1: \documentclass[prl,aps,showpacs,twocolumn,unsortedaddress]{revtex4}
2: \usepackage{graphics,bm}
3: \usepackage{amssymb}
4: \usepackage{epsfig}
5: \usepackage{epsf}
6:
7: \begin{document}
8:
9: \title{Antiferromagnetic Metal Spintronics}
10:
11: \author{A. S. N\'u\~nez}
12: \email{alnunez@physics.utexas.edu}
13: \homepage{http://www.ph.utexas.edu/~alnunez}
14:
15: \author{R.A. Duine}
16: \email{duine@physics.utexas.edu}
17: \homepage{http://www.ph.utexas.edu/~duine}
18:
19: \author{A.H. MacDonald}
20: \email{macd@physics.utexas.edu}
21: \homepage{http://www.ph.utexas.edu/~macdgrp}
22:
23: \affiliation{The University of Texas at Austin, Department of
24: Physics, 1 University Station C1600, Austin, TX 78712-0264}
25:
26:
27: \date{\today}
28:
29: \begin{abstract}
30: Spintronics in ferromagnetic metals is built on a complementary
31: set of phenomena in which magnetic configurations influence
32: transport coefficients and transport currents alter magnetic
33: configurations. In this Letter we propose that corresponding
34: effects occur in circuits containing antiferromagnetic metals. The
35: critical current for switching can be smaller in the
36: antiferromagnetic case because of the absence of shape anisotropy
37: and because spin torques act through the entire volume of an
38: antiferromagnet. Our findings suggest that current-induced order
39: parameter dynamics can be used to coarsen the microstructure of
40: antiferromagnetic thin films.
41: \end{abstract}
42: \vskip2pc
43:
44: \maketitle
45:
46: \noindent {\em Introduction} --- Spintronics in ferromagnetic
47: metals\cite{Review} is based on one hand on the dependence of
48: resistance on magnetic microstructure \cite{GMRrefs}, and on the
49: other hand on the ability to alter magnetic microstructures with
50: transport currents
51: \cite{Slon,Berger,Tsoi1,Tsoi2,Sun,SMT-exp,Chien,MSU}. These
52: effects are often largest and most robust in circuits containing
53: ferromagnetic nanoparticles that have a spatial extent smaller
54: than a domain wall width and therefore largely coherent
55: magnetization dynamics. In this Letter we point out that similar
56: effects occur in circuits containing antiferromagnetic metals. The
57: systems that we have in mind are antiferromagnetic transition
58: metals similar to Cr\cite{FawcettI} and its alloys\cite{FawcettII}
59: or the rock salt structure intermetallics \cite{ExchangeBias} used
60: as exchange bias materials which are well described by
61: time-dependent mean-field-theory in its density-functional
62: theory\cite{Gunnarrson} setting.
63:
64: Our proposal that currents can alter the micromagnetic state of an
65: antiferromagnet may seem surprising since spin-torque effects in
66: ferromagnets \cite{TransferTheory} are usually discussed in terms
67: of conservation of total spin, a quantity that is not related to
68: the staggered moment order parameter of an antiferromagnet. Our
69: arguments are based on a microscopic picture of
70: spin-torques\cite{nunez2004} in which they are viewed as a
71: consequence of changes in the exchange-correlation effective
72: magnetic fields experienced by all quasiparticles in the transport
73: steady state. A spin torque that drives the staggered-moment
74: orientation $\mathbf{n}$ must also be staggered, and will be produced\cite{nunez2004}
75: by the exchange potential due to an {\em unstaggered}
76: transport electron spin-density in the plane perpendicular to
77: $\mathbf{n}$. The required alteration in torque is produced by the
78: alternating moment orientations in the antiferromagnet
79: rather than the transport electron exchange
80: field. As we now explain the transverse
81: spin-densities necessary for a staggered torque
82: occur generically in circuits containing
83: antiferromagnetic elements.
84:
85: The key observations behind our theory
86: concern the scattering properties of a single channel containing
87: non-collinear antiferromagnetic elements with a staggered exchange
88: field that varies periodically along the channel and is
89: commensurate with an underlying lattice that has inversion
90: symmetry. For an antiferromagnetic element that is invariant
91: under simultaneous spatial and staggered moment inversion it
92: follows from standard one-dimensional scattering theory
93: \cite{Melo} considerations that transmission through an individual
94: antiferromagnetic element is spin-independent, and that the
95: spin-dependent reflection amplitude from the antiferromagnet or
96: any period thereof has the form ${\bf r} = r_s \mathbf{1} + r_t \;
97: {\bf n} \cdot \vec{\tau}$, where ${\bf n}$ is the order parameter orientation
98: and $\vec{\tau}$ are the
99: Pauli spin matrices; $r_s$ and $r_t$ are proportional to sums and
100: differences of reflection amplitudes for incident spins oriented
101: along and opposite to the staggered moment. The reflection
102: amplitude for a spinors incident from opposite sides differ by changing
103: the sign of ${\bf n}$ and the transmission amplitudes are identical.
104: It then follows from composition rules for
105: transmission and reflection amplitudes in a compound circuit
106: containing paramagnetic source and drain electrodes and two
107: antiferromagnetic elements with staggered moment orientations
108: ${\bf n}_1$ and ${\bf n}_2$ separated by a paramagnetic spacer
109: (see Fig.~\ref{fig:cartoon}) that the transport electron
110: spin-density in the ${\bf n}_1 \times {\bf n}_2$ direction is
111: periodic in the antiferromagnets. (We define the direction of ${\bf
112: n}_i$ to be the direction of the local moment opposite the
113: spacer.) The spin-torques that appear in this type of circuit
114: therefore {\em act through the entire volume of each
115: antiferromagnet}.
116:
117: A proof of this property will be presented elsewhere. Here we
118: illustrate the potential consequences of this property by using
119: non-equilibrium Greens function techniques to evaluate
120: antiferromagnetic giant magnetoresistance (AGMR) effects and
121: layer-dependent spin-torques in model two-dimensional circuits
122: containing paramagnetic and antiferromagnetic elements. We focus
123: on the most favorable case in which the antiferromagnet has a
124: single ${\bf Q}$ spin-density-wave state with ${\bf Q}$ in the
125: current direction. In the following we first explain the model
126: system that we study and the non-equilibrium Greens function
127: calculation that we use to evaluate magnetoresistance and
128: spin-torque effects. We conclude that under favorable
129: circumstances, both effects can be as large as the ones that occur
130: in ferromagnets. We then estimate typical critical current for
131: switching an antiferromagnet. Finally, we discuss some of the
132: challenges that stand in the way of realizing these effects
133: experimentally.
134:
135: \noindent {\em Antiferromagnetic giant magnetoresistance} --- We
136: start by analyzing the simplified two-dimensional lattice model of
137: an antiferromagnetic heterostructure characterized by
138: near-neighbor hopping, transverse translational invariance, and
139: spin-dependent on-site energies, that is illustrated in
140: Fig.~\ref{fig:cartoon}:
141: \begin{eqnarray}\label{eq:singlepartHamiltonian}
142: \mathcal{H}_k &=& - t \sum_{\langle i,j \rangle,\sigma}
143: c^\dagger_{k,i,\sigma}\;
144: c^{\phantom{\dagger}}_{k,j,\sigma}+ \mbox{h.c.} \nonumber \\
145: &+&\sum_{i,\sigma,\sigma^\prime} \left[ ( \epsilon_i+\epsilon_k )
146: \delta_{\sigma,\sigma^\prime}- \Delta_i \hat
147: \mathbf{{\Omega}}_i\cdot\vec{\tau}_{\sigma,\sigma^\prime} \right]
148: c^\dagger_{k,i,\sigma}\; c^{\phantom{\dagger}}_{k,i,\sigma'}~.
149: \end{eqnarray}
150: Here, $k$ denotes the transverse wave number, $t$ the hopping
151: amplitude and $\epsilon_k$ the transverse kinetic energy. The second term in
152: Eq.~(\ref{eq:singlepartHamiltonian}) describes the exchange
153: coupling $\Delta_i$ of electrons to antiferromagnetically ordered
154: local moments $\hat {\mathbf \Omega}_i = (-)^{i} \mathbf{n}$ that alternate
155: in each antiferromagnet. In the paramagnetic
156: regions of these model systems $\Delta_i = 0$. The on-site
157: energies $\epsilon_i$ are allowed to change across a
158: heterojunction.
159:
160: \begin{figure}
161: \vspace{-0.5cm}
162: \centerline{\epsfig{figure=fig_1.eps,width=9.0cm}}
163: \caption{ The model heterostructure for which we perform our
164: calculations.
165: } \label{fig:cartoon}
166: \end{figure}
167:
168: We use the non-equilibrium Greens function formalism to describe
169: the transport of quasiparticles across the magnetic
170: heterostructure. The essential physical properties of the system
171: are encoded in the real time Greens function
172: \cite{caroli1972,dattabook}, defined by the ensemble average,
173: $G^{<}_{\sigma,i;\sigma^\prime,j} (k;t,t^\prime)= {\mathbf i}
174: \langle c^\dagger_{k,i,\sigma} (t)\;
175: c^{\phantom{\dagger}}_{k,j,\sigma}(t^\prime)\rangle$, from which
176: the (spin) current and (spin) density can be evaluated. To
177: determine the model's AGMR effect, we calculate the transmission
178: coefficient as a function of the angle $\theta$ between
179: orientations $\hat {\mathbf \Omega}_i$ on opposite sides of the
180: spacer. In Fig.~\ref{fig:transmission} the transmission
181: coefficient is shown for specific values of the number of layers
182: $N$ and $M$, in the first and second antiferromagnet. The AGMR
183: effect can be traced to the interference between spin-current
184: carrying electron spinors reflected by the facing layers. (This is
185: also the origin of spin transfer.) For the model we study the AGMR
186: depends on the orientation of the layers opposite the spacer in
187: the usual way, {\em i.e.} the resistance is highest for
188: $\theta=\pi$ and lowest for $\theta=0$. Also, we find that the
189: AGMR ratio, defined as the absolute difference between the maximum
190: and minimum value of the transmission coefficient normalized to
191: the minimum, saturates as a function of the length of the
192: antiferromagnets.
193:
194: \begin{figure}
195: \centerline{\epsfig{figure=fig_2.eps,width=7.0cm}}
196: \caption{Landauer-Buttiker conductance as a function of the angle
197: $\theta$ between the magnetization orientations $\hat {\mathbf
198: \Omega}_i$ on opposite sides of the paramagnetic spacer layer.
199: There is a sizable giant magnetoresistance effect, with larger
200: conductance at smaller $\theta$ and weak dependence on layer
201: thicknesses. These results were obtained for $\Delta/t=1$ and
202: $\epsilon_i=0$. } \label{fig:transmission}
203: \end{figure}
204:
205:
206: \noindent {\em Current-driven switching of an antiferromagnet} ---
207: To address the possibility of current-induced switching of an
208: antiferromagnet we evaluate spin transfer torques in the second
209: antiferromagnet. The spin transfer torque originates from the
210: contribution made by transport electrons to the
211: exchange-correlation effective magnetic field and is
212: given\cite{nunez2004} by ${\mathbf \Gamma}=\Delta_i \hat {\mathbf
213: \Omega}_i \times \langle {\bf s}_i\rangle/\hbar$, where $\langle
214: {\bf s}_i\rangle$ is the nonequilibrium expectation value of the
215: quasiparticle spin. We distinguish the spin-torque component in
216: the plane spanned by ${\bf n}_1$ and ${\bf n}_2$ and the component
217: out of this plane. In Fig.~\ref{fig:localST} we show the in-plane
218: and out-of-plane transport-induced spin torques. As anticipated the
219: in-plane spin transfer torque in this model is {\em exactly}
220: staggered and is therefore extremely effective in driving
221: order-parameter dynamics. We have checked numerically that
222: staggered in-plane spin-transfer torques that do not decay also
223: occur in continuum toy models of an antiferromagnet with
224: piece-wise constant and sinusoidal exchange fields. These
225: persistent spin torques are a generic property of
226: antiferromagnetic circuits related to the absence of
227: spin-splitting in the Bloch bands. The
228: staggered in-plane spin-transfer is produced by an out-of-plane
229: spin density that is {\em exactly} constant in our lattice model
230: antiferromagnet and exactly periodic in a continuum model
231: antiferromagnet.
232:
233: \begin{figure}
234: \centerline{\epsfig{figure=fig_3.eps, width=7cm}} \caption{Local
235: spin-transfer torques in the down-stream antiferromagnet. The
236: in-plane spin transfer is staggered and therefore effective in
237: driving coherent order parameter dynamics. The out-of-plane
238: spin-transfer is locally up to one order of magnitude larger, but
239: is ineffective because it is not staggered. These results were
240: obtained for $\Delta/t=1$, $\epsilon_i=0$, $\theta=\pi/2$, $N=50$,
241: and $M=50$.} \label{fig:localST}
242: \end{figure}
243:
244: If the exchange-interactions that stabilize the antiferromagnetic
245: state are very strong, the magnetization dynamics of each antiferromagnetic
246: element will be coherent and respond only to the staggered component of each spin-torque.
247: In Fig.~\ref{fig:STvsQ} we show the total staggered torque acting
248: on the downstream antiferromagnet, as a function of the angle
249: $\theta$. Clearly, the out-of-plane component of the torque is
250: small compared to the in-plane component. Fig.~\ref{fig:g} shows the derivative of the spin transfer
251: torque per unit current with respect to $\theta$, which we denote
252: $Mg(\theta)$, at $\theta=\pi$. As we will see, the critical
253: current for reversal is inversely proportional to this quantity.
254:
255: \begin{figure}
256: \vspace{-0.5cm} \centerline{\epsfig{figure=fig_4.eps,width=7.0cm}}
257: \caption{Total spin transfer torque action on the downstream antiferromagnet,
258: as a function of $\theta$. We used the parameters $\Delta/t=1$ and $\epsilon_i=0$.} \label{fig:STvsQ}
259: \end{figure}
260:
261: \begin{figure}
262: \vspace{-0.5cm}
263: \centerline{\epsfig{figure=fig_5.eps,width=7.0cm}}
264: \caption{Derivative of the total spin transfer torque per unit current, $M g(\theta=\pi)$, acting on the downstream antiferromagnet
265: with respect to the angle $\theta$ at $\theta=\pi$ as a function of $M$. We used the parameters $\Delta/t=1$ and $\epsilon_i=0$.} \label{fig:g}
266: \end{figure}
267:
268:
269: Having demonstrated the presence of spin transfer torques in a
270: heterostructure containing two antiferromagnetic elements, we
271: estimate the critical current for switching the second
272: antiferromagnet assuming that the first is pinned. To illustrate our ideas, we use the crystalline
273: anisotropy energy density for Cr \cite{FawcettI,fenton1978}, given
274: by
275: \begin{equation}
276: \label{eq:Eanis}
277: E( {\bf n})= K_1 (\hat {\bf z} \cdot {\bf n})^2 + K_2 (\hat {\bf x} \cdot {\bf
278: n})^2 (\hat {\bf y} \cdot {\bf
279: n})^2~,
280: \end{equation}
281: where ${\bf n}$ is a unit vector in the direction of the staggered
282: moment. The first term changes sign at the spin flop transition
283: \cite{FawcettI}, and forces the staggered moment to be either
284: parallel or perpendicular to the ordering vector ${\bf Q}$. At
285: room temperature ${\bf Q} \bot {\bf n}$, and for the geometry in
286: Fig.~\ref{fig:cartoon} we have ${\bf Q} \parallel \hat {\bf z}$
287: leading to the first term in Eq.~(\ref{eq:Eanis}). The term
288: involving $K_2$ describes cubic anisotropy in the plane
289: perpendicular to ${\bf Q}$.
290:
291: As we have seen, the spin transfer torques act cooperatively
292: throughout the entire antiferromagnet. Therefore, we can focus our
293: description on one ferromagnetic layer within the antiferromagnet,
294: since the antiferromagnet ordering will be preserved as the
295: antiferromagnet switches. Within this approach, the dynamics of
296: the staggered moment of the second antiferromagnet is analogous to
297: the ferromagnetic case, and its equation of motion reads
298: \begin{eqnarray}
299: \label{eq:eom}
300: \frac{d {\bf n}_2}{dt} &=& {\bf n}_2 \times \left[ -\frac{\gamma}{M_s} \frac{\partial E({\bf n}_2)}{\partial {\bf n}_2}\right]
301: + g (\theta) \omega_j {\bf n}_2 \times ({\bf n}_1 \times {\bf n}_2) \nonumber \\
302: &&- \alpha {\bf
303: n}_2 \times \frac{d{\bf n}_2}{dt}~.
304: \end{eqnarray}
305: Here, $\gamma\simeq \mu_{\rm B}/\hbar$ denotes the
306: gyromagnetic ratio, and $M_s \simeq \mu_{\rm B}/a^3$ denotes the
307: saturated staggered moment density, where $a \simeq 0.3$ nm
308: denotes the lattice constant of Cr. The term involving $\omega_j \equiv
309: \gamma \, \hbar j \, /(2 e a M_s)$, with $j$ the current density and $e$
310: the electron charge, describes the in-plane spin transfer torque.
311: We neglect the out-of-plane component because, as we have seen, it
312: averages to a small value. Moreover, the out-of-plane
313: component of the spin torque competes with the anisotropy, whereas
314: the in-plane component competes with the damping term. For this reason
315: it turns out that, even in ferromagnets, the in-plane component of
316: the spin torque is most important in determining the critical current for
317: current-driven switching. The last term in Eq.~(\ref{eq:eom}) describes the usual Gilbert
318: damping, with a dimensionless damping constant for which we take
319: the typical value $\alpha=0.1$ \cite{fenton1978}. The anisotropy
320: constants are given by $K_1=10^3$ J m$^{-3}$ and $K_2=10$ J
321: m$^{-3}$ \cite{fenton1978}.
322:
323: A linear stability analysis of Eq.~(\ref{eq:eom}) shows that for
324: the optimal situation ${\bf n}_1 = -\hat {\bf x}$, the fixed point
325: ${\bf n}_2 = \hat {\bf x}$ becomes unstable if
326: \begin{equation}
327: \label{eq:jc} j \equiv j_c = \frac{e \alpha a}{g(\pi) \hbar}
328: \left( K_1+8 K_2\right) \simeq 10^5 {\rm A~cm}^{-2},
329: \end{equation}
330: where the value for $g(\pi)$ is found to be $g(\pi)\simeq 0.05$.
331:
332: This critical current is smaller than the typical
333: value for switching an ferromagnet primarily because
334: the spin transfer torques act cooperatively throughout the
335: entire antiferromagnet and also because of the absence of shape anisotropy.
336: Using the model of Eq.~(\ref{eq:eom}) we also find that depending on the
337: applied current, the staggered moment ${\bf n}_2$ can relax to stable
338: fixed points at ${\bf n}_2 = \pm \hat {\bf y}$ or completely
339: reverse its direction.
340:
341: \noindent {\em Discussion and conclusions} --- The calculations we
342: have performed are in the ballistic regime, and we expect the AGMR
343: and spin transfer torque effect to occur only in sufficiently
344: clean samples. Since both effects rely on interferences,
345: however, we do not expect that disorder will make it impossible to
346: realize the effect in typical nanoscale layered systems. Initial experimental
347: explorations of this effect might be most easily interpreted in clean epitaxially
348: grown materials. The complicated antiferromagnetic domain structure, known to
349: play a complex role in exchange biasing materials \cite{nogues1999}, might
350: cause AGMR and antiferromagnetic spin transfer to be smaller
351: than expected on the basis of our calculation. We point out that it might
352: be possible to use the effects discussed here to coarsen the domain structure
353: of antiferromagnetic thin films. We therefore
354: expect that the metallic materials used for exchange biasing are
355: generally a good starting point in searching for materials
356: displaying these antiferromagnetic spintronics phenomena. The materials
357: combinations that will exhibit the effects we have in mind most strongly
358: depend on a large variety of considerations and can be identified by a
359: combination of experimental and theoretical work which follows in the
360: footsteps of the successful ferromagnetic metals materials research. Finally
361: we remark that related effects occur in hybrid circuits containing both
362: antiferromagnetic and ferromagnetic elements.
363:
364: In conclusion we propose that the experimental and theoretical
365: study of the influence of current on microstructure in circuits
366: containing antiferromagnetic elements will reveal interesting new
367: physics only partly anticipated in this Letter, and that
368: microstructure changes can be sensed by resistance changes. It is
369: a pleasure to thank Olle Heinonen, Chris Palmstrom, and Maxim Tsoi
370: for helpful remarks. This work was supported by the National
371: Science Foundation under grants DMR-0115947 and DMR-0210383, by a
372: grant from Seagate Corporation, and by the Welch Foundation.
373:
374:
375: \begin{thebibliography}{00}
376:
377: \bibitem{Review} S.A. Wolf {\it et al.}, Science {\bf 294}, 1488 (2001).
378:
379: \bibitem{GMRrefs} M. Baibich {\it et al.}, \prl {\bf 61}, 2472
380: (1988); J. Barnas {\it et al.}, \prb {\bf 42}, 8110 (1990).
381:
382: \bibitem{Slon} J.C. Slonczewski, J. Magn. Magn. Mater. {\bf 159}, L1
383: (1996).
384:
385: \bibitem{Berger} L. Berger, Phys. Rev. B {\bf 54}, 9353 (1996).
386:
387: \bibitem{Tsoi1} M. Tsoi {\em et al.}, Phys. Rev. Lett. {\bf 80},
388: 4281 (1998);
389: M. Tsoi \textit{et al.}, Nature {\bf 406}, 46 (2000);
390:
391: \bibitem{Tsoi2} M. Tsoi {\em et al.}, Phys. Rev. Lett. {\bf 89},
392: 246803 (2002).
393:
394: \bibitem{Sun} J.Z. Sun, J. Magn. Magn. Mater. {\bf 202}, 157
395: (1999).
396:
397: \bibitem{SMT-exp} E. B. Myers {\em et al.}, Science {\bf 285}, 867
398: (1999); J.A. Katine {\em et al.}, Phys. Rev. Lett. {\bf 84}, 4212
399: (2000); E.B. Myers, {\it et al.}, Phys. Rev. Lett. {\bf 89},
400: 196801 (2002); S.I. Kiselev, {\em et al.}, Nature {\bf 425}, 380
401: (2003); W.H. Rippard, M.R. Pufall, and T.J. Silva,
402: Appl. Phys. Lett. {\bf 82}, 1260 (2003);
403: F. B. Mancoff {\em et al.}, Appl. Phys. Lett. {\bf 83}, 1596
404: (2003).
405:
406: \bibitem{Chien} Y. Ji, C.L. Chien and M. D. Stiles,
407: Phys. Rev. Lett. {\bf 90}, 106601 (2003).
408:
409: \bibitem{MSU} S. Urazhdin {\it et al.}, Phys. Rev. Lett. {\bf 91},
410: 146803 (2003)
411:
412:
413:
414: \bibitem{FawcettI} E. Fawcett, Rev. Mod. Phys. {\bf 60}, 209 (1988).
415:
416: \bibitem{FawcettII} E. Fawcett, Rev. Mod. Phys. {\bf 66}, 25 (1994).
417:
418: \bibitem{ExchangeBias} A.E. Berkowitz and K. Takano, J. Magn. Magn. Mater. {\bf 200}, 552 (1999).
419:
420: \bibitem{Gunnarrson} O. Gunnarsson and B. I. Lundqvist,
421: Phys. Rev. B {\bf 13}, 4274 (1976).
422:
423: \bibitem{TransferTheory} J. Z. Sun, Phys. Rev. B {\bf 62}, 570 (2000 );
424: A. Brataas, Y. V. Nazarov, and G.E.W. Bauer, Phys. Rev. Lett. {\bf
425: 84}, 2481 (2000); X. Waintal {\it et al.}, Phys. Rev. B {\bf 62}
426: 12317 (2000); X. Waintal and P.W. Brouwer, Phys. Rev. B {\bf 63},
427: 220407 (2001); C. Heide, Phys. Rev. B {\bf 65}, 054401 (2002); M.
428: Stiles and A. Zangwill, Phys. Rev. B {\bf 65}, 014407 (2002);
429: J.-E. Wegrowe, Appl. Phys. Lett. {\bf 80}, 3775 (2002); S. Zhang,
430: P.M. Levy, and A. Fert, Phys. Rev. Lett. {\bf 88}, 236601 (2002);
431: G.E.W. Bauer, Y. Tserkovnyak, D. Huertas-Hernando, and A. Brataas,
432: Phys. Rev. B {\bf 67}, 094421 (2003); M.L. Polianski and P.W.
433: Brouwer, \prl {\bf 92}, 026602 (2004); A. Shapiro, P. M.
434: Levy, and S. Zhang, Phys. Rev. B {\bf 67}, 104430 (2003); A. Fert
435: {\em et al.}, J. Magn. Magn. Mater. {\bf 272}, (2004); Ya. B.
436: Bazaliy, B.A. Jones, and Shou-Cheng Zhang, Phys. Rev. B {\bf 69},
437: 094421 (2004).
438:
439: \bibitem{nunez2004} A.S. N\'u\~nez and A.H. MacDonald,
440: cond-mat/0403710.
441:
442: \bibitem{Melo} P.A. Mello and N. Kumar, {\it Quantum Transport in Mesoscopic
443: Systems}, Oxford University Press (2004).
444:
445: \bibitem{caroli1972} C. Caroli {\em et al.}, J. Phys. C: Solid State Physics {\bf 5},
446: 21 (1972).
447:
448: \bibitem{dattabook} S. Datta, {\em Electronic Transport in Mesoscopic Systems}, Cambridge University Press
449: (1995).
450:
451:
452: \bibitem{fenton1978} E. W. Fenton, J. Phys. F: Metal Phys. {\bf
453: 8}, 689 (1978).
454:
455: \bibitem{nogues1999} J. Nogues and I.K. Schuller, J. Magn. Magn.
456: Mater. {\bf 192}, 203 (1999).
457:
458:
459:
460:
461:
462:
463: \end{thebibliography}
464:
465:
466:
467: \end{document}
468: