compared to the
617: Fe$^{3+}$ ($d^{5}$) ion should point in the opposite direction of
618: the rest of the ``$d^{5}$'' electrons. Therefore, one can expect the
619: interaction between the Fe$^{2+}$ and Fe$^{3+}$ to be ferromagnetic
620: based on a kinetic energy argument, analogous to the double exchange
621: mechanism in manganites.\cite{54} However, since this compound is
622: insulating at all temperatures, such ``extra'' electron cannot be
623: mobile, but presumably resides in a resonance state between the two
624: neighboring Fe ions.
625: 
626: 
627: One of the important degrees of freedom in this system, often
628: overlooked, is the orbital degrees of freedom. Since the Fe ions all
629: have trigonal bipyramidal crystal field environment, the 3d orbitals
630: will split into three levels. The highest energy level will be
631: $d_{z^{2}}$, and there will be two doubly degenerate orbitals
632: ($d_{xy}$ and $d_{x^{2}-y^{2}}$; $d_{xz}$ and $d_{yz}$) at lower
633: energies. Following the Hunds rule, one can easily see that
634: Fe$^{3+}$ ion has $d^{5}$ configuration and is isotropic. However,
635: Fe$^{2+}$ ion has $d^{6}$ configuration and the lowest energy
636: orbitals will have orbital degeneracy. In order to break this
637: orbital degeneracy, there could be a cooperative Jahn-Teller
638: distortion of Fe$^{2+}$ ions, once the charge order sets in. Taken
639: together, it is conceivable that two neighboring Fe$^{2+}$ and
640: Fe$^{3+}$ ions form a ``dimer'', sharing a minority spin electron.
641: In fact, such bond dimerization scenario has been considered in
642: their study of mixed valence B-site Fe ions in Fe$_3$O$_4$ by Seo
643: and coworkers.\cite{55} Such bond dimerization will break the
644: orbital degeneracy, and make the dimer spins form a highly
645: frustrated Kagome lattice, which may be responsible for the observed
646: spin glass behavior. Further structural investigation of this system
647: is required to address this conjecture.
648: 
649: Another important issue is the oxygen non-stoichiometry. If there
650: exists oxygen non-stoichiometry, this will necessarily affect the
651: ratio of Fe$^{2+}$ and Fe$^{3+}$ spins, leading to disorder in
652: charge order, and possibly disorder in the exchange interactions. In
653: fact, our preliminary studies suggest that the oxygen contents seem
654: to change magnetic properties quite dramatically, possibly
655: explaining different magnetic behavior of $\rm LuFe_2O_4$ compounds
656: reported in the literature.
657: 
658: 
659: Another possibility is the charge-ordering model itself. If the
660: charge ordering at 300 K is not of second-order kind, and has some
661: relaxational component, such as charge glass type ordering, then
662: this will naturally affect the spin ordering part. For example, one
663: can imagine charge-ordered domains exhibiting relaxor type behavior
664: and influence dielectric and magnetic behavior at lower
665: temperatures. Our preliminary x-ray scattering experiments also
666: suggest that the charge ordering in this system is quite complicated
667: and has nontrivial temperature dependence.
668: 
669: Another surprising aspect of the spin-glass transition is its large
670: temperature scale. Almost all known spin-glass systems occur at low
671: temperatures. This is due to the fact that the lower critical
672: dimension of the spin glass transition is believed to be between 2
673: and 3. Since a 3D SG system is very close to the lower critical
674: dimension, its ordering temperature is very close to absolute zero.
675: In that sense this observation of spin glass behavior at
676: temperatures above 200 K is not only unusual, but very surprising
677: for a quasi 2D system. Again, if the glassy nature of the system
678: arises from the charge sector, this may provide a natural
679: explanation.
680: 
681: %==============================================================================
682: 
683: \section{Summary}
684: 
685: The magnetic properties of $\rm LuFe_2O_4$ single crystals were
686: investigated with DC magnetization and AC susceptibility. Based on
687: the dynamic scaling of AC susceptibility and the behavior of
688: non-linear susceptibility, it is suggested that $\rm LuFe_2O_4$ goes
689: through first a ferrimagnetic ordering at 236 K, and then
690: subsequently goes through a reentrant spin-glass transition at $\sim
691: 228.5$ K. Typical properties of spin glass system, such as aging,
692: memory, and rejuvenation have been also observed in this low
693: temperature phase. The field dependence of the spin glass transition
694: temperature is described well by the well-known de~Almeida-Thouless
695: theory. It was also observed that the ferrimagnetic transition
696: temperature shows quite sizable field dependence. In order to
697: understand the origin of the spin glass behavior in this compound,
698: possibilities based on the frustration, oxygen non-stoichiometry,
699: and glassy charge ordering have been discussed.
700: 
701: 
702: %==============================================================================
703: 
704: \begin{acknowledgements}
705: 
706: We would like to thank  David Ellis, S. M. Shapiro, G. Xu, J.
707: Brittain, A. Gershon and H. Zhang for invaluable discussions. The
708: work at University of Toronto was supported by Natural Sciences and
709: Engineering Research Council of Canada, Canadian Foundation for
710: Innovation, Ontario Innovation Trust, and Early Researcher Award by
711: Ontario Ministry of Research and Innovation. The work at Brookhaven
712: was supported by the U. S. Department of Energy, Office of Science.
713: 
714: \end{acknowledgements}
715: 
716: 
717: %\bibliography{lfo}
718: \begin{thebibliography}{44}
719: \expandafter\ifx\csname
720: natexlab\endcsname\relax\def\natexlab#1{#1}\fi
721: \expandafter\ifx\csname bibnamefont\endcsname\relax
722:   \def\bibnamefont#1{#1}\fi
723: \expandafter\ifx\csname bibfnamefont\endcsname\relax
724:   \def\bibfnamefont#1{#1}\fi
725: \expandafter\ifx\csname citenamefont\endcsname\relax
726:   \def\citenamefont#1{#1}\fi
727: \expandafter\ifx\csname url\endcsname\relax
728:   \def\url#1{\texttt{#1}}\fi
729: \expandafter\ifx\csname urlprefix\endcsname\relax\def\urlprefix{URL
730: }\fi \providecommand{\bibinfo}[2]{#2}
731: \providecommand{\eprint}[2][]{\url{#2}}
732: 
733: \bibitem[{\citenamefont{Kimizuka et~al.}(1990)\citenamefont{Kimizuka,
734:   Muromachi, and Siratori}}]{1}
735: \bibinfo{author}{\bibfnamefont{N.}~\bibnamefont{Kimizuka}},
736:   \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{Muromachi}},
737:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Siratori}},
738:   \emph{\bibinfo{title}{Handbook on the Physics and Chemistry of Rare Earths}}
739:   (\bibinfo{publisher}{Elsevier Science}, \bibinfo{address}{Amsterdam},
740:   \bibinfo{year}{1990}), vol.~\bibinfo{volume}{13}, pp.
741:   \bibinfo{pages}{283--384}.
742: 
743: \bibitem[{\citenamefont{Kimizuka and Katsura}(1975)}]{2}
744: \bibinfo{author}{\bibfnamefont{N.}~\bibnamefont{Kimizuka}} \bibnamefont{and}
745:   \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Katsura}},
746:   \bibinfo{journal}{J. Solid State Chem.} \textbf{\bibinfo{volume}{13}},
747:   \bibinfo{pages}{176} (\bibinfo{year}{1975}).
748: 
749: \bibitem[{\citenamefont{Hamilton}(1958)}]{30}
750: \bibinfo{author}{\bibfnamefont{W.~C.} \bibnamefont{Hamilton}},
751:   \bibinfo{journal}{Phys. Rev.} \textbf{\bibinfo{volume}{110}},
752:   \bibinfo{pages}{1050} (\bibinfo{year}{1958}).
753: 
754: \bibitem[{\citenamefont{Garc\'{\i}a and Sub\'{\i}as}(2004)}]{31}
755: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Garc\'{\i}a}} \bibnamefont{and}
756:   \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Sub\'{\i}as}},
757:   \bibinfo{journal}{J. Phys: Condens. Matter} \textbf{\bibinfo{volume}{16}},
758:   \bibinfo{pages}{R145} (\bibinfo{year}{2004}).
759: 
760: \bibitem[{\citenamefont{Tomioka et~al.}(1995)\citenamefont{Tomioka, Asamitsu,
761:   Moritomo, Kuwahara, and Tokura}}]{32}
762: \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Tomioka}},
763:   \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Asamitsu}},
764:   \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Moritomo}},
765:   \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Kuwahara}}, \bibnamefont{and}
766:   \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Tokura}},
767:   \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{74}},
768:   \bibinfo{pages}{5108} (\bibinfo{year}{1995}).
769: 
770: \bibitem[{\citenamefont{Yamada et~al.}(1997)\citenamefont{Yamada, Shinichiro,
771:   and Ikeda}}]{3}
772: \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Yamada}},
773:   \bibinfo{author}{\bibfnamefont{N.}~\bibnamefont{Shinichiro}},
774:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{N.}~\bibnamefont{Ikeda}},
775:   \bibinfo{journal}{J. Phys. Soc. Japan} \textbf{\bibinfo{volume}{66}},
776:   \bibinfo{pages}{3733} (\bibinfo{year}{1997}).
777: 
778: \bibitem[{\citenamefont{Ikeda et~al.}(2005)\citenamefont{Ikeda, Ohsumi, Ohwada,
779:   Ishii, Inami, Kakurai, Murakami, K.Yoshii, Mori, Horibe et~al.}}]{4}
780: \bibinfo{author}{\bibfnamefont{N.}~\bibnamefont{Ikeda}},
781:   \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Ohsumi}},
782:   \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Ohwada}},
783:   \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Ishii}},
784:   \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Inami}},
785:   \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Kakurai}},
786:   \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Murakami}},
787:   \bibinfo{author}{\bibnamefont{K.Yoshii}},
788:   \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Mori}},
789:   \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Horibe}},
790:   \bibnamefont{et~al.}, \bibinfo{journal}{Nature}
791:   \textbf{\bibinfo{volume}{436}}, \bibinfo{pages}{1136} (\bibinfo{year}{2005}).
792: 
793: \bibitem[{\citenamefont{Zhang et~al.}(2007)\citenamefont{Zhang, Yang, Ma, Tian,
794:   and Li}}]{56}
795: \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Zhang}},
796:   \bibinfo{author}{\bibfnamefont{H.~X.} \bibnamefont{Yang}},
797:   \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Ma}},
798:   \bibinfo{author}{\bibfnamefont{H.~F.} \bibnamefont{Tian}}, \bibnamefont{and}
799:   \bibinfo{author}{\bibfnamefont{J.~Q.} \bibnamefont{Li}},
800:   \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{98}},
801:   \bibinfo{pages}{247602} (\bibinfo{year}{2007}).
802: 
803: \bibitem[{\citenamefont{Xiang and Whangbo}(2007)}]{57}
804: \bibinfo{author}{\bibfnamefont{H.~J.} \bibnamefont{Xiang}} \bibnamefont{and}
805:   \bibinfo{author}{\bibfnamefont{M.-H.} \bibnamefont{Whangbo}},
806:   \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{98}},
807:   \bibinfo{pages}{246403} (\bibinfo{year}{2007}).
808: 
809: \bibitem[{\citenamefont{Subramanian et~al.}(2006)\citenamefont{Subramanian, He,
810:   Chen, Rogado, Calvarese, and Sleight}}]{58}
811: \bibinfo{author}{\bibfnamefont{M.~A.} \bibnamefont{Subramanian}},
812:   \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{He}},
813:   \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Chen}},
814:   \bibinfo{author}{\bibfnamefont{N.~S.} \bibnamefont{Rogado}},
815:   \bibinfo{author}{\bibfnamefont{T.~G.} \bibnamefont{Calvarese}},
816:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{A.~W.}
817:   \bibnamefont{Sleight}}, \bibinfo{journal}{Adv. Mater.}
818:   \textbf{\bibinfo{volume}{18}}, \bibinfo{pages}{1737} (\bibinfo{year}{2006}).
819: 
820: \bibitem[{\citenamefont{Tanaka et~al.}(1979)\citenamefont{Tanaka, Kato,
821:   Kimizuka, and Siratori}}]{5}
822: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Tanaka}},
823:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Kato}},
824:   \bibinfo{author}{\bibfnamefont{N.}~\bibnamefont{Kimizuka}}, \bibnamefont{and}
825:   \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Siratori}},
826:   \bibinfo{journal}{J. Phys. Soc. Japan} \textbf{\bibinfo{volume}{47}},
827:   \bibinfo{pages}{1737} (\bibinfo{year}{1979}).
828: 
829: \bibitem[{\citenamefont{Tanaka et~al.}(1982)\citenamefont{Tanaka, Akimitsu,
830:   Inada, Kimizuka, Shindo, and Siratori}}]{6}
831: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Tanaka}},
832:   \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Akimitsu}},
833:   \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Inada}},
834:   \bibinfo{author}{\bibfnamefont{N.}~\bibnamefont{Kimizuka}},
835:   \bibinfo{author}{\bibfnamefont{I.}~\bibnamefont{Shindo}}, \bibnamefont{and}
836:   \bibinfo{author}{\bibfnamefont{N.}~\bibnamefont{Siratori}},
837:   \bibinfo{journal}{Solid State Commun.} \textbf{\bibinfo{volume}{44}},
838:   \bibinfo{pages}{687} (\bibinfo{year}{1982}).
839: 
840: \bibitem[{\citenamefont{Nakagawa et~al.}(1979)\citenamefont{Nakagawa, Inazumi,
841:   Kimizuka, and Siratori}}]{7}
842: \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Nakagawa}},
843:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Inazumi}},
844:   \bibinfo{author}{\bibfnamefont{N.}~\bibnamefont{Kimizuka}}, \bibnamefont{and}
845:   \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Siratori}},
846:   \bibinfo{journal}{J. Phys. Soc. Japan} \textbf{\bibinfo{volume}{47}},
847:   \bibinfo{pages}{1369} (\bibinfo{year}{1979}).
848: 
849: \bibitem[{\citenamefont{Ikeda et~al.}(2002)\citenamefont{Ikeda, Mori, Kohn,
850:   Mizumaki, and Akao}}]{8}
851: \bibinfo{author}{\bibfnamefont{N.}~\bibnamefont{Ikeda}},
852:   \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Mori}},
853:   \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Kohn}},
854:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Mizumaki}}, \bibnamefont{and}
855:   \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Akao}},
856:   \bibinfo{journal}{Ferroelectrics} \textbf{\bibinfo{volume}{272}},
857:   \bibinfo{pages}{309} (\bibinfo{year}{2002}).
858: 
859: \bibitem[{\citenamefont{Nakagawa et~al.}(1981)\citenamefont{Nakagawa, Kishi,
860:   Hiroyoshi, Kimizuka, and Siratori}}]{9}
861: \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Nakagawa}},
862:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Kishi}},
863:   \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Hiroyoshi}},
864:   \bibinfo{author}{\bibfnamefont{N.}~\bibnamefont{Kimizuka}}, \bibnamefont{and}
865:   \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Siratori}},
866:   \emph{\bibinfo{title}{Ferrites, Pro. 3rd Int. Conf. Ferrites, Kyoto}}
867:   (\bibinfo{publisher}{CAPJ}, \bibinfo{address}{Tokyo}, \bibinfo{year}{1981}),
868:   p. \bibinfo{pages}{115}.
869: 
870: \bibitem[{\citenamefont{Iida et~al.}(1993)\citenamefont{Iida, Tanaka, Nakagawa,
871:   Funahashi, Kimizuka, and Takekawa}}]{10}
872: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Iida}},
873:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Tanaka}},
874:   \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Nakagawa}},
875:   \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Funahashi}},
876:   \bibinfo{author}{\bibfnamefont{N.}~\bibnamefont{Kimizuka}}, \bibnamefont{and}
877:   \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Takekawa}},
878:   \bibinfo{journal}{J. Phys. Soc. Japan} \textbf{\bibinfo{volume}{62}},
879:   \bibinfo{pages}{1723} (\bibinfo{year}{1993}).
880: 
881: \bibitem[{\citenamefont{Nagai et~al.}()\citenamefont{Nagai, Matsuda, Ishii,
882:   Kakurai, Kito, Ikeda, and Yamada}}]{38}
883: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Nagai}},
884:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Matsuda}},
885:   \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Ishii}},
886:   \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Kakurai}},
887:   \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Kito}},
888:   \bibinfo{author}{\bibfnamefont{N.}~\bibnamefont{Ikeda}}, \bibnamefont{and}
889:   \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Yamada}},
890:   \bibinfo{note}{unpublished}.
891: 
892: \bibitem[{\citenamefont{Christianson et~al.}()\citenamefont{Christianson,
893:   Lumsden, Angst, Yamani, Tian, Jin, Payzant, Nagler, Sales, and
894:   Mandrus}}]{neutron-new}
895: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Christianson}},
896:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Lumsden}},
897:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Angst}},
898:   \bibinfo{author}{\bibfnamefont{Z.}~\bibnamefont{Yamani}},
899:   \bibinfo{author}{\bibfnamefont{W.}~\bibnamefont{Tian}},
900:   \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Jin}},
901:   \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{Payzant}},
902:   \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Nagler}},
903:   \bibinfo{author}{\bibfnamefont{B.}~\bibnamefont{Sales}}, \bibnamefont{and}
904:   \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Mandrus}},
905:   \eprint{arXiv:0711.3560v1}.
906: 
907: \bibitem[{\citenamefont{Iida et~al.}(1990)\citenamefont{Iida, Takekawa, and
908:   Kimizuka}}]{33}
909: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Iida}},
910:   \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Takekawa}}, \bibnamefont{and}
911:   \bibinfo{author}{\bibfnamefont{N.}~\bibnamefont{Kimizuka}},
912:   \bibinfo{journal}{J. Cryst Growth} \textbf{\bibinfo{volume}{102}},
913:   \bibinfo{pages}{398} (\bibinfo{year}{1990}).
914: 
915: \bibitem[{\citenamefont{Mattsson et~al.}(1995)\citenamefont{Mattsson, Jonsson,
916:   Nordblad, ArugaKatori, and Ito}}]{23}
917: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Mattsson}},
918:   \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Jonsson}},
919:   \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Nordblad}},
920:   \bibinfo{author}{\bibfnamefont{H.} \bibnamefont{ArugaKatori}},
921:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Ito}},
922:   \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{74}},
923:   \bibinfo{pages}{4305} (\bibinfo{year}{1995}).
924: 
925: \bibitem[{\citenamefont{Gunnarsson et~al.}(1992)\citenamefont{Gunnarsson,
926:   Svedlindh, Andersson, Nordblad, Lundgren, ArugaKatori, and Ito}}]{29}
927: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Gunnarsson}},
928:   \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Svedlindh}},
929:   \bibinfo{author}{\bibfnamefont{J.~O.} \bibnamefont{Andersson}},
930:   \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Nordblad}},
931:   \bibinfo{author}{\bibfnamefont{L.}~\bibnamefont{Lundgren}},
932:   \bibinfo{author}{\bibfnamefont{H.} \bibnamefont{ArugaKatori}},
933:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Ito}},
934:   \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{46}},
935:   \bibinfo{pages}{8227} (\bibinfo{year}{1992}).
936: 
937: \bibitem[{\citenamefont{Hohenberg and Halperin}(1977)}]{13}
938: \bibinfo{author}{\bibfnamefont{P.~C.} \bibnamefont{Hohenberg}}
939:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{B.~I.}
940:   \bibnamefont{Halperin}}, \bibinfo{journal}{Rev. Mod. Phys}
941:   \textbf{\bibinfo{volume}{49}}, \bibinfo{pages}{435} (\bibinfo{year}{1977}).
942: 
943: \bibitem[{\citenamefont{Mathieu et~al.}(2004)\citenamefont{Mathieu, Akahoshi,
944:   Asamitsu, Tomioka, and Y.Tokura}}]{28}
945: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Mathieu}},
946:   \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Akahoshi}},
947:   \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Asamitsu}},
948:   \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Tomioka}}, \bibnamefont{and}
949:   \bibinfo{author}{\bibnamefont{Y.Tokura}}, \bibinfo{journal}{Phys. Rev. Lett.}
950:   \textbf{\bibinfo{volume}{93}}, \bibinfo{pages}{227202}
951:   (\bibinfo{year}{2004}).
952: 
953: \bibitem[{\citenamefont{Mydosh}(1993)}]{14}
954: \bibinfo{author}{\bibfnamefont{J.~A.} \bibnamefont{Mydosh}},
955:   \emph{\bibinfo{title}{Spin Glass: An Experimental Introduction}}
956:   (\bibinfo{publisher}{Taylor $\&$ Francis}, \bibinfo{address}{London},
957:   \bibinfo{year}{1993}).
958: 
959: \bibitem[{\citenamefont{Vincent and Hammann}(1987)}]{15}
960: \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{Vincent}} \bibnamefont{and}
961:   \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Hammann}},
962:   \bibinfo{journal}{J. Phys. C} \textbf{\bibinfo{volume}{20}},
963:   \bibinfo{pages}{2659} (\bibinfo{year}{1987}).
964: 
965: \bibitem[{\citenamefont{Campbell}(1988)}]{16}
966: \bibinfo{author}{\bibfnamefont{I.~A.} \bibnamefont{Campbell}},
967:   \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{37}},
968:   \bibinfo{pages}{9800} (\bibinfo{year}{1988}).
969: 
970: \bibitem[{\citenamefont{Ogielski}(1985)}]{17}
971: \bibinfo{author}{\bibfnamefont{A.~T.} \bibnamefont{Ogielski}},
972:   \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{32}},
973:   \bibinfo{pages}{7384} (\bibinfo{year}{1985}).
974: 
975: \bibitem[{\citenamefont{Ogielski and Morgenstern}(1985)}]{18}
976: \bibinfo{author}{\bibfnamefont{A.~T.} \bibnamefont{Ogielski}} \bibnamefont{and}
977:   \bibinfo{author}{\bibfnamefont{I.}~\bibnamefont{Morgenstern}},
978:   \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{54}},
979:   \bibinfo{pages}{928} (\bibinfo{year}{1985}).
980: 
981: \bibitem[{\citenamefont{Ramirez et~al.}(1990)\citenamefont{Ramirez, Espinosa,
982:   and Cooper}}]{44}
983: \bibinfo{author}{\bibfnamefont{A.~P.} \bibnamefont{Ramirez}},
984:   \bibinfo{author}{\bibfnamefont{G.~P.} \bibnamefont{Espinosa}},
985:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{A.~S.}
986:   \bibnamefont{Cooper}}, \bibinfo{journal}{Phys. Rev. Lett.}
987:   \textbf{\bibinfo{volume}{64}}, \bibinfo{pages}{2070} (\bibinfo{year}{1990}).
988: 
989: \bibitem[{\citenamefont{Gingras et~al.}(1996)\citenamefont{Gingras, Stager,
990:   Gaulin, Raju, and Greedan}}]{45}
991: \bibinfo{author}{\bibfnamefont{M.~J.~P.} \bibnamefont{Gingras}},
992:   \bibinfo{author}{\bibfnamefont{C.~V.} \bibnamefont{Stager}},
993:   \bibinfo{author}{\bibfnamefont{B.~D.} \bibnamefont{Gaulin}},
994:   \bibinfo{author}{\bibfnamefont{N.~P.} \bibnamefont{Raju}}, \bibnamefont{and}
995:   \bibinfo{author}{\bibfnamefont{J.~E.} \bibnamefont{Greedan}},
996:   \bibinfo{journal}{J. Appl. Phys} \textbf{\bibinfo{volume}{79}},
997:   \bibinfo{pages}{6170} (\bibinfo{year}{1996}).
998: 
999: \bibitem[{\citenamefont{Rivadulla et~al.}(2004)\citenamefont{Rivadulla,
1000:   L\'{o}pez-Quintela, and Rivas}}]{46}
1001: \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Rivadulla}},
1002:   \bibinfo{author}{\bibfnamefont{M.~A.} \bibnamefont{L\'{o}pez-Quintela}},
1003:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Rivas}},
1004:   \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{93}},
1005:   \bibinfo{pages}{167206} (\bibinfo{year}{2004}).
1006: 
1007: \bibitem[{\citenamefont{Bouchiat}(1986)}]{52}
1008: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Bouchiat}},
1009:   \bibinfo{journal}{J.Phys. (Paris)} \textbf{\bibinfo{volume}{47}},
1010:   \bibinfo{pages}{71} (\bibinfo{year}{1986}).
1011: 
1012: \bibitem[{\citenamefont{Wakimoto et~al.}(2000)\citenamefont{Wakimoto, Ueki, Endoh, and Yamada}}]{53}
1013: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Wakimoto}},
1014:   \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Ueki}},
1015:   \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Endoh}}, \bibnamefont{and}
1016:   \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Yamada}},
1017:   \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{62}},
1018:   \bibinfo{pages}{3547} (\bibinfo{year}{2000}).
1019: 
1020: \bibitem[{\citenamefont{Nair and Banerjee}(2003)}]{Nair03}
1021: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Nair}} \bibnamefont{and}
1022:   \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Banerjee}},
1023:   \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{68}},
1024:   \bibinfo{pages}{094408} (\bibinfo{year}{2003}).
1025: 
1026: \bibitem[{\citenamefont{Suzuki and Suzuki}(2006)}]{50}
1027: \bibinfo{author}{\bibfnamefont{I.~S.} \bibnamefont{Suzuki}} \bibnamefont{and}
1028:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Suzuki}},
1029:   \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{73}},
1030:   \bibinfo{pages}{94448} (\bibinfo{year}{2006}).
1031: 
1032: \bibitem[{\citenamefont{Sato et~al.}(2001)\citenamefont{Sato, Ando, Ogawa,
1033:   Morimoto, and Ito}}]{51}
1034: \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Sato}},
1035:   \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Ando}},
1036:   \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Ogawa}},
1037:   \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Morimoto}}, \bibnamefont{and}
1038:   \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Ito}},
1039:   \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{64}},
1040:   \bibinfo{pages}{184432} (\bibinfo{year}{2001}).
1041: 
1042: \bibitem[{\citenamefont{Fischer and Hertz}(1991)}]{Fischer-book}
1043: \bibinfo{author}{\bibfnamefont{K.~H.} \bibnamefont{Fischer}} \bibnamefont{and}
1044:   \bibinfo{author}{\bibfnamefont{J.~A.} \bibnamefont{Hertz}},
1045:   \emph{\bibinfo{title}{Spin glasses}} (\bibinfo{publisher}{Cambridge
1046:   University Press}, \bibinfo{address}{Cambridge, UK}, \bibinfo{year}{1991}).
1047: 
1048: \bibitem[{\citenamefont{Lundgren et~al.}(1985)\citenamefont{Lundgren, Nordblad,
1049:   Svedlindh, and Beckman}}]{35}
1050: \bibinfo{author}{\bibfnamefont{L.}~\bibnamefont{Lundgren}},
1051:   \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Nordblad}},
1052:   \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Svedlindh}},
1053:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{O.}~\bibnamefont{Beckman}},
1054:   \bibinfo{journal}{J. Appl. Phys.} \textbf{\bibinfo{volume}{57}},
1055:   \bibinfo{pages}{3371} (\bibinfo{year}{1985}).
1056: 
1057: \bibitem[{\citenamefont{Vincent et~al.}(2000)\citenamefont{Vincent, Dupuis,
1058:   Alba, Hammann, and Bouchaud}}]{42}
1059: \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{Vincent}},
1060:   \bibinfo{author}{\bibfnamefont{V.}~\bibnamefont{Dupuis}},
1061:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Alba}},
1062:   \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Hammann}}, \bibnamefont{and}
1063:   \bibinfo{author}{\bibfnamefont{J.~P.} \bibnamefont{Bouchaud}},
1064:   \bibinfo{journal}{Europhys. Lett.} \textbf{\bibinfo{volume}{50}},
1065:   \bibinfo{pages}{674} (\bibinfo{year}{2000}).
1066: 
1067: \bibitem[{\citenamefont{Jonason et~al.}(1998)\citenamefont{Jonason, Vincent,
1068:   Hammann, Bouchaud, and Nordblad}}]{43}
1069: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Jonason}},
1070:   \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{Vincent}},
1071:   \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Hammann}},
1072:   \bibinfo{author}{\bibfnamefont{J.~P.} \bibnamefont{Bouchaud}},
1073:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Nordblad}},
1074:   \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{81}},
1075:   \bibinfo{pages}{3243} (\bibinfo{year}{1998}).
1076: 
1077: \bibitem[{\citenamefont{de. Almeida and Thouless}(1978)}]{19}
1078: \bibinfo{author}{\bibfnamefont{J.~R.~L.} \bibnamefont{de. Almeida}}
1079:   \bibnamefont{and} \bibinfo{author}{\bibfnamefont{D.~J.}
1080:   \bibnamefont{Thouless}}, \bibinfo{journal}{J. Phys. A}
1081:   \textbf{\bibinfo{volume}{11}}, \bibinfo{pages}{983} (\bibinfo{year}{1978}).
1082: 
1083: \bibitem[{\citenamefont{Katori and Ito}(1994)}]{22}
1084: \bibinfo{author}{\bibfnamefont{H.~A.} \bibnamefont{Katori}} \bibnamefont{and}
1085:   \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Ito}}, \bibinfo{journal}{J.
1086:   Phys. Soc. Japan} \textbf{\bibinfo{volume}{63}}, \bibinfo{pages}{3122}
1087:   (\bibinfo{year}{1994}).
1088: 
1089: \bibitem[{\citenamefont{Zener}(1951)}]{54}
1090: \bibinfo{author}{\bibfnamefont{C.}~\bibnamefont{Zener}},
1091:   \bibinfo{journal}{Phys. Rev.} \textbf{\bibinfo{volume}{82}},
1092:   \bibinfo{pages}{403} (\bibinfo{year}{1951}).
1093: 
1094: \bibitem[{\citenamefont{Seo et~al.}(2002)\citenamefont{Seo, Ogata, and
1095:   Fukuyama}}]{55}
1096: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Seo}},
1097:   \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Ogata}}, \bibnamefont{and}
1098:   \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Fukuyama}},
1099:   \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{65}},
1100:   \bibinfo{pages}{85107} (\bibinfo{year}{2002}).
1101: 
1102: \end{thebibliography}
1103: 
1104: 
1105: \end{document}
1106: