1: % ****** Start of file apssamp.tex ******
2: %
3: % This file is part of the APS files in the REVTeX 4 distribution.
4: % Version 4.0 of REVTeX, August 2001
5: %
6: % Copyright (c) 2001 The American Physical Society.
7: %
8: % See the REVTeX 4 README file for restrictions and more information.
9: %
10: % TeX'ing this file requires that you have AMS-LaTeX 2.0 installed
11: % as well as the rest of the prerequisites for REVTeX 4.0
12: %
13: % See the REVTeX 4 README file
14: % It also requires running BibTeX. The commands are as follows:
15: %
16: % 1) latex apssamp.tex
17: % 2) bibtex apssamp
18: % 3) latex apssamp.tex
19: % 4) latex apssamp.tex
20: %
21: \documentclass[prb,twocolumn,showpacs,preprintnumbers,amsmath,amssymb]{revtex4}
22: %\documentclass[preprint,showpacs,preprintnumbers,amsmath,amssymb]{revtex4}
23:
24: % Some other (several out of many) possibilities
25: %\documentclass[preprint,aps]{revtex4}
26: %\documentclass[preprint,aps,draft]{revtex4}
27: %\documentclass[prb]{revtex4}% Physical Review B
28:
29: \usepackage{graphicx}% Include figure files
30: \usepackage{psfig}
31: \usepackage{dcolumn}% Align table columns on decimal point
32: \usepackage{bm}% bold math
33:
34: %\nofiles
35:
36: \begin{document}
37:
38: %\preprint{APS/123-QED}
39:
40: \title{Selective dilution and magnetic properties of
41: La$_{0.7}$Sr$_{0.3}$Mn$_{1-x}$\textit{M}$^\prime_x$O$_3$ (\textit{M}$^\prime$ = Al, Ti)}% Force line breaks with \\
42: \author {D. N. H. Nam,$^{1,2}$ L. V. Bau,$^{1,3,5}$ N. V.
43: Khiem,$^{4}$ N. V. Dai,$^{1}$ L. V. Hong,$^{1}$ N. X. Phuc,$^{1}$ R.
44: S. Newrock,$^{2}$ and P. Nordblad$^{5}$}
45: \affiliation{$^1$Institute
46: of Materials Science, VAST, 18 Hoang-Quoc-Viet, Hanoi, Vietnam \\
47: $^{2}$Department of Physics, University of Cincinnati, OH
48: 45221-0011, USA\\$^3$Department of Science and Technology, Hongduc
49: University, Thanhhoa, Vietnam\\$^4$Department of Natural Sciences,
50: Hongduc University, Thanhhoa, Vietnam\\$^5$The \AA ngstr\"{o}m
51: Laboratory, Uppsala University, Box 534, SE 751-21 Uppsala, Sweden}
52: \date{\today}% It is always \today, today,
53: % but any date may be explicitly specified
54: \begin{abstract}
55: The magnetic lattice of mixed-valence Mn ions in
56: La$_{0.7}$Sr$_{0.3}$MnO$_3$ is selectively diluted by partial
57: substitution of Mn by Al or Ti. The ferromagnetic transition
58: temperature and the saturation moment decreases with substitution in
59: both series. The volume fraction of the non-ferromagnetic phases
60: evolves non-linearly with the substitution concentration and faster
61: than theoretically expected. By presenting the data in terms of
62: selective dilutions, the reduction of $T_\mathrm{c}$ is found to be
63: scaled by the relative ionic concentrations and is consistent with a
64: prediction derived from molecular-field theory.
65: \end{abstract}
66: \pacs{75.10.Hk, 75.30.Cr, 75.30.Et, 75.47.Lx}% PACS, the Physics and Astronomy
67: % Classification Scheme.
68: %\keywords{Suggested keywords}%Use showkeys class option if keyword
69: %display desired
70: \maketitle
71: \section{\label{sec:intro}Introduction}
72: Perovskite manganites with the general composition ($R$,$A$)MnO$_3$
73: ($R$: rare earth, $A$: alkali elements) are attractive to scientists
74: not only because of their potential applications but also due to
75: their very rich and intriguing physics. While most of the pristine
76: compounds $R$MnO$_3$ are (insulating) antiferromagnets due to
77: antiferromagnetic (AF) superexchange (SE) interactions between
78: Mn$^{3+}$ ions, the introduction of divalent alkali cations such as
79: Sr$^{2+}$, Ca$^{2+}$, or Ba$^{2+}$ into the composition converts an
80: adapted number of Mn$^{3+}$ to Mn$^{4+}$ that in turn gives rise to
81: the $\mathrm{Mn}^{4+}-\mathrm{O}^{2-}-\mathrm{Mn}^{3+}$
82: ferromagnetic (FM) double-exchange (DE) \cite{Zener} interaction.
83: The presence of DE couplings can turn an insulating
84: antiferromagnetic manganite into a ferromagnet with metallic
85: conductivity. It has been widely accepted that, along with lattice
86: distortions \cite{Millis1,Millis2, Hwang} and phase segregation
87: phenomena, \cite{Dagotto} the DE mechanism plays a very important
88: role in governing the properties of manganites.
89:
90: Perovskite manganites have been intensively studied in the last
91: decade since the discovery of the Colossal Magneto-Resistance (CMR)
92: phenomenon. \cite{Helmolt} Chemical substitution has been widely
93: used as a conventional method to uncover the underlying physics and
94: to search for compositions with novel properties. Results for both
95: $R$- and Mn-site substitution have been quite well documented in the
96: literatures. For manganites with a Mn$^{3+}$/Mn$^{4+}$ ratio of 7/3,
97: substitution of different elements at the rare-earth site has shown
98: that magnetic and transport properties of many manganites depend
99: rather systematically on the average ionic size, $\langle
100: r_{A}\rangle=\sum_{i}y_{i}r_{i}$, and the ionic size mismatch
101: defined as $\sigma^2=\sum_{i}y_{i}r_{i}^2-\langle r_{A}\rangle^2$
102: (where $y_i$ is the fraction and $r_i$ the ionic radius of the $i$th
103: species occupying the $R$-site), \cite{Hwang, Teresa} i.e. on the
104: degree of GdFeO$_3$-type lattice distortion and lattice disorder.
105: However, despite many attempts at modifying the magnetic lattice of
106: Mn ions by direct substitution at the Mn-sites, the complexity
107: caused by too many factors that govern the properties of the
108: materials, no universal features have been revealed to date. In this
109: paper, we report that the reduction of the ferromagnetic ordering
110: transition temperature $T_\mathrm{c}$ of
111: La$_{0.7}$Sr$_{0.3}$Mn$_{1-x}$\textit{M}$^\prime_x$O$_3$ ($M^\prime=
112: \mathrm{Al}$, Ti) scales with the relative substitution
113: concentrations and is consistent with a prediction from
114: molecular-field theory (MFT) of the dilution of a magnetic lattice.
115: The substitutions with Al and Ti are \emph{selective} in nature
116: because Al$^{3+}$ only substitutes for Mn$^{3+}$ and Ti$^{4+}$ only
117: for Mn$^{4+}$. Another advantage of these substitutions is that
118: because Al$^{3+}$ and Ti$^{4+}$ ions do not carry a magnetic moment,
119: they are also expected not to participate in the magnetic
120: interaction.
121: \section{\label{sec:exper}Experiment}
122: In this paper,
123: La$_{0.7}$Sr$_{0.3}$Mn$_{1-x}$\textit{M}$^\prime_x$O$_3$ is denoted
124: as LSMA$_x$ for \textit{M}$^\prime$ = Al and LSMT$_x$ for
125: \textit{M}$^\prime$ = Ti. All the samples were prepared using a
126: conventional solid state reaction method. Pure ($\geq$99.99\%) raw
127: powders with appropriate amounts of La$_2$O$_3$, SrCO$_3$, MnO$_2$,
128: Al$_2$O$_3$, and TiO$_2$ are thoroughly ground, mixed, pressed into
129: pellets and then calcined at several processing steps with
130: increasing temperatures from 900 $^\mathrm{o}$C to 1200
131: $^\mathrm{o}$C and with intermediate grindings and pelletizations.
132: The products are then sintered at 1370 $^\mathrm{o}$C for 48 h in
133: ambient atmosphere. The final samples are obtained after a very slow
134: cooling process from the sintering to room temperature with an
135: annealing step at 700 $^\mathrm{o}$C for few hours. Room-temperature
136: X-ray diffraction patterns (measured by a SIEMENS-D5000 with
137: Cu-$K_\alpha$ radiation) show that all of the samples are
138: crystallized in perovskite rhombohedral structures with almost no
139: sign of secondary phases or remnants of the starting materials. The
140: crystal structures obtained for the samples are in agreement with
141: earlier structural studies. \cite{Qin,Hu,Kallel,Kim,Troyanchuk}
142: Transport and magnetotransport measurements were carried out in a
143: non-commercial cryostat using the standard 4-probe method. Magnetic
144: and magnetization measurements were performed in a Quantum Design
145: MPMS SQUID magnetometer and (sometimes) by a Vibrating Sample
146: Magnetometer (VSM).
147: \begin{figure}
148: % Requires \usepackage{graphicx}
149: \vspace{0.1in}
150: \includegraphics[width=3.3in]{Fig1eps.eps}
151: \caption{Field-cooled magnetization as a function of temperature for
152: (a)
153: $\mathrm{La}_{0.7}\mathrm{Sr}_{0.3}\mathrm{Mn}_{1-x}\mathrm{Al}_x\mathrm{O}_3$
154: and (b)
155: $\mathrm{La}_{0.7}\mathrm{Sr}_{0.3}\mathrm{Mn}_{1-x}\mathrm{Ti}_x\mathrm{O}_3$,
156: measured in $H = 100$ G. The zero-field-cooled
157: \textit{M}$_\mathrm{ZFC}(T)$ ($\blacktriangledown$) of the $x = 0.3$
158: sample is added for further discussions.}\label{Figure1}
159: \end{figure}
160: \section {Results and Discussion}
161: Magnetic, transport, and magnetotransport characterization of
162: LSMA$_x$ and LSMT$_x$ compounds has been reported previously by
163: other authors and can be referenced in a number of publications.
164: \cite{Qin,Hu,Kallel,Kim,Troyanchuk,Sawaki} Although consistent
165: tendencies as to the variation of e.g. the transition temperature
166: and saturation moment with doping concentration have been found,
167: there is significant scatter in the data in-between the different
168: studies. In the current work, the essential characteristics are
169: carefully measured and reexamined.
170: \subsection{Temperature dependent characterization}
171: Temperature dependent magnetization measurements, $M(T)$, in both
172: zero-field-cooled (ZFC) and field-cooled (FC) protocols, are carried
173: out for all the samples. The FC $M(T)$ curves presented in Fig. 1
174: clearly indicate that the substitution of Al or Ti for Mn causes the
175: ferromagnetic-paramagnetic (PM) transition temperature
176: $T_\mathrm{c}$ to drop drastically. The reduction of $T_\mathrm{c}$
177: in the case of Ti substitution is much more substantial. The FM-PM
178: transition is very sharp at small $x$ concentrations for both doping
179: series but becomes broader with increasing $x$. A transition is
180: observed for all the samples, even with LSMT$_{0.3}$ where Mn$^{4+}$
181: ions are supposed to be completely absent. However, as is implied by
182: the ZFC and FC $M(T)$ curves in Fig. \ref{Figure1}, the
183: LSTMT$_{0.3}$ sample is not a true ferromagnet, nor a pure spin
184: glass as has been suggested in Ref. 12 for this composition. This
185: last conclusion is also corroborated by the fact that our
186: frequency-dependent ac-susceptibility measurements
187: $\chi_\mathrm{ac}(T,\omega)$ (not shown) do not indicate a dynamic
188: phase transition. The $M(T)$ curves for this sample probably suggest
189: the existence of an AF background state with a weak ferromagnetic
190: component --- possible indications of a canted antiferromagnet, as
191: was proposed in Ref. 10. The $T_\mathrm{c}$ vs. $x$ data extracted
192: from the $M(T)$ curves for all samples are plotted in Fig.
193: \ref{Figure4} and will be discussed later in detail.
194:
195: Temperature dependent transport measurements for both series show
196: that the resistivity, $\rho$, strongly increases with increasing $x$
197: while the metal-insulator (MI) transition usually observed at a
198: temperature $T_\mathrm{p}$ near (but lower than) $T_\mathrm{c}$ is
199: shifted to lower temperatures. The MI transition is observed in all
200: of the LSMA$_x$ samples while it can only be observed in LSMT$_x$
201: for $x$ up to 0.1. For LSMT$_{0.2}$, the MI transition is no longer
202: observed, but there still exists a slope change in the $\rho(T)$
203: curve at a temperature below $T_\mathrm{c}$ signalling a magnetic
204: contribution from the FM phase to the conductivity. In agreement
205: with the magnetic behavior, the LSMT$_{0.3}$ sample only shows
206: insulating characteristics.
207: \begin{figure}
208: % Requires \usepackage{graphicx}
209: \vspace{0.1in}
210: \includegraphics[width=3.3in]{Fig2eps.eps}
211: \caption{$\mathrm{La}_{0.7}\mathrm{Sr}_{0.3}\mathrm{Mn}_{1-x}\mathrm{Al}_x\mathrm{O}_3$:
212: Theoretical $M_\mathrm{s}$ vs. $x$ calculated from equation
213: \ref{eqn0} (bold curve without symbols) and experimental
214: $M_\mathrm{s}$ data (thin curve with $\odot$ symbols) extracted from
215: $M(H)$ measurements at 5 K (inset).}\label{Figure2}
216: \end{figure}
217: \subsection{Magnetization characterization}
218: \begin{figure}
219: % Requires \usepackage{graphicx}
220: \vspace{0.1in}
221: \includegraphics[width=3.3in]{Fig3eps.eps}
222: \caption{$\mathrm{La}_{0.7}\mathrm{Sr}_{0.3}\mathrm{Mn}_{1-x}\mathrm{Ti}_x\mathrm{O}_3$:
223: Theoretical $M_\mathrm{s}$ vs. $x$ calculated from equation
224: \ref{eqn0} (bold curve without symbols) and experimental
225: $M_\mathrm{s}$ data (thin curve with $\odot$ symbols) extracted from
226: $M(H)$ measurements at 5 K (inset).}
227: \label{Figure3}
228: \end{figure}
229: The theoretical zero-temperature spin-only saturation magnetization
230: (in emu/g) of the LSMA$_x$ and LSMT$_x$ compounds follow
231: \begin{eqnarray}
232: M_\mathrm{s}=\left\{ \begin{array}{ll}
233: (3.7-4x)10^3/(40.548-5.006x) & \mbox{if $M^\prime=\mathrm{Al}$}\\
234: (3.7-3x)10^3/(40.548-1.801x) & \mbox{if
235: $M^\prime=\mathrm{Ti}$}.
236: \end{array}
237: \right.
238: \label{eqn0}
239: \end{eqnarray}
240: For the LSMA$_x$ series, magnetization measurements at $T = 5$ K
241: (the inset of Fig. 2) indicate that the samples are substantially
242: saturated in an applied field of just above 1 T. At higher fields,
243: the LSMA$_0$ and LSMA$_{0.05}$ samples exhibit flat $M(H)$
244: dependencies as expected for conventional ferromagnets at high
245: fields and low temperatures. Closer inspection on the $M(H)$ curves
246: for $x\geq0.1$ shows, however, that they are quite linear in the
247: high-field regime up to 4.5 T with a small slope which increases
248: with $x$. The small slope would come from the suppression of thermal
249: fluctuations of the magnetization by the magnetic field. However,
250: the evolution of the slope of the $M(H)$ curves with $x$ in the high
251: field regime may signal a magnetic contribution from certain Mn ions
252: that do not take part in the ferromagnetic phase. In addition, for
253: the whole series, the measured magnetization in magnetic fields up
254: to 4.5 T does not reach the theoretical magnetization value and even
255: deviates further with increasing $x$. Based on these features, it is
256: presumable that segregation into FM and non-FM phases occurs in
257: these samples. The experimental values of the saturation
258: magnetization $M_\mathrm{s}$ of the FM phase presented in Fig.
259: \ref{Figure2} (and also Fig. \ref{Figure3} for Ti substitution) are
260: determined by extrapolating the linear part of the $M(H)$ curves in
261: the high field regime to $H = 0$. The difference between
262: experimental and theoretical $M_\mathrm{s}$ is then attributed to
263: the non-FM contributions, $M_\mathrm{non-FM}$.
264:
265: Very similar results are also obtained for the LSMT$_x$ series, as
266: can be seen in Fig. \ref{Figure3}. Previous observations of the
267: decrease of $M_\mathrm{s}$ in Al and Ti substitutions in manganites
268: have been published and interpreted by several other authors.
269: \cite{Kallel,Kim,Troyanchuk,Blasco,Cao} The decrease of
270: $M_\mathrm{s}$ was attributed to the dilution and the weakening of
271: FM exchange couplings as in general. Kallel \textit{et al.}
272: \cite{Kallel} suggest even a change of the spin state or the orbital
273: ordering of the Mn ions. The deviation between the experimental
274: $M_\mathrm{s}$ and its theoretical values has not been adequately
275: considered. In a study on
276: La$_{0.7}$Ca$_{0.3}$Mn$_{1-x}$Al$_{x}$O$_3$ by neutron diffraction,
277: the effective magnetic moment per Mn ion was found far smaller than
278: the saturation moment and that was explained by assumptions of
279: magnetic inhomogeneity and structural disorder.\cite{Blasco} Our
280: interpretation for the $M_\mathrm{s}$ deviation in terms of magnetic
281: phase segregation is somehow closer to that proposed for
282: La$_{0.7}$Ca$_{0.3}$Mn$_{1-x}M^\prime_x$O$_3$ ($M^\prime$= Ti, Ga)
283: compounds by Cao \textit{et al.}, \cite{Cao} where the authors
284: suggest that Ti$^{4+}$ and Ga$^{3+}$ ions generate around them
285: non-ferromagnetic (paramagnetic or possibly antiferromagnetic)
286: regions.
287:
288: As pointed out above, there is a growing deviation of the FM
289: saturation magnetization from the calculated value with increasing
290: substitution concentration. This observation indicates that the
291: substitution not only reduces the total number of Mn ions but also
292: raises the number of non-FM Mn ions at the expense of the FM phase.
293: Since the Mn$^{3+}$/Mn$^{4+}$ ionic concentration ratio is driven
294: away from the optimal value of 7/3 by the substitution, a certain
295: amount of the Mn ions may become \emph{redundant} with respect to
296: the DE couplings. I.e., the selective substitution on one Mn ionic
297: species produces the redundancy on the other Mn species. Those
298: redundant ions contribute to the non-FM phase. With small values of
299: $x$, the redundant ions are thus mostly Mn$^{4+}$ for $M^\prime=
300: \mathrm{Al}$ while they are Mn$^{3+}$ for the Ti substitution. Based
301: on this assumption the concentration of non-FM Mn ions can
302: approximately be derived from $M_\mathrm{non-FM}$. For examples,
303: with $x = 0.2$, the estimated redundant concentration of Mn$^{4+}$
304: is 0.187 for $M^\prime=\mathrm{Al}$ and the corresponding
305: concentration of Mn$^{3+}$ is 0.064 for $M^\prime=\mathrm{Ti}$. With
306: sufficiently high substitution concentrations, contributions to the
307: the non-FM phase may also arise from the Mn ions of both species
308: that are isolated by the non-magnetic ones.
309: \begin{figure}
310: % Requires \usepackage{graphicx}
311: \vspace{0.1in}
312: \includegraphics[width=3.3in]{Fig4eps.eps}
313: \caption{La$_{0.7}$Sr$_{0.3}$Mn$_{1-x}$\textit{M}$^\prime_x$O$_3$:
314: Variation of $T_\mathrm{c}$ with substitution concentration $x$
315: ($0\leq x\leq0.2$) for $M^\prime=\mathrm{Al}$ ($\odot$) and Ti
316: ($\boxdot$).}
317: \label{Figure4}
318: \end{figure}
319: \subsection {Effects of selective dilution and the MFT approximation}
320: Figure \ref{Figure4} shows the dependence of the FM-PM phase
321: transition temperature $T_\mathrm{c}$ on substitution concentration,
322: derived from the low-field $M(T)$ data in Fig. \ref{Figure1}. The
323: $T_\mathrm{c}$ values are determined by the temperatures at which
324: the $\partial M/\partial T$ curves peak. The parent compound,
325: $\mathrm{La}_{0.7}\mathrm{Sr}_{0.3}\mathrm{MnO}_3$, has
326: $T_\mathrm{c} = 364.4$ K, in good agreement with previously reported
327: data.\cite{Coey} $T_\mathrm{c}$ decreases monotonically as $x$
328: increases in both cases. Explanations for the decrease of
329: $T_\mathrm{c}$ with Al and Ti substitution in manganites have been
330: suggested by several authors. \cite{Sawaki, Qin, Kallel, Kim} One
331: explanation is simply that the decrease of $T_\mathrm{c}$ is due to
332: the suppression of long-range FM order of the localized
333: $t_\mathrm{2g}$ spins by local breakdown of the exchange couplings
334: where the substitution occurs. \cite{Qin,Sawaki} Hu \textit{et al.}
335: \cite{Hu} assumed that the substituted Ti$^{4+}$ ions tends to
336: demolish the DE $\mathrm{Mn}^{3+}-\mathrm{O}^{2-}-\mathrm{Mn}^{4+}$
337: bonds and lowers the hole carrier concentration, thus suppressing
338: the DE interaction and lowering $T_\mathrm{c}$. Kallel \textit{et
339: al.} \cite{Kallel} suggested that the presence of Ti in LSMT$_x$
340: favors SE interaction and suppresses the DE mechanism.
341: Significantly, Kim \textit{et al.} \cite{Kim} recently found that
342: the Ti substitution in LSMT$_x$ increases the
343: $\mathrm{Mn}-\mathrm{O}-\mathrm{Mn}$ bond length and reduces the
344: bond angle. Based on structural data, the authors calculated the
345: variation of $e_\mathrm{g}$-electron bandwidth $W$, finding a
346: decrease of $W$ with $x$, and related it to the decrease of
347: $T_\mathrm{c}$. It is worth noting that the ionic size of Mn$^{4+}$
348: ($0.530$ \AA) is smaller than that of Ti$^{4+}$ ($0.605$ \AA) while
349: Mn$^{3+}$ ($0.645$ \AA) is larger than Al$^{3+}$ ($0.535$
350: \AA).\cite{Shannon} As a result, the effects of Al and Ti
351: substitution on the structure, and hence $W$, may not be the same
352: and even opposite in the two cases. However, in reality,
353: $T_\mathrm{c}$ has been found always to decrease with the
354: substitution. A detailed study on the structure of LSMA$_x$ may help
355: justify the cause.
356:
357: As is seen in Fig. \ref{Figure4}, the effect of substitution with Ti
358: on the reduction of $T_\mathrm{c}$ (defined by $\Delta
359: T_\mathrm{c}=T_\mathrm{c}$($x$)$-T_\mathrm{c}$(0)) is as much as
360: more than twice of that of Al substitution. However, as mentioned
361: above, because the substitution is selective, in order to compare
362: their effects, instead of using $x$ as the common variable, the data
363: should be presented as functions of relative substitution
364: concentrations, defined as $n_\mathrm{p} = x/0.7$ when $M^\prime =
365: \mathrm{Al}$ and $n_\mathrm{p} = x/0.3$ when $M^\prime =
366: \mathrm{Ti}$. Physically, $n_\mathrm{p}$ is the average
367: concentration of $M^\prime$ per Mn site of the
368: selectively-substituted Mn ionic species (Mn$^{3+}$ for $M^\prime =
369: \mathrm{Al}$, or Mn$^{4+}$ for $M^\prime = \mathrm{Ti}$), or the
370: probability that site is occupied by an $M^\prime$ ion. Strikingly,
371: as displayed in Figure \ref{Figure5}, the
372: $T_\mathrm{c}(n_\mathrm{p})$ data for $M^\prime = \mathrm{Al}$ and
373: Ti collapse onto one curve which follows very well the linear line
374: predicted by the molecular-field theory (see below). It is also
375: surprising that the linear behavior of $T_\mathrm{c}(n_\mathrm{p})$
376: of LSMT$_x$ sustains in a very wide range of $n_\mathrm{p}$ and has
377: the tendency to reduce to zero at $n_\mathrm{p}=1$.
378:
379: For further understanding and analyzes of the results, we use a
380: mean-field approximation to derive the relation between
381: $T_\mathrm{c}$ and $n_\mathrm{p}$. According to the Heisenberg
382: model, the potential energy of exchange interactions of any
383: particular magnetic ion $i$ with the other ions $j$ is given by $U_i
384: = -2S_i\cdot\sum_{j\neq i}J_{ij}S_j$ where $S_i$, $S_j$ are the
385: spins of the $i$th and $j$th ions respectively; and $J_{ij}$ the
386: exchange integral. $U_i$ can be rewritten as
387: \begin{equation}
388: U_i=-2\frac{\mu_i}{g^2}\cdot\sum_{j\neq
389: i}J_{ij}\mu_j=-2\frac{\mu_i\cdot M}{Ng^2}\sum_{j\neq i}J_{ij}
390: \label{eqn1}
391: \end{equation}
392: where $M$ and $N$ are the magnetic moment and the number of magnetic
393: ions per unit volume, respectively and $\mu_i$, $\mu_j$ are the
394: magnetic moment of the $i$th and $j$th ions. For simplicity, $\mu_j$
395: is replaced with the average magnetic moment per site
396: $\langle\mu\rangle=\frac{M}{N}$. Assuming that the $i$th ion
397: interacts only with its nearest-neighbor ions, but with $n$
398: different exchange coupling constants $J_{i\alpha}$ each involves
399: $z_{i\alpha}$ ions, then instead of summing over the ions $j$, the
400: summation is made over the interactions. We can then recast Eq.
401: \ref{eqn1} as
402: \begin{equation}
403: U_i=-2\frac{\mu_i\cdot
404: M}{Ng^2}\sum_{\alpha}^{n}z_{i\alpha}J_{i\alpha}. \label{eqn2}
405: \end{equation}
406: According to the MFT, $U_i=-\mu_i\cdot B_\mathrm{M}=-\mu_i\cdot
407: \lambda M$ and $T_\mathrm{c}=\lambda C$, where $B_\mathrm{M}$ is the
408: molecular field, and $C$ and $\lambda$ are, respectively, the Curie
409: and Weiss constants, Eq. \ref{eqn2} becomes
410: \begin{equation}
411: T_\mathrm{c}=\frac{2S_i\left(S_i+1\right)}{3k_\mathrm{B}}\sum_{\alpha}^{n}z_{i\alpha}J_{i\alpha}.
412: \label{eqn3}
413: \end{equation}
414: For a simple system where there is only one species of magnetic ion
415: and hence one kind of exchange interaction, Eq. \ref{eqn3} equals
416: the standard expression
417: \begin{equation}
418: T_\mathrm{c}=\frac{2S\left(S+1\right)zJ}{3k_\mathrm{B}}.
419: \label{eqn4}
420: \end{equation}
421: When the system is diluted by a substituting non-magnetic element,
422: and if the substitution is completely random with respect to its
423: lattice site, the dilution would result in a proportional dependence
424: $z(n_\mathrm{p})=z(0)(1-n_\mathrm{p})$ where $0\leq
425: n_\mathrm{p}\leq1$ and $z(0)$ refers to the undiluted system.
426: Replacing $z(n_\mathrm{p})$ for $z$ in Eq. \ref{eqn4} yields a
427: linear dependence of $T_\mathrm{c}$ on $n_\mathrm{p}$ as illustrated
428: in Figure \ref{Figure5}.
429: \begin{figure}
430: % Requires \usepackage{graphicx}
431: \vspace{0.1in}
432: \includegraphics[width=3.3in]{Fig5eps.eps}
433: \caption{La$_{0.7}$Sr$_{0.3}$Mn$_{1-x}$\textit{M}$^\prime_x$O$_3$:
434: $T_\mathrm{c}$ is presented as a function of relative concentration
435: $n_\mathrm{p} = x/0.7$ for $M^\prime=\mathrm{Al}$ ($\odot$) or
436: $n_\mathrm{p} = x/0.3$ for $M^\prime=\mathrm{Ti}$ ($\boxdot$). The
437: broken line presents a prediction from molecular-field theory. The
438: data for $\mathrm{La_{0.7}Sr_{0.3}Mn_{0.9}Al_{0.07}Ti_{0.03}O_{3}}$
439: ($n_\mathrm{p}=0.19$) and
440: $\mathrm{La_{0.7}Sr_{0.3}Mn_{0.8}Al_{0.14}Ti_{0.06}O_{3}}$
441: ($n_\mathrm{p}=0.36$) ($\blacktriangle$) are added for further
442: discussions (see text for details).}
443: \label{Figure5}
444: \end{figure}
445:
446: Nevertheless, Eq. \ref{eqn4} is perhaps not relevant to our
447: manganite systems at first because of the mixed-valence of the Mn
448: ions and the coexistence of different kinds of interaction including
449: SE AF interactions
450: $\mathrm{Mn}^{3+}-\mathrm{O}^{2-}-\mathrm{Mn}^{3+}$,
451: $\mathrm{Mn}^{4+}-\mathrm{O}^{2-}-\mathrm{Mn}^{4+}$, and the DE FM
452: interaction $\mathrm{Mn}^{3+}-\mathrm{O}^{2-}-\mathrm{Mn}^{4+}$ with
453: corresponding exchange constants denoted as $J_{\mathrm{SE1}}$,
454: $J_{\mathrm{SE2}}$, and $J_{\mathrm{DE}}$, and nearest-neighbor
455: interacting ions number $z_{\mathrm{SE1}}$, $z_{\mathrm{SE2}}$, and
456: $z_{\mathrm{DE}}$, respectively. According to Eq. \ref{eqn3}, a
457: linear behavior is expected to be observed for
458: $T_\mathrm{c}\left(n_\mathrm{p}\right)$ only if (i) the exchange
459: coupling constants are unchanged and (ii) the dilution either
460: affects $z_{i\alpha}$ in a proportional manner such that
461: $z_{i\alpha}\left(n_\mathrm{p}\right) \propto
462: \left(1-n_\mathrm{p}\right)z_{i\alpha}(0)$ or does not affect it at
463: all. Supposing that $J_\mathrm{DE}$, $J_\mathrm{SE1}$,
464: $J_\mathrm{SE2}$ do not change with substitution, within the
465: linearity regime of $T_\mathrm{c}(n_\mathrm{p})$, it is presumable
466: that the Al (or Ti) substitution leaves $z_\mathrm{SE2}$ (or
467: $z_\mathrm{SE1}$) intact but possibly changes both $z_\mathrm{DE}$
468: and $z_\mathrm{SE1}$ (or $z_\mathrm{SE2}$) proportionally to
469: $1-n_\mathrm{p}$. The most remarkable feature of the
470: $T_\mathrm{c}(n_\mathrm{p})$ variation in Fig. \ref{Figure5} is that
471: $T_\mathrm{c}$ has a tendency to go to zero when $n_\mathrm{p}=1$.
472: This would simply mean that the $zJ$ product of DE couplings is
473: totally dominant over those of the SE ones. Considering the fact
474: that at low dilution concentrations, since the Mn$^{3+}$ and
475: Mn$^{4+}$ concentrations are not too different, $z_\mathrm{DE}$,
476: $z_\mathrm{SE1}$, and $z_\mathrm{SE2}$ are possibly comparable, thus
477: it is reasonable to suppose that the SE couplings $J_\mathrm{SE1}$,
478: $J_\mathrm{SE2}$ both are negligibly small in these systems. This
479: could be one reason behind the fact that
480: $\mathrm{La}_{0.7}\mathrm{Sr}_{0.3}\mathrm{MnO}_3$ is a unique
481: manganite in the sense that it has the highest $T_\mathrm{c}$
482: amongst the perovskite manganites. In a situation when the SE
483: coupling constants are significant, because either $z_\mathrm{SE1}$
484: or $z_\mathrm{SE2}$ does not change by selective substitution,
485: according to Eq. \ref{eqn3}, $T_\mathrm{c}(n_\mathrm{p})$ may still
486: have a linear dependence but should have a tendency to intercept
487: $n_\mathrm{p}$-axis at a certain value $n_\mathrm{p}<1$.
488:
489: Apart from changes of the $z$ values, the substitution of Al or Ti
490: for Mn could cause some additional effects. Because they are
491: nonmagnetic, the magnetic coupling will be broken at any site they
492: occupy, leading to a weakening of long-range FM ordering established
493: by the dominant DE interactions and a deterioration of the metallic
494: conductivity. The selectivity of the substitution also drives the
495: Mn$^{3+}$/Mn$^{4+}$ ratio away from the optimal 7/3 value
496: contributing to developing the non-FM phase. The differences in
497: ionic sizes between Mn and Al and Ti could modify the crystal
498: structure, especially the angle and length of the
499: $\mathrm{Mn}-\mathrm{O}-\mathrm{Mn}$ couplings.\cite{Garcia} All
500: these factors contribute to the degradation of the ferromagnetism
501: and metallicity of the manganite system. Nevertheless, the linear
502: behavior of $T_\mathrm{c}(n_\mathrm{p})$ observed in Fig.
503: \ref{Figure5} implies that, within the substitution ranges, those
504: effects are dominated by the effect of dilution. In addition, there
505: is a significant difference between the two substitutions. While the
506: amount of Al substitution reduces the number of hopping electrons by
507: an equivalent amount, Ti substitution does not affect it at all. The
508: shortage of hopping electrons would possibly explain the large
509: number of Mn$^{4+}$ redundancy in the LSMA$_x$ compounds. We believe
510: that this difference has a link to the reason why the linear
511: behavior of $T_\mathrm{c}(n_\mathrm{p})$ of LSMA$_x$ occurs in a
512: much narrower range of dilution ($n_\mathrm{p}\leq0.25$) than that
513: of LSMT$_x$ where $T_\mathrm{c}(n_\mathrm{p})$ is found linear up to
514: $n_\mathrm{p} = 0.67$. It is worth noting that
515: $T_\mathrm{c}(n_\mathrm{p})$ would not follow the linear behavior up
516: to $n_\mathrm{p} = 1$ as predicted by the MFT because there should
517: exist a percolation threshold $n_\mathrm{c}$, above which clustering
518: occurs and the ferromagnetic network collapses into only short-range
519: ordering and superparamagnetism.
520:
521: In the case when both Al and Ti are substituted for Mn with
522: according $n_\mathrm{p}(\mathrm{Al})$ and
523: $n_\mathrm{p}(\mathrm{Ti})$, supposing that the ferromagnetic double
524: exchange is totally dominant in the system, the dilution
525: concentration of the whole system is determined as $n_\mathrm{p} =
526: 1-\left(1-n_\mathrm{p}(\mathrm{Al})\right)\left(1-n_\mathrm{p}(\mathrm{Ti})\right)$.
527: To check the validity of the MFT analysis in this case, the
528: $T_\mathrm{c}(n_\mathrm{p})$ values of the
529: $\mathrm{La_{0.7}Sr_{0.3}Mn_{0.9}Al_{0.07}Ti_{0.03}O_{3}}$
530: ($n_\mathrm{p}=0.19$) and
531: $\mathrm{La_{0.7}Sr_{0.3}Mn_{0.8}Al_{0.14}Ti_{0.06}O_{3}}$
532: ($n_\mathrm{p}=0.36$) compounds are also added to Figure
533: \ref{Figure5}. The data fit fairly well the MFT prediction. However,
534: when the superexchange interactions are taken into account, the
535: physical scenario for this case will be more complicated and may
536: need further detailed investigations.
537: \section{Conclusion}
538: We have reexamined the magnetic and transport properties of the
539: La$_{0.7}$Sr$_{0.3}$Mn$_{1-x}M^\prime_x$O$_3$ system, where Mn is
540: selectively substituted by $M^\prime = \mathrm{Al}$ or Ti; either
541: Mn$^{3+}$ or Mn$^{4+}$ is selectively substituted for, but
542: $M^\prime$ is randomly distributed in the Mn network. The analyzes
543: of the $M(H)$ measurements and saturation magnetization revealed
544: that the substitution appears to not only merely dilute the magnetic
545: lattice of Mn ions, but also induce a redundancy of Mn ions. We have
546: introduced the \emph{selective dilution concentration}
547: $n_\mathrm{p}$ and discovered that in certain ranges of
548: $n_\mathrm{p}$, depending on particular $M^\prime$, $T_\mathrm{c}$
549: scales very well with $n_\mathrm{p}$ in linearity, being in good
550: agreement with the MFT approximation. The tendency of $T_\mathrm{c}$
551: to reduce to zero at $n_\mathrm{p}=1$ suggests a dominant role of
552: the DE mechanism in this system; the SE interaction is effectively
553: negligible. Remarkably, the linear behavior of
554: $T_\mathrm{c}(n_\mathrm{p})$ is observed in a very wide range of
555: $n_\mathrm{p}$ in LSMT$_x$ making this system an excellent candidate
556: for studies on the effects of dilution where $T_\mathrm{c}$ could be
557: almost independently tuned without side effects. It is also proposed
558: that the reduction in the number of hopping $e_\mathrm{g}$ electrons
559: is one cause for the difference in the effects of dilution between
560: the LSMT$_x$ and LSMA$_x$ compounds. The MFT analyzes are found to
561: some extend also valid for the case when Al and Ti are both
562: substituted for Mn.
563: \begin{acknowledgments}
564: This work has been performed partly under the sponsorship of a
565: collaborative project between the Institute of Materials Science
566: (VAST, Vietnam) and Uppsala University (Sweden). Two of us would
567: like to thank the University of Cincinnati and the National Science
568: Foundation for support. Dr. L. V. Bau would like to acknowledge the
569: financial support from the Ph.D Training Program of the Ministry of
570: Education and Training of Vietnam, and is in debt to the
571: collaboration and training project between the IMS and Hongduc
572: University.
573: \end{acknowledgments}
574: \begin{references}
575: \bibitem{Zener}
576: C. Zener, Phys. Rev. {\bf 82}, 403 (1951).
577: \bibitem{Millis1}
578: A. J. Millis, P. B. Littlewood, and B. I. Shraiman, Phys. Rev. Lett. {\bf 74}, 5144 (1995).
579: \bibitem{Millis2}
580: A. J. Millis, R. Mueller, B. I. Shraiman, Phys. Rev. Lett. {\bf 77}, 175 (1996).
581: \bibitem{Hwang}
582: H. Y. Hwang, S. W. Cheong, P. G. Radaelli, M. Marezio, and B. Batlogg, Phys. Rev. Lett. {\bf 75}, 914 (1995).
583: \bibitem{Dagotto}
584: E. Dagotto, T. Hotta, and A. Moreo, Phys. Rep. {\bf 344}, 1 (2001).
585: \bibitem{Helmolt}
586: R. V. Helmolt, J. Wecker, B. Holzapfel, L. Schultz, and K. Samwer, Phys. Rev. Lett. {\bf 71}, 2331 (1993).
587: \bibitem{Teresa}
588: J. M. de Teresa, M. R. Ibarra, J. Garc\'{i}a, J. Blasco, C. Ritter, P. A. Algarabel, C. Marquina, and A. del Moral, Phys. Rev. Lett. 76, 3392 (1996).
589: \bibitem{Qin}
590: H. Qin, J. Hu, J. Chen, H. Niu, and L. Zhu, J. Magn. Magn. Mater. {\bf 263}, 249 (2003).
591: \bibitem{Hu}
592: J. Hu, H. Qin, J. Chen, and Z. Wang, Mater. Sci. Eng. B {\bf 90}, 146 (2002).
593: \bibitem{Kallel}
594: N. Kallel, G. Dezanneau, J. Dhahri, M. Oumezzine, and H. Vincent, J. Magn. Magn. Mater. {\bf 261}, 56 (2003).
595: \bibitem{Kim}
596: M. S. Kim, J. B. Yang, Q. Cai, X. D. Zhou, W. J. James, W. B. Yelon, P. E. Parris, D. Buddhikot, and S. K. Malik, Phys. Rev. B {\bf 71}, 014433 (2005).
597: \bibitem{Troyanchuk}
598: I. O. Troyanchuk, M. V. Bushinsky, H. Szymczak, K. B\"{a}rner, and A. Maignan, Eur. Phys. J. B {\bf 28}, 75 (2002).
599: \bibitem{Sawaki}
600: Y. Sawaki, K. Takenaka, A. Osuka, R. Shiozaki, and S. Sugai, Phys. Rev. B {\bf 61}, 11588 (2000).
601: \bibitem{Blasco}
602: J. Blasco, J. Garc\'{i}a, J. M. de Teresa, M. R. Ibarra, J. Perez, P. A. Algarabel, and C. Marquina, Phys. Rev. B {\bf 55}, 8905 (1997).
603: \bibitem{Cao}
604: D. Cao, F. Bridges, M. Anderson, A. P. Ramirez, M. Olapinski, M. A. Subramanian, C. H. Booth, and G. H. Kwei, Phys. Rev. B {\bf 64}, 184409 (2001).
605: \bibitem{Coey}
606: J. M. D. Coey, M. Viret, and S. von Moln\'{a}r, Adv. Phys. {\bf 48}, 167 (1999).
607: \bibitem{Shannon}
608: R. D. Shannon, Acta Cryst. A {\bf 32}, 751 (1976).
609: \bibitem{Garcia}
610: J. L. Garc\'{i}a-Mu\~{n}oz, J. Fontcuberta, M. Suaaidi, and X. Obradors, J. Phys.: Condens. Matter {\bf 8}, L787 (1996).
611:
612: %\bibitem{Kanamori}
613: % J. Kanamori, J. Phys. Chem. Solids {\bf 10}, 87 (1959).
614: \end{references}
615: \end{document}
616: %
617: % ****** End of file apssamp.tex ******
618: