1: %% ****** Start of file template.aps ****** %
2: %%
3: %%
4: %% This file is part of the APS files in the REVTeX 4 distribution.
5: %% Version 4.0 of REVTeX, August 2001
6: %%
7: %%
8: %% Copyright (c) 2001 The American Physical Society.
9: %%
10: %% See the REVTeX 4 README file for restrictions and more information.
11: %%
12: %
13: % This is a template for producing manuscripts for use with REVTEX 4.0
14: % Copy this file to another name and then work on that file.
15: % That way, you always have this original template file to use.
16: %
17: % Group addresses by affiliation; use superscriptaddress for long
18: % author lists, or if there are many overlapping affiliations.
19: % For Phys. Rev. appearance, change preprint to twocolumn.
20: % Choose pra, prb, prc, prd, pre, prl, prstab, or rmp for journal
21: % Add 'draft' option to mark overfull boxes with black boxes
22: % Add 'showpacs' option to make PACS codes appear
23: % Add 'showkeys' option to make keywords appear
24: %\documentclass[aps,prl,preprint,groupedaddress]{revtex4}
25: %\documentclass[aps,prl,preprint,groupedaddress]{revtex4}
26: %\documentclass[aps,prl,preprint,superscriptaddress]{revtex4}
27: %\documentclass[aps,prl,preprint,superscriptaddress]{revtex4}
28: \documentclass[aps,prl,preprint,groupedaddress]{revtex4}
29: %\documentclass[aps,prl,twocolumn,superscriptaddress,showpacs]{revtex4}
30:
31: \usepackage{graphicx}
32:
33: \def\gamnas{{Ga$_{1-x}$Mn$_{x}$As}}
34: \def\gacras{{Ga$_{1-x}$Cr$_{x}$As}}
35: \def\gacrn{{Ga$_{1-x}$Cr$_{x}$N}}
36:
37: % You should use BibTeX and apsrev.bst for references
38: % Choosing a journal automatically selects the correct APS
39: % BibTeX style file (bst file), so only uncomment the line
40: % below if necessary.
41: \bibliographystyle{apsrev}
42:
43: \begin{document}
44: \special{papersize=8.5 in, 11 in}
45:
46: % Use the \preprint command to place your local institutional report
47: % number in the upper righthand corner of the title page in preprint mode.
48: % Multiple \preprint commands are allowed.
49: % Use the 'preprintnumbers' class option to override journal defaults
50: % to display numbers if necessary
51: %\preprint{}
52:
53: %Title of paper
54: \title{Role of Disorder in Mn:GaAs, Cr:GaAs, and Cr:GaN}
55:
56: % repeat the \author .. \affiliation etc. as needed
57: % \email, \thanks, \homepage, \altaffiliation all apply to the current
58: % author. Explanatory text should go in the []'s, actual e-mail
59: % address or url should go in the {}'s for \email and \homepage.
60: % Please use the appropriate macro foreach each type of information
61:
62: % \affiliation command applies to all authors since the last
63: % \affiliation command. The \affiliation command should follow the
64: % other information
65: % \affiliation can be followed by \email, \homepage, \thanks as well.
66:
67:
68: \author{J.~L. Xu}
69: \affiliation{Arizona State University, Tempe, AZ, 85284}
70:
71: \author{M. van Schilfgaarde}
72: %\homepage[]{Your web page}
73: %\thanks{}
74: \affiliation{Arizona State University, Tempe, AZ, 85284}
75: \email[]{Mark.vanSchilfgaarde@asu.edu}
76:
77: \author{G.~D. Samolyuk}
78: \affiliation{Ames Laboratory, Iowa State University, Ames, IA, 50011}
79:
80: %Collaboration name if desired (requires use of superscriptaddress
81: %option in \documentclass). \noaffiliation is required (may also be
82: %used with the \author command).
83: %\collaboration can be followed by \email, \homepage, \thanks as well.
84: %\collaboration{}
85: %\noaffiliation
86: %Abstract: DE-6
87: \date{\today}
88:
89: \begin{abstract}
90:
91: We present calculations of magnetic exchange interactions and
92: critical temperature $T_c$ in \gamnas, \gacras\ and \gacrn. The
93: local spin density approximation is combined with a
94: linear-response technique to map the magnetic energy onto a
95: Heisenberg hamiltonion, but no significant further approximations
96: are made. Special quasi-random structures in large unit cells
97: are used to accurately model the disorder. $T_c$ is computed
98: using both a spin-dynamics approach and the cluster variation
99: method developed for the classical Heisenberg model.
100:
101: We show the following: $(i)$ configurational disorder results in
102: large dispersions in the pairwise exchange interactions; $(ii)$
103: the disorder strongly reduces $T_c$; $(iii)$ clustering in the
104: magnetic atoms, whose tendency is predicted from total-energy
105: considerations, further reduces $T_c$. Additionally the exchange
106: interactions $J(R)$ are found to decay exponentially with
107: distance $R^3$ on average; and the mean-field approximation is
108: found to be a very poor predictor of $T_c$, particularly when
109: $J(R)$ decays rapidly. Finally the effect of spin-orbit coupling
110: on $T_c$ is considered. With all these factors taken into
111: account, $T_c$ is reasonably predicted by the local spin-density
112: approximation in MnGaAs without the need to invoke compensation
113: by donor impurities.
114:
115: \end{abstract}
116:
117: \pacs{75.50.Pp, 75.30.Mb, 71.15.Mb}
118:
119: \maketitle
120:
121: \def\tc{$T_c$}
122:
123: Dilute magnetic semiconductors (DMS), i.e. semiconductors doped
124: with low concentrations of magnetic impurities (usually Cr, Mn,
125: or Co), have attracted much interest because of their potential
126: application to spintronics\cite{Ohno96,Ohno99} . \gamnas\ is the most
127: widely studied DMS, and it continues to attract interest because
128: it is one of the few DMS where it is generally agreed that the
129: magnetism is carrier-mediated. (This is important in spintronics
130: because the magnetic state can be manipulated by electrical or
131: optical means.)
132:
133: In recent years Curie temperatures in \gamnas\ have risen
134: steadily, reaching $\sim$170K for $x${}$\sim$0.08 when grown in
135: thin films annealed at low temperature
136: \cite{Chiba03,Ku03,Edmonds04}. It is generally believed defects
137: (probably Mn interstitials) migrate out of the as-deposited films
138: during the anneal, largely eliminating donor defects that hamper
139: ferromagnetism. Since most practical applications of spintronics
140: require room-temperature operation, a crucial question is then,
141: what is the ultimate limit to $T_c$ in the DMS compounds, and in
142: \gamnas\ in particular?
143:
144: This question was first addressed by Dietl in his now classic
145: paper\cite{Dietl00}, where he predicted a wide range of $T_c$ in
146: tetrahedrally coordinated alloys. This stimulated a great deal
147: of interest, although there is a growing consensus that most of
148: the claims of that paper were artifacts of the assumptions in his
149: original model. On the other hand, Akai\cite{Akai98} first used
150: the local spin-density approximation (LSDA) to estimate $T_c$
151: within the Coherent Potential Approximation (CPA) in (In,Mn)As;
152: he argued that a double exchange mechanism was a more appropriate
153: description of the magnetism than the $pd$ exchange assumed by
154: Dietl. Since then LSDA calculations of exchange interactions
155: have been performed by a variety of groups
156: \cite{mark01,Sandratskii02,Bouzerar03,Kudrnovsky04,Erwin03,Sato03},
157: usually extracting exchange parameters by calculating total
158: energies of a fixed atomic but multiple-spin configurations, or
159: by a linear-response technique within the CPA.
160:
161: To date, disorder has almost always been neglected or treated
162: within some mean-field (MF) approximation (MFA), either in the
163: computation of the exchange parameters themselves, or in the
164: subsequent analysis of magnetization $M(T)$ at
165: finite-temperature, or both (though better treatments
166: within $k\cdot p$ theory has been
167: reported \cite{Schliemann01}). The LSDA+MF predict a rather high
168: $T_c$ for \gamnas\ (typically 350$\sim$400~K for $x${}$\sim$0.08 \cite{Sato03}).
169: The large discrepancy with experiment (at least in Mn:GaAs) is
170: usually attributed to the very large numbers of compensating
171: defects in real samples, which reduce $T_c$
172: \cite{Chiba03,Ku03,Edmonds04}. The situation remains somewhat
173: uncertain because the number of defects still remaining in the
174: best samples to date is not known.
175:
176: This Letter addresses the issue of the ultimate limit to $T_c$ in
177: some DMS alloys (focusing on Mn:GaAs) by adopting relatively
178: rigorous approach to the calculation of the magnetic exchange
179: interactions and $T_c$. Random alloys are approximated by large
180: (128-250 atom) supercells where special quasirandom structures
181: (SQS) \cite{zunger90} are used for the cation sublattice. Using
182: a linear-response technique within the LSDA and the
183: linear-muffin-tin orbitals method\cite{licht87,mark99}, the
184: magnetic energy is mapped onto a Heisenberg form\cite{notea}
185: \begin{equation}
186: H = -\sum_{ij} J(R_{ij}) \, {\hat e_i}\cdot{\hat e_j}
187: \label{eq:heisenberg}
188: \end{equation}
189: where the sum is over all pairs $ij$ of magnetic atoms. To model
190: $M(T)$ and $T_c$, Eq.(\ref{eq:heisenberg}) is treated classically
191: and integrated using a spin-dynamics (SD) technique\cite{sdyn96};
192: alternatively $M(T)$ is estimated by the cluster variation method
193: (CVM)\cite{kikuchi} adapted\cite{notecvm} to solve
194: Eq.(\ref{eq:heisenberg}). Thus it is evaluated without recourse
195: to empirical parameters or to the MFA. We show that the widely
196: used MFA turns out be a very poor predictor of $M(T)$ in these
197: disorded, dilute alloys, dramatically overestimating $T_c$.
198:
199: With SQS we can rather precisely mimic a fully random
200: configuration, but it is also possible to consider configurations
201: that deviate from random. This can be important because LSDA
202: predicts a strong attractive interaction between magnetic
203: elements \cite{mark01}, which implies a tendency towards
204: clustering. In brief, we show that
205: \begin{itemize}
206: \item
207: the disorder induces large fluctuations in
208: $J_{ij}\equiv{}J(R_{ij})$ for every connecting vector
209: $R_{ij}$;
210: \item The fluctuations in $J_{ij}$ {\em{reduce}} $T_c$ relative
211: to the configurationally averaged $\overline
212: J_{ij}=\left<J_{ij}\right>$;
213: \item clustering {\em{reduces}} $T_c$, while ordering of the
214: magnetic elements {\em{increases}} $T_c$.
215: \end{itemize}
216:
217: \begin{figure}[ht]
218: \centering
219: \includegraphics[width=8.5cm]{jij.eps}
220: \caption{Pair exchange interactions $J(R_{ij})$, in mRy, for Mn:GaAs and
221: Cr:GaN at two different concentrations as a function of
222: $R_{ij}^3$. $R_{ij}$ is measured in units of the lattice
223: constant $a$.}
224: \label{fig:jij}
225: \end{figure}
226:
227: Fig.~\ref{fig:jij} shows $J_{ij}$ computed for an ensemble of
228: 108-cation (216-atom) random supercells following the method of
229: Ref.\cite{mark99}, for \gamnas\ and \gacrn\ alloys at $x$=4.6\%
230: and $x$=8.3\%. $3\times3\times3$ $k$-points were used, enabling
231: the calculation of $J$ to very distant neighbors. We chose these
232: two alloys because they are approximately representative of
233: limiting cases. For Cr:GaN, the GaN host has a wide bandgap, and
234: the Cr $t_2$ level falls near midgap. It broadens into an
235: impurity band with 1/3 occupancy, and is believed to be
236: responsible for the ferromagnetic exchange. For Mn:GaAs, most of
237: the weight of the Mn $t_2$-derived state falls below the valence
238: band maximum. A second $t_2$ impurity band about 0.1~eV above
239: the valence band maximum is mainly responsible for the
240: ferromagnetic exchange coupling in this case; the strength of
241: $J(R)$ depends critically on the amount of Mn character in this
242: band\cite{Mahadevan04}. Katayama-Yoshida used the $x$-dependence
243: of $J_0=\sum_R\overline J(R)$ (computed within the CPA) to
244: identify the ferromagnetism obtained from LSDA with model
245: theories\cite{Sato03}. Within the CPA, $J_0\sim{}x^{1/2}$ for Cr:GaN, which
246: corresponds to a double-exchange model, while Mn:GaAs displays
247: character intermediate between $J_0\sim{}x^{1/2}$ and the $pd$
248: exchange ($J_0\sim{}x$) usually assumed by $k\cdot p$ models
249: \cite{MacDonald99,Dietl00}.
250:
251: Comparing Cr:GaN to Mn:GaAs, Cr:GaN shows substantially stronger
252: nearest-neighbor (NN) interactions, owing to its small lattice
253: constant; however $\overline J(R_{ij})$ decays much more rapidly
254: with $R_{ij}$. This is because the wave function overlap between
255: transition metal $d$ states decays much more rapidly for midgap states than
256: near band-edge states. Evident also is the large dispersion in
257: $J_{ij}$ for fixed $R_{ij}$ (note $J$ is drawn on a log scale):
258: the root-mean square fluctuations $\Delta J_{ij} =
259: \sqrt{\left<J_{ij}^2-\bar{J}_{ij}^2\right>}$ are roughly
260: comparable to ${\overline J}$. However $\Delta J_{ij}$ {\em
261: increases} with $x$, and is substantially larger for the wide-gap
262: case (Cr:GaN). Note that there is little evidence in either
263: Cr:GaN or Mn:GaAs for oscillatory RKKY-like behavior, which in
264: the simplest approximation predicts $J(R)\sim \cos(2k_F{}R)/R^3$.
265: Instead, ${\overline{J}(R)}$ decays roughly exponentially in
266: $R^3$, corresponding to a Fermi surface with imaginary wave
267: number, as would obtain if the coupling were described by
268: tunneling via a disordered impurity band\cite{Berciu01}.
269:
270: %$J$ is further modulated by a sizeable dependence on crystal
271: %orientation which can be understood in terms of band-structure
272: %effects, and also by a large amount of randomness originating from the
273: %disorder.
274:
275: %As we will show in detail elsewhere, there is a systematic
276: %dependence of $J$ for the entire class of Mn- and Cr- doped III-V
277: %semiconductors. More generally the dependence of $J(R_{ij})$
278: %(average value and dispersion), on materials system can
279: %reasonably be accounted for on the host lattice constant, host
280: %bandgap, and the position of the transition-metal Cr or Mn $t_2$
281: %state relative to the host valence-band maximum. Some of these
282: %dependences have been noted by other researchers [Yoshida, Bruno,
283: %Kudnrovsky, Uppsalla, Dederichs, Erwin], but it is clear that the
284: %large dispersions shown in Fig.~\ref{fig:jij} require a careful
285: %accounting of the disorder for a quantitative theory.
286:
287: We now apply Eq.(\ref{eq:heisenberg}) to compute $M(T)$, focusing
288: on $T_c$. Mean-field theory, which estimates the effective field
289: at each site from the average field contributed by other sites,
290: predicts $T_c$ well above room temperature both in Mn:GaAs and
291: Cr:GaN\cite{Newman03,Sato03}. In spite of the
292: rather strong differences in the form of $J(R)$
293: (Fig.~\ref{fig:jij}), mean-field theory predicts that Mn:GaAs and
294: Cr:GaN have roughly similar $T_c$ for
295: $x${}$\sim$0.08\cite{Newman03}. This is because the NN
296: interaction in the latter case is strongest, but the $J$ decays
297: faster with $R$, leading to a comparable mean-field\cite{notemf}
298: estimate $\overline{T}^{MFA}_c$.
299:
300: But it should be evident from Fig.~\ref{fig:jij} that the MFA is
301: of questionable reliability. First, it is well known that for
302: dilute alloys there is a percolation threshold for the onset of
303: ferromagnetism. (The threshold in the present case cannot be
304: readily mapped to known models because $J(R)$ is nonneglible for
305: a rather large number of neighbors.) Moreover, the large
306: fluctuations $\Delta J(R)$ may strongly affect $T_c$, especially
307: since $\Delta J(R)$ itself is purely a function of the
308: environment\cite{mark01}, and consequently of the local
309: percolation path.
310:
311: To obtain a precise estimate for $M(T)$ and $T_c$, we adopt a
312: spin-dynamics approach\cite{sdyn96}. A 200 atom SQS structure
313: (250 atom for the 4\% alloy) was used to mimic the random alloy.
314: From the TM atoms in the SQS structure, a supercell containing
315: $\sim 2000$ Mn or Cr atoms was constructed to make a simulation
316: cell for prosecuting spin-dynamical simulations. Following the
317: method described in Ref.\cite{sdyn96}, the Landau-Lifshitz (L-L)
318: equation was integrated numerically at a fixed temperature
319: allowing the system to equilibrate, followed by a simulation for
320: $\sim 2\times10^6$ atomic units. The L-L equations were
321: integrated with the Bulirsch-Stoer method. As the L-L equation
322: is a first-order equation, global deamons were used for the heat
323: bath\cite{sdyn96}, to ensure ergodic behavior. The average
324: magnetization $\overline{M}(T)$ was computed as a function of
325: temperature, and $T_c$ was estimated from the inflection point in
326: $\overline{M}(T)$. Owing to finite-size effects and the
327: stochastic character of the simulation, $T_c$ could be determined
328: to a precision of $\sim$5\%.
329:
330: Also we employed a CVM approach recently adapted to the classical
331: Heisenberg hamiltonion\cite{notecvm}. This
332: relatively simple scheme has been found to be accurate in simple
333: 3$d$ magnets, overestimating $T_c$ by $\sim$5\% (similar to the
334: usual CVM for the Ising hamiltonion\cite{vaks99}). We can check the validity
335: both methods in the DMS case by comparing their predictions of
336: $T_c$. Fig~\ref{fig:tc} shows $T_c$ determined by both methods
337: for \gamnas\ and \gacras\ : agreement between the two methods is
338: $\sim$10\%, which is quite satisfactory considering the
339: complexity of the $J_{ij}$. $\overline{T}^{MFA}_c$ is also
340: shown: evidently the MFA rather badly overestimates $T_c$.
341: $\overline{T}^{MFA}_c>T_c$ by $\sim$200K in the Mn:GaAs alloy,
342: and by a somewhat larger amount in Cr:GaAs. The discrepancy is
343: still more dramatic in Cr:GaN (not shown); we find $T_c<50$K for
344: all concentrations studied while $\overline{T}^{MFA}_c\sim
345: 600$K\cite{Newman03}. Indeed we have found this generally to be
346: the case when $J(R)$ decays rapidly or when $\Delta J(R)/J(R)$ is
347: not small.
348:
349: These results stand in stark contrast to the $\sim$15\%
350: discrepancy between $T^{MFA}_c$ and $T_c$ typically found in
351: simple metals. The reason is easily understood by considering
352: the effective field a mean-field atom sees, $\vec
353: H^{eff}_i=\sum_jJ_{ij}{\hat e}_j$. From the exponential decay of
354: $J(R)$, it is evident that $H_i$ will be dominated by the nearest
355: neighbors. But for dilute alloys, near-neighbors are not
356: sufficent to form a percolation path. This is immediately
357: evident in the extreme case of a NN pair of magnetic atoms well
358: separated from any other magnetic atoms: the contribution to
359: $T^{MFA}_c$ from this pair would be high, even though the pair
360: would actually contribute nothing to ferromagnetism.
361:
362: % This picture explains
363: %naturally why a tight-binding analysis \cite{Berciu01} using MFA
364: %predicted that disorder {\em increases} $T_c$. It is also clear
365: %that the faster the rate of decay in $J$, the wider the
366: %discrepancy between $T^{MFA}_c$ and $T_c$.
367:
368: In Ref.~\cite{Bouzerar03} a small discrepancy between $T^{MFA}_c$
369: and a more sophisticated calculation for $T_c$ was reported. In
370: that calculation the CPA was used to construct an average
371: $\overline{J}_{ij}$ and $M(T)$ modeled by constructing a fcc
372: lattice of magnetic atoms, using concentrated-weighted
373: $\overline{J}_{ij}$ for the exchange parameters. It would seem
374: that their conclusions are an artifact of the neglect of
375: configurational disorder (except in the computation of
376: $\overline{J}_{ij}$). Better would be to estimate
377: $\overline{J}_{ij}$ within the CPA, and then construct a {\em
378: disordered} simulation cell using the $\overline{J}_{ij}$ to
379: estimate $M(T)$. Still this approach neglects fluctuations
380: $\Delta J$, which as we have seen are comparable to
381: $\overline{J}_{ij}$ itself. To assess the effect of
382: fluctuations, we repeated the calculation for $T_c$ within the
383: CVM, replacing the environment-specific ${J}_{ij}$ with the
384: configurationally averaged $\overline{J}_{ij}$. For \gamnas\ at
385: $x$=0.08, the effect of disorder ($\overline{J}_{ij}\to{J}_{ij}$)
386: was to reduce $T_c$ by 50~K. (It is interesting that the MFA
387: predicts the {\em opposite} trend, because of an artificial
388: tendency for $M^{MFA}(T)$ to track whichever site $i$ has the
389: largest $\vec H^{eff}_i$. Then $T^{MFA}_c-\overline{T}^{MFA}_c$
390: is positive\cite{notemf} and increases with $\Delta J/J$. This
391: explains why a tight-binding+MF analysis\cite{Berciu01} predicted
392: that disorder {\em increases} $T_c$.)
393:
394:
395: \begin{figure}[ht]
396: \centering
397: \includegraphics[width=6.0cm]{tc.eps}
398: \caption{Dependence of $T_c$ (K) on $x$ in \gamnas\ and \gacras.
399: Solid lines: $T_c$ computed from the MF
400: $\overline{T}^{MFA}_c$\protect\cite{notemf}. Dotted line:
401: $T_{c}$ extracted from spin-dynamics simulations of
402: Eq.~\ref{eq:heisenberg}. Diamonds: $T_{c}$ computed from the
403: Heisenberg Cluster Variation Method.
404: \label{fig:tc}
405: }
406: \end{figure}
407:
408: We next consider the effects of nonrandomness. As noted above,
409: real DMS alloys should exhibit some clustering owing to the
410: attractive interaction between magnetic elements\cite{mark01}.
411: The true situation is complicated by the nonequilibrium growth
412: required to stabilize the alloy in the zincblende structure.
413: Nevertheless the Mn-Mn or Cr-Cr binding energy is
414: calculated\cite{mark01} to be an order of magnitude larger than
415: the growth temperature ($\sim$250K), and some pairing or other
416: clustering should be expected, particularly since films must be
417: annealed to obtain good $T_c$. There is some experimental
418: evidence for a tendency to cluster\cite{Sullivan03}.
419:
420: The effect of clustering on $T_c$ in Ga$_{0.92}$Mn$_{0.08}$As was
421: studied by a simple model. To characterize the configurational
422: disorder we adopt the standard Ising formalism, and assign
423: $\sigma=\pm 1$ to each cation site (+1 for Mn and $-$1 for Ga).
424: The random (SQS) configuration was constructed by searching for
425: configurations which best approximate the ideal random
426: configuration for pair correlation functions
427: $P_{R_{ij}}=\left<\sigma_i\sigma_j\right>$ (and some higher-order
428: correlation functions) up to some fixed distance. For a random
429: configuration, $P_R=(2x-1)^2$ independent of $R$. To parameterize
430: the clustering in a simple manner, we adopted the NN pair
431: correlation function $P_1$ as a measure of clustering. Starting from
432: an initial SQS configuration, a simulated annealing cycle was
433: performed by generating a set of site configurations with increasing
434: $P_1$, corresponding to longer annealing times (For simplicity,
435: $P_n(n>1)$ was optimized to be $(2x-1)^2$ for each configuration.)
436: $J_{ij}$ and $T_c$ were computed by the CVM and MFA\cite{notemf} as a function
437: of $P_1$; see Fig.~\ref{fig:tccluster}. $T_c$ is rather strongly {\em
438: reduced} with increasing $P_1$. This is perhaps not surprising
439: since increased clustering implies more distant average
440: separation between atoms, which is deleterious to links in the
441: percolation path. Even within the MFA $T_c$ changes slightly,
442: albeit for a different reason. In that case, there is an
443: increase in NN pairs, which would increase $T_c$, but at
444: the same time there is some increase in the likelihood of {\em
445: three-} and higher body neighbors. The presence of a third
446: neighbor has the effect of {\em reducing} the pairwise $J_{ij}$
447: \cite{mark01}, and is the origin of the factor-of-three
448: variations in the NN $J$ in Fig.~\ref{fig:jij}.
449:
450: We also considered the {\em ordered} limit, by putting 1 Mn in a
451: 24-atom unit cell, corresponding to $x$=0.083. In this case
452: $P_1$ decreases to 2/3, and $T_c$ increases to 350K (see
453: Fig.~\ref{fig:tccluster}). Thus we conclude that ordering {\em
454: increases} $T_c$, while clustering {\em decreases} $T_c$.
455: Perhaps not suprisingly, the MFA $T_c$ approaches the CVM result
456: in the ordered case, since percolation is less critical.
457:
458:
459: \begin{figure}[ht]
460: \centering
461: \includegraphics[width=6.5cm]{tc-cluster.eps}
462: \caption{Dependence of $T_c$~(K) on the pair correlation function
463: $P_1$ in Ga$_{0.92}$Mn$_{0.08}$As. The random (SQS) configuration
464: corresponds to $P_1$=0.7056. (Two SQS structures were calculated.)
465: Diamonds show $T_c$ computed with CVM; circles show
466: $\overline{T}^{MFA}_c$, and triangles show ${T}^{MFA}_c$. The
467: point at $P_1=2/3$ corresponds to the ordered compound.
468: }
469: \label{fig:tccluster}
470: \end{figure}
471:
472: To conclude, we have shown that ferromagnetism is very sensitive
473: to configurational disorder in DMS alloys, and that with proper
474: treatment of disorder $T_c$ is reasonably predicted by the LSDA
475: for \gamnas, without needing to invoke compensating defects. We
476: briefly consider two important sources of error from elements
477: missing in the theory. First, spin-orbit coupling strongly
478: reduces $T_c$ in $k\cdot{}p$ models. We estimated its effect by
479: computing the change in $\overline{T}^{MFA}_c$ when the
480: $L\cdot{}S$ coupling is added to the LSDA hamiltonion. For
481: Ga$_{0.92}$Mn$_{0.08}$As, $\overline{T}^{MFA}_c$ was reduced by
482: $\sim$10\%. Finally, the LSDA itself will overestimate $T_c$
483: somewhat \cite{Mahadevan04}. In a future work we will present a
484: reliable parameter-free theory that corrects the principal errors
485: in LSDA---most importantly the Mn $d$ character at $E_F$---and
486: quantify the extent to which the LSDA overestimates $T_c$.
487: Finally, we conclude that the present calculations represent a
488: rather strict upper bound to $T_c$, and that for random or
489: clustered \gamnas\ alloys, $T_c>250$~K is unlikely.
490:
491: %\newpage
492: % Create the reference section using BibTeX:
493:
494: This work was supported by the Office of Naval Research.
495: %\bibliography{dms}
496:
497: \begin{thebibliography}{28}
498: \expandafter\ifx\csname natexlab\endcsname\relax\def\natexlab#1{#1}\fi
499: \expandafter\ifx\csname bibnamefont\endcsname\relax
500: \def\bibnamefont#1{#1}\fi
501: \expandafter\ifx\csname bibfnamefont\endcsname\relax
502: \def\bibfnamefont#1{#1}\fi
503: \expandafter\ifx\csname citenamefont\endcsname\relax
504: \def\citenamefont#1{#1}\fi
505: \expandafter\ifx\csname url\endcsname\relax
506: \def\url#1{\texttt{#1}}\fi
507: \expandafter\ifx\csname urlprefix\endcsname\relax\def\urlprefix{URL }\fi
508: \providecommand{\bibinfo}[2]{#2}
509: \providecommand{\eprint}[2][]{\url{#2}}
510:
511: \bibitem[{\citenamefont{Ohno et~al.}(1996)\citenamefont{Ohno, Shen, Matsukura,
512: Oiwa, Endo, Katsumoto, and Iye}}]{Ohno96}
513: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Ohno}},
514: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Shen}},
515: \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Matsukura}},
516: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Oiwa}},
517: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Endo}},
518: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Katsumoto}},
519: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Iye}},
520: \bibinfo{journal}{Applied Physics Letters} \textbf{\bibinfo{volume}{69}},
521: \bibinfo{pages}{363} (\bibinfo{year}{1996}).
522:
523: \bibitem[{\citenamefont{Ohno et~al.}(1999)\citenamefont{Ohno, Young, Beschoten,
524: Matsukura, Ohno, and Awschalom}}]{Ohno99}
525: \bibinfo{author}{\bibfnamefont{Y.}~\bibnamefont{Ohno}},
526: \bibinfo{author}{\bibfnamefont{D.~K.} \bibnamefont{Young}},
527: \bibinfo{author}{\bibfnamefont{B.}~\bibnamefont{Beschoten}},
528: \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Matsukura}},
529: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Ohno}}, \bibnamefont{and}
530: \bibinfo{author}{\bibfnamefont{D.~D.} \bibnamefont{Awschalom}},
531: \bibinfo{journal}{Nature} \textbf{\bibinfo{volume}{402}},
532: \bibinfo{pages}{790} (\bibinfo{year}{1999}).
533:
534: \bibitem[{\citenamefont{Chiba et~al.}(2003)\citenamefont{Chiba, Takamura,
535: Matsukura, and Ohno}}]{Chiba03}
536: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Chiba}},
537: \bibinfo{author}{\bibfnamefont{K.}~\bibnamefont{Takamura}},
538: \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Matsukura}},
539: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Ohno}},
540: \bibinfo{journal}{Appl. Phys. Lett.} \textbf{\bibinfo{volume}{82}},
541: \bibinfo{pages}{3020} (\bibinfo{year}{2003}).
542:
543: \bibitem[{\citenamefont{Ku et~al.}(2003)\citenamefont{Ku, Potashnik, Wang,
544: Chun, Schiffer, Samarth, Seong, Mascarenhas, Johnston-Halperin, Myers
545: et~al.}}]{Ku03}
546: \bibinfo{author}{\bibfnamefont{K.~C.} \bibnamefont{Ku}},
547: \bibinfo{author}{\bibfnamefont{S.~J.} \bibnamefont{Potashnik}},
548: \bibinfo{author}{\bibfnamefont{R.~F.} \bibnamefont{Wang}},
549: \bibinfo{author}{\bibfnamefont{S.~H.} \bibnamefont{Chun}},
550: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Schiffer}},
551: \bibinfo{author}{\bibfnamefont{N.}~\bibnamefont{Samarth}},
552: \bibinfo{author}{\bibfnamefont{M.~J.} \bibnamefont{Seong}},
553: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Mascarenhas}},
554: \bibinfo{author}{\bibfnamefont{E.}~\bibnamefont{Johnston-Halperin}},
555: \bibinfo{author}{\bibfnamefont{R.~C.} \bibnamefont{Myers}},
556: \bibnamefont{et~al.}, \bibinfo{journal}{Appl. Phys. Lett.}
557: \textbf{\bibinfo{volume}{82}}, \bibinfo{pages}{2302} (\bibinfo{year}{2003}).
558:
559: \bibitem[{\citenamefont{Edmonds et~al.}(2004)\citenamefont{Edmonds, Bogusawski,
560: Wang, Campion, Novikov, Farley, Gallagher, Foxon, Sawicki, Dietl
561: et~al.}}]{Edmonds04}
562: \bibinfo{author}{\bibfnamefont{K.~W.} \bibnamefont{Edmonds}},
563: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Bogusawski}},
564: \bibinfo{author}{\bibfnamefont{K.~Y.} \bibnamefont{Wang}},
565: \bibinfo{author}{\bibfnamefont{R.~P.} \bibnamefont{Campion}},
566: \bibinfo{author}{\bibfnamefont{S.~N.} \bibnamefont{Novikov}},
567: \bibinfo{author}{\bibfnamefont{N.~R.~S.} \bibnamefont{Farley}},
568: \bibinfo{author}{\bibfnamefont{B.~L.} \bibnamefont{Gallagher}},
569: \bibinfo{author}{\bibfnamefont{C.~T.} \bibnamefont{Foxon}},
570: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Sawicki}},
571: \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Dietl}},
572: \bibnamefont{et~al.}, \bibinfo{journal}{Phys. Rev. Lett.}
573: \textbf{\bibinfo{volume}{92}}, \bibinfo{pages}{037201}
574: (\bibinfo{year}{2004}).
575:
576: \bibitem[{\citenamefont{Dietl et~al.}(2002)\citenamefont{Dietl, Ohno,
577: Matsukura, Cibert, and Ferrand}}]{Dietl00}
578: \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Dietl}},
579: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Ohno}},
580: \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Matsukura}},
581: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Cibert}}, \bibnamefont{and}
582: \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Ferrand}},
583: \bibinfo{journal}{Science} \textbf{\bibinfo{volume}{287}},
584: \bibinfo{pages}{1019} (\bibinfo{year}{2002}).
585:
586: \bibitem[{\citenamefont{Akai}(1998)}]{Akai98}
587: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Akai}}, \bibinfo{journal}{Phys.
588: Rev. Lett.} \textbf{\bibinfo{volume}{81}}, \bibinfo{pages}{3002}
589: (\bibinfo{year}{1998}).
590:
591: \bibitem[{\citenamefont{van Schilfgaarde and Mryasov}(2001)}]{mark01}
592: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{van Schilfgaarde}}
593: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{O.}~\bibnamefont{Mryasov}},
594: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{63}},
595: \bibinfo{pages}{233205} (\bibinfo{year}{2001}).
596:
597: \bibitem[{\citenamefont{Sandratskii and Bruno}(2002)}]{Sandratskii02}
598: \bibinfo{author}{\bibfnamefont{L.~M.} \bibnamefont{Sandratskii}}
599: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Bruno}},
600: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{66}},
601: \bibinfo{pages}{134435} (\bibinfo{year}{2002}).
602:
603: \bibitem[{\citenamefont{Bouzerar et~al.}(2003)\citenamefont{Bouzerar,
604: Kudrnovsk\'y, Bergqvist, and Bruno}}]{Bouzerar03}
605: \bibinfo{author}{\bibfnamefont{G.}~\bibnamefont{Bouzerar}},
606: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Kudrnovsk\'y}},
607: \bibinfo{author}{\bibfnamefont{L.}~\bibnamefont{Bergqvist}},
608: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Bruno}},
609: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{68}},
610: \bibinfo{pages}{081203(R)} (\bibinfo{year}{2003}).
611:
612: \bibitem[{\citenamefont{Kudrnovsk\'y et~al.}(2004)\citenamefont{Kudrnovsk\'y,
613: Turek, Drchal, Maca, Weinberger, , and Bruno}}]{Kudrnovsky04}
614: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Kudrnovsk\'y}},
615: \bibinfo{author}{\bibfnamefont{I.}~\bibnamefont{Turek}},
616: \bibinfo{author}{\bibfnamefont{V.}~\bibnamefont{Drchal}},
617: \bibinfo{author}{\bibfnamefont{F.}~\bibnamefont{Maca}},
618: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Weinberger}}, ,
619: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Bruno}},
620: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{69}},
621: \bibinfo{pages}{115208} (\bibinfo{year}{2004}).
622:
623: \bibitem[{\citenamefont{Erwin and Hellberg}(2003)}]{Erwin03}
624: \bibinfo{author}{\bibfnamefont{S.~C.} \bibnamefont{Erwin}} \bibnamefont{and}
625: \bibinfo{author}{\bibfnamefont{C.~S.} \bibnamefont{Hellberg}},
626: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{68}},
627: \bibinfo{pages}{245206} (\bibinfo{year}{2003}).
628:
629: \bibitem[{\citenamefont{Sato et~al.}(2003)\citenamefont{Sato, Katayama-Yoshida,
630: and Dederics}}]{Sato03}
631: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Sato}},
632: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Katayama-Yoshida}},
633: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Dederics}},
634: \bibinfo{journal}{Europhys. Lett.} \textbf{\bibinfo{volume}{61}},
635: \bibinfo{pages}{403} (\bibinfo{year}{2003}).
636:
637: \bibitem[{\citenamefont{Schliemann et~al.}(2001)\citenamefont{Schliemann,
638: K{\"o}nig, Lin, and MacDonald}}]{Schliemann01}
639: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{Schliemann}},
640: \bibinfo{author}{\bibfnamefont{J.}~\bibnamefont{K{\"o}nig}},
641: \bibinfo{author}{\bibfnamefont{H.-H.} \bibnamefont{Lin}}, \bibnamefont{and}
642: \bibinfo{author}{\bibfnamefont{A.~H.} \bibnamefont{MacDonald}},
643: \bibinfo{journal}{Appl. Phys. Lett.} \textbf{\bibinfo{volume}{78}},
644: \bibinfo{pages}{1550} (\bibinfo{year}{2001}).
645:
646: \bibitem[{\citenamefont{Zunger et~al.}(1990)\citenamefont{Zunger, Wei,
647: Ferreira, and Bernard}}]{zunger90}
648: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Zunger}},
649: \bibinfo{author}{\bibfnamefont{S.-H.} \bibnamefont{Wei}},
650: \bibinfo{author}{\bibfnamefont{L.~G.} \bibnamefont{Ferreira}},
651: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{J.~E.}
652: \bibnamefont{Bernard}}, \bibinfo{journal}{Phys. Rev. Lett.}
653: \textbf{\bibinfo{volume}{65}}, \bibinfo{pages}{353} (\bibinfo{year}{1990}).
654:
655: \bibitem[{\citenamefont{Liechtenstein et~al.}(1987)\citenamefont{Liechtenstein,
656: Katsnelson, Antropov, and Gubanov}}]{licht87}
657: \bibinfo{author}{\bibfnamefont{A.~I.} \bibnamefont{Liechtenstein}},
658: \bibinfo{author}{\bibfnamefont{M.~I.} \bibnamefont{Katsnelson}},
659: \bibinfo{author}{\bibfnamefont{V.~P.} \bibnamefont{Antropov}},
660: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{V.~A.}
661: \bibnamefont{Gubanov}}, \bibinfo{journal}{J. Magn. Magn. Mater.}
662: \textbf{\bibinfo{volume}{67}}, \bibinfo{pages}{65} (\bibinfo{year}{1987}).
663:
664: \bibitem[{\citenamefont{van Schilfgaarde and Antropov}(1999)}]{mark99}
665: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{van Schilfgaarde}}
666: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{V.~P.}
667: \bibnamefont{Antropov}}, \bibinfo{journal}{J. Appl. Phys.}
668: \textbf{\bibinfo{volume}{85}}, \bibinfo{pages}{4827} (\bibinfo{year}{1999}).
669:
670: \bibitem[{not({\natexlab{a}})}]{notea}
671: \bibinfo{note}{Because the TM magnetic moments are large, the local-moment
672: picture and long-wave approximation is expected to be rather accurate; see
673: \cite{vladimir03}}.
674:
675: \bibitem[{\citenamefont{Antropov et~al.}(1996)\citenamefont{Antropov,
676: Katsnelson, Harmon, van Schilfgaarde, and Kusnezov}}]{sdyn96}
677: \bibinfo{author}{\bibfnamefont{V.~P.} \bibnamefont{Antropov}},
678: \bibinfo{author}{\bibfnamefont{M.~I.} \bibnamefont{Katsnelson}},
679: \bibinfo{author}{\bibfnamefont{B.~N.} \bibnamefont{Harmon}},
680: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{van Schilfgaarde}},
681: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{D.}~\bibnamefont{Kusnezov}},
682: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{54}},
683: \bibinfo{pages}{1019} (\bibinfo{year}{1996}).
684:
685: \bibitem[{\citenamefont{Kikuchi}(1951)}]{kikuchi}
686: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Kikuchi}},
687: \bibinfo{journal}{Phys. Rev.} \textbf{\bibinfo{volume}{81}},
688: \bibinfo{pages}{988} (\bibinfo{year}{1951}).
689:
690: \bibitem[{not({\natexlab{b}})}]{notecvm}
691: \bibinfo{note}{In contrast to the Ising model\cite{vaks99}, the classical
692: Heisenberg model has continuous rotational degrees of freedom which leads to
693: a more complicated form for the partition function: $Z_{ij} =
694: \sum_{l}(2l+1)i_l(\psi^{ij}_i/T)i_l(J_{ij}/T)i_l(\psi^{ij}_j/T)$, where
695: $i_l(z)=\sqrt{(\pi/2z)}I_{l+1/2}(z)$ and $I_n(z)$ is the modified Bessel
696: function of the first kind. Cluster methods can include strong correlation
697: among atoms, which the MFA neglects. The CVM accurately describes
698: interactions within the clusters (pair clusters in our case). The
699: interactions of the cluster with the environment is described by
700: renormalizing fields $\psi^{ij}$.}
701:
702: \bibitem[{\citenamefont{Mahadevan and Zunger}(2004)}]{Mahadevan04}
703: \bibinfo{author}{\bibfnamefont{P.}~\bibnamefont{Mahadevan}} \bibnamefont{and}
704: \bibinfo{author}{\bibfnamefont{A.}~\bibnamefont{Zunger}},
705: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{69}},
706: \bibinfo{pages}{115211} (\bibinfo{year}{2004}).
707:
708: \bibitem[{\citenamefont{Jungwirth et~al.}(1999)\citenamefont{Jungwirth,
709: Atkinson, Lee, and MacDonald}}]{MacDonald99}
710: \bibinfo{author}{\bibfnamefont{T.}~\bibnamefont{Jungwirth}},
711: \bibinfo{author}{\bibfnamefont{W.~A.} \bibnamefont{Atkinson}},
712: \bibinfo{author}{\bibfnamefont{B.~H.} \bibnamefont{Lee}}, \bibnamefont{and}
713: \bibinfo{author}{\bibfnamefont{A.~H.} \bibnamefont{MacDonald}},
714: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{59}},
715: \bibinfo{pages}{981} (\bibinfo{year}{1999}).
716:
717: \bibitem[{\citenamefont{Berciu and Bhatt}(2001)}]{Berciu01}
718: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{Berciu}} \bibnamefont{and}
719: \bibinfo{author}{\bibfnamefont{R.~N.} \bibnamefont{Bhatt}},
720: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{87}},
721: \bibinfo{pages}{107203} (\bibinfo{year}{2001}).
722:
723: \bibitem[{\citenamefont{Wu et~al.}(2003)\citenamefont{Wu, Liu, Gu, Singh, Budd,
724: van Schilfgaarde, McCartney, Smith, and Newman}}]{Newman03}
725: \bibinfo{author}{\bibfnamefont{S.}~\bibnamefont{Wu}},
726: \bibinfo{author}{\bibfnamefont{H.}~\bibnamefont{Liu}},
727: \bibinfo{author}{\bibfnamefont{L.}~\bibnamefont{Gu}},
728: \bibinfo{author}{\bibfnamefont{R.}~\bibnamefont{Singh}},
729: \bibinfo{author}{\bibfnamefont{L.}~\bibnamefont{Budd}},
730: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{van Schilfgaarde}},
731: \bibinfo{author}{\bibfnamefont{M.}~\bibnamefont{McCartney}},
732: \bibinfo{author}{\bibfnamefont{D.~J.} \bibnamefont{Smith}}, \bibnamefont{and}
733: \bibinfo{author}{\bibfnamefont{N.}~\bibnamefont{Newman}},
734: \bibinfo{journal}{Appl. Phys. Lett.} \textbf{\bibinfo{volume}{82}},
735: \bibinfo{pages}{3047} (\bibinfo{year}{2003}).
736:
737: \bibitem[{not({\natexlab{c}})}]{notemf}
738: \bibinfo{note}{Mean-field theory for the Heisenberg model approximates the
739: coupling to neighboring sites $j$ by an effective field
740: $\vec{H}^{eff}_i=\sum_j J_{ij}\hat{e}_j$ at site $i$, and entails the
741: solution of a set of coupled nonlinear equations in $H_i$ and $\hat{e}_i$.
742: Alternatively, we can replace $J_{ij}$ with its average $\overline{J}_{ij}$
743: and then calculate $M(T)$ for the $\overline{J}_{ij}$. We call $T_c$ computed
744: by the two approaches respectively, $T^{MFA}_c$ and $\overline{T}^{MFA}_c$.
745: The latter can be easily evaluated as $\overline{T}^{MFA}_c=2/3
746: \left<\sum_{ij}J_{ij}\right>$.}
747:
748: \bibitem[{\citenamefont{Vaks and Samolyuk}(1999)}]{vaks99}
749: \bibinfo{author}{\bibfnamefont{V.~G.} \bibnamefont{Vaks}} \bibnamefont{and}
750: \bibinfo{author}{\bibfnamefont{G.~D.} \bibnamefont{Samolyuk}},
751: \bibinfo{journal}{Sov. Phys. : JETP} \textbf{\bibinfo{volume}{88}},
752: \bibinfo{pages}{89} (\bibinfo{year}{1999}).
753:
754: \bibitem[{\citenamefont{Sullivan et~al.}(2003)\citenamefont{Sullivan, Boishin,
755: Whitman, Hanbicki, Jonker, and Erwin}}]{Sullivan03}
756: \bibinfo{author}{\bibfnamefont{J.~M.} \bibnamefont{Sullivan}},
757: \bibinfo{author}{\bibfnamefont{G.~I.} \bibnamefont{Boishin}},
758: \bibinfo{author}{\bibfnamefont{L.~J.} \bibnamefont{Whitman}},
759: \bibinfo{author}{\bibfnamefont{A.~T.} \bibnamefont{Hanbicki}},
760: \bibinfo{author}{\bibfnamefont{B.~T.} \bibnamefont{Jonker}},
761: \bibnamefont{and} \bibinfo{author}{\bibfnamefont{S.~C.} \bibnamefont{Erwin}},
762: \bibinfo{journal}{Phys. Rev. B} \textbf{\bibinfo{volume}{68}},
763: \bibinfo{pages}{235324} (\bibinfo{year}{2003}).
764:
765: \end{thebibliography}
766:
767: \end{document}
768:
769: