0707.3048/G2.tex
1: 
2: 
3: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%555
4: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
5: \documentclass[twocolumn,showpacs,prb,preprintnumbers,amsmath,amssymb]{revtex4}
6: % Some other (several out of many) possibilities\begin{normalsize}
7: %\documentclass[preprint,aps]{revtex4}
8: %\documentclass[preprint,aps,draft]{revtex4}
9: %\documentclass[prb]{revtex4}% Physical Review B
10: 
11: \usepackage{graphicx,graphics}% Include figure files
12: \usepackage{dcolumn}% Align table columns on decimal point
13: \usepackage{bm}% bold math
14: \usepackage{epsfig}
15: 
16: \begin{document}
17: 
18: 
19: 
20: \title{
21: Fine structure of the local pseudogap and Fano effect\\ 
22: for superconducting electrons near a zigzag graphene edge 
23: }
24: 
25: \author{Grigory Tkachov}
26: \affiliation{
27: Max Planck Institute for the Physics of Complex Systems, 01187 Dresden, Germany
28: }
29: 
30: %\date{\today}% It is always \today, today,
31:              %  but any date may be explicitly specified
32: 
33: \begin{abstract}
34: Motivated by recent scanning tunneling experiments 
35: on zigzag-terminated graphene 
36: this paper investigates an interplay of evanescent 
37: and extended quasiparticle states in 
38: the local density of states (LDOS) near a zigzag edge 
39: using the Green's function of the Dirac equation. 
40: A model system is considered where the local electronic structure 
41: near the edge influences transport of both normal 
42: and superconducting electrons via a Fano resonance. 
43: In particular, the temperature enhancement of the 
44: critical Josephson current and $0-\pi$ transitions are predicted. 
45: \end{abstract}
46: 
47: \pacs{73.23.Ad,74.50.+r,74.78.Na}
48: 
49: \maketitle
50: 
51: 
52: {\bf Introduction.}- 
53: Experimental evidence~\cite{Kostja05,Zhang05} 
54: for massless Dirac-like quasiparticles in graphene - 
55: a carbon monolayer with the hexagonal structure - 
56: has stimulated vigorous interest in electronic properties 
57: of this system 
58: (e.g. Refs.~\onlinecite{Gusynin05,Peres06,Brey06,McCann06,Nomura06,
59: Aleiner06,Tworzydlo06,Beenakker06,Cheianov07}). 
60: The unit cell of graphene contains two atoms 
61: each belonging to a triangular sublattice, 
62: and the low-energy states are described by a two-dimensional Dirac equation 
63: where the role of spin is assumed 
64: by the sublattice degree of freedom (pseudospin)~\cite{DiVincenzo84,Ando05}. 
65: Similar to relativistic spin-half particles in two dimensions, 
66: the graphene bulk density of states has a linear pseudogap~\cite{Kostja05} 
67: around zero energy $E=0$.
68: Natural boundaries can however give rise 
69: to additional spectral branches 
70: such as the low-energy edge states~\cite{Fujita96,Wakabayashi99}.
71: They are localized near a zigzag-shaped edge, 
72: whose outermost sites all belong to the same sublattice [Fig.~\ref{LDOSfig}(a)], 
73: and originate from the effective pseudospin "polarization" 
74: due to vanishing of one of the pseudospinor components as required by 
75: particle conservation~\cite{Brey06}.
76: Recent scanning tunneling experiments~\cite{Koba05,Niimi06} 
77: report a singular enhancement of the LDOS near zigzag boundaries 
78: attributed to the edge states. 
79: 
80: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
81: %%%%%%%%%%%%%% Figure1 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
82: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
83: \begin{figure}[b]
84: \epsfxsize=0.8\hsize
85: \epsffile{LDOS.eps}
86: \caption{(Color online)
87: (a) Schematic view of a zigzag graphene ribbon 
88: terminated by atomic lines belonging to different sublattices, 
89: conventionally denoted as A and B.   
90: (b)
91: LDOS vs. energy at different distances from the edge: 
92: (A) $d=4a$, (B) $d=10a$, (C) $d=30a$, (D)  $d=10000a$; 
93: where $a=0.246\, {\rm nm}$ is graphene's lattice constant,  
94: $\nu_a=1/8\pi\hbar va$ and $E_a=\hbar v/2a$.
95: (c)
96: LDOS vs. distance from the edge for different energies:
97: (A) $E=0.15E_a$, (B) $E=0.3E_a$, (C) $E=0.5E_a$. 
98: The delta function in Eq.~(\ref{LDOS}) is approximated 
99: by a Lorenzian $o/\pi(E^2+o^2)$ with $o=0.03E_a$.}
100: \label{LDOSfig}
101: \end{figure}
102: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
103: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
104: 
105: The measurements~\cite{Koba05,Niimi06} also revealed 
106: another peculiarity of the energy dependence of the LDOS -  
107: a fine oscillatory structure superimposed on the pseudogap 
108: with the amplitude enhanced at larger energies~\cite{Niimi06}. 
109: The origin of this behavior is still unaccounted for,  
110: although subsequent publications have studied the graphene LDOS,  
111: e.g. numerical simulations of Ref.~\onlinecite{Koba06} 
112: found damped spatial oscillations of the LDOS. 
113: The present study indends to show that both findings are   
114: consistent with the picture of interfering Dirac electron waves near a zigzag edge. 
115: To demonstrate this point, the one-particle Green's function 
116: of the Dirac equation was calculated for clean graphene with a zigzag edge described 
117: by the boundary condition of Ref.~\onlinecite{Brey06}. 
118: Then, the following expression for the LDOS  
119: $\nu(E,d)$, as a function of energy $E$ and distance $d$ from the edge, 
120: was obtained
121: \!
122: \begin{equation}
123: \nu(E,d)=|E|\frac{1+J_0\left(\frac{2Ed}{\hbar v}\right)}{\pi (2\hbar v)^2}-
124: \frac{ J_1\left(\left|\frac{2Ed}{\hbar v}\right|\right) }{ 4\pi\hbar v d }
125: +\frac{\delta(E)}{4\pi
126:  d^2}.
127: \label{LDOS}
128: \end{equation} 
129: %
130: Here the delta-functional term results from the dispersionless zero-energy edge state 
131: whereas the oscillating components given by the Bessel functions $J_0(2Ed/\hbar v)$ 
132: and $J_1(2|E|d/\hbar v)$ are due to interfering waves formed of the states 
133: belonging to the Dirac spectrum 
134: ($v$ and $\hbar$ are the electron velocity and Planck's constant).  
135: For $2|E|d/\hbar v\gg 1$ the amplitude of the oscillations is 
136: proportional to $\sqrt{|E|/d}$ [see also Figs.~\ref{LDOSfig}(b) and (c)], 
137: which qualitatively agrees with both the experiment~\cite{Niimi06} 
138: and numerical simulations~\cite{Koba06}. 
139: 
140: Another issue this study focuses on is the 
141: connection between the local electronic structure 
142: of zigzag-terminated graphene and Fano scattering~\cite{Fano}. 
143: Unlike earlier works [e.g. Ref.~\onlinecite{Waka00}] where 
144: the Fano effect was due to resonant flux states in finite-size ribbons, 
145: here the Fano resonance is studied in a nanowire side-coupled
146: to half-infinite graphene, by analogy with similar quantum-dot structures~\cite{Kobayashi02,Johnson04,Fuhrer06,Miriam07},
147: and originates from a single dispersionless edge state. 
148: Also, unlike Ref.~\onlinecite{Waka00}, the main focus here is on  
149: the transport of correlated electrons in Josephson nanowires.  
150: The Fano effect is predicted to cause quite unusual behaviors
151: of the critical current such as enhancement by temperature 
152: and, under certain conditions, $0-\pi$ transitions 
153: similar to those in ferromagnetic Josephson junctions~\cite{Buzdin82,Ryazanov01,Kontos01,Bauer04,Golubov04,Buzdin05}.  
154: In the context of the Josephson effect in graphene nanostructures 
155: (e.g. Refs.~\onlinecite{Titov06,Zareyan06}) 
156: these issues have not yet been addressed.   
157: 
158: 
159: {\bf Green's function of a zigzag ribbon.}- 
160: Assuming no scattering between the two valleys, $K$ 
161: and $K^\prime$, of graphene's Brillouin zone~\cite{Ando05}, 
162: one only needs to calculate the Green's function in one of them, 
163: e.g. $K$, where the Dirac equation reads    
164: $
165: [\sigma_0E + i\hbar v(\sigma_x\partial_x+\sigma_y\partial_y)]G=
166: \sigma_0\delta(x-x^\prime)\delta(y-y^\prime).	
167: $
168: %
169: Here the (retarded) Green's function matrix 
170: $G_{jk}$ with $j,k=A,B$, 
171: Pauli $\sigma_{x,y}$ and unity $\sigma_0$ matrices all act in pseudospin space. 
172: It suffices to solve the pair of equations for $G_{AA}$ and $G_{BA}$. 
173: After expanding in plane waves ${\rm e}^{ikx}$, 
174: the equations for the Fourier components are 
175: $
176: G_{BA|k}=(\hbar v/E)(k+\partial_y)G_{AA|k}$ 
177: and 
178: $
179: [\partial^2_y-q^2]G_{AA|k}=(E/\hbar^2 v^2)\delta(y-y^\prime)$
180: with $\quad q^2=k^2-(E/\hbar v)^2$. 
181: The solution can be sought in the form 
182: $G_{AA|k}(y,y^\prime)=a(y^\prime){\rm e}^{-qy}+b(y^\prime){\rm e}^{qy}
183: -E{\rm e}^{ -q|y-y^\prime| }/2\hbar^2 v^2q$,
184: where the last term is the Green's function of an unbounded system,  
185: and the coefficients $a(y^\prime)$ and $b(y^\prime)$ are to be found from the  
186: boundary conditions~\cite{Brey06} 
187: $G_{BA|k}|_{y=0}=(k+\partial_y)G_{AA|k}|_{y=0}=0$ and $G_{AA|k}|_{y=w}=0$.    
188: This yields the following result 
189: \!
190: \begin{eqnarray}
191: 	&&
192: 	G_{AA|k}(y,y^\prime)=E/2\hbar^2 v^2q
193: 	\times
194: 	\nonumber\\
195: 	&&
196: 	\times\left\{
197: 	\frac{k[ \cosh q(w-|y-y^\prime|)-\cosh q(w-y-y^\prime) ]}{q\cosh qw
198:  -k\sinh qw}-\right.
199: 	\nonumber\\
200: 	&&
201: 	\left.
202: 	-\frac{q[ \sinh q(w-|y-y^\prime|)+\sinh q(w-y-y^\prime) ]}{q\cosh qw
203:  -k\sinh qw}
204: 	\right\}.
205: 	\label{G_k}
206: \end{eqnarray}
207: %
208: The poles of $G_{AA|k}$, given by the equation $q=k\tanh qw$ 
209: (cf. Ref.~\onlinecite{Brey06}), determine the excitation spectrum.
210: As known~\cite{Fujita96,Wakabayashi99}, 
211: it has an almost flat branch merging with the Fermi level $E=0$  
212: corresponding to a state exponentially decaying from the edge into the interior. 
213: This can be easily seen from Eq.~(\ref{G_k}) in the limit $w\to\infty$:   
214: \!
215: \begin{equation}
216: 	G_{AA|k}(y,y^\prime)=-\frac{E{\rm e}^{-q|y-y^\prime|}}{2\hbar^2 v^2q}
217: 	+\frac{(q+k)^2{\rm e}^{-q(y+y^\prime)}}{2qE}.
218: 	\label{G_k_inf}
219: \end{equation}
220: %
221: The pole $E=0$ describes a dispersionless edge state existing for $k>0$. 
222: From Eq.~(\ref{G_k_inf}) an exact position representation for 
223: the Green's function 
224: $G_{AA}(xy,xy^\prime)=\int_{-\infty}^{\infty}dkG_{AA|k}(y,y^\prime)/(2\pi)$ 
225: can be obtained as
226: \!
227: \begin{eqnarray}
228: 	&&
229: 	G_{AA}(xy,xy^\prime)=
230: 	\frac{EY_0(k_E|y-y^\prime|)-i|E|J_0(k_E|y-y^\prime|)}{(2\hbar v)^2}
231: 	\nonumber\\
232: 	&&
233: 	+\frac{EY_0(k_E(y+y^\prime))-i|E|J_0(k_E(y+y^\prime))}{(2\hbar v)^2}
234: 	\nonumber\\
235: 	&&
236: 	-\frac{2EY_1(k_E(y+y^\prime))-2i|E|J_1(k_E(y+y^\prime))}{(2\hbar
237:  v)^2k_E(y+y^\prime)},
238: 	\label{G}
239: \end{eqnarray}
240: %
241: where $J_n(z)$ and $Y_n(z)$ ($n=0,1$) are, respectively, the Bessel and
242: Neumann functions, and $k_E=\sqrt{ E^2 }/\hbar v$. 
243: The LDOS [Eq.~(\ref{LDOS})] is obtained via 
244: $\nu(E,d)=-(1/\pi){\rm Im}G_{AA}(xy, xy^\prime)|_{y=y^\prime=d}$,
245: taking into account the pole of the function $Y_1$. 
246: We note that the interference of the waves 
247: incident at and reflected from the edge with small momenta $|k|\leq|E|/\hbar v$ 
248: produces spatial oscillations of the LDOS with the period much bigger than 
249: the lattice constant $a$ [Fig.~\ref{LDOSfig}(c)], i.e. 
250: well within the scanning tunneling microscop resolution.  
251: It is also instructive to examine Eq.~(\ref{G})
252: near the edge where it assumes a universal form, 
253: \!
254: \begin{equation}
255: G_{AA}(xy,xy^\prime)\approx 1/\pi E (y+y^\prime)^2, \qquad y,y^\prime\to 0,
256: \label{G1}
257: \end{equation}
258: %
259: independent of material parameters. 
260: To regulate the divergence at $y=y^\prime=0$, 
261: due to the effective continuum description,  
262: it is convenient to introduce the cutoff 
263: $G_{AA}(x0,x0)\approx 1/4\pi E d^2_c$ with $d_c\sim a$. 
264: 
265: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
266: %%%%%%%%%%%%%% Figure3 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
267: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
268: \begin{figure}[t]
269: %\begin{center}
270: \epsfxsize=1\hsize
271: \epsffile{Geo.eps}
272: %\end{center}
273: \caption{
274: (a) Schematic view of a zigzag-edge graphene ribbon with 
275: a side tunnel contact to a nanowire
276: connecting electron reservoirs 1 and 2. 
277: The contact is assumed point-like, i.e. 
278: its size is much bigger than the interatomic distances, 
279: but smaller than the electronic mean free paths in both wire and graphene. 
280: (b) Zero-temperature conductance $g$ (in units of $e^2/\pi\hbar$) vs. 
281: bias voltage $V$ for spin-degenerate ($h=0$) and spin-split
282: ($h=3\Gamma$) edge state in graphene. 
283: }
284: \label{Geo}
285: \end{figure}
286: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
287: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
288:    
289: {\bf Fano scattering off a zigzag edge.}-
290: The behavior of the Green's function near the edge 
291: can have a direct impact on charge transport.  
292: Let us consider a quasi-one-dimensional wire 
293: (with conventional quasiparticle spectrum) coupled in parallel
294: to a zigzag graphene edge via 
295: a point-like tunnel barrier [Fig.~\ref{Geo}(a)]. 
296: The contact is modelled by a real-space tunneling Hamiltonian of the form 
297: $H_T=\psi^\dagger_w(0)({\cal T}\psi_A({\bf r}_0)
298:   +{\cal T}^\prime\psi^\prime_A({\bf r}_0))+{\rm h.c.},\,
299:   {\bf r}_0=(0,0),$
300: where the electron operator in the wire $\psi^\dagger_w(0)$  
301: at the contact point $x=0$ is coupled to those in graphene on sublattice A in
302: both valleys $K$, $\psi_A({\bf r}_0)$ and  $K^\prime$, $\psi^\prime_A({\bf r}_0)$ 
303: with the matrix elements ${\cal T}$ and ${\cal T}^\prime$.   
304: To describe electron scattering caused by the contact, 
305: I use the equations-of-motion method and calculate 
306: the retarded Green's function in the wire 
307: \!
308: \begin{equation}
309: G_w(x,x^\prime)=G_w^{(0)}( x,x^\prime )+
310: 	        \frac{G_w^{(0)}(x,0) \Sigma\, G_w^{(0)}(0,x^\prime)}
311: 	             {1-G_w^{(0)}(0,0)\Sigma}.
312: \label{Gw}
313: \end{equation}
314: %
315: Here 
316: $G_w^{(0)}(x,x^\prime)={\rm e}^{ik_w|x-x^\prime|}/i\hbar v_F$ 
317: is the Green's function in the absence of tunneling 
318: ($k_w\approx k_F + E/\hbar v_F$ with $v_F$ and $k_F$ 
319: being the Fermi velocity and wave number in the channel),   
320: $\Sigma =|{\cal T}|^2G_{AA}({\bf r}_0,{\bf r}_0)
321:          +|{\cal T}^\prime|^2G^\prime_{AA}({\bf r}_0,{\bf r}_0)$
322: is the tunneling self-energy, and 
323: $G^\prime_{AA}({\bf r}_0,{\bf r}_0)$ is the Green's function 
324: in valley $K^\prime$ coinciding with $G_{AA}({\bf r}_0,{\bf r}_0)$ [Eq.~(\ref{G1})].
325: The transmission amplitude~\cite{Fisher81} between the reservoirs is
326: $t=i\hbar v_F G_w\left(\frac{L}{2},-\frac{L}{2}\right)={\rm e}^{ik_wL}E/(E+i\Gamma)$,
327: where $\Gamma =(|{\cal T}|^2+|{\cal T}^\prime|^2)/4\pi d_c^2\hbar v_F$ 
328: determines the tunneling rate $\Gamma/\hbar$ between the systems.
329: The Fano-like transmission antiresonance at $E=0$
330: manifests complete backscattering of an electron wave incoming 
331: from one of the reservoirs.  
332: It is due to destructive interference 
333: between the electron wave directly transmitted through the wire 
334: (without tunneling) and the wave transmitted 
335: via tunneling through the graphene edge state whose energy 
336: is pinned to the Fermi level in the wire.   
337: It is straightforward to generalize the analysis 
338: to a spin-split edge state with energies 
339: $\mp h$ for spin projections $\alpha=\pm 1/2$. 
340: In this case, we have 
341: \!
342: \begin{equation}
343: t_\alpha(E)={\rm e}^{ik_wL}(E+2\alpha h)/(E+2\alpha h +i\Gamma).
344: \label{t}
345: \end{equation}
346: %
347: Figure~\ref{Geo}(b) shows the voltage dependence of the zero-temperature
348: Landauer conductance 
349: $g(V)=e^2/(2\pi\hbar)\sum_{\alpha=\pm 1/2}|t_\alpha(eV)|^2$. 
350: For $h\not =0$ the conductance dip is split due to 
351: the spin-filtering effect discussed earlier~\cite{Torio04,Lee06} 
352: in the context of possible applications in spintronics~\cite{Fabian04}. 
353: 
354: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
355: %%%%%%%%%%%%%%%%%%%%%%%%%% Figure5 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
356: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
357: \begin{figure}[t]
358: %\begin{center}
359: \epsfxsize=0.6\hsize
360: \epsffile{IcT.eps}
361: %\end{center}
362: \caption{
363: Critical current vs. temperature for a spin-degenerate edge state:
364: $h=0$, $\gamma=0.1E_L$, $\Delta=10E_L$. 
365: The current is normalized to the value $I_c(T_{min})$ 
366: where $T_{min}=0.05E_L/k_B$ is the lowest temperature for which 
367: the condition $\gamma\leq\pi k_BT\ll \Delta$ of weak proximity effect still holds.}
368: \label{IcT}
369: \end{figure}
370: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
371: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
372: 
373: {\bf Fano effect in a Josephson junction.}-
374: Let us finally discuss the case of superconducting reservoirs  
375: supporting an equilibrium Josephson current.  
376: The Josephson coupling is maintained
377: due to the Andreev process~\cite{Andreev} 
378: whereby an electron is retro-reflected as a Fermi-sea hole 
379: from one of the superconductors with the subsequent hole-to-electron 
380: conversion in the other one. 
381: Such an Andreev reflection circle facilitates a Cooper pair transfer 
382: between the superconductors.
383: Since both electron and hole also experience normal scattering inside the junction, 
384: the transmission antiresonance is expected 
385: to strongly influence the Josephson current. 
386: It is convenient to use the approach of Refs.~\onlinecite{Beenakker,Brouwer} 
387: relating the supercurrent to the scattering amplitudes 
388: via a sum over the Matsubara frequencies $\omega_n=(2n+1)\pi k_BT$ as follows
389: \begin{equation}
390: I_c=-4ek_BT/\hbar\sum_{n\geq 0,\alpha}
391: {\rm a}^2_{\alpha}(E)t_{\alpha}(E)t^*_{-\alpha}(-E)|_{E=i\omega_n},
392: \label{Ic}
393: \end{equation} 
394: where $t^*_{-\alpha}(-E)$ is the hole transmission amplitude 
395: corresponding to the time-reversed counterpart of the electron 
396: Hamiltonian~\cite{deGennes,Beenakker},
397: and ${\rm a}_{\alpha}(E)$ is the Andreev reflection amplitude at the
398: point contacts to superconductors 1 and 2. 
399: Equation~(\ref{Ic}) is applicable for arbitrary $t_{\alpha}(E)$ 
400: as long as ${\rm a}^2_{\alpha}$ is small enough so that 
401: one can neglect higher order Andreev processes. 
402: In this case the Josephson current-phase relation is sinusoidal 
403: with $I_c$ [Eq.~(\ref{Ic})] being the critical value of the current. 
404: 
405: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
406: %%%%%%%%%%%%%% Figure6 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
407: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
408: \begin{figure}[t]
409: %\begin{center}
410: \epsfxsize=0.5\hsize
411: \epsffile{IcH.eps}
412: %\end{center}
413: \caption{
414: Critical current vs. spin-splitting energy:
415: $T=0.1E_L/k_B$, $\gamma=0.1E_L$, $\Delta=10E_L$.}
416: \label{IcH}
417: \end{figure}
418: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
419: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
420: 
421: In contacts to conventional Bardeen-Cooper-Schrieffer (BSC) superconductors, 
422: the Andreev process is well described by the scattering model of Ref.~\onlinecite{Blonder82}. 
423: However, in many practical cases superconducting contacts to low-dimensional 
424: systems can hardly be regarded as BCS-like ones. 
425: Proximity-effect contacts to 
426: semiconductor nanowires~\cite{vanDam,Tinkham} 
427: and carbon nanotubes~\cite{Jorgensen,Tsuneta} 
428: are important examples of such a situation. 
429: In this case a thin normal-metal layer is inserted between the superconductor and
430: the wire to ensure a good electrical contact.
431: In proximity-effect point contacts the Andreev scattering amplitude can be expressed 
432: in terms of the quasiclassical condensate ${\cal F}_{\alpha}(\omega_n)$ 
433: and quasiparticle ${\cal G}_{\alpha}(\omega_n)$ 
434: Green's functions of the normal layer as~\cite{GolKup,Volkov}
435: ${\rm a}_{\alpha}(\omega_n)=i{\cal F}_{\alpha}/(1+{\cal G}_{\alpha})$.
436: I will adopt this approach and make use of McMillan's expressions~\cite{McMillan,McGolubov} 
437: for the Green's functions:
438: ${\cal F}_{\alpha}=\Delta_n/\sqrt{\omega_n^2+\Delta_n^2}$, 
439: ${\cal G}_{\alpha}=(\omega_n/\Delta_n){\cal F}_{\alpha}$,
440: and 
441: $\Delta_n=\gamma\Delta/(\gamma+\sqrt{\omega_n^2+\Delta^2})$,
442: where $\Delta$ is the superconductor's pairing energy and $\gamma$
443: is McMillan's parameter controlling the strength 
444: of the proximity effect in the normal layer and, hence, the Andreev
445: reflection amplitude 
446: ${\rm a}_{\alpha}(\omega_n)=i\Delta_n/( \omega_n
447:  +\sqrt{\omega^2_n + \Delta^2_n} )$.  
448: For a weak proximity effect with $\gamma\leq\pi k_BT\ll\Delta$, 
449: the amplitude ${\rm a}^2_{\alpha}$ is small~\cite{Tkachov07} 
450: and equation~(\ref{Ic}) assumes the form
451: \!
452: \begin{equation}
453: I_c=\frac{8ek_BT}{\hbar}\sum\limits_{n\geq 0}
454: \frac{\Delta^2_n{\rm e}^{-\omega_n/E_L}}
455: {[\omega_n +\sqrt{\omega^2_n + \Delta^2_n} ]^2}
456: {\rm Re}\frac{( h+i\omega_n )^2}{[ h+i(\omega_n +\Gamma)  ]^2},
457: \nonumber
458: \end{equation} 
459: %
460: where the exponential factor results from the dynamical phase $2EL/\hbar v_F$
461: accumulated in the Andreev circle, introducing the Thouless energy  
462: $E_L=\hbar v_F/2L$. 
463: 
464: 
465: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
466: %%%%%%%%%%%%%% Figure7 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
467: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
468: \begin{figure}[t]
469: %\begin{center}
470: \epsfxsize=0.7\hsize
471: \epsffile{IcT_H.eps}
472: %\end{center}
473: \caption{
474: Critical current vs. temperature 
475: for a spin-polarized edge state:
476: $h=0.5E_L$, $\gamma=0.1E_L$, $\Delta=10E_L$.
477: }
478: \label{IcH_T}
479: \end{figure}
480: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
481: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
482: 
483: Figure~\ref{IcT} shows the temperature dependence of $I_c$
484: for the spin-degenerate case ($h=0$). 
485: In the absence of tunneling ($\Gamma=0$) it is just a monotonic
486: exponential decrease. However, for $\Gamma\not=0$ the interplay 
487: of the transmission antiresonance and the exponential suppression 
488: gives rise to a maximum at finite $T$.
489: Spin splitting of the graphene edge state lifts 
490: the Fano resonance condition $E=0$ for both electron 
491: and Andreev reflected hole. 
492: Therefore, for relatively weak tunneling coupling, 
493: when $\pi k_BT\leq\Gamma<E_L$,
494: the critical current increases with $h$ [Fig.~\ref{IcH}], 
495: a behavior quite unusual for Josephson junctions.   
496: Surprisingly, for stronger tunneling coupling ($\Gamma>E_L$) 
497: the function $I_c(h)$ becomes nonmonotonic with a rather broad region 
498: $h_1\leq |h|\leq h_2\approx \Gamma$ 
499: where $I_c$ is negative. 
500: The lower boundary $h_1\approx \pi k_BT$ is set by the temperature 
501: and is much smaller than all $E_L$, $\Gamma$ and $\Delta$. 
502: The supercurrent reversal is a consequence of the spin-dependent phases
503: acquired by both electron and hole due to scattering 
504: off the spin-polarized graphene edge, with   
505: negative values of $I_c$ implying a built-in $\pi$-phase difference 
506: in the ground state of a Josephson junction~\cite{Bulaevskii77} 
507: as opposed to $0$-phase difference for $I_c>0$. 
508: The $0-\pi$ transition can be driven by temperature as shown in Fig.~\ref{IcH_T}. 
509: Such a $\pi$ state is known to occur 
510: in ferromagnetic junctions 
511: where the condensate function oscillates 
512: in space~\cite{Buzdin82,Ryazanov01,Kontos01,Bauer04} 
513: (see, also recent reviews~\onlinecite{Golubov04,Buzdin05}). 
514: The author is not aware of any earlier work predicting $0-\pi$ transitions 
515: due to spin-dependent Fano scattering.
516: The main condition for this mechanism to work, i.e. $|h|>\pi k_BT$, 
517: can be met in the millikelvin region at modest external magnetic fields. 
518: 
519: The author thanks G. Cuniberti, M. Hentschel and C. Strunk for discussions. 
520: This work was partially funded by the European Union grant CARDEQ 
521: under Contract No. FP6-IST-021285-2. 
522: 
523: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
524: \begin{thebibliography}{00}
525: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
526: \bibitem{Kostja05}
527: K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, M. I. Katsnelson,
528: I. V. Grigorieva, S. V. Dubonos, and A. A. Firsov, Nature (London)
529: {\bf 438}, 197 (2005).
530: 
531: \bibitem{Zhang05}
532: Y. Zhang, Y.-W. Tan, H. L. Stormer, and P. Kim, {\em ibid.} {\bf
533:  438}, 201 (2005).
534: 
535: \bibitem{Gusynin05}
536: V. P. Gusynin and S. G. Sharapov, Phys. Rev. Lett. {\bf 95}, 146801 (2005).
537:  
538: \bibitem{Peres06}
539: N. M. R. Peres, F. Guinea, and A. H. Castro Neto, Phys. Rev. B {\bf 73}, 125411 (2006).
540: 
541: \bibitem{Brey06}
542: L. Brey and H. A. Fertig, {\em ibid.} {\bf 73}, 235411 (2006). 
543: 
544: 
545: \bibitem{McCann06}
546: E. McCann and V. I. Falko, Phys. Rev. Lett. {\bf 96}, 086805 (2006).
547: 
548: \bibitem{Nomura06}
549: K. Nomura and A. H. MacDonald, {\em ibid.} {\bf 96}, 256602 (2006).
550: 
551: \bibitem{Aleiner06}
552: I. L. Aleiner and K. B. Efetov, {\em ibid.} {\bf 97}, 236801
553:  (2006).
554: 
555: \bibitem{Tworzydlo06}
556: J. Tworzydlo, B. Trauzettel, M. Titov, A. Rycerz, 
557: and C. W. J. Beenakker, 
558: {\em ibid.} {\bf 96}, 246802 (2006).
559: 
560: \bibitem{Beenakker06}
561: C. W. J. Beenakker, {\em ibid.} 97, 067007 (2006).
562: 
563: \bibitem{Cheianov07}
564: V.V. Cheianov, V. Fal'ko, B. L. Altshuler, Science {\bf 315}, 1252 (2007).
565: 
566: 
567: \bibitem{DiVincenzo84}
568: D. P. DiVincenzo and E. J. Mele, Phys. Rev. B {\bf 29}, 1685 (1984).
569: 
570: \bibitem{Ando05}
571: T. Ando, J. Phys. Soc. Jpn. {\bf 74}, 777 (2005).
572: 
573: \bibitem{Fujita96}
574: M. Fujita, K. Wakabayashi, K. Nakada, and K. Kusakabe, 
575: {\em ibid.} {\bf 65}, 1920 (1996);
576: K. Nakada, M. Fujita, G. Dresselhaus, and M. S. Dresselhaus, 
577: Phys. Rev. B {\bf 54}, 17954 (1996).
578: 
579: \bibitem{Wakabayashi99}
580: K. Wakabayashi, M. Fujita, H. Ajiki, and M. Sigrist, {\em ibid.} {\bf 59}, 8271 (1999).
581: 
582: \bibitem{Koba05}
583: Y. Kobayashi, K. I. Fukui, T. Enoki, K. Kusakabe, and Y. Kaburagi,
584: {\em ibid.} {\bf 71}, 193406 (2005).
585: 
586: \bibitem{Niimi06}
587: Y. Niimi, T. Matsui, H. Kambara, K. Tagami, M. Tsukada, and H.
588: Fukuyama, {\em ibid.} {\bf 73}, 085421 (2006).
589: 
590: \bibitem{Koba06}
591: Y. Kobayashi, K. I. Fukui, T. Enoki, and K. Kusakabe, {\em ibid.} {\bf 73}, 125415 (2006).
592: 
593: \bibitem{Fano}
594: U. Fano, Phys. Rev. {\bf 124}, 1866 (1961).
595: 
596: \bibitem{Waka00}
597: K. Wakabayashi and M. Sigrist, Phys. Rev. Lett. 84, 3390 (2000).
598: 
599: \bibitem{Kobayashi02}
600: K. Kobayashi, H. Aikawa, S. Katsumoto, and Y. Iye, {\em ibid.} {\bf 88}, 256806 (2002).
601: 
602: \bibitem{Johnson04}
603: A. C. Johnson, C. M. Marcus, M. P. Hanson, and A. C. Gossard, {\em ibid.} {\bf 93}, 106803 (2004).
604: 
605: \bibitem{Fuhrer06}
606: A. Fuhrer, P. Brusheim, T. Ihn, M. Sigrist, K. Ensslin, W. Wegscheider, and M. Bichler, 
607: Phys. Rev. B {\bf 73}, 205326 (2006).
608: 
609: \bibitem{Miriam07}
610: M. del Valle, R. Gutierrez, C. Tejedor, and G. Cuniberti,	
611: Nature Nanotechnology {\bf 2}, 176 (2007).
612: 
613: 
614: \bibitem{Buzdin82}
615: A. I. Buzdin, L. N. Bulaevskii, and S. V. Panyukov, 
616: Pis'ma Zh. Eksp. Teor. Fiz. {\bf 35}, 147 (1982) [JETP Lett. {\bf 35}, 178 (1982)].
617: 
618: \bibitem{Ryazanov01}
619: V. V. Ryazanov, V. A. Oboznov, A. Yu. Rusanov, A. V. Veretennikov, A. A. Golubov, and J. Aarts, 
620: Phys. Rev. Lett. {\bf 86}, 2427 (2001). 
621: 
622: \bibitem{Kontos01}
623: T. Kontos, M. Aprili, J. Lesueur, and X. Grison, {\em ibid.} {\bf 86}, 304 (2001).
624: 
625: \bibitem{Bauer04}
626: A. Bauer, J. Bentner, M. Aprili, M. L. Della Rocca, M. Reinwald, W. Wegscheider, and C. Strunk, 
627: {\em ibid.} {\bf 92}, 217001 (2004). 
628: 
629: \bibitem{Golubov04}
630: A. A. Golubov, M. Yu. Kupriyanov, and E. Il'ichev, Rev. Mod. Phys. {\bf 76}, 411 (2004).
631: 
632: \bibitem{Buzdin05}
633: A. I. Buzdin, {\em ibid.} {\bf 77}, 935 (2005).
634: 
635: \bibitem{Titov06}
636: M. Titov and C. W. J. Beenakker, 
637: Phys. Rev. B {\bf 74}, 041401(R) (2006).
638: 
639: \bibitem{Zareyan06}
640: Ali G. Moghaddam and Malek Zareyan, {\em ibid.} {\bf 74}, 241403(R) (2006).
641: 
642: 
643: \bibitem{Fisher81}
644: D. S. Fisher and P. A. Lee, {\em ibid.} {\bf 23}, 6851 (1981).
645: 
646: \bibitem{Torio04}
647: M. E. Torio, K. Hallberg, S. Flach, A. E. Miroshnichenko, and M. Titov, Eur. Phys. J. B {\bf 37}, 399 (2004).
648: 
649: \bibitem{Lee06}
650: M. Lee and C. Bruder, Phys. Rev. B {\bf 73}, 085315 (2006).
651: 
652: \bibitem{Fabian04}
653: I. \u{Z}uti\'{c}, J. Fabian, and S. Das Sarma, Rev. Mod. Phys. {\bf 76}, 323 (2004).
654: 
655: 
656: 
657: \bibitem{Andreev}
658: A.F. Andreev, Zh. Eksp. Teor. Fiz. {\bf 46}, 1823 (1964) 
659: [Sov. Phys. JETP {\bf 19}, 1228 (1964)].
660: 
661: \bibitem{Beenakker}
662: C.W.J. Beenakker, Phys. Rev. Lett. {\bf 67}, 3836 (1991);
663: in {\em Transport Phenomena in Mesoscopic Systems}, 
664: edited by H. Fukuyama and T. Ando, p. 235 (Springer, Berlin 1992).
665: 
666: \bibitem{Brouwer}
667: P.W. Brouwer and C.W.J. Beenakker, 
668: Chaos, Solitons and Fractals {\bf 8}, 1249 (1997).
669: 
670: \bibitem{deGennes}
671: P.G. de Gennes, Rev. Mod. Phys {\bf 36}, 225 (1964).
672: 
673: \bibitem{Blonder82}
674: G.E. Blonder, M. Tinkham, and T.M. Klapwijk, 
675: Phys. Rev. B {\bf 25}, 4515 (1982).
676: 
677: 
678: \bibitem{vanDam}
679: J. A. van Dam, Y. V. Nazarov, E. P. A. M. Bakkers, S. De Franceschi, L.
680: P. Kouwenhoven, Nature {\bf 442}, 667 (2006). 
681: 
682: \bibitem{Tinkham}
683: Jie Xiang, A. Vidan, M. Tinkham, R. M. Westervelt  
684: and Charles M. Lieber, Nature Nanotechnology {\bf 1}, 208 (2006).
685: 
686: \bibitem{Jorgensen}
687: H.I. Jorgensen, K. Grove-Rasmussen, T. Novotny, K. Flensberg,
688: and P.E. Lindelof, Phys. Rev. Lett. {\bf 96}, 207003 (2006).
689: 
690: \bibitem{Tsuneta}
691: T. Tsuneta, L. Lechner, and P. J. Hakonen, {\em ibid.} {\bf 98},
692:  087002 (2007).
693: 
694: 
695: \bibitem{GolKup}
696: A.A. Golubov and M.Yu. Kupriyanov, Physica C {\bf 259}, 27 (1996).
697: 
698: \bibitem{Volkov}
699: A. F. Volkov and A. V. Zaitsev, Phys. Rev. B {\bf 53}, 9267 (1996).
700: 
701: \bibitem{McMillan}
702: W. L. McMillan, Phys. Rev. {\bf 175}, 537 (1968).
703: 
704: \bibitem{McGolubov}
705: A. A. Golubov, E. P. Houwman, J. G. Gijsbertsen, V. M. Krasnov, J.
706:  Flokstra, H. Rogalla, and M. Yu. Kupriyanov, 
707: Phys. Rev. B {\bf 51}, 1073 (1995).
708: 
709: \bibitem{Tkachov07}
710: G. Tkachov and K. Richter, {\em ibid.} {\bf 75}, 134517 (2007),
711: discusses magnetic pair breaking as 
712: a means of controlling the Andreev amplitude in quantum dot Josephson junctions.
713: 
714: \bibitem{Bulaevskii77}
715: L. N. Bulaevskii, V. V. Kuzii, and A. A. Sobyanin, 
716: Pis'ma Zh. Eksp. Teor. Fiz. {\bf 25}, 314 (1977) 
717: [JETP Lett. {\bf 25}, 290 (1977)].
718: 
719: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
720: \end{thebibliography}
721: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
722: 
723: 
724: 
725: 
726: \end{document}
727: 
728: 
729: % ****** End of file apssamp.tex ******
730: 
731: 
732: 
733: 
734: 
735: 
736: 
737: 
738: 
739: 
740: 
741: 
742: 
743: 
744: