1: \documentclass[12pt]{iopart}
2: \usepackage{iopams}
3: \usepackage{graphics}
4: \usepackage[next]{inputenc}
5: \usepackage[dvips]{epsfig}
6: \begin{document}
7: \title{Transport of charge and spin in the weak link between misoriented $PrOs_4Sb_{12}$ superconductors}
8: \author{G. Rashedi$^{1}$\footnote[7]{rashedy@iasbs.ac.ir} and Yu. A. Kolesnichenko$^{1,2}$}
9: \address{$^1$ Institute for Advanced Studies in Basic Sciences, 45195-159,
10: Zanjan, Iran\\ $^2$ B.Verkin Institute for Low Temperature Physics
11: Engineering of National Academy of Sciences of Ukraine, 47,
12: Lenin ave , 61103, Kharkov, Ukraine}
13: \date{\today}
14: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
15: \begin{abstract}
16: Recently, the ``$(p+h)-$wave'' form of pairing symmetry has been
17: proposed for the superconductivity in $PrOs_{4}Sb_{12}$ compound
18: [Parker D, Maki K and Haas S, \textbf{cond-mat/0407254}]. In the
19: present paper, a stationary Josephson junction as a weak-link
20: between $PrOs_{4}Sb_{12}$ triplet superconductors is theoretically
21: investigated. The quasiclassical Eilenberger equations are
22: analytically solved. The spin and charge current-phase diagrams
23: are plotted and the effect of misorientation between crystals on
24: the spin current, and spontaneous and Josephson currents is
25: studied. It is found that such experimental investigations of the
26: current-phase diagrams can be used to test the pairing symmetry in
27: the above-mentioned superconductors. It is shown that this
28: apparatus can be applied as a polarizer for the spin current.
29: \end{abstract}
30: \pacs{74.50.+r, 74.20.Rp, 72.25.-b, 74.70.Pq, 74.70.Tx}
31: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
32: \maketitle
33: \section{Introduction}
34: The pairing symmetry of the recently discovered superconductor
35: compound $PrOs_{4}Sb_{12}$ is an interesting topic of research in
36: the field of superconductivity \cite{Maki1,Maki2,Maki3}.
37: Superconductivity in this compound was discovered in papers
38: \cite{Bauer1,Bauer2,Kotegawa} and two different phases ($A$ and
39: $B$) have been considered for this kind of superconductor in
40: Refs.\cite{Maki2,Nakajima}. Although authors of \cite{Maki2} at
41: first considered the spin-singlet ``$(s+g)-$wave'' pairing
42: symmetry for this superconductor, later it was specified that the
43: spin-triplet is the real pairing symmetry of the $PrOs_{4}Sb_{12}$
44: complex \cite{Maki1,Tou}. Using the Knight shift in NMR
45: measurement authors of paper \cite{Tou}, estimated the
46: spin-triplet pairing symmetry for the superconductivity in $PrOs
47: _{4}Sb_{12}$. Consequently, the ``$(p+h)-$wave'' model of the
48: order parameter was proposed for the pairing symmetry of the
49: superconductivity in
50: $PrOs_{4}Sb_{12}$ compound, recently \cite{Maki1}. In the paper \cite{Maki1}%
51: , the self-consistent equation for the superconducting gap $\Delta
52: \left( T\right) $ (BCS gap equation) has been solved for the
53: finite temperature $T$ numerically and for the temperatures $T$
54: close to zero and the critical
55: temperature $T_{c}$ analytically. For this compound, using the ``$(p+h)-$%
56: wave'' symmetry for the order parameter vector (gap function), the
57: value of the $\Delta \left( 0\right) $ has been obtained for both
58: $A$ and $B$-phases, in terms of the critical temperature. In
59: addition, the dependence of $\Delta \left( T\right) $ in the
60: temperature limit of $T\rightarrow 0$ and $T\rightarrow T_{c}$
61: have been obtained. Authors of paper \cite{Maki1} have
62: investigated the temperature dependence of critical field,
63: specific heat and heat conductivity. Also, the Josephson effect in
64: the point contact between triplet superconductors with $f-$wave
65: triplet pairing has been studied in Ref.\cite{Mahmoodi}. In this
66: paper the effect of misorientation on the charge transport has
67: been studied and a spontaneous current tangential to the interface
68: between the $f$-wave superconductors has been observed.
69: Additionally, the spin-current in the weak-link between the
70: $f$-wave superconductors has been investigated in our paper
71: \cite{Rashedi1}. In the paper \cite{Rashedi1}, this kind of
72: weak-link device has been proposed as the filter for polarization
73: of the spin-current. These weak-link structures have been used to
74: demonstrate the order parameter symmetry in Ref.
75: \cite{Stefanakis}.\newline In the present paper, the ballistic
76: Josephson weak-link via an interface between two bulks of
77: ``$(p+h)-$wave'' superconductor with different orientations of the
78: crystallographic axes is investigated. It is shown that the spin
79: and charge current-phase diagrams are totally different
80: from the current-phase diagrams of the point-contacts between conventional ($%
81: s$-wave) superconductors \cite{Kulik}, high $T_{c}$ ($d$-wave)
82: superconductors \cite{Coury} and from the charge and spin-current
83: phase diagrams in the weak-link between the $f$-wave
84: superconductors \cite{Mahmoodi,Rashedi1}. We have found that in
85: the weak-link structure between the ``$(p+h)-$wave''
86: superconductors, the spontaneous current parallel to the
87: interface, as the characteristic of unconventional
88: superconductivity, can be present. The effect of misorientation on
89: the spontaneous, Josephson and spin currents for the different
90: models of the paring symmetry ($A-$ and $B-$ phases in
91: Fig.\ref{phases}) are investigated. It is possible to find the
92: value of the phase difference, at which the Josephson current is
93: zero, but the spontaneous current tangential to the interface is
94: present. In some configurations and at the zero phase difference,
95: the Josephson current is not zero but has a finite value. This
96: finite value corresponds to a spontaneous phase difference, which
97: is related to the misorientation between the gap vectors. Finally,
98: it is observed that at certain values of the phase difference
99: $\phi ,$ at which the charge current is zero, the spin current is
100: present and vise versa. In addition, in the configuration in which
101: both gap vectors are directed along the $\hat{\mathbf{c}}\perp
102: \hat{\mathbf{n}}$ axis ($\hat{\mathbf{n}}$ is the normal to the
103: interface unit vector), only the normal to the interface spin
104: current $\hat{\mathbf{s}}_{\hat{\mathbf{n}}}$ can be present and
105: the other terms of the spin current are absent ($\hat{\mathbf{s}}$
106: is the spin vector of electrons).
107: \begin{figure}[h]
108: \includegraphics[width=5cm]{a-phase.eps}
109: \includegraphics[width=5cm]{direction.eps}
110: \includegraphics[width=5cm]{b-phase.eps}
111: \caption{$A-$phase (left), $B-$phase (right) order parameters and
112: direction of $\mathbf{a}$, $\mathbf{b}$, $\mathbf{c}$ and
113: $\mathbf{\hat{z}}$ unit vectors (middle) \cite{Maki1}. $A-$ and
114: $B-$ phases are high field (high temperature) and low field (low
115: temperature) phases, respectively \cite{Goryo}.}
116: \end{figure}\label{phases}Consequently, this structure can be used as a filter for
117: polarization of the spin transport. Furthermore, our analytical
118: and numerical calculations have shown that the misorientation is
119: the origin of the spin current.\newline
120:
121: The organization of the rest of this paper is as follows. In
122: Sec.\ref{section2} we describe our configuration, which has been
123: investigated. For a non-self-consistent model of the order
124: parameter, the quasiclassical Eilenberger equations
125: \cite{Eilenberger} are solved and suitable Green functions have
126: been obtained analytically. In Sec.\ref{section3} the obtained
127: formulas for the Green functions have been used for calculation
128: the charge and spin current densities at the interface. An
129: analysis of numerical results will be done in Sec.\ref{section4}.
130: The paper will be finished with some conclusions in
131: Sec.\ref{section5}.
132:
133: \section{Formalism and Basic Equations}
134: \label{section2}
135: We consider a model of a flat interface $y=0$ between two misoriented ``$%
136: (p+h)-$wave'' superconducting half-spaces (Fig.\ref{fig1}) as a
137: ballistic Josephson junction. In the quasiclassical ballistic
138: approach, in order to calculate the current, we use
139: ``transport-like'' equations \cite{Eilenberger} for the
140: energy integrated Green matrix $\breve{g}\left( \mathbf{\hat{v}}_{F},%
141: \mathbf{r},\varepsilon _{m}\right) $
142: \begin{equation}
143: \mathbf{v}_{F}\nabla \breve{g}+\left[ \varepsilon _{m}\breve{\sigma}_{3}+i%
144: \breve{\Delta},\breve{g}\right] =0, \label{Eilenberger}
145: \end{equation}
146: and the normalization condition $\breve{g}\breve{g}=\breve{1}$, where $%
147: \varepsilon _{m}=\pi T(2m+1)$ are discrete Matsubara energies $m=0,1,2...$, $%
148: T$ is the temperature, $\mathbf{v}_{F}$ is the Fermi velocity and $\breve{%
149: \sigma}_{3}=\hat{\sigma}_{3}\otimes \hat{I}$ in which $\hat{\sigma}%
150: _{j}\left( j=1,2,3\right) $ are Pauli matrices. Also the matrix
151: structure of the off-diagonal self energy $\breve{\Delta}$ in the
152: Nambu space is
153: \begin{equation}
154: \breve{\Delta}=\left(
155: \begin{array}{cc}
156: 0 & \mathbf{d}\hat{\mathbf{\sigma }}i\hat{\sigma}_{2} \\
157: i\hat{\sigma}_{2}\mathbf{{d^{\ast }}\hat{\sigma}} & 0
158: \end{array}
159: \right),
160: \end{equation}
161: \label{order parameter} where, the gap vector $\mathbf{d}$, is the
162: three dimensional counterpart of the BCS energy-gap function
163: $\Delta$ for the case of triplet superconductivity.
164: \begin{figure}[tbp]
165: \includegraphics[width=0.5\columnwidth]{fig1.eps}
166: \caption{Scheme of a flat interface between two superconducting
167: bulks, which are misoriented by angle $\protect\alpha $. In
168: geometry (i), the $ab$-plane on the right side is rotated as much
169: as $\alpha$ around the $c$-axis and in geometry (ii) the $c$-axis
170: on the right side is rotated around the $b$-axis. $ab-$planes and
171: $c-$axis have been shown in Fig.\ref{phases}.} \label{fig1}
172: \end{figure}
173: The Green matrix $\breve{g}$ can be written in the form:
174: \begin{equation}
175: \breve{g}=\left(
176: \begin{array}{cc}
177: g_{1}+\mathbf{g}_{1}\mathbf{\hat{\sigma}} & \left( g_{2}+\mathbf{g}_{2}\hat{%
178: \mathbf{\sigma }}\right) i\hat{\sigma}_{2} \\
179: i\hat{\sigma}_{2}\left( g_{3}+\mathbf{g}_{3}\hat{\mathbf{\sigma
180: }}\right)
181: &i\hat{\sigma}_{2}(g_{1}-\mathbf{g}_{1}\hat{\mathbf{\sigma
182: }})i\hat{\sigma}_{2}
183: \end{array}
184: \right).
185: \end{equation}
186: \label{Green's function} Here,
187: $g_{1}+\mathbf{g}_{1}\mathbf{\hat{\sigma}}$ and
188: $(g_{2}+\mathbf{g}_{2}\hat{\mathbf{\sigma }})i\hat{\sigma}_{2}$
189: [$i\hat{\sigma}_{2}(g_{3}+\mathbf{g}_{3}\hat{\mathbf{\sigma }})$]
190: are normal and anomalous Green function matrix, respectively. The
191: terms of $g_i$ and $\mathbf{g}_{ij}$ are coefficients of Green
192: function matrix expansion in terms of the unit matrix and Pauli
193: $2\times 2$ matrices in the Nmabu space. Also, the terms $g_1$ and
194: $\mathbf{g_1}$ determine the charge and spin current densities
195: through the equations (\ref{charge-current}) and
196: (\ref{spin-current}), respectively and $\mathbf{g_2}$ and
197: $\mathbf{g_3}$ are used to determine the gap vector, $\mathbf{d}$,
198: using the self consistent relations (Eq.\ref{self-consistent} and
199: it's conjugate). It is remarkable that, in this paper, the unitary
200: states, for which $\mathbf{d\times d}^{\ast }=0,$ is investigated.
201: Also, the unitary states vectors $\mathbf{d}_{1,2}$ can be written
202: as
203: \begin{equation}
204: \mathbf{d}_{n}=\mathbf{\Delta }_{n}\exp i\psi _{n},
205: \end{equation}
206: where $\mathbf{\Delta }_{1,2}$ are the real vectors in the left
207: and right sides of the junction.\ The gap (order parameter) vector
208: $\mathbf{d}$ has to be determined from the self-consistency
209: equation:
210: \begin{equation}
211: \mathbf{d}\left( \mathbf{\hat{v}}_{F},\mathbf{r}\right) =2\pi
212: TN\left(
213: 0\right) \sum_{m}\left\langle V\left( {\mathbf{\hat{v}}}_{F},{\mathbf{\hat{v}%
214: }}_{F}^{\prime }\right) \mathbf{g}_{2}\left(
215: {\mathbf{\hat{v}}}_{F}^{\prime },\mathbf{r},\varepsilon
216: _{m}\right) \right\rangle \label{self-consistent}
217: \end{equation}
218: where $V\left(
219: {\mathbf{\hat{v}}}_{F},{\mathbf{\hat{v}}}_{F}^{\prime }\right) $,
220: is a potential of pairing interaction, $\left\langle
221: ...\right\rangle $ stands for averaging over the directions of an
222: electron
223: momentum on the Fermi surface ${\mathbf{\hat{v}}}_{F}^{\prime }$ and $%
224: N\left( 0\right) $ is the electron density of states at the Fermi
225: level of energy. Solutions to Eqs.(\ref{Eilenberger}) and
226: (\ref{self-consistent}) must satisfy the conditions for Green
227: functions and vector $\mathbf{d}$ in the bulks of the
228: superconductors far from the interface as follow:
229: \begin{eqnarray}
230: \breve{g}\left( \pm \infty \right) &=&\frac{\varepsilon _{m}\breve{\sigma}%
231: _{3}+i\breve{\Delta}_{2,1}}{\sqrt{\varepsilon _{m}^{2}+\left|
232: \mathbf{\Delta
233: }_{2,1}\right| ^{2}}}; \label{Bulk solution} \\
234: \mathbf{d}\left( \pm \infty \right) &=&\mathbf{\Delta }_{2,1}\left( \mathbf{%
235: \hat{v}}_{F}\right) \exp \left( \mp \frac{i\phi }{2}+i\psi
236: _{2,1}\right) , \label{Bulk order parameter}
237: \end{eqnarray}
238: where $\phi $ is the external phase difference between the order
239: parameters of the bulks. Eqs. (\ref{Eilenberger}) and
240: (\ref{self-consistent}) have to be supplemented by the continuity
241: conditions at the interface between superconductors. For all
242: quasiparticle trajectories, the Green functions satisfy the
243: boundary conditions both in the right and left bulks as well as at
244: the interface.\newline The system of equations
245: (\ref{Eilenberger})and (\ref{self-consistent}) can be solved only
246: numerically.
247: \begin{figure}[tbp]
248: \includegraphics[width=0.5\columnwidth]{fig2.eps}
249: \includegraphics[width=0.5\columnwidth]{fig3.eps}
250: \caption{Component of charge current normal to the interface
251: (Josephson current) versus the phase difference $\protect\phi $
252: for $A$ and $B-$phases, $T/T_{c}=0.08$ and different
253: misorientations. Part (a) is plotted for geometry (i) and part (b)
254: for geometry (ii). Charge currents are given in units of
255: $j_{0}=\frac{\protect\pi }{2}eN(0)v_{F}\Delta _{0}(0).$}
256: \label{fig2}
257: \end{figure}
258: For unconventional superconductors such solution
259: requires the information of the function $V\left( {\mathbf{\hat{v}}}_{F},{%
260: \mathbf{\hat{v}}}_{F}^{\prime }\right) $. This information, as
261: that of the nature of unconventional superconductivity in novel
262: compounds, in most cases is unknown. Usually, the spatial
263: variation of the gap vector and its dependence on the momentum
264: direction can be separated in the form of $\Delta
265: ({\mathbf{\hat{v}}}_{F},y)=\Delta ({\mathbf{\hat{v}}}_{F})\Psi
266: (y)$. It has
267: been shown that the absolute value of a self-consistent order parameter and $%
268: \Psi (y)$ are suppressed near the interface and at the distances
269: of the order of the coherence length, while its dependence on the
270: direction in the momentum space ($\Delta
271: ({\mathbf{\hat{v}}}_{F})$) remains unaltered \cite{Barash}.
272: Consequently, this suppression doesn't influence the Josephson
273: effect drastically. This suppression of the order parameter keeps
274: the current-phase dependence unchanged but, it changes the
275: amplitude value of the current. For example, it has been verified
276: in Refs.\cite{Coury} for the junction between unconventional
277: $d$-wave, in Ref.\cite{Barash} for the case of ``$f$-wave''
278: superconductors and in Ref.\cite{Viljas} for pinholes in $^{3}He$
279: that, there is a good qualitative agreement between
280: self-consistent and non-self-consistent results. Also, it has been
281: observed that the results of the non-self-consistent investigation
282: of $D-N-D$ structure in \cite{Faraii} are coincident with the
283: experimental results of \cite{Freamat} and the results of the
284: non-self-consistent model in \cite{Yip} are similar to the
285: experiment \cite{Backhaus}. Consequently, despite the fact that
286: self-consistent numerical results cannot be applied directly for a
287: quantitative analysis of the real experiment, only a qualitative
288: comparison of calculated and experimental current-phase relations
289: is possible. In our calculations, a simple model of the constant
290: order parameter up to the interface is considered and the pair
291: breaking and the scattering on the interface are ignored. We
292: believe that under these strong assumptions our results describe
293: the real situation qualitatively. In the framework of such model,
294: the analytical expressions for the current can be obtained for an
295: arbitrary form of the order parameter.
296: \section{Analytical results.}
297: \label{section3} The solution of Eqs.(\ref{Eilenberger}) and
298: (\ref{self-consistent}) allows us to calculate the charge and spin
299: current densities. The expression for the charge current is:
300: \begin{equation}
301: \mathbf{j}_{e}\left( \mathbf{r}\right) =2i\pi eTN\left( 0\right)
302: \sum_{m}\left\langle \mathbf{v}_{F}g_{1}\left( \mathbf{\hat{v}}_{F},\mathbf{r%
303: },\varepsilon _{m}\right) \right\rangle, \label{charge-current}
304: \end{equation}
305: and for the spin current we have:
306: \begin{equation}
307: \mathbf{j}_{s_{i}}\left( \mathbf{r}\right) =2i\pi (\frac{\hbar
308: }{2})TN\left(
309: 0\right) \sum_{m}\left\langle \mathbf{v}_{F}\left( \mathbf{{\hat{e}}}_{i}%
310: \mathbf{g}_{1}\left( \mathbf{\hat{v}}_{F},\mathbf{r},\varepsilon
311: _{m}\right) \right) \right\rangle \label{spin-current}
312: \end{equation}
313: where, $\mathbf{{\hat{e}}}_{i}\mathbf{=}\left( \hat{\mathbf{x}},\hat{\mathbf{%
314: y}},\hat{\mathbf{z}}\right).$ We assume that the order parameter
315: does not depend on the coordinates and in each half-space it
316: equals its value (\ref {Bulk order parameter}) far from the
317: interface in the left or right bulks.
318: \begin{figure}[tbp]
319: \includegraphics[width=0.5\columnwidth]{fig4.eps}
320: \includegraphics[width=0.5\columnwidth]{fig5.eps}
321: \caption{Components of charge current tangential to the interface
322: versus the phase difference $\protect\phi $ for $A$ and
323: $B-$phases, geometry (ii), $T/T_{c}=0.08$ and the different
324: misorientations. Part (a) is plotted for $x-$component and part
325: (b) for $z-$component.} \label{fig3}
326: \end{figure}
327: For such a model, the current-phase dependence of a Josephson
328: junction can be calculated analytically. It enables us to analyze
329: the main features of current-phase dependence for different models
330: of the order parameter of ``$(p+h)-$wave'' superconductivity.
331: The Eilenberger equations (\ref{Eilenberger}) for Green functions $\breve{g}$%
332: , which are supplemented by the condition of continuity of
333: solutions across the interface, $y=0$, and the boundary conditions
334: at the bulks, are solved for a non-self-consistent model of the
335: order parameter analytically. Two diagonal terms of the Green
336: matrix which determine the current densities at the interface,
337: $y=0$, are shown below. For the term relating to the charge
338: current we obtain:
339: \begin{equation}
340: g_{1}\left( 0\right) =\frac{\varepsilon _{m}(\Omega _{1}+\Omega
341: _{2})\cos \beta +i\eta (\Omega _{1}\Omega _{2}+\varepsilon
342: _{m}^{2})\sin \beta }{i\eta \varepsilon _{m}(\Omega _{1}+\Omega
343: _{2})\sin \beta +(\Omega _{1}\Omega
344: _{2}+\varepsilon _{m}^{2})\cos \beta +\mathbf{\Delta }_{1}\mathbf{\Delta }%
345: _{2}}, \label{charge-term}
346: \end{equation}
347: and for the case of spin current we have:
348: $$\mathbf{g_{1}}\left(0\right)=\mathbf{\Delta
349: }_{1}\times \mathbf{\Delta}_{2}$$
350: \begin{equation}
351: \hspace{-2.5cm}\frac{(B-1)^{2}(\eta \Omega
352: _{1}+\varepsilon_{m})(\eta \Omega _{2}+\varepsilon _{m})\exp
353: (i\beta )-(B+1)^{2}(\eta \Omega _{2}-\varepsilon _{m})(\eta \Omega
354: _{1}-\varepsilon _{m})\exp (-i\beta
355: )}{{2\eta(A+B)\left|\mathbf{\Delta }_{1}\right|
356: ^{2}\left|\mathbf{\Delta }_{2}\right| ^{2}}}
357: \end{equation}\label{spin-term}
358: where $\eta =sgn\left( v_{y}\right) $, $\Omega
359: _{n}=\sqrt{\varepsilon _{m}^{2}+\left| \mathbf{\Delta }_{n}\right|
360: ^{2}}$, $\beta =\psi _{1}-\psi _{2}+\phi $,
361: \begin{equation}
362: B=\frac{\eta \varepsilon _{m}(\Omega _{1}+\Omega _{2})\cos \beta
363: +i(\Omega _{1}\Omega _{2}+\varepsilon _{m}^{2})\sin \beta }{i\eta
364: \varepsilon _{m}(\Omega _{1}+\Omega _{2})\sin \beta +(\Omega
365: _{1}\Omega _{2}+\varepsilon _{m}^{2})\cos \beta +\mathbf{\Delta
366: }_{1}\mathbf{\Delta }_{2}} \label{mathematics-term}
367: \end{equation}
368: and
369: \begin{equation}
370: A=\frac{\mathbf{\Delta }_{1}\mathbf{\Delta
371: }_{2}}{2}\left[\frac{(B-1)\exp (i\beta )}{(\eta \Omega
372: _{1}-\varepsilon _{m})(\eta \Omega _{2}-\varepsilon
373: _{m})}+\frac{(B+1)\exp (-i\beta )}{(\eta \Omega _{1}+\varepsilon
374: _{m})(\eta \Omega _{2}+\varepsilon _{m})}\right].
375: \end{equation}
376: Also, $n=1,2,$ label the left and right half-spaces respectively.
377: We consider a rotation $\breve{R}$ only in the right
378: superconductor (see, Fig.\ref{fig1}), i.e.,
379: $\mathbf{d}_{2}(\hat{\mathbf{k}})=\breve{R}\mathbf{d}_{1}(\breve{R}^{-1}\hat{%
380: \mathbf{k}});$ $\hat{\mathbf{k}}$ is the unit vector in the
381: momentum space. The crystallographic $c$-axis in the left
382: half-space is selected parallel to the partition between the
383: superconductors (along the $z$-axis in
384: Figs.\ref{phases},\ref{fig1}). To
385: illustrate the results obtained by computing the formula (\ref{charge-term}%
386: ), we plot the current-phase diagrams for different models of the ``$%
387: (p+h)-$wave'' pairing symmetry (\ref{A-phase},\ref{B-phase}) and
388: for two different geometries. These geometries correspond to the
389: different orientations of the crystals on the right and left sides
390: of the interface (Fig.\ref{fig1}):\newline
391: (i) The basal $ab$-plane on the right side has been rotated around the $c$%
392: -axis by $\alpha $; $\hat{\mathbf{c}}_{1}\Vert
393: \hat{\mathbf{c}}_{2}$.\newline (ii) The $c$-axis on the right side
394: is rotated around the $b$-axis by $
395: \alpha $ ($y$-axis in Fig.\ref{fig1}); $\hat{\mathbf{b}}_{1}\Vert \hat{\mathbf{b}}%
396: _{2} $.\newline Further calculations require a certain model of
397: gap vector (order parameter vector) $\mathbf{d}$.
398: \section{Analysis of numerical results}
399: \label{section4} In the present paper, two forms of
400: ``$(p+h)-$wave'' unitary gap vector $\mathbf{d}$ in
401: $PrOs_{4}Sb_{12}$ are considered. The first model to explain the
402: properties of the $A$-phase of $PrOs_{4}Sb_{12}$ is (left side of
403: Fig.\ref{phases}):
404: \begin{equation}
405: \mathbf{d}=\Delta _{0}(T)(k_{x}+ik_{y})\frac{3}{2}(1-\hat{k}_{x}^{4}-\hat{k}%
406: _{y}^{4}-\hat{k}_{z}^{4})\hat{\mathbf{z}}. \label{A-phase}
407: \end{equation}
408: The coordinate axes
409: $\hat{\mathbf{x}},\hat{\mathbf{y}},\hat{\mathbf{z}}$ are chosen along the
410: crystallographic axes $\hat{\mathbf{a}},%
411: \hat{\mathbf{b}},\hat{\mathbf{c}}$ on the left side of Fig.\ref{fig1}; $%
412: \widehat{\mathbf{k}}$ is the unit vector along $\mathbf{v}_{F}$.
413: The scalar function $\Delta _{0}=$ $\Delta _{0}\left( T\right)$ describes the dependence of the gap vector $%
414: \mathbf{d}$ on the temperature $T$. The second model to describe
415: the gap vector of the $B$-phase of $PrOs_{4}Sb_{12}$ is (right
416: side of Fig.\ref{phases}):
417: \begin{equation}
418: \mathbf{d}=\Delta
419: _{0}(T)(k_{x}+ik_{y})(1-\hat{k}_{z}^{4})\hat{\mathbf{z}}.
420: \label{B-phase}
421: \end{equation}
422: \begin{figure}[tbp]
423: \includegraphics[width=0.5\columnwidth]{fig7.eps}
424: \includegraphics[width=0.5\columnwidth]{fig8.eps}
425: \caption{Components of spin current ($s_{y}$) tangential to the
426: interface versus the phase difference $\protect\phi $ for geometry
427: (ii), $\frac{T}{T_{c}}=0.08$ and different misorientations between
428: the $A$ and $B-$phase of ``$(p+h)-$wave.'' Part (a) is plotted for
429: $x-$component
430: and part (b) for $z-$component. Spin currents are given in units of
431: $j_{s0}=\frac{\protect\pi }{4}\hbar N(0)v_{F}\Delta _{0}(0).$}
432: \label{fig4}
433: \end{figure}
434: Our numerical calculations are done at the low temperatures,
435: $T/T_{c}=0.08$, and we have used the formulas $\ln (\Delta
436: (T)/\Delta (0))=-\frac{7\pi \zeta (3)}{8}(\frac{T}{\Delta
437: (0)})^{3}$ for the $A-$phase and $\ln (\Delta (T)/\Delta
438: (0))=-\frac{135\pi \zeta (3)}{512}(\frac{T}{\Delta (0)})^{3}$ for
439: the $B-$phase, from the paper \cite{Maki3}, for temperature
440: dependence of the gap functions $\Delta _{0}(T)$ at the low
441: temperatures ($T<<T_{c}$). Also, in the paper \cite{Maki3} the
442: value of $\Delta _{0}(T)$ has been calculated in terms of the
443: critical temperature for both $A$ and $B-$ phases. They are
444: $\Delta (0)/T_{c}=2.34$ and $\Delta (0)/T_{c}=1.93$ for $A$ and
445: $B-$ phases, respectively. Using these two models of order
446: parameters (\ref{A-phase}, \ref{B-phase}) and solution to the
447: Eilenberger equations (\ref{charge-term}, \ref{spin-term}), we
448: have calculated the current density at the interface numerically.
449: These numerical results are listed below:\newline 1) In part (a)
450: of Fig.\ref{fig2}, the component of current normal to the
451: interface of
452: current, which is known as the Josephson current, is plotted for both $A$ and $%
453: B-$phases, geometry (i), misorientations $\alpha =\pi /4$ and
454: $\alpha =\pi /6 $. It is observed that the critical values of
455: current for the $B-$phase is larger than the $A-$phase. Also,
456: unlike Josephson junction between the conventional
457: superconductors, here, at $\phi =0$, the current is not zero. The
458: current is zero at the phase difference value $\phi =\phi _{0}$,
459: which depends on the misorientation between the gap vectors. In
460: Fig.\ref{fig2}, the value of the spontaneous phase difference
461: $\phi _{0}$ is close to misorientation $\alpha .$\newline
462: 2) In part (b) of Fig.\ref{fig2}, the Josephson current is plotted for both $A$ and $B-$%
463: phases, geometry (ii) and different misorientations. Again, the
464: maximum value of the current for the $B-$phase is larger than for
465: $A-$phase. Increasing the misorientation between the gap vectors,
466: the maximum value of the current decreases. It is demonstrated that at the phase difference values $%
467: \phi =0$, $\phi =\pi $ and $\phi =2\pi $, the Josephson current is
468: zero while both spontaneous and spin currents are not zero and
469: have a finite value. Increasing the misorientation between the gap
470: vectors decreases the derivative ($\frac{dj_{y}}{d\phi }$) of the
471: current with respect to the phase difference close to $\phi =\pi
472: $, which is known as the \textbf{SQUID} sensitivity and it is
473: important from the application point of view.\newline 3) In
474: Fig.\ref{fig3}, the tangential components of the charge current
475: ($x$ and $z-$components) in terms of the phase difference $\phi $
476: are plotted.
477: \begin{figure}[tbp]
478: \includegraphics[width=0.5\columnwidth]{fig9.eps}
479: \includegraphics[width=0.5\columnwidth]{fig10.eps}
480: \caption{Component of charge and spin currents ($j_{y}$ and $%
481: j_{sy}$) normal to the interface versus the phase difference
482: $\protect\phi $ for geometry (ii), $\frac{T}{T_{c}}=0.08$, $\protect\alpha =\frac{\protect\pi }{6}$ and $%
483: \protect\alpha =\frac{\protect\pi }{4}.$ Part (a) is plotted for
484: $A-$phase and part (b) for $B-$phase.} \label{fig5}
485: \end{figure}
486: It is seen that at $\phi =0$, $\phi =\pi $ and $\phi =2\pi ,$ at
487: which the Josephson current is zero, the parallel spontaneous
488: currents have finite values. Although the normal component
489: of charge current (see part (b) of Fig.\ref{fig2}) is an odd
490: function of the phase difference with respect to the line of $\phi
491: =\pi $ while the parallel charge currents for this geometry (ii)
492: are even functions of the phase difference with respect to $\phi
493: =\pi $ (compare part (b) of Fig.\ref{fig2} with
494: Fig.\ref{fig3}).\newline
495: 4) In Fig.\ref{fig4}, the tangential components of the spin
496: ($s_{y}$) current are plotted in terms of the phase difference,
497: for geometry (ii) and different misorientations. By increasing the
498: misorientation the maximum value of the spin current increases. In
499: spite of the charge current for this state, the spin current at
500: the phase differences $\phi =0$, $\phi =\pi $ and $\phi =2\pi $ is
501: zero exactly(compare Fig.\ref{fig3} with Fig.\ref{fig4}).\newline
502: 5) In Fig.\ref{fig5}, the normal component of charge and spin
503: current ($j_{y}$ and $j_{s_{y}}$) are plotted for different
504: misorientations and $A$ and $B-$phases respectively. An
505: interesting case in our observations, is the finite value of
506: normal spin current at $\phi =0$, $\phi =\pi $ and $\phi =2\pi $
507: at which the normal charge current ($j_{y}$) is zero (see
508: Fig.\ref{fig5}).
509: \newline
510: 6) In part (a) of Fig.\ref{fig6}, the
511: Josephson current is plotted in terms of the phase difference for the case of $p-$wave, and $A$ and $B-$phases of ``$(p+h)-$%
512: wave''. The $p-$wave pairing symmetry as the first candidate for
513: the superconducting state in Sr$_{2}$RuO$_{4}$ is as follows
514: \cite{Rice}:
515: \begin{equation}
516: \mathbf{d}=\Delta _{0}(T)(k_{x}+ik_{y})\hat{\mathbf{z}}
517: \label{p-wave}
518: \end{equation}
519: It is observed that the maximum value of the Josephson current
520: ($j_{y}$) of the
521: junction between the $p-$wave superconductors, is larger than for the $B-$%
522: phase of ``$(p+h)-$wave'' and the Josephson current of $B-$phase
523: is larger than its value for the $A-$phase counterpart. Also, the
524: place of the zero of the charge current for these three types of
525: superconductor (geometry (i)) is the same. It is at the
526: spontaneous phase difference which is close to the misorientation
527: $\phi _{0}=\alpha $ (see, part (a) of Fig.\ref{fig2} and part (a)
528: of Fig.\ref{fig6}).\newline 7) In part (b) of Fig.\ref{fig6}, the
529: normal component of the spin current is plotted for $p-$wave, $A$
530: and $B-$phases of ``$(p+h)-$wave'' pairing
531: symmetries and for a specified value of misorientation $\alpha =\frac{\pi }{4%
532: }$. In both (a) and (b) parts of Fig.\ref{fig6}, the maximum value of
533: the current of junction between the $p-$wave superconductors is larger
534: than for the $B-$phase. For $B-$phase the current has the maximum
535: value, which is larger than the value for junction between the
536: ``$(p+h)-$wave'' superconductors in the $A-$phase. This different
537: characteristic of the current-phase diagrams enables us to distinguish
538: between the three states.
539: Also, it is observed that at phase differences $\phi =0$,
540: $\phi =\pi $ and $\phi =2\pi $, the spin current has a finite
541: value which may have its maximum value. This is a counterpart of
542: part (b) of Fig.\ref{fig2}, in which the charge currents are zero at the
543: mentioned values of the phase difference, but the spin current has
544: a finite value.\newline
545:
546: Furthermore, our analytical and numerical calculations have shown
547: that the origin of the spin current is misorientation between the
548: gap vectors [cross product in Eq.(\ref{spin-term})]. Because the
549: geometry (i) is a rotation by $ \alpha $ around the
550: $\mathbf{{\hat{z}}-}$axis and both of the left and right gap
551: vectors are in the same direction, the cross product between gap
552: vectors and consequently, the spin current for geometry (i) is
553: zero.
554: \begin{figure}[tbp]
555: \includegraphics[width=0.5\columnwidth]{fig6.eps}
556: \includegraphics[width=0.5\columnwidth]{fig11.eps}
557: \caption{Component of charge and spin currents normal to the
558: interface versus the phase difference $\protect\phi $,
559: $\frac{T}{T_{c}}=0.08$, $\protect\alpha =\frac{\protect\pi }{4}$,
560: $A-$phase, $B-$phase and $p-$wave pairing symmetry. Part (a) is
561: plotted for charge current and geometry (i) and part (b) is for
562: the case of spin current and geometry (ii)} \label{fig6}
563: \end{figure}
564: It is shown that in this structure ($y-$direction is normal
565: to the interface, $c-$axis is selected in the
566: $z-$direction and rotation is done around the $y-$direction) only
567: the current of the $s_{y}$ flows and other terms of the spin
568: current are absent. So, this kind of weak-link experiment can be
569: used as a filter for the polarization of spin transport. Since the
570: spin is a vector, the spin current is a tensor and we
571: have the current of spin $s_{y}$ in the three $\mathbf{{\hat{x}}}$, $\mathbf{%
572: {\hat{y}}}$ and $\mathbf{{\hat{z}}}$ directions.
573: \section{Conclusions}
574: \label{section5} We have theoretically studied the the spin and
575: charge transport in the ballistic Josephson junction in the model
576: of an ideal transparent interface between two misoriented
577: $PrOs_{4}Sb_{12}$ crystals with ``$(p+h)-$wave'' pairing symmetry,
578: which are subject to a phase difference $\phi $. Our analysis has
579: shown that the different misorientations and different models of
580: the gap vectors influence the spin and charge currents. This has
581: been shown for the charge current in the point contact between two
582: bulks of``$f-$wave'' superconductors in \cite{Mahmoodi} and for
583: the spin current in the weak link between ``$f$-wave''
584: superconductors in
585: \cite{Rashedi1}. In this paper, it is shown that the
586: misorientation of the superconductors leads to a spontaneous phase
587: difference that corresponds to the zero Josephson current and to
588: the minimum of the weak link energy. This phase difference depends
589: on the misorientation angle. We have found a spontaneous charge
590: current tangential to the interface which is not equal to zero in
591: the absence of the Josephson current generally. It has been found
592: that the spin current is the result of the misorientation between
593: the gap vectors. Furthermore, it is observed that a certain model
594: of the gap vectors and geometries can be applied to polarize the
595: spin transport. Finally, as an interesting and new result, it is
596: observed that at certain values of the phase difference $\phi $,
597: the charge-current vanishes while the spin-current flows, although
598: the carriers of both spin and charge are the same (electrons). The
599: spatial variation of the phase of the order parameter plays a role
600: as the origin of the charge current and, similarly, due to the
601: broken $G^{spin-orbit}$ symmetry, a spatial difference of the gap
602: vectors in two half-spaces is the cause of spin currents. This is
603: because there is a position-dependent phase difference between
604: ``spin up'' and ``spin down'' Cooper pairs and, although the total
605: charge current vanishes, there can be a net transfer of the spin.
606: Therefore, in our system, there is a discontinuous jump between
607: the gap vectors and, consequently the spin currents should
608: generally be present. For instance, if up-spin states and
609: down-spin states have a velocity in the opposite direction, the
610: charge currents cancel each other whereas the spin current is
611: being transported.
612: Mathematically speaking, $\mathbf{{j_{charge}}={j_{\uparrow }}+{j_{\downarrow }},{j_{spin}}=%
613: {j_{\uparrow }}-{j_{\downarrow }}}$, so it is possible to find the
614: state in which one of these current terms is zero and the other
615: term has a finite value \cite{Maekawa}. In conclusion, the spin
616: current in the absence of the charge current can be observed and
617: vice versa.
618: \section*{References}
619: \begin{thebibliography}{99}
620: \bibitem{Maki1} Parker D, Maki K and Haas S, cond-mat/0407254.
621:
622: \bibitem{Maki2} Maki K, Won H, Thalmeier P, Yuan Q, Izawa K, and
623: Matsuda Y, 2003 Europhys. Lett. \textbf{64} (4), 496.
624:
625: \bibitem{Maki3} Maki K, Haas S, Parker D, Won H, Izawa K and Matsuda Y,
626: cond-mat/0406492.
627:
628: \bibitem{Bauer1} Bauer E.D, Frederick N.A, Ho P.-C, Zapf V.S, and Maple M.B, 2002
629: Phys. Rev. B \textbf{65}, R100506.
630:
631: \bibitem{Bauer2} Vollmer R, Fai$\beta $t A, Pfleiderer C, L\"{o}hneysen H.v, Bauer E.D, P.-C. Ho,
632: Zapf V and Maple M.B, 2003 Phys. Rev. Lett. \textbf{90}, 57001.
633:
634: \bibitem{Kotegawa} Kotegawa H, Yogi M, Imamura Y, Kawasaki Y,
635: Zheng G -Q, Kitaoka K, Ohsaki S, Sugawara H, Aoki Y, and Sato H, 2003
636: Phys. Rev. Lett. \textbf{90}, 027001.
637:
638: \bibitem{Nakajima} Izawa K, Nakajima Y, Goryo J, Matsuda Y, Osaki S,
639: Sugawara H, Sato H, Thalmeier P, and Maki K, 2003 Phys. Rev. Lett. \textbf{90}%
640: , 117001.
641:
642: \bibitem{Tou} Tou H, Ishida K, Kitaoka Y, cond-mat/0308562.
643:
644: \bibitem{Mahmoodi} Mahmoodi R, Shevchenko S N and Kolesnichenko Yu A, 2002
645: Fiz. Nizk.Temp.\textbf{28}, 262 [2002 Sov. J. Low Temp. Phys.
646: \textbf{28} ,184].
647:
648: \bibitem{Rashedi1} Rashedi G and Kolesnichenko Yu A, (preprint).
649:
650: \bibitem{Stefanakis} Stefanakis N, 2002 Phys. Rev. B \textbf{65}, 064533.
651:
652: \bibitem{Kulik} Kulik I O and Omelyanchouk A N, 1978 Fiz. Nizk. Temp.,\textbf{4%
653: }, 296 [1978 Sov. J. Low Temp. Phys., \textbf{4}, 142].
654:
655: \bibitem{Coury} Amin M H S, Coury M, Rashkeev S N, Omelyanchouk A N, and
656: Zagoskin A M, 2002 Physica B, \textbf{318}, 162.
657:
658: \bibitem{Goryo} Goryo J, Phys. Rev. B \textbf{67}, 184511 (2003).
659:
660: \bibitem{Eilenberger} Eilenberger G, 1968 Z. Phys., \textbf{214}, 195.
661:
662: \bibitem{Barash} Barash Yu S, Bobkov A M, and Fogelstr\"{o}m M, 2001 Phys.
663: Rev. B \textbf{64}, 214503.
664:
665: \bibitem{Viljas} Viljas J K, cond-mat/0004246.
666:
667: \bibitem{Faraii} Faraii Z and Zareyan M, 2004 Phys. Rev. B \textbf{69}, 014508.
668:
669: \bibitem{Freamat} Freamat M, Ng K -W, 2003 Phys. Rev. B \textbf{68}, 060507.
670:
671: \bibitem{Yip} Yip S -K, 1999 Phys. Rev. Lett. \textbf{83}, 3864.
672:
673: \bibitem{Backhaus} Backhaus S, Pereverzev S, Simmonds R W, Loshak A,
674: Davis J C, and Packard R E, 1998 Nature \textbf{392}, 687.
675:
676: \bibitem{Maekawa} Maekawa S, 2004 J. Magn. Magn. Mater. \textbf{272}, 1459.
677:
678: \bibitem{Rice} Rice T M and Sigrist M, 1995 J. Phys.: Condens. Matter\textbf{7}
679: , L643.
680: \end{thebibliography}
681:
682: \end{document}
683: