1: %\documentclass[preprint,aps]{revtex4}
2: %\documentclass[preprint,aps,draft]{revtex4}
3: %\documentclass[prb]{revtex4}% Physical Review B
4: %\documentclass[prl]{revtex4}
5: \documentclass[twocolumn,showpacs,preprintnumbers,amsmath,amssymb]{revtex4}
6: %\documentclass[preprint,showpacs,preprintnumbers,amsmath,amssymb]{revtex4}
7:
8: \usepackage{graphicx}% Include figure files
9: \usepackage{dcolumn}% Align table columns on decimal point
10: \usepackage{bm}% bold math
11:
12: %\nofiles
13:
14: \begin{document}
15:
16: %\preprint{APS/123-QED}
17:
18: \title{Andreev Reflection and Pair Breaking Effects at the
19: Superconductor/Magnetic Semiconductor Interface}
20:
21: \author{R.P. Panguluri,$^1$ K.C. Ku,$^2$ T. Wojtowicz,$^{3,4}$
22: X. Liu,$^3$ J.K. Furdyna,$^3$ Y.B. Lyanda-Geller,$^5$
23: N. Samarth,$^2$ and B. Nadgorny$^1$}
24: \affiliation{$^1$Department of Physics and Astronomy, Wayne State
25: University, Detroit, MI 48201\\
26: $^2$Department of Physics and Materials Research Institute,\\
27: The Pennsylvania State University, University Park, PA 16802\\
28: $^3$Department of Physics, University of Notre Dame, Notre Dame, IN 46556\\
29: $^4$Institute of Physics, Polish Academy of Sciences, 02-668 Warsaw, Poland\\
30: $^5$Department of Physics, Purdue University, West Lafayette, IN
31: 47907}
32:
33: \date{\today}% It is always \today, today,
34: % but any date may be explicitly specified
35:
36: \begin{abstract}
37: We investigate the applicability of spin polarization measurements
38: using Andreev reflection in a point contact geometry in heavily doped
39: dilute magnetic semiconductors, such as (Ga,Mn)As. While we observe
40: conventional Andreev reflection in non-magnetic (Ga,Be)As epilayers,
41: our measurements indicate that in ferromagnetic (Ga,Mn)As epilayers
42: with comparable hole concentration the conductance spectra can only be
43: adequately described by a broadened density of states and a reduced
44: superconducting gap. We suggest that these pair-breaking effects stem
45: from inelastic scattering in the metallic impurity band of (Ga,Mn)As
46: and can be explained by introducing a finite quasiparticle lifetime or
47: a higher effective temperature. For (Ga,Mn)As with 8\% Mn concentration
48: and 140~K Curie temperature we evaluate the spin polarization to be
49: $83\pm 17\%$.
50: \end{abstract}
51:
52: \pacs{72.25.Dc,72.25.Mk,74.45.+c}
53: \maketitle
54:
55:
56: {}The advance of semiconductor spintronics has revived a long-standing
57: interest in understanding the coupling of charge and spin in
58: semiconductors \cite{ArOr}. Ferromagnetic semiconductors \cite{Dietl},
59: \cite{Ohno/Samarth} are of central importance to semiconductor spintronics
60: since they have a conductivity compatible with that of conventional
61: semiconductors and the potential for a high intrinsic spin
62: polarization, thus providing promising conditions for efficient spin
63: injection into conventional semiconductors. In this context the
64: ferromagnetic semiconductor (Ga,Mn)As with Curie temperatures
65: routinely reproducible in the range of $140 \lesssim T_C \lesssim
66: 170$K \cite{Ku}, \cite{Edmonds} stands out as
67: a well-studied model system \cite{Dietl}, \cite{Ohno/Samarth}. Furthermore,
68: (Ga,Mn)As has been successfully incorporated into a variety of spin
69: injection and spin transport devices \cite{GaMnAsjunction}. Although
70: measurements of the carrier (hole) spin polarization in this material
71: are immediately relevant to contemporary efforts in semiconductor
72: spintronics, systematic experiments probing this important quantity
73: are just beginning \cite{Braden}.
74:
75: Spin polarization measurements can be carried out using the tunneling
76: geometry in superconductor (S)/insulator (I)/ferromagnet (FM)
77: structures \cite{Tedrow}; however, attempts to use this technique in
78: (Ga,Mn)As have thus far been unsuccessful \cite{Braden}. In spite of some
79: theoretical problems \cite{Kelly}, Andreev reflection (AR) in a FM/S
80: contact \cite{Soulen,Buhrman} provides a viable alternative to tunneling
81: for measurements of the spin polarization ($P$) in a variety of
82: materials, including ferromagnetic metals and metallic oxides
83: \cite{Zutic}. Recently, AR measurements in planar junctions of Ga/(Ga,Mn)As
84: have estimated a value of $ P\sim 85\%$ for samples with 5\% Mn and $T_C=65$~K
85: \cite{Braden}. Despite extensive attempts to make epitaxial S/(Ga,Mn)As
86: planar junctions with a variety of superconductors, useful data has
87: been obtained only in a limited number of Ga/(Ga,Mn)As samples
88: \cite{Braden}, suggesting extreme sensitivity to
89: the nature of the
90: heterointerface. Additionally, the planar geometry has unavoidable
91: limitations imposed by constraints on the materials growth, limiting
92: the post-growth modifications of the sample characteristics \cite{capping}
93: and thus restricting AR measurements to (Ga,Mn)As samples with
94: relatively low Curie temperatures ($T_C \sim 65$~K). Finally, the conductance
95: spectra in Ref. \cite{Braden} have been explained using a distribution of
96: the energy gaps in the Ga superconducting film. An alternative
97: interpretation of this data has also been suggested, involving a
98: distortion of the density of states in a superconductor \cite{Kant}. In
99: this context, the measurements of spin polarization in (Ga,Mn)As in
100: the conventional point contact Andreev reflection (PCAR) geometry are
101: vital for both extending the range of sample parameters as well as for
102: resolving different interpretation of the data.
103:
104: In this Letter, we use PCAR to evaluate $P$ in (Ga,Mn)As epitaxial
105: layers with a high Curie temperature, $T_C=140$~K. In order to develop a
106: reliable interpretation of our study of (Ga,Mn)As, we first apply the
107: PCAR technique to a non-magnetic analogue of (Ga,Mn)As -- (Ga,Be)As --
108: with doping concentrations similar to those in the ferromagnetic
109: semiconductor ($ p \sim 10^{21} {\rm cm}^{-3}$). Our PCAR studies of
110: (Ga,Be)As yield
111: the data that is well described by a {\it conventional} weak coupling
112: Blonder-Tinkham-Klapwijk (BTK) model \cite{BTK}. In contrast, the PCAR
113: experiments in ferromagnetic (Ga,Mn)As cannot be described by a simple
114: BTK model modified for the spin-polarized case \cite{Mazin}. The (Ga,Mn)As
115: data indicate a significant broadening of the density of states (DOS)
116: accompanied by a reduction of the bulk superconducting gap $\Delta_b$. We note
117: that these observations are not an intrinsic characteristic of
118: ferromagnetic semiconductors: for instance, PCAR measurements of
119: (In,Mn)Sb -- a higher mobility ferromagnetic semiconductor --
120: consistently yield the bulk superconducting gap with no DOS broadening
121: \cite{InMnSb}. This suggests that our observations in (Ga,Mn)As stem from
122: inelastic scattering in a low mobility ferromagnetic semiconductor.
123:
124: Since the pioneering work of Kastalsky {\it et al.} \cite{Kastalsky}
125: most studies
126: of AR in semiconductors have been carried out in a 2D geometry. This
127: is not surprising, as serious problems are anticipated for AR
128: experiments in a superconductor-semiconductor (S/Sm) junction in a 3D
129: geometry due to the high resistivity of semiconductors at low
130: temperatures and the presence of a Schottky barrier in most S/Sm
131: contacts. The Schottky barrier fundamentally limits the accuracy of
132: spin polarization measurements in ferromagnetic semiconductors by
133: strongly decreasing the probability of AR. To avoid both these
134: problems, we use heavily doped (Ga,Be)As and (Ga,Mn)As semiconductors
135: with metallic type conductivity and thus thin Schottky barriers, which
136: make highly transparent S/Sm junctions \cite{Schottky}.
137:
138: The models of
139: ferromagnetism in (Ga,Mn)As discussed in the literature invoke either
140: free valence hole \cite{Dietl,McDonald}, or impurity bands
141: \cite{DasSarma/Bhatt}. Both Mn and Be are nominally acceptors, with binding
142: energies of 113 and 28~meV, respectively. As a result, at low doping
143: concentrations transport in both (Ga,Be)As and (Ga,Mn)As at our
144: characteristic experimental temperature $T=1$~K must be described by
145: impurity bands. At high doping levels of $N \sim 10^{21} {\rm
146: cm}^{-3}$ we assume
147: that impurity disorder yields spacial modulation near the top of the
148: valence band, which results in a metallic state with ``ballistic''
149: propagation through the contact for (Ga,Be)As, and ``diffusive''
150: propagation for (Ga,Mn)As. Note that the energy scale determined by
151: kinetic and potential energies $e^2N^{1/3}/\varepsilon$ and
152: $\hbar^2N^{2/3}/2m$,
153: respectively, is of the order of 100~meV, where $\varepsilon$ is the dielectric
154: constant and $m$ is the effective mass. This is especially important for
155: (Be,Ga)As, which behaves as a conventional heavily doped semiconductor
156: in which the valence band is modulated by the impurity potential.
157:
158: A number of 230~nm thick (Ga,Be)As samples with hole concentrations $p
159: = 8\times 10^{20} {\rm cm}^{-3}$ and $p = 5\times 10^{20} {\rm
160: cm}^{-3}$ were grown by low-temperature (LT)
161: molecular beam epitaxy (Riber 32 R\&D) on semi-insulating (001) GaAs
162: substrates. The 15~nm thick (Ga,Mn)As samples with a Mn composition of
163: 8\% were grown in an EPI 930 system on n+, epi-ready (001) GaAs
164: substrates using conditions described elsewhere \cite{Ku}. Post-growth
165: annealing of the (Ga,Mn)As samples at 250 $^\circ$C yielded $T_C =140$~K and a
166: resistivity $\rho \sim 2 m\Omega \cdot {\rm cm}$ at 4.2~K \cite{Ku}. A
167: point contact is established between the
168: sample and a mechanically polished Sn tip. Conductance ($dI/dV$) curves
169: were measured with the standard lock-in technique, as described in
170: detail in Ref. \cite{MnAs}, allowing us to monitor the characteristics of
171: same point contact from $\sim 1$~K to the critical temperature $T_c= 3.7$~K
172: of the Sn tip.
173:
174: To study the properties of AR in non-magnetic semiconductors, we have
175: measured a series of temperature dependencies for a large number of
176: different Sn/(Ga,Be)As point contacts. In Fig.~\ref{fig1}, we show the
177: evolution of $dI/dV$ for two typical contacts in a sample with a hole
178: concentration $p = 8\times 10^{20} {\rm cm}^{-3}$ and a residual
179: resistivity $\rho \sim 150 \mu \Omega \cdot {\rm cm}$.
180: \begin{figure}
181: \includegraphics{fig1final}
182: \caption{\label{fig1}Two Sn/(Ga,Be)As contacts at different reduced
183: temperatures $t = T/T_c$ analyzed with the model of
184: Ref. \cite{Mazin}. Left panel: Contact resistance $R_c= 35\Omega$,
185: $Z \sim 0.45$; right panel: $R_c = 28\Omega$, $Z\sim 0.8$. A small dip
186: above $\Delta$ is due
187: to the proximity effect. }
188: \end{figure}
189: Each
190: of the $dI/dV$ curves is analyzed independently using the model of
191: Ref. \cite{Mazin}, with the interface transparency of a contact
192: characterized by a dimensionless parameter $Z$. In this analysis, we use
193: the measured physical temperature of the contact and the corresponding
194: value of the Bardeen-Cooper-Schrieffer (BCS) gap $\Delta_b$. This procedure
195: results in a range of $Z$ for different contacts $0.4<Z<0.8$, with $Z$
196: practically temperature-independent for a given contact. The
197: resistivity and the measured carrier concentration for (Ga,Be)As yield
198: a mean free path $l\sim 10$~nm; this is comparable to the contact size
199: $d\sim 10$~nm
200: and suggests that the measurements occur in the ballistic transport
201: regime. Although complications may arise from the need to match wave
202: functions of different symmetry from (Ga,Be)As and Sn \cite{Kelly}, we will
203: attempt to describe the system phenomenologically following
204: Ref. \cite{BTK,ZuticSM}. For (Ga,Be)As we assume that the impurity and
205: the valence bands overlap, and thus we can still use light and heavy
206: holes to estimate the minimum $Z$-values, $Z=[(r-1)^2/4r]^{1/2}$,
207: which are due to the Fermi
208: velocity mismatch $r$ between the superconductor (Sn) and (Ga,Be)As,
209: $r=v_{\rm Sn}/v_{\rm GaAs}$. Simple estimates of $r$ for light and
210: heavy holes, $r_l \sim m_{lh}(n_{\rm Sn}/n_{\rm GaAs})^{1/3}) \sim 1.7$
211: and $r_h \sim m_{hh}(n_{\rm Sn}/n_{\rm GaAs})^{1/3}) \sim 4$, result
212: in $Z_{hh} \sim 0.8$ and $Z_{lh} \sim 0.3$,
213: in good agreement with the $Z$-values obtained from analyzing $dI/dV$
214: curves. The
215: experimental zero-bias conductance for the two contacts shown in
216: Fig.~\ref{fig1}
217: and the two corresponding curves obtained {\it independently} from the BTK
218: model with $Z=0.45$ and $Z=0.8$ are shown in Fig.~\ref{fig2}.
219: The surprisingly good agreement
220: between the data and the BTK model indicate the ``canonical'' AR,
221: typically observed in all-metal systems \cite{BTK}. Similar results were
222: obtained for (Ga,Be)As with $p= 5\times 10^{20}{ \rm cm}^{-3}$ and for
223: (In,Be)Sb
224: \cite{InMnSb}. These results - in conjunction with the estimates of the
225: $Z$-values based on the Fermi velocity mismatch - suggest that the use of
226: highly doped semiconductors can minimize the role of the Schottky
227: barrier in these measurements. Simple estimates yield the Schottky
228: barrier thickness in this system of the order of several \AA, confirming
229: these conclusions. This is also consistent with the experimentally
230: observed symmetric and linear $I - V$ characteristics above
231: $\Delta_b$.
232: \begin{figure}
233: \includegraphics{fig2final}
234: \caption{\label{fig2}Zero bias conductance for the two contacts shown
235: in Fig.~\ref{fig1}. Dashed lines show the exact results obtained
236: from the BTK model for $Z = 0.45$ and $Z = 0.8$ respectively. }
237: \end{figure}
238:
239: After
240: demonstrating that we can thoroughly understand the PCAR measurements
241: in the heavily-doped (Ga,Be)As, we now turn to the ferromagnetic
242: semiconductor (Ga,Mn)As with comparable carrier concentration. A
243: number of Sn contacts with (Ga,Mn)As epilayers have been
244: investigated. Qualitatively, all the contacts appear similar and
245: exhibit a zero-bias conductance that is significantly smaller than the
246: conductance at $V \gg \Delta/e$, suggesting a high spin polarization in
247: (Ga,Mn)As. However, analyzing the data is much more difficult compared
248: to both the non-magnetic case of (Ga,Be)As and the magnetic case of
249: (In,Mn)Sb \cite{InMnSb}. We find that AR in (Ga,Mn)As does not fit a model
250: of Ref. \cite{Mazin}, as all the experimental curves show a strong
251: broadening of the DOS and reduction of the superconducting gap (see
252: Fig.~\ref{fig3}).
253: \begin{figure}
254: \includegraphics{fig3final}
255: \caption{\label{fig3}Fits to the modified BTK model for Sn/(Ga,Mn)As
256: contact with $R_c= 68 \Omega$ measured at $T =
257: 1.2$~K. $T^{\ast}=5.2$~K was used for both fits. A reduced gap
258: provides a better fit (left panel). The variation of $\Delta$ results in
259: different $P$. }
260: \end{figure}
261:
262: Our results on (Ga,Mn)As as well as (In,Mn)Sb \cite{InMnSb}
263: indicate that transport processes occurring in the semiconductor are
264: most likely responsible for the observed effects \cite{field}. In the
265: ballistic regime $l>d$, typical for most AR experiments, the conductance is
266: determined by the transparency of the FM/S interface with the bulk
267: superconducting gap $\Delta_b$. For (In,Mn)Sb \cite{InMnSb}, our
268: estimates show that
269: holes are all in the ballistic regime. Thus the (In,Mn)Sb data is
270: easily interpreted in terms of a ballistic model \cite{Mazin}. In strong
271: contrast, transport in the (Ga,Mn)As impurity bands is ``diffusive'',
272: with $l \sim 2$~nm. While diffusive transport in AR experiments in
273: metals can be
274: described by conventional theory \cite{Mazin}, in (Ga,Mn)As with $d
275: \sim 150$~nm the holes
276: spend significant time $t^\ast \sim d^2/D \sim 10^{-10}$~s within the
277: contact area, where $D$ is the
278: diffusivity. Hence the holes experience enhanced inelastic and
279: spin-flip scattering \cite{Alt-Ar,highspin}, with $t^\ast$
280: comparable to the
281: hole scattering time $\tau_\varepsilon \sim 10^{-9} - 10^{-10}$~s due
282: to acoustic phonons and hole-hole
283: interaction. However, the observed broadened DOS and reduced
284: $\Delta_b$ both
285: require processes with shorter characteristic times $\sim 10^{-11}$~s, possibly
286: inelastic scattering off magnetic ions. We note that, while inelastic
287: scattering can explain the DOS broadening and gap reduction of the
288: superconductor, spin-flip scattering provides additional channels for
289: the Andreev current, introducing an uncertainty in spin polarization
290: measurements \cite{Falko}.
291:
292: We describe pair-breaking effects in the
293: Sn/(Ga,Mn)As contact using an empirical approach \cite{Dynes}, wherein we
294: account for inelastic scattering via an effective temperature, $T^\ast$, and a
295: reduced superconducting gap $\Delta$. Our approach provides a good description
296: of the experimental data, as seen in Fig.~\ref{fig3}. However, as we are unable
297: to evaluate $\Delta$ and $T^\ast$ from first principles, our model leads to the
298: uncertainty in $\Delta$ and a broadened DOS. This in turn yields a
299: fairly large
300: uncertainty in the extracted values of $P$ for (Ga,Mn)As, $P=83 \pm
301: 17\%$. We have
302: observed a qualitatively similar -- but quantitatively less
303: significant -- gap reduction and DOS broadening in ongoing PCAR
304: studies of the ferromagnetic semiconductor (Ga,Mn)Sb, whose mobility
305: is between that of (Ga,Mn)As and (In,Mn)Sb \cite{GaMnSb}. In that case, the
306: accuracy in determining the spin polarization is significantly better,
307: $P=57 \pm 5\%$
308: \cite{GaMnSb}. We note that, while we have a single critical temperature of
309: Sn, $T_c=3.7$~K, the observed spectra are qualitatively quite similar to
310: those in Ref. \cite{Braden}, suggesting that they may also be explained by
311: the gap reduction and DOS broadening. Our observations are consistent
312: with our conjecture that the PCAR measurements may suffer from
313: inelastic scattering effects that enhance the uncertainty in measuring
314: the spin polarization, particularly in highly spin-polarized materials
315: characterized by unconventional transport mechanism, such as
316: (Ga,Mn)As. This explanation is also in agreement with a recent
317: experiment in superconductor/normal metal nanostructures, in which Pt
318: impurities have been deliberately introduced at the
319: superconductor/normal metal interface to enhance inelastic scattering
320: \cite{privateBuhrman}.
321:
322: {\it Acknowledgments}: We thank R.A. Buhrman,
323: A.A. Golubov, E. Demler, V.I. Fal'ko, I.I. Mazin, S.A.Wolf, and
324: I. Zutic for useful discussions. The work is supported by the DARPA
325: SpinS through ONR grant N00014-02-1-0886 and NSF Career grant 0239058
326: (B.N.), by ONR grants N00014-99-1-0071, --0716, and N00014-99-1-1093
327: (N.S), and by the DARPA SpinS and NSF-NIRT Grant DMR02-01519 (J.K.F.).
328:
329: \begin{thebibliography}{99}
330:
331: \bibitem{ArOr} A. G.Aronov, {\it JEPT Lett.} {\bf 24}, 32 (1976); {\it
332: Optical
333: Orientation}, F. Meier
334: and B.P. Zakharchenya Eds., North Holland, New York, 1984.
335: \bibitem{Dietl} T. Dietl,
336: {\it Semicond. Sci. Technol.} {\bf 17}, 377 (2002);
337: \bibitem{Ohno/Samarth} H. Ohno in {\it Semiconductor
338: Spintronics and Quantum Computation}, Eds. D. D. Awschalom, D. Loss,
339: and N. Samarth, Springer-Verlag Berlin (2002), p. 1; N. Samarth in
340: {\it Solid State Physics}, {\bf Vol. 58}, Eds. H. Ehrenreich and F. Spaepan,
341: {\it Acad. Press} (2004), p.1.
342: \bibitem{Ku} K. C. Ku, {\it et al.}, {\it Appl. Phys. Lett.} {\bf 82}, 2302
343: (2003);
344: \bibitem{Edmonds} K. W. Edmonds {\it et al.}, {\it Phys. Rev. Lett.} {\bf
345: 92}, 037201 (2004).
346: \bibitem{GaMnAsjunction}Y. Ohno {\it et al.}, {\it Nature} {\bf 402}, 790 (1999);
347: M. Tanaka and Y. Higo,
348: {\it Phys. Rev. Lett.} {\bf 87}, art. no. (2001); S. H. Chun {\it et al.},
349: {\it Phys. Rev. B}
350: {\bf 66}, 100408(R) (2002); R. Mattana {\it et al.}, {\it Phys. Rev. Lett.} (2003).
351: \bibitem{Braden} J. G. Braden, {\it et al.}, {\it Phys. Rev. Lett.}
352: {\bf 91}, 056602 (2003); J.G. Braden,
353: {\it et al.}, {\it Phys. Rev. Lett.} {\bf 93}, 169704 (2004).
354: \bibitem{Tedrow}P.M. Tedrow and
355: R. Meservey, {\it Phys. Rep.} {\bf 238},173 (1994).
356: \bibitem{Kelly} K.Xia, {\it et al.},
357: {\it Phys. Rev. Lett.} {\bf 89}, 166603 (2002).
358: \bibitem{Beenakker} M.J.M. de Jong and
359: C.W.J. Beenakker, {\it Phys. Rev. Lett.} {\bf 74}, 1657 (1995).
360: \bibitem{Soulen} R.J. Soulen, et
361: al., {\it Science} {\bf 282}, 85 (1998).
362: \bibitem{Buhrman}S.K. Upadhyay, {\it et al.},
363: {\it Phys. Rev. Lett.} {\bf 81}, 3247 (1998).
364: \bibitem{Zutic} I. Zutic, J. Fabian, and S. Das
365: Sarma, {\it Rev. Mod. Phys.} {\bf 76}, 323 (2004).
366: \bibitem{capping} M.B. Stone {\it et al.},
367: {\it Appl. Phys. Lett.} {\bf 83}, 4568 (2003).
368: \bibitem{Kant} C.H Kant, {\it et al.},
369: {\it Phys. Rev. Lett.} {\bf 93},169703 (2004).
370: \bibitem{BTK} G. E. Blonder, M. Tinkham, and
371: T.M. Klapwijk, {\it Phys. Rev. B} {\bf 25}, 4515 (1982).
372: \bibitem{Mazin} I.I. Mazin,
373: A.A. Golubov, and B. Nadgorny, {\it Journ. Appl. Phys.} {\bf 89}, 7576 (2001).
374: \bibitem{InMnSb} R. P. Panguluri, {\it et al.}, {\it
375: Appl. Phys. Lett.} {\bf 84}, 4947 (2004).
376: \bibitem{Kastalsky} A. Kastalsky, {\it et al.}, {\it
377: Phys. Rev. Lett.} {\bf 67}, 3026
378: (1991).
379: \bibitem{Schottky} P.A. Barnes
380: and A.Y. Choi, {\it Appl. Phys. Lett.} {\bf 33}, 651 (1978).
381: \bibitem{McDonald} J. Konig, {\it et al.},
382: {\it Phys. Rev. Lett.} {\bf 84}, 5628 (2000).
383: \bibitem{DasSarma/Bhatt} P.M. Bercia and R.N. Bhatt,
384: {\it Phys. Rev. Lett.} {\bf 87},107203 (2000); A. Chattopadhyay, {\it et al.},
385: {\it Phys. Rev. Lett.} {\bf 87}, 227202 (2001); E.J. Singley {\it et al.},
386: {\it Phys. Rev. Lett.} {\bf 89}, 097203 (2002).
387: \bibitem{MnAs} R. P. Panguluri, {\it et al.},
388: {\it Phys. Rev. B} {\bf 68}, R201307 (2003).
389: \bibitem{ZuticSM} I. Zutic, and S. Das Sarma,
390: {\it Phys. Rev. B} {\bf 60}, R163222 (1999).
391: \bibitem{field} Local magnetic fields from a ferromagnet may lead to
392: suppression of superconductivity near the interface (e.g. Y. Miyoshi
393: {\it et al.}, {\it cond-mat} 0410264). This is possible if magnetic moment in
394: (Ga,Mn)As has an out-of-plane component. However, due to a compressive
395: strain, magnetization in 8\% Mn(Ga,Mn)As epilayers is in plane.
396: \bibitem{Alt-Ar} B.L. Altshuler and A.G. Aronov, Electron-Electron
397: Interactions in
398: Disordered Systems., Eds. A.L. Efros and M. Pollak, {\it Elsevier Science
399: Publishers}, Amsterdam, 1985.
400: \bibitem{highspin} For a highly spin polarized (Ga,Mn)As,
401: quasiparticles with stay in the contact area even longer, until they
402: find a partner of opposite spin to form a Cooper pair.
403: \bibitem{Falko} Another
404: possible mechanism of inelastic scattering is due to magnons, see
405: G. Tkachov, E. McCann, and V.I. Fal'ko, {\it Phys. Rev. B} {\bf 65}, 024519
406: (2001). This, however, should lead to a significant asymmetry of the
407: characteristics , which has not been observed.
408: \bibitem{Dynes} R.C. Dynes,
409: V. Narayanamurti, and J.P. Garno, {\it Phys. Rev. Lett.} {\bf 41}, 1509 (1978).
410: \bibitem{GaMnSb} R.P. Panguluri {\it et al.}, unpublished.
411: \bibitem{privateBuhrman} R.A. Buhrman, private
412: communication.
413:
414: \end{thebibliography}
415:
416:
417: \end{document}
418:
419:
420: