1: \documentclass[prb,twocolumn,showpacs,preprintnumbers,amsmath,amssymb]{revtex4}
2: \usepackage{graphicx}% Include figure files
3: \usepackage{dcolumn}% Align table columns on decimal point
4:
5: \newcommand{\rhoCEF}{\mbox{$\rho_{\mathrm{CEF}}$}}
6:
7: \begin{document}
8:
9: \title{Magnetoresistance of Pr$_{1-x}$La$_x$Os$_4$Sb$_{12}$: Disentangling
10: local crystalline-electric-field physics and lattice effects}
11:
12: \author{C.\ R.\ Rotundu}
13: \affiliation{ Department of Physics, University of Florida
14: P.O.\ Box 118440, Gainesville, Florida 32611--8440, USA}
15:
16: \author{K.\ Ingersent}
17: \affiliation{ Department of Physics, University of Florida
18: P.O.\ Box 118440, Gainesville, Florida 32611--8440, USA}
19:
20: \author{B.\ Andraka}
21: \email{andraka@phys.ufl.edu}
22: \affiliation{ Department of Physics, University of Florida
23: P.O.\ Box 118440, Gainesville, Florida 32611--8440, USA}
24:
25: \date{\today}
26:
27: \begin{abstract}
28: Resistivity measurements were performed on Pr$_{1-x}$La$_x$Os$_4$Sb$_{12}$
29: single crystals at temperatures down to 20\,mK and in fields up to 18\,T.
30: The results for dilute-Pr samples ($x=0.3$ and 0.67) are consistent with
31: model calculations performed assuming a singlet crystalline-electric-field
32: (CEF) ground state. The residual resistivity of these crystals features a
33: smeared step centered around 9\,T, the predicted crossing field for the
34: lowest CEF levels. The CEF contribution to the magnetoresistance has a
35: weaker-than-calculated dependence on the field direction, suggesting that
36: interactions omitted from the CEF model lead to avoided crossing in the
37: effective levels of the Pr$^{3+}$ ion. The dome-shaped magnetoresistance
38: observed for $x = 0$ and 0.05 cannot be reproduced by the CEF model, and
39: likely results from fluctuations in the field-induced antiferroquadrupolar
40: phase.
41: \end{abstract}
42:
43: \pacs{74.25.Ha, 74.70.Tx}
44:
45: \maketitle
46:
47: \section{Introduction}
48:
49: PrOs$_4$Sb$_{12}$, the first discovered Pr-based heavy fermion and
50: superconductor,\cite{Bauer} remains a focus of extensive theoretical and
51: experimental investigation. Its significance lies in the fact that the origin
52: of the heavy-fermion behavior is associated with non-Kramers $f$-electron ions,
53: for which the conventional Kondo effect seems unlikely. Our previous
54: specific-heat results in magnetic fields\cite{Rotundu} established that the
55: crystalline-electric-field (CEF) ground state is a nonmagnetic $\Gamma_1$
56: singlet. The field dependence of the CEF Schottky anomaly for fields greater
57: than 14\,T is clearly inconsistent with the alternative scenario of a
58: nonmagnetic $\Gamma_3$ doublet ground state. This conclusion was independent
59: of whether the exact $T_h$ point-group symmetry or the higher (approximate)
60: $O_h$ symmetry was assumed for the Pr sites.\cite{Takegahara}
61: The singlet nature of the CEF ground state was subsequently confirmed by
62: inelastic neutron scattering measurements and their analysis within the
63: $T_h$ symmetry scheme.\cite{Goremychkin}
64:
65: Despite the overwhelming evidence in favor of a singlet CEF ground state,
66: there are experimental results for PrOs$_4$Sb$_{12}$ that seem to be
67: better understood in terms of a doublet ground state.
68: For example, the magnetoresistance\cite{Bauer,Frederick,Lacerda,Sugawara} at
69: 1.4\,K exhibits a dome-like shape that is consistent with model calculations
70: of the CEF resistivity for a $\Gamma_3$ ground state and inconsistent with
71: similar calculations for a $\Gamma_1$ ground state.\cite{Frederick}
72: However, the CEF resistivity is a single-ion property that might be strongly
73: affected in PrOs$_4$Sb$_{12}$ by lattice coherence and by strong quadrupolar
74: and exchange interactions.
75: To probe this possibility, we have performed magnetoresistance measurements
76: on single-crystal Pr$_{1-x}$La$_x$Os$_4$Sb$_{12}$, in which lattice
77: translational symmetry is broken and intersite effects should be weaker than
78: in the pure compound.
79: Based on previously published magnetic susceptibility and specific heat
80: results,\cite{Rotundu2} we do not expect significant changes in CEF energies
81: (and eigenstates) of Pr upon doping with La. In addition, we have extended
82: magnetoresistance measurements of the undoped material down to 20\,mK.
83:
84: We find that the magnetoresistance of pure PrOs$_4$Sb$_{12}$ at 20\,mK is
85: inconsistent with model calculations for either the $\Gamma_3$ or the
86: $\Gamma_1$ CEF ground state, and conclude that the dome feature most probably
87: results from fluctuations in the field-induced antiferroquadrupolar (AFQ)
88: phase. On dilution of the Pr lattice with La, the dome in the magnetoresistance
89: is replaced by a smeared step that is consistent with the picture of a
90: $\Gamma_1$ singlet CEF ground state but not with a $\Gamma_3$ doublet.
91: The dependence of the $f$-electron contribution to the magnetoresistance
92: on the direction of the magnetic field is smaller than is predicted
93: theoretically based on a CEF model. This discrepancy suggests that
94: interactions omitted from the CEF model lead to avoided crossing in the
95: effective levels of the Pr$^{3+}$ ion.
96:
97: \section{Methods}
98:
99: Results are presented below for Pr$_{1-x}$La$_x$Os$_4$Sb$_{12}$ with
100: four different La concentrations: $x=0$, 0.05, 0.3, and 0.67.
101: For $x=0$, 0.3, and 0.67, we grew large single crystals (cubes with masses
102: as large as 50\,mg) on which accurate magnetic susceptibility measurements
103: were performed up to 300\,K in order to extract the room-temperature
104: paramagnetic effective moment.
105: In each case, this moment was within 10\% of that expected for Pr$^{3+}$.
106: (For the undoped compound, this finding contradicts a wide range
107: of values reported in literature.)
108: The superconducting transition temperatures $T_c$ of the large single crystals
109: and of smaller resistivity bars, also obtained in the same growths, were
110: checked via ac susceptibility measurements. A good agreement between $T_c$
111: values of large and small crystals confirmed the stoichiometry assigned
112: to samples used in this study. The residual resistivity ratio (the ratio of
113: the resistance at room temperature to that extrapolated to $T=0$) was
114: RRR $= 100$, 50, 180, and 170 for $x=0$, 0.05, 0.3, and 0.7, respectively.
115: The value RRR $= 100$ exceeds those reported
116: previously\cite{Bauer,Sugawara,Sugawara2} for pure PrOs$_4$Sb$_{12}$,
117: indicative of the high quality of our samples.
118: The $x=0.05$ crystal was from the batch for which results were
119: reported in Ref.~\onlinecite{Rotundu2}.
120:
121: The resistivity was measured by a conventional four-probe technique. The
122: estimated uncertainty in the determination of the absolute value of the
123: resistivity was 30\% due to the unfavorable geometry of the crystals.
124: Within this uncertainty, the resistivity at room temperature was the same
125: in all cases. In the plots below, we have scaled the resistivity
126: of each sample to a zero-field room-temperature value of 300\,$\mu\Omega$cm,
127: in the range reported previously. It is important to emphasize that this
128: scaling procedure is in no way essential for the conclusions of the paper,
129: which are based on the temperature and field dependence of the resistivity
130: of a given sample.
131:
132: We calculated the CEF contribution \rhoCEF\ to the electrical
133: resistivity via the method applied by Fisk and Johnston\cite{Fisk} to the
134: resistivity of PrB$_6$ and by Frederick and Maple\cite{Frederick} to the
135: magnetoresistance of PrOs$_4$Sb$_{12}$.
136: This method focuses on a single Pr ion, neglects intersite effects, and takes
137: no account of the direction of the current relative to the crystal axes or to
138: the magnetic field.
139: Our calculations started with one or other of two forms for $\hat{H}_0$,
140: the CEF Hamiltonian for Pr$^{3+}$ in zero magnetic field:
141: that (corresponding in the $O_h$-symmetry notation of Ref.\ \onlinecite{Lea}
142: to $W=-2.97$\,K and $x=-0.7225$) deduced\cite{Frederick} by fitting the
143: temperature dependence of the zero-field resistivity of PrOs$_4$Sb$_{12}$;
144: or the Hamiltonian (described in the $T_h$-symmetry notation of
145: Ref.\ \onlinecite{Takegahara} by $W=3.0877$\,K, $x=0.45991$, and $y=0.10503$)
146: determined from elastic\cite{Kohgi} and inelastic\cite{Goremychkin} neutron
147: scattering.
148: Henceforth, we refer to these cases as the ``doublet'' and ``singlet'' CEF
149: scheme, respectively, according to the ground-state degeneracy of $\hat{H}_0$.
150:
151: The CEF states in a magnetic field $\mathbf{H}$ were obtained by diagonalizing
152: the Hamiltonian $\hat{H}_0 + g \mu_B \mathbf{H}\cdot\hat{\mathbf{J}}$, where
153: $g = 4/5$ is the Land\'{e} $g$ factor for $Pr^{3+}$ and $\hat{\mathbf{J}}$ is
154: the $f$-electron angular momentum operator. The CEF resistivity is completely
155: determined by these CEF states, the temperature $T$, and two constants
156: $\rho_{\mathrm{ex}}$ and $\rho_A$, which parametrize, respectively, the overall
157: strengths of magnetic exchange and the aspherical Coulomb interaction between
158: the $4f$ and conduction electrons. Following Ref. \onlinecite{Frederick}, we
159: took $\rho_{\mathrm{ex}}=\rho_A = 0.25\,\mu\Omega$cm. However, our conclusions
160: are insensitive to the particular choice of constants.
161:
162: \section{Results and Discussion}
163:
164: \begin{figure}
165: \includegraphics[width=3.0in]{fig1.eps}
166: \caption{\label{rho:low-T}%
167: Resistivity vs temperature for PrOs$_4$Sb$_{12}$ in three different magnetic
168: fields, with both current and field along the (001) direction.}
169: \end{figure}
170:
171: Figure \ref{rho:low-T} shows the resistivity of undoped PrOs$_4$Sb$_{12}$ for
172: three representative fields, with both current and magnetic field oriented
173: along the (001) direction.
174: The results are similar to those reported by other groups.\cite{Bauer,Maple}
175: Below 200--300\,mK the resistivity saturates but has a strong field variation.
176: The residual resistivity $\rho_0$ can be obtained using the previously
177: noted\cite{Bauer,Maple} temperature variation at fixed field:
178: $\rho(T)=\rho_0+ B T^n$ with $n>2$.
179: Within the precision of our measurement, there is no difference between
180: $\rho_0$ obtained in this manner and $\rho(T=\mathrm{20\,mK})$.
181:
182: \begin{figure}
183: \includegraphics[width=3.0in]{fig2.eps}
184: \caption{\label{rho_0:pure}%
185: Residual resistivity vs magnetic field for PrOs$_4$Sb$_{12}$,
186: with both current and field along the (001) direction.
187: Arrows indicate boundaries between paramagnetic and
188: field-induced ordered phases.}
189: \end{figure}
190:
191: \begin{figure}
192: \includegraphics[width=3.3in]{fig3.eps}
193: \caption{\label{rho:2_schemes}%
194: Theoretical CEF resistivity at 20\,mK vs magnetic field calculated within the
195: singlet (upper panel) and doublet (lower panel) CEF schemes, for fields
196: along (001) ($\scriptstyle\blacksquare$) and (011) ($\triangledown$).}
197: \end{figure}
198:
199:
200: This residual resistivity (or resistivity at 20\,mK), when plotted against
201: magnetic field (Fig.\ \ref{rho_0:pure}), has a dome shape centered around
202: 9--10\,T.
203: Such a dome-shaped magnetoresistance has been reported previously at the
204: somewhat higher temperatures of 1.4\,K (Ref.~\onlinecite{Frederick}) and
205: 0.36\,K (Ref.~\onlinecite{Sugawara}). Two explanations for this dome have
206: been considered: field-induced long-range antiferroquadrupolar (AFQ) order,
207: and crossing of the lowest CEF levels. It is striking that $\rho_0(H)$ rises
208: sharply at the AFQ boundaries, indicated by arrows in Fig.\ \ref{rho_0:pure},
209: and peaks around 10\,T, where the AFQ transition temperature is highest.
210: However, Frederick and Maple have shown (see Fig.\ 2 of
211: Ref.\ \onlinecite{Frederick}) that the width, peak position, and height of
212: the dome in the magnetoresistance of PrOs$_4$Sb$_{12}$ at 1.4\,K are
213: reproduced quite well by the single-ion CEF resistivity computed within the
214: doublet scheme. (By contrast, \rhoCEF\ for the singlet scheme shows a step-like
215: increase in the vicinity of the crossing field at which the lowest $T_h$
216: $\Gamma_4^{(2)}$ level falls in energy below the $\Gamma_1$ singlet.)
217:
218: We find that neither CEF scheme accounts satisfactorily for the
219: magnetoresistance measured at 20\,mK (Fig.~\ref{rho_0:pure}), which shows a
220: dome of similar width to that at 1.4\,K. Irrespective of the CEF scheme,
221: \rhoCEF\ for fields oriented along the (001) direction (square symbols in
222: Fig.~\ref{rho:2_schemes}) is discontinuous at the crossing field and
223: essentially flat at higher fields. It therefore seems that the low-temperature
224: magnetoresistance of PrOs$_4$Sb$_{12}$ is dominated by effects beyond those
225: considered in the single-ion CEF model.
226:
227: \begin{figure}
228: \includegraphics[width=3.0in]{fig4.eps}
229: \caption{\label{rho:x=0.05}%
230: (color online)
231: Longitudinal magnetoresistance of Pr$_{0.95}$La$_{0.05}$Os$_4$Sb$_{12}$ at
232: 20, 310, and 660\,mK, for current and field along the (001) direction.}
233: \end{figure}
234:
235: We now turn to the effects of La doping.
236: Figure \ref{rho:x=0.05} shows the magnetoresistance of
237: Pr$_{0.95}$La$_{0.05}$Os$_4$Sb$_{12}$ at 20, 310, and 660\,mK. Similarly to
238: PrOs$_4$Sb$_{12}$, there is a negligible temperature variation of the
239: resistivity below 300\,mK in fields above-critical for superconductivity
240: (as evidenced by the overlap of the 20-mK and 310-mK isotherms). However,
241: the shape of the dome for $x=0.05$ is much less symmetric about the peak field
242: than its $x=0$ counterpart.
243: Between 2\,T and 10\,T, $\rho$(T=20 mK) for the doped sample increases by
244: over 80\%, compared to a 25\% increase for the undoped material, whereas
245: the resistivity drop above 10\,T is greater in percentage terms for $x=0$.
246:
247: \begin{figure}
248: \includegraphics[width=3.3in]{fig5.eps}
249: \caption{\label{rho:x=0.3@20mK}%
250: Magnetoresistance of Pr$_{0.7}$La$_{0.3}$Os$_4$Sb$_{12}$ at 20mK for three
251: different orientations of the magnetic field. The current direction was (001).}
252: \end{figure}
253:
254: The magnetoresistance becomes qualitatively different at higher La doping.
255: Figure \ref{rho:x=0.3@20mK} shows the 20-mK magnetoresistance of
256: Pr$_{0.7}$La$_{0.3}$Os$_4$Sb$_{12}$ for fields along (001), (011), and (010);
257: in each case, the current passed along the (001) direction.
258: All three isotherms exhibit a pronounced but wide step, centered near 9--10\,T,
259: superimposed on a linear background. In the investigated field range this
260: $x=0.3$ material does not exhibit the dome structure characteristic of $x=0$
261: and 0.05. For each curve, $\rho$ versus $H$ is approximately linear above
262: 13\,T. The resistivity of the non-$f$-electron analog LaOs$_4$Sb$_{12}$,
263: measured at 0.36\,K, has a quite large and approximately linear field
264: dependence.\cite{Sugawara} Furthermore, the directional dependence of the
265: magnetoresistance of LaOs$_4$Sb$_{12}$---$d\rho/dH$ being larger along
266: (011) than along (001)---is in agreement with the trend of the linear background
267: in Fig. \ref{rho:x=0.3@20mK}. It thus seems that the differences between the
268: high-field slopes of $\rho(H)$ in Fig. \ref{rho:x=0.3@20mK} can be attributed
269: primarily to non-$f$-electron contributions to the magnetoresistance.
270: Subtracting such linear parts results in very similar curves (not shown) for
271: all three field directions. We conclude that the $f$-electron magnetoresistance
272: in this moderately doped material is nearly isotropic.
273:
274: Compared to the cases $x=0$ and $x=0.05$, the low-temperature magnetoresistance
275: of Pr$_{0.7}$La$_{0.3}$Os$_4$Sb$_{12}$ for fields along (001) is much closer to
276: that given by the CEF model. The 20-mK measurements (Fig.\ \ref{rho:x=0.3@20mK})
277: are more consistent with the singlet CEF scheme than with the doublet scheme,
278: in that the latter predicts a sharp peak that is absent in the data. A second
279: and stronger argument in favor of the singlet scheme is provided by the
280: near-isotropy of the $f$-electron magnetoresistance noted in the previous
281: paragraph. Figure \ref{rho:2_schemes} plots the CEF resistivity for fields along
282: (001) and (011). The doublet scheme (lower panel in Fig.\ \ref{rho:2_schemes})
283: predicts a highly anisotropic \rhoCEF\ stemming from the fact that the lowest
284: two CEF levels cross at 8.5\,T along (001), but instead diverge along (011).
285: In the singlet CEF scheme (upper panel in Fig.\ \ref{rho:2_schemes}), the
286: anisotropy is much smaller because the lowest CEF levels cross at 8.6\,T along
287: (001), while along the (011) direction they anticross at 8.3\,T with a minimum
288: gap of only 0.7\,K. Unlike the doublet CEF scheme, the singlet scheme does a
289: reasonable job of reproducing the measured (011) magnetoresistance.
290: However, it underestimates the width of the step for fields along (001), and
291: hence still overestimates the anisotropy in the $f$-electron magnetoresistance
292: extracted from Fig.\ \ref{rho:x=0.3@20mK}.
293:
294: \begin{figure}
295: \includegraphics[width=3.3in]{fig6.eps}
296: \caption{\label{rho:x=0.3@310mK}%
297: Measured magnetoresistance of Pr$_{0.7}$La$_{0.3}$Os$_4$Sb$_{12}$ at
298: 310\,mK for current and field along the (001) direction, and theoretical CEF
299: resistivity \rhoCEF\ for the same temperature and field direction.}
300: \end{figure}
301:
302: \begin{figure}
303: \includegraphics[width=3.3in]{fig7.eps}
304: \caption{\label{rho:x=0.67@350mK}%
305: Magnetoresistance of Pr$_{0.33}$La$_{0.67}$Os$_4$Sb$_{12}$ at
306: 350\,mK for current and field along the (001) direction.}
307: \end{figure}
308:
309: Figure \ref{rho:x=0.3@310mK} shows that at the higher temperature of 310\,mK,
310: there is much better agreement between the magnetoresistance of
311: Pr$_{0.7}$La$_{0.3}$Os$_4$Sb$_{12}$ along (001) and and \rhoCEF\ calculated
312: within the singlet CEF scheme. At this temperature, the thermal smearing
313: of the step in \rhoCEF\ matches quite well the width of the rise
314: in the measured magnetoresistance. The results of similar calculations for
315: the doublet scheme (not shown, but very similar to the 350-mK results in
316: Fig.\ 2 of Ref.\ \onlinecite{Frederick}) are in gross disagreement with the
317: measurement.
318:
319: The character of the magnetoresistance seems to be little changed by further
320: La dilution. The longitudinal magnetoresistance for $x=0.67$ was investigated
321: down to 0.35\,K and in fields to 14\,T. The magnetoresistance at the lowest
322: temperature (Fig. \ref{rho:x=0.67@350mK}) exhibits essentially identical
323: magnetic field dependence to that for $x=0.3$.
324:
325: The main difference between the $f$-electron magnetoresistance inferred for
326: dilute-Pr alloys and that calculated in the singlet CEF scheme relates to the
327: low-temperature width of the step along the (001) field direction. The CEF
328: model predicts an almost discontinuous jump of the magnetoresistance at 20\,mK
329: at the level-crossing field, while the rise in the measured magnetoresistance
330: of Pr$_{0.7}$La$_{0.3}$Os$_4$Sb$_{12}$ takes place over 3--4\,T.
331: Since there is better agreement between the measured and theoretical
332: magnetoresistance along the (011) direction, where level anticrossing is
333: expected, we speculate that interactions omitted from the CEF model mix the
334: lowest levels, preventing any crossing even along high-symmetry field
335: directions and thereby producing isotropic magnetoresistance.
336: These interactions are most likely nonlocal. We note, however, that a
337: mean-field treatment of intersite magnetic and quadrupolar interactions
338: between Pr ions did not find avoided crossing.\cite{Kohgi,Rotundu}
339:
340: Figure \ref{rho:x=0.3@310mK} shows that the measured midpoint field for the
341: magnetoresistance rise in Pr$_{0.7}$La$_{0.3}$Os$_4$Sb$_{12}$ is about 1\,T
342: higher than is predicted based on the CEF level scheme determined
343: for pure PrOs$_4$Sb$_{12}$. This perhaps points to a small shift in the CEF
344: levels upon doping to 30\% La. The similarity between the magnetoresistance
345: steps observed for $x=0.3$ and 0.67 suggests that there is little further
346: evolution of the CEF energies over this doping range. A weak dependence of
347: CEF levels on La doping is in agreement with our specific-heat measurements
348: of the Schottky anomaly in lightly doped alloys,\cite{CEF} and is also
349: supported by the nearly invariant temperature of the maximum in the magnetic
350: susceptibility.\cite{Rotundu2}
351:
352: The contrast between the resistivity versus field curves for the $x=0$ and
353: $x=0.3$ samples is striking. The obvious differences between these two
354: compositions are the presence of a field-induced ordered phase for the pure
355: material and the absence of translational symmetry in the diluted case.
356: Our previous specific heat measurements\cite{CEF} indicate that the
357: field-induced ordered (AFQ) phase disappears somewhere near $x=0.2$.
358: Since the model for the CEF resistivity are single-site in character,
359: it seems likely that the dome-shaped magnetoresistance observed for
360: $x=0$ and 0.05 is associated with the presence of long-range order
361: in these materials, perhaps through enhanced scattering of conduction
362: electrons caused by fluctuations in the AFQ order parameter.
363:
364: In summary, we have shown that a simple CEF model accounts quite well
365: for the $f$-electron contribution to the magnetoresistance of
366: Pr$_{1-x}$La$_x$Os$_4$Sb$_{12}$ with $x = 0.3$ and 0.67. The weak dependence
367: of this contribution on field direction is consistent with the existence of a
368: singlet CEF in zero magnetic field, with avoided level crossing in applied
369: fields. At the lowest temperatures, the magnetoresistance for $x=0$ and
370: $x=0.05$ is inconsistent with the results of CEF model calculations.
371: This discrepancy is attributed to the long-range order present in the
372: pure and lightly doped materials.
373:
374: \begin{acknowledgments}
375:
376: Stimulating discussions with P.\ Kumar are acknowledged.
377: This work has been supported by U.S. Department of Energy (DOE) Grant
378: No.\ DE-FG02-99ER45748, by National Science Foundation (NSF) Grant No.\
379: DMR-0312939, and by the National High Magnetic Field Laboratory.
380: \end{acknowledgments}
381:
382: \begin{thebibliography}{12}
383:
384: \bibitem{Bauer}
385: E.\ D.\ Bauer, N.\ A.\ Frederick, P.-C.\ Ho, V.\ S.\ Zapf, and M.\ B.\ Maple,
386: Phys.\ Rev.\ B \textbf{65}, 100506(R) (2002).
387:
388: \bibitem{Rotundu}
389: C.\ R.\ Rotundu, H.\ Tsujii, Y.\ Takano, B.\ Andraka, H.\ Sugawara, Y.\ Aoki,
390: and H.\ Sato,
391: Phys.\ Rev.\ Lett.\ \textbf{92}, 037203 (2004).
392:
393: \bibitem{Takegahara}
394: K.\ Takegahara, H.\ Harima, and A.\ Yanase,
395: J.\ Phys.\ Soc.\ Jpn.\ \textbf{70}, 1190 (2001).
396:
397: \bibitem{Goremychkin}
398: E.\ A.\ Goremychkin, R.\ Osborn, E.\ D.\ Bauer, M.\ B.\ Maple,
399: N.\ A.\ Frederick, W.\ M.\ Yuhasz, F.\ M.\ Woodward, and J.\ W.\ Lynn,
400: Phys.\ Rev.\ Lett.\ \textbf{93}, 157003 (2004).
401:
402: \bibitem{Frederick}
403: N.\ A.\ Frederick and M.\ B.\ Maple,
404: J.\ Phys.: Condens.\ Matter \textbf{15}, 4789 (2003).
405:
406: \bibitem{Lacerda}
407: P.-C.\ Ho, N.\ A.\ Frederick, V.\ S.\ Zapf, E.\ D.\ Bauer, T.\ D.\ Do,
408: M.\ B.\ Maple, A.\ D.\ Christianson, and A.\ H.\ Lacerda,
409: Phys.\ Rev.\ B \textbf{67}, 180508(R) (2003).
410:
411: \bibitem{Sugawara}
412: H.\ Sugawara, S.\ Osaki, E.\ Kuramochi, M.\ Kobayashi, S.\ R.\ Saha,
413: T.\ Namiki, Y.\ Aoki, and H.\ Sato,
414: Phys.\ Rev.\ B \textbf{72}, 014519 (2005).
415:
416: \bibitem{Rotundu2}
417: C.\ R.\ Rotundu, P.\ Kumar, and B.\ Andraka,
418: Phys.\ Rev.\ B \textbf{73}, 014515 (2006).
419:
420: % Not cited
421: %\bibitem{Kadowaki}
422: %K.\ Kadowaki and S.\ B.\ Woods,
423: %Solid State Commun.\ \textbf{58}, 507 (1986).
424:
425: \bibitem{Sugawara2}
426: H.\ Sugawara, S.\ Osaki, S.\ R.\ Saha, Y.\ Aoki, H.\ Sato, Y.\ Inada,
427: H.\ Shishido, R.\ Settai, Y.\ \-{O}nuki, H.\ Harima, and K.\ Oikawa,
428: Phys.\ Rev.\ B \textbf{66}, 220504(R) (2002).
429:
430: \bibitem{Fisk}
431: Z.\ Fisk and D.\ C.\ Johnston,
432: Solid State Commun.\ \textbf{22}, 359 (1977).
433:
434: \bibitem{Lea}
435: K.\ R.\ Lea, M.\ J.\ M.\ Leask, and W.\ P.\ Wolf,
436: J.\ Phys.\ Chem.\ Solids \textbf{23}, 1381 (1962).
437:
438: \bibitem{Kohgi}
439: M.\ Kohgi, K.\ Iwasa, M.\ Nakajima, N.\ Metoki, S.\ Araki, N.\ Bernhoeft,
440: J.-M.\ Mignot, A.\ Gukasov, H.\ Sato, Y.\ Aoki, and H.\ Sugawara,
441: J.\ Phys.\ Soc.\ Jpn.\ \textbf{72}, 1002 (2003).
442:
443: \bibitem{Maple}
444: M.\ B.\ Maple, P.-C.\ Ho, V.\ S.\ Zapf, N.\ A.\ Frederick, E.\ D.\ Bauer,
445: W.\ M.\ Yuhasz, F.\ M.\ Woodward, and J.\ W.\ Lynn,
446: J.\ Phys.\ Soc.\ Jpn.\ Suppl.\ \textbf{71}, 23 (2002).
447:
448: \bibitem{CEF}
449: C.\ R.\ Rotundu and B.\ Andraka, in
450: Low Temperature Physics: 24th International Conference on Low Temperature
451: Physics, edited by Y.\ Takano, et al.\
452: (AIP Conference Proceedings, Melville, NY, 2006).
453:
454: \end{thebibliography}
455: \end{document}
456: