1: \documentclass[aps,twocolumn]{revtex4}
2: %\documentclass[aps,preprint]{revtex4}
3: \newcommand{\mb}[1]{ { \mbox{\boldmath{$#1$}}} }
4: \usepackage{epsf}
5: \sloppy
6:
7: \begin{document}
8: \title{Meservey-Tedrow-Fulde effect in a quantum dot embedded \\
9: between metallic and superconducting electrodes}
10:
11: \author{T.\ Doma\'nski$^{1}$, A.\ Donabidowicz$^{1}$,
12: and K.I.\ Wysoki\'nski$^{1,2}$}
13:
14: \affiliation{$^{1}$Institute of Physics,
15: M.\ Curie Sk\l odowska University,
16: 20-031 Lublin, Poland\\
17: $^{2}$Max Planck Institut f\"ur
18: Physik komplexer Systeme,
19: D-01187 Dresden, Germany}
20:
21: \begin{abstract}
22: Magnetic field applied to the quantum dot coupled between one
23: metallic and one superconducting electrode can produce a similar
24: effect as has been experimentally observed by Meservey, Tedrow
25: and Fulde [Phys.\ Rev.\ Lett.\ {\bf 25}, 1270 (1970)] for the
26: planar normal metal -- superconductor junctions. We investigate
27: the tunneling current and show that indeed the square root
28: singularities of differential conductance exhibit the Zeeman
29: splitting near the gap edge features $V = \pm\Delta/e$. Since
30: magnetic field affects also the in-gap states of quantum dot
31: it furthermore imposes a hyperfine structure on the anomalous
32: (subgap) Andreev current which has a crucial importance for
33: a signature of the Kondo resonance.
34: \end{abstract}
35: \date{\today}
36:
37:
38: \maketitle
39:
40: \section{Introduction}
41:
42: Already in early days of the tunneling spectroscopy it has been
43: shown that magnetic field $B$ (which couples to spin of the charge
44: carriers) is in superconductors responsible for splitting the square
45: root singularities of the tunneling conductance \cite{Meservey-70} by
46: the Zeeman energy $2\mu_{B}B$, where $\mu_B$ is the Bohr magneton.
47: This Meservey-Tedrow-Fulde (MTF) effect has been observed
48: experimentally in the thin superconducting aluminum films applying
49: parallel magnetic field so that orbital diamagnetic effects could
50: be avoided. Similar qualitative results have been recently noticed
51: in the measurements of $c$-axis tunneling for the layered high
52: temperature superconducting compounds \cite{Alvarez06}.
53:
54: We argue that the MTF effect should be also feasible in various
55: nanostructures consisting of a quantum dot (QD) placed between
56: one metallic and one superconducting electrode. Zero-dimensional
57: character of QDs in a natural way eliminates the influence of
58: orbital effects therefore magnetic field would affect the charge
59: transport only through the Zeeman term. This can in turn
60: manifest itself in the differential conductance. Roughly speaking,
61: the charge current flows if an external bias $V$ exceeds the energy
62: gap $\Delta$ (necessary to break the Cooper pairs into individual
63: electrons) thereof the resulting conductance has a low voltage
64: onset near the gap edges $eV=\pm \Delta$. In presence of a magnetic
65: field these gap edge singularities are going to split (see section III).
66:
67: More detailed analysis of the charge tunneling \cite{Blonder}
68: involves however also the additional (anomalous) channels due
69: to mixing of the particle and hole excitations in superconductors.
70: In particular, even at subgap voltages $|eV|\leq \Delta$ the
71: mechanism of Andreev reflections provides a finite contribution
72: to the conductance. Since the Andreev mechanism is very sensitive
73: to location of the in-gap QD states \cite{Beenakker92,Claughton95,
74: Yeyati97,Zhao99,Sun99} and the on-dot correlations \cite{Fazio-98,
75: Clerk-00,Cuevas-01,Krawiec-04,Oguri-04,Tanaka-07,Domanski-07,
76: Taddei-07,Gezzi08,Hecht-08} we shall explore the influence
77: of magnetic field on such subgap conductance. In section IV
78: we discuss a hyperfine structure for the Andreev conductance
79: neglecting the correlations. In the next section V we extend
80: our study taking into account a finite value of the on-dot
81: repulsion $U$. We show that appearance of the low temperature
82: Kondo resonance enhances the zero bias conductance and this
83: feature undergoes the Zeeman splitting when magnetic field
84: is applied.
85:
86: As concerns some practical aspects, there have been considered the
87: proposals for using the magnetic field tuned Andreev scattering as
88: an efficient cooling mechanism in two dimensional electron gas
89: - superconductor nanostructure \cite{Giazotto06}. There is also
90: considered a possibility to use the, so called, Andreev quantum
91: dot as a magnetic flux detector \cite{Sadovsky07}.
92:
93:
94: \section{The model}
95:
96: For a general description of transport phenomena through a nanoscopic
97: island placed between external leads one should consider a quantized
98: multilevel structure of QD \cite{Aleiner-02}. However, in the case
99: when a level spacing is smaller in comparison to QD hybridization
100: with the electrodes one can restrict to a simplified picture
101: of the Anderson model \cite{Claughton95,
102: Yeyati97,Fazio-98,Clerk-00,Cuevas-01}
103: %
104: \begin{eqnarray}
105: \hat{H} &=& \hat{H}_{N} + \hat{H}_{S} + \sum_{\sigma}
106: \epsilon_{d,\sigma} \hat{d}^{\dagger}_{\sigma} \hat{d}_{\sigma}
107: + U \; \hat{n}_{d \uparrow} \hat{n}_{d \downarrow}
108: \nonumber \\
109: &+& \sum_{{\bf k},\sigma } \sum_{{\beta}=N,S}
110: \left( V_{{\bf k} \beta} \; \hat{d}_{\sigma}^{\dagger}
111: \hat{c}_{{\bf k} \sigma \beta } + V_{{\bf k} \beta}^{*}
112: \; \hat{c}_{{\bf k} \sigma, \beta }^{\dagger} \hat{d}_{\sigma}
113: \right) .
114: \label{model}
115: \end{eqnarray}
116: Operators $d_{\sigma}$ ($d_{\sigma}^{\dagger}$) denote the annihilation
117: (creation) of electron whose energy level is $\varepsilon_{d,\sigma}$
118: and $U$ is the on-dot Coulomb repulsion between opposite spin electrons.
119: The last terms describe hybridization of QD with the normal
120: ($\beta\!=\!N$) and superconducting ($\beta\!=\!S$) electrodes.
121: Magnetic field eventually shifts the QD level by $\varepsilon_{d,\sigma}
122: =\varepsilon_{d} - g_{\sigma}\mu_{B}B$, where the spin-dependent
123: coefficients are defined as $g_\uparrow\!=\!1$ and $g_\downarrow\!=\!-1$.
124:
125: Hamiltonian of the normal (metallic) lead is taken as $\hat{H}_{N}
126: \!=\! \sum_{{\bf k},\sigma} \xi_{{\bf k}N}^{\sigma} \hat{c}_{{\bf k}
127: \sigma N}^{\dagger} \hat{c}_{{\bf k} \sigma N}$ whereas for the
128: superconducting electrode we choose the usual BCS form
129: $\hat{H}_{S} \!=\!\sum_{{\bf k},\sigma} \xi_{{\bf k}S}^{\sigma}
130: \hat{c}_{{\bf k} \sigma S }^{\dagger} \hat{c}_{{\bf k} \sigma S}
131: \!-\! \sum_{\bf k} \left( \Delta \hat{c}_{{\bf k} \uparrow S }
132: ^{\dagger} \hat{c}_ {-{\bf k} \downarrow S }^{\dagger} +
133: \mbox{h.c.} \right)$ with an isotropic energy gap $\Delta$.
134: The relative energies $\xi_{{\bf k}\beta}^{\sigma}\!=\!
135: (\varepsilon_{{\bf k}\beta} \!-\!g _{\sigma} \mu_{B}B)
136: \!-\!\mu_{\beta}$ are measured from the chemical potentials
137: $\mu_{\beta}$. We shall focus on the wide band limit
138: $|V_{{\bf k}\beta}| \! \ll \! D$ (where $-D\!\leq\!
139: \varepsilon_{{\bf k}\beta} \! \leq \! D$) and consider a small
140: external voltage $V$, which detunes the chemical potentials
141: by $\mu_{N}\!-\!\mu_{S}=eV$ inducing the charge flow through
142: N-QD-S junction. We assume $|eV|$ to be much smaller than
143: level spacings typical for the realistic QDs \cite{Aleiner-02}
144: so that applicability of the model (\ref{model}) can be justified.
145:
146: Let us start by establishing the QD Green's function in
147: the equilibrium situation, i.e.\ for $V\!=\!0$. Fourier transform
148: of the retarded Green's functions can be formally expressed
149: by the Dyson equation
150: \begin{eqnarray}
151: {\mb G}_{\sigma}(\omega)^{-1} &\equiv &
152: \left[ \begin{array}{cc}
153: \langle\langle \hat{d}_{\sigma} ;\hat{d}_{\sigma}^{\dagger}
154: \rangle\rangle_{\omega} & \langle\langle\hat{d}_{\sigma};
155: \hat{d}_{\sigma} \rangle\rangle_{\omega} \\ \langle\langle
156: \hat{d}_{-\sigma}^{\dagger}; \hat{d}_{\sigma}^{\dagger}
157: \rangle\rangle_{\omega} & \langle \langle \hat{d}_{-\sigma}
158: ^{\dagger} ;\hat{d}_{-\sigma} \rangle\rangle_{\omega}
159: \end{array}\right]^{-1} \label{GF} \\ &=&
160: \left[ \begin{array}{cc}
161: \omega\!-\!\varepsilon_{d,\sigma} & 0 \\ 0 &
162: \omega\!+\!\varepsilon_{d,-\sigma}\end{array}\right]
163: - {\mb \Sigma}_{d,\sigma}^{0}(\omega)
164: - {\mb \Sigma}_{d,\sigma}^{U}(\omega)
165: \nonumber
166: \end{eqnarray}
167: where ${\mb \Sigma}_{d,\sigma}^{0}$ denotes the selfenergy of
168: noninteracting QD ($U\!=\!0$) and ${\mb \Sigma}_{d,\sigma}^{U}$
169: accounts for the correlation effects. For a simple understanding
170: of the MTF effect it would be helpful to focus first on the
171: uncorrelated QD when the selfenergy is known exactly.
172: Further corrections due to ${\mb \Sigma}_{d,\sigma}^{U}$
173: contribute a renormalization of the spectral function
174: \cite{Tanaka-07} whose impact on the charge transport will
175: be discussed separately in section V.
176:
177: For convenience we introduce the hybridization coupling
178: $\Gamma_{\beta} \equiv 2\pi\sum_{\bf k} |V_{{\bf k}\beta}
179: |^{2} \delta(\omega\!-\!\varepsilon_{{\bf k\beta}})$ and
180: define the following spin-dependent energy $\tilde{\omega}
181: _{\sigma}\!=\!\omega\!+\!g_{\sigma} \mu_{B}B$. Imaginary
182: part of the selfenergy ${\mb \Sigma}_{d,\sigma}^{0}$ for
183: $|\tilde{\omega}_{\sigma}|\!\leq\!\Delta$ is given by
184: $\mbox{Im}{\mb \Sigma}_{d,\sigma}^{0}(\omega)=-\frac{
185: \Gamma_{N}}{2}{\mb 1}$ while at large energies $D \! >
186: \! |\tilde{\omega}_{\sigma}| \! > \! \Delta$ it takes
187: the following form \cite{Krawiec-04,Domanski-07}
188: %
189: \begin{eqnarray}
190: &&\mbox{Im}{\mb \Sigma}_{d,\sigma}^{0}(\omega)= \nonumber \\
191: &&-\frac{1}{2} \left[ \begin{array}{cc}
192: \Gamma_{N}+\Gamma_{S}\frac{|\tilde{\omega}_{\sigma}|}
193: {\sqrt{\tilde{\omega}_{\sigma}^{2}-\Delta^{2}}} &
194: \Gamma_{S} \frac{\Delta \; \mbox{sgn}(\tilde{\omega}_{\sigma})}
195: {\sqrt{\tilde{\omega}_{\sigma}^{2}-\Delta^{2}}} \\
196: \Gamma_{S} \frac{\Delta \; \mbox{sgn}(\tilde{\omega}_{\sigma})}
197: {\sqrt{\tilde{\omega}_{\sigma}^{2}-\Delta^{2}}} &
198: \Gamma_{N}+\Gamma_{S}\frac{|\tilde{\omega}_{\sigma}|}
199: {\sqrt{\tilde{\omega}_{\sigma}^{2}-\Delta^{2}}}
200: \end{array} \right] .
201: \label{imag_free}
202: \end{eqnarray}
203: %
204: The corresponding real parts can be determined using the
205: Kramers-Kr\"onig relations.
206:
207: Imaginary part of the selfenergy ${\mb \Sigma}_{d,\sigma}^{0}$
208: has thus the square root singularities at energies $\omega\!=\!
209: \pm\Delta \pm \mu_{B}B$, so in presence of magnetic field there
210: are altogether 4 such points. They show up as kinks in
211: the spectral function $\rho_{d}(\omega) =\sum_{\sigma}
212: \rho_{d,\sigma}(\omega)$, where
213: \begin{eqnarray}
214: \rho_{d,\sigma}(\omega) = - \frac{1}{\pi} \mbox{Im}
215: \langle\langle \hat{d}_{\sigma} ;\hat{d}_{\sigma}^{\dagger}
216: \rangle\rangle_{\omega+i0^{+}} .
217: \label{rho_sigma}
218: \end{eqnarray}
219: We shall see below that appearance of such characteristic
220: points leads to the MTF effect observed in the tunneling
221: conductance.
222:
223:
224: \section{Meservey-Tedrow-Fulde effect}
225:
226: To compute the tunneling current we adopt the formalism outlined
227: in the previous studies \cite{Fazio-98,Sun99,Krawiec-04} extending
228: it here on a situation with the spin sensitive transport due to
229: magnetic field. The steady charge current is defined as $I(V)\!=
230: \! -e\frac{d}{dt}\langle \sum_{{\bf k},\sigma} \hat{c}_{{\bf k}
231: \sigma N}^{\dagger} \hat{c}_{{\bf k} \sigma N} \rangle\!=\!
232: e\frac{d}{dt}\langle \sum_{{\bf k},\sigma} \hat{c}_{{\bf k}
233: \sigma S}^{\dagger} \hat{c}_{{\bf k} \sigma S} \rangle$. We
234: carry out the time derivative and determine the expectation
235: value using the nonequilibrium Keldysh Green's functions.
236:
237: % ---------- Figure ---------- %
238: \begin{figure}
239: {\epsfxsize=9cm \centerline{\epsffile{figure1.eps}}}
240: \caption{The differential conductance $G(V)$ versus
241: bias voltage $V$ for N-QD-S junction. Notice a splitting
242: of the gap-edge singularities around $eV\!=\!\pm\Delta$
243: induced by magnetic field $B$. We used for computations
244: $\varepsilon_{d}\!=\!0$, $U\!=\!0$, $\Gamma_{N}\!=\!
245: \Delta$, $\Gamma_{S}\!=\!0.1\Delta$, $T=0.01\Delta$
246: assuming $\Delta\!=\!0.1D$.}
247: \label{MT_plot}
248: \end{figure}
249: % ------------------------------- %
250:
251: In analogy to the standard Blonder-Tinkham-Klawijk theory
252: \cite{Blonder} we express the current as composed
253: of two contributions
254: %
255: \begin{eqnarray}
256: I(V) = I_{1}(V) + I_{A}(V) .
257: \label{total}
258: \end{eqnarray}
259: %
260: The first part $I_{1}(V)$ stands for a contribution which
261: at low temperatures appears practically outside the energy
262: gap $|eV|\!\geq\!\Delta$. Its magnitude is expressed by
263: the Landauer-type formula
264: %
265: \begin{eqnarray}
266: I_{1}(V) = \frac{e}{h} \sum_{\sigma} \int d\omega \;
267: T_{1,\sigma}(\omega) \; \left[ f(\omega\!+\!eV)\!-\!
268: f(\omega) \right] ,
269: \end{eqnarray}
270: %
271: where $f(\omega)\!=\!\left[1+\mbox{exp}(\omega/k_BT) \right]
272: ^{-1}$. The transmittance $T_{1,\sigma}(\omega)$ is nonvanishing
273: only outside the energy gap $|\tilde{\omega}_{\sigma}| \geq
274: \Delta$ and is given by the following parts of the retarded
275: Green's functions \cite{Sun99,Krawiec-04}
276: %
277: \begin{eqnarray}
278: T_{1,\sigma}(\omega) &=& \frac{\Gamma_{N} \Gamma_{S} \;
279: |\tilde{\omega}_{\sigma}|}{\sqrt{\tilde{\omega}_{\sigma}^{2}
280: -\Delta^{2}}} \left( \left| \langle \langle \hat{d}_{\sigma};
281: \hat{d}_{\sigma}^{\dagger} \rangle \rangle_{\omega}
282: \right|^{2} + \left| \langle \langle \hat{d}_{\sigma};
283: \hat{d}_{-\sigma} \rangle \rangle_{\omega} \right|^{2}
284: \right) \nonumber \\
285: &-& \frac{2 \Gamma_{N} \Gamma_{S} \Delta}{\sqrt{\tilde
286: {\omega}_{\sigma}^{2}-\Delta^{2}}} \; \mbox{Re} \left\{
287: \langle \langle \hat{d}_{\sigma}; \hat{d}_{\sigma}^{\dagger}
288: \rangle \rangle_{\omega} \;\; \langle \langle \hat{d}_{\sigma};
289: \hat{d}_{-\sigma} \rangle \rangle_{\omega}^{*} \right\} .
290: \label{T1}
291: \end{eqnarray}
292: %
293:
294:
295: % ---------- Figure ---------- %
296: %\begin{figure}
297: %{\epsfxsize=9cm \centerline{\epsffile{MT_effect.eps}}}
298: %\caption{The same as in figure \ref{MT_plot} but for
299: %several values of magnetic field.}
300: %\label{MT_plot2}
301: %\end{figure}
302: % ------------------------------- %
303:
304: The second part in (\ref{total}) originates from the mechanism
305: of Andreev reflections \cite{Blonder,Fazio-98,Krawiec-04}
306: %
307: \begin{eqnarray}
308: I_{A}(V) = \frac{e}{h} \sum_{\sigma} \int d\omega
309: T_{A,\sigma}(\omega) \left[ f(\omega\!+\!eV)\!-\!
310: f(\omega\!-\!eV) \right] .
311: \end{eqnarray}
312: %
313: Its transmittance is finite even inside the energy gap
314: \cite{Sun99,Krawiec-04}
315: %
316: \begin{eqnarray}
317: T_{A,\sigma}(\omega) = \Gamma_{N}^{2}
318: \left| \langle \langle \hat{d}_{\sigma}; \hat{d}_{-\sigma}
319: \rangle \rangle_{\omega} \right|^{2} .
320: \label{T_A}
321: \end{eqnarray}
322: %
323: Physically such process occurs when an incident electron from
324: $N$ electrode (of arbitrary energy) is converted into a pair
325: on QD (with a simultaneous reflection of a hole) and
326: it propagates in $S$ electrode as a Cooper pair. This
327: anomalous Andreev current is closely related to the
328: off-diagonal order parameter induced in the QD
329: (proximity effect) \cite{Domanski-07,Tanaka-07}.
330:
331: Figure (\ref{MT_plot}) illustrates the influence of magnetic
332: field on the total differential conductance $G(V)\!=\!\frac{d}
333: {dV}I(V)$ obtained for N-QD-S junction. We clearly notice
334: the Zeeman splitting of the square root singularities resembling
335: the former experimental observation for N-I-S (I-insulator)
336: junction \cite{Meservey-70}. However, in a present case
337: the conductance does not saturate to a finite value far
338: outside the gap $|eV| \gg \Delta$ because the QD spectrum
339: spreads only nearby $\varepsilon_{d}$ (usually in realistic
340: multilevel QDs there would be seen the quantum oscillations
341: of $G(V)$ \cite{Aleiner-02}). The in-gap features related
342: to the Andreev current are discussed in the next section.
343:
344: \section{Magnetic field effect on the Andreev current}
345:
346: % ---------- Figure ---------- %
347: \begin{figure}
348: {\epsfxsize=7.5cm \centerline{\epsffile{figure2a.eps}}}
349: {\epsfxsize=8cm \centerline{\epsffile{figure2b.eps}}}
350: \caption{Zeeman splitting of the bound Andreev states for
351: the QD located in the center of superconducting gap
352: $\varepsilon_{d}=0$. Upper panel illustrates the density
353: of states $\rho_{d}(\omega)$ and the bottom figure shows
354: differential conductance of the in-gap current. For
355: computations we used $\Gamma_{N}\!=\!0.1\Gamma_{S}$,
356: $\mu_{B}B\!=\!0.1\Gamma_{S}$ assuming $\Gamma_{S}=0.01D$
357: and $U\!=\!0$.}
358: \label{bound_states}
359: \end{figure}
360: % ------------------------------- %
361:
362: % ---------- Figure ---------- %
363: \begin{figure}[t]
364: \epsfxsize=10cm \centerline{\epsffile{figure3.eps}}
365: \caption{Differential conductance $G_{A}(V)$ of the in-gap
366: Andreev current as a function of the bias voltage $V$ and
367: the QD level $\varepsilon_{d}$. We used for computations
368: $\Gamma_{S}=0.01D$, $\Gamma_{N}=0.1\Gamma_S$, $T=0.01\Gamma_{S}$
369: and set the magnetic field $\frac{1}{2}\mu_{B}B=0.1\Gamma_{S}$.
370: The conductance is expressed in units of $4e^{2}/h$.}
371: \label{GA_vs_ed}
372: \end{figure}
373: % ------------------------------- %
374:
375: The mechanism of Andreev reflections transmits the charge current
376: even for the subgap voltages. To focus solely on this anomalous
377: current it is convenient to consider the extreme limit $\Delta
378: \rightarrow \infty$ as proposed by Tanaka et al \cite{Tanaka-07}.
379: In such case $I_{1}$ can be completely discarded from our analysis.
380: Using (\ref{imag_free}) we obtain the selfenergy
381: ${\mb \Sigma}_{d,\sigma}^{0}$ simplified to
382: \cite{Fazio-98,Domanski-07,Tanaka-07}
383: %
384: \begin{eqnarray}
385: {\mb \Sigma}_{d,\sigma}^{0}(\omega) = -\; \frac{1}{2}
386: \left[ \begin{array}{cc}
387: i \Gamma_{N} & \Gamma_{S} \\
388: \Gamma_{S} & i \Gamma_{N}
389: \end{array} \right]
390: \label{static}
391: \end{eqnarray}
392: %
393:
394: Upon neglecting the Coulomb correlations one can analytically determine
395: the Green's function (\ref{GF}), where the spin dependent spectral
396: function (\ref{rho_sigma}) acquires the BCS structure \cite{Tanaka-07}
397: %
398: \begin{eqnarray}
399: \rho_{d,\sigma}(\omega) &=& \frac{1}{2} \left[ 1 +
400: \frac{\varepsilon_{d}}{E_{d}} \right] \frac{\frac{1}{\pi} \;
401: \Gamma_{N}/2}{(\tilde{\omega}_{\sigma}\!-\!E_{d})^{2}+
402: (\Gamma_{N}/2)^{2}} \nonumber \\ &+& \frac{1}{2}
403: \left[ 1 - \frac{\varepsilon_{d}}{E_{d}} \right]
404: \frac{\frac{1}{\pi} \;\Gamma_{N}/2}
405: {(\tilde{\omega}_{\sigma}\!+\!E_{d})^{2}
406: +(\Gamma_{N}/2)^{2}}
407: \label{rho_ingap}
408: \end{eqnarray}
409: %
410: with a quasiparticle energy $E_{d}=\sqrt{\varepsilon_{d}^{2}
411: +(\Gamma_{S}/2)^{2}}$. The in-gap QD states (often referred as
412: {\em Andreev bound states}) form around $\pm E_{d} \! \pm \!
413: \mu_{B}B$ as illustrated in the upper panel of figure
414: \ref{bound_states}. Their line broadening is given by
415: $\Gamma_{N}/2$ and in absence of magnetic field the particle-hole
416: splitting is controlled by $\Gamma_{S}$ \cite{Tanaka-07,Domanski-07}
417: (the dashed line in figure \ref{bound_states}). Magnetic field
418: further enforces the Zeeman splitting of these in-gap states.
419:
420: Above mentioned behavior has an indirect effect on the
421: off-diagonal parts of the Green's function (\ref{GF}) which
422: in turn determine the Andreev transmittance. In the
423: limit $\Delta \rightarrow \infty$ (\ref{T_A}) reduces to
424: %
425: \begin{eqnarray}
426: T_{A,\sigma}(\omega) = \frac{\Gamma_{N}^{2}
427: \left( \Gamma_{S}/2 \right)^{2} }
428: {\left[ (\tilde{\omega}_{\sigma}\!-\!E_{d})^{2}
429: \!+\!(\Gamma_{N}/2)^{2} \right] \left[
430: (\tilde{\omega}_{\sigma}\!+\!E_{d})^{2}
431: \!+\!(\Gamma_{N}/2)^{2}\right] }
432: \nonumber \\
433: \label{T_A_ingap}
434: \end{eqnarray}
435: %
436: The subgap Andreev conductance $G_{A}(V)=\frac{d}{dV} I_{A}(V)$
437: is thus characterized by a four peak structure as shown in
438: the bottom panel of figure \ref{bound_states}. Obviously
439: the weights of particle and hole peaks of the spectral
440: function (\ref{rho_ingap}) as well as their weights in
441: the Andreev transmittance (\ref{T_A_ingap}) depend on the QD
442: level $\varepsilon_{d}$. Variation of the Andreev conductance
443: with respect to ($V$, $\varepsilon_{d}$) is plotted in figure
444: \ref{GA_vs_ed}. We can notice that optimal conditions for
445: the subgap current occur when the QD level is located near
446: the energy gap center, otherwise the proximity effect is
447: less efficient.
448:
449: % ---------- Figure ---------- %
450: \begin{figure}[t]
451: \epsfxsize=10cm \centerline{\epsffile{figure4.eps}}
452: \caption{Differential conductance $G_{A}(V)$ of the in-gap
453: Andreev current as a function of the bias voltage $V$ and
454: magnetic field $B$ for the QD level $\varepsilon_{d}\!=\!0$ and
455: $\Gamma_{N}\!=\!0.1\Gamma_S$, $T=0.01\Gamma_{S}$ .
456: Dark areas denote the regions where $G_{A}$ approaches
457: the value $4e^{2}/h$.}
458: \label{GA_vs_B}
459: \end{figure}
460: % ------------------------------- %
461:
462: On top of the particle-hole structure seen in the Andreev states
463: there is an additional Zeeman splitting brought by magnetic field.
464: In figure \ref{GA_vs_B} we sketch the Andreev conductance in
465: ($V$,$B$) plane for $\varepsilon_{d}\!=\!0$, where the dark areas
466: correspond to a maximal value $4e^{2}/h$. There appears a
467: characteristic diamond shape marking the positions of such
468: maximal conductance $G_{A}(V,B)$. We believe that this
469: hyperfine structure could be probed experimentally.
470:
471: To complete the discussion of the subgap Andreev current
472: we briefly comment on a possible influence of an asymmetry
473: between the hybridization couplings $\Gamma_{N}$, $\Gamma_{S}$.
474: We explore for this purpose the zero bias conductance $G_{A}
475: (V\!=\!0)$. At low temperature we find from equation
476: (\ref{T_A_ingap}) that
477: %
478: \begin{eqnarray}
479: &&G_{A}(0) = \label{GA_zero} \\&& \frac{4e^{2}}{h} \frac{\Gamma_{N}^{2} \;
480: \left( \Gamma_{S}/2 \right)^{2} }{\left[ (\mu_{B}B
481: \!-\!E_d)^{2}\!+\!\left( \frac{\Gamma_{N}}{2} \right)^{2}
482: \right]\!\left[(\mu_{B}B\!+\!E_d)^{2}\!+\!\left(
483: \frac{\Gamma_{N}}{2}\right)^{2}\right] } .
484: \nonumber
485: \end{eqnarray}
486: %
487: In figure \ref{GA_zeroV} we show the influence of magnetic
488: field on the zero bias Andreev conductance for several
489: values of the asymmetry rate $\Gamma_{N}/\Gamma_{S}$. If
490: $\Gamma_{N}/\Gamma_{S}\! \ll \! 1$ then a line-broadening
491: of the Andreev states diminishes so in consequence the
492: particle and hole peaks become well separated. Under such
493: conditions the subgap conductance has maxima around the
494: quasiparticle states at $\pm\Gamma_{S}/2$ (where the ideal
495: conductance $4e^{2}/h$ is reached). Let us recall, that in
496: absence of magnetic field the equation (\ref{GA_zero})
497: reproduces for $\varepsilon_{d}\!=\!0$ the well known result
498: $G_{A}(0)\! =\! \frac{4e^{2}}{h} \left( \frac{2\Gamma_{N}
499: \Gamma_{S}}{\Gamma_{S}^{2}\!+\! \Gamma_{N}^{2}}\right)^{2}$
500: \cite{Tanaka-07}. For the symmetric coupling $\Gamma_S
501: \!=\!\Gamma_N$ it yields $G_{A}(0)\! = \! 4e^{2}/h$
502: \cite{Fazio-98}.
503:
504: % ---------- Figure ---------- %
505: \begin{figure}
506: \epsfxsize=8cm \centerline{\epsffile{figure5.eps}}
507: \caption{The zero bias differential conductance $G_{A}(0)$
508: as a function the magnetic field $B$ for several relative
509: values of $\Gamma_{N}/\Gamma_{S}$. We used for computations
510: $\varepsilon_{d}\!=\!0$ assuming $T\!\rightarrow\!0$.}
511: \label{GA_zeroV}
512: \end{figure}
513: % ------------------------------- %
514:
515:
516: \section{Influence of the Coulomb correlations}
517:
518: In the limit $\Delta\!\rightarrow\!\infty$ the selfenergy
519: ${\mb \Sigma}_{d,\sigma}^{0}$ becomes a static quantity (\ref{static})
520: therefore the role of superconducting lead can be exactly replaced
521: by the on-dot gap parameter $\Delta_{d}=\Gamma_{S}/2$. Instead of
522: (\ref{model}) we can thus use the following auxiliary Hamiltonian
523: %
524: \begin{eqnarray}
525: \hat{H} &=& \hat{H}_{N} + \sum_{{\bf k},\sigma }
526: \left( V_{{\bf k} N} \; \hat{d}_{\sigma}^{\dagger}
527: \hat{c}_{{\bf k} \sigma \beta } + \mbox{h.c.} \right)
528: + \sum_{\sigma} \epsilon_{d,\sigma} \hat{d}^{\dagger}_{\sigma}
529: \hat{d}_{\sigma} \nonumber \\ & + &
530: \left( \Delta_{d} \hat{d}_{\uparrow}^{\dagger}
531: \hat{d}_{\downarrow}^{\dagger} + \mbox{h.c.} \right)
532: + U \; \hat{n}_{d \uparrow} \hat{n}_{d \downarrow} ,
533: \label{Tanaka_model}
534: \end{eqnarray}
535: %
536: which turns out to be very convenient for investigating the
537: correlations. Tanaka and coworkers \cite{Oguri-04,Tanaka-07} were
538: able to rigorously prove that the selfenergy ${\mb \Sigma}_{d,
539: \sigma}^{U}$ must have a diagonal structure due to invariance
540: of $U \hat{n}_{d \uparrow} \hat{n}_{d \downarrow}$ term on
541: the Bogoliubov-Valatin transformation.
542:
543: In the remaining part of this section we shall focus on the subgap
544: Andreev current transmitted through the correlated QD. The matrix
545: Green's function (\ref{GF}) simplifies in the limit $\Delta \!
546: \rightarrow \!\infty$ to the following (exact) structure
547: %
548: \begin{eqnarray}
549: && {\mb G}_{\sigma}(\omega)= \label{arbitrary_U} \\
550: &&\left( \begin{array}{cc} \omega\!-\!\varepsilon_{d,\sigma}\!-\!
551: \Sigma_{N,\sigma}(\omega) & \frac{1}{2}\Gamma_{S} \\
552: \frac{1}{2}\Gamma_{S} & \omega\!+\!\varepsilon_{d,-\sigma}
553: \!+\!\Sigma_{N,-\sigma}^{*}(-\omega)\end{array}\right)^{-1} .
554: \nonumber
555: \end{eqnarray}
556: %
557: Influence of the correlations have been so far analyzed for
558: the Hamiltonian (\ref{model}) using various techniques
559: \cite{Fazio-98,Clerk-00,Cuevas-01,Krawiec-04,Oguri-04,Tanaka-07,
560: Domanski-07}. Here we estimate the diagonal selfenergy $\Sigma_{N,
561: \sigma}(\omega)$ within (\ref{Tanaka_model}) by the equation
562: of motion method \cite{EOM,TDAD-08}
563: \begin{widetext}
564: %
565: \begin{eqnarray}
566: \omega\!-\!\varepsilon_{d,\sigma}\!-\!\Sigma_{N,\sigma}(\omega)\!=\!
567: \frac{[\omega\!-\!\varepsilon_{d,\sigma}\! -\!\Sigma^{0}_{d,\sigma}
568: (\omega)][\omega\!-\!\varepsilon_{d,\sigma}\! -\!U\!-\!
569: \Sigma^{0}_{d,\sigma}(\omega)\!-\!\Sigma^{3}_{d,\sigma}(\omega)]
570: \!+\!U \Sigma^{1}_{d,\sigma}(\omega)} {\omega-\varepsilon_{d,\sigma}
571: - \Sigma^{0}_{d,\sigma}(\omega)-\Sigma^{3}_{d,\sigma}(\omega)
572: -U[1-\langle \hat{n}_{d,-\sigma}\rangle ]}
573: \label{ansatz}
574: \end{eqnarray}
575: %
576: where $\Sigma^{\nu=1,3}_{d,\sigma}(\omega)$ are
577: given by \cite{EOM}
578: %
579: \begin{eqnarray}
580: \Sigma^{\nu}_{d,\sigma}(\omega) & = & \sum_{\bf k}
581: |V_{{\bf k} N}|^{2} \left(\frac{1}{\omega\!+\!\xi_{{\bf k} N}
582: \!-\!\varepsilon_{d,-\sigma}\!-\!\varepsilon_{d,\sigma}\!-\! U}
583: + \frac{1}{\omega\!-\!\xi_{{\bf k} N}\!+\!\varepsilon_{d,-\sigma}
584: \!-\!\varepsilon_{d,\sigma}} \right) \; \left[ f(\omega,T)
585: \right]^{\frac{3 - \nu }{2}} .
586: \label{sigma1and3}
587: \end{eqnarray}
588: %
589: \end{widetext}
590:
591: Approximation (\ref{ansatz},\ref{sigma1and3}) qualitatively
592: reproduces the following properties caused by on-dot correlations:
593: (i) the charging effect and (ii) a possible appearance of the Kondo
594: resonance for temperatures smaller than
595: $T_{K}\!=\!\frac{\sqrt{U\Gamma_{N}}}{2} \mbox{exp}\{ \pi \varepsilon_{d}
596: \left( \varepsilon_{d} \! + \! U \right) / U \Gamma_{N}\}$. The
597: latter one is related to screening of the quantum dot spin by
598: itinerant electrons of the metallic lead. In the case when energy
599: level $\varepsilon_{d}$ is located slightly below $\mu_{N}$ the
600: hybridization $V_{{\bf k}N}$ induces effectively antiferromagnetic
601: interaction between the QD and metallic lead. In consequence the
602: bound singlet state can be formed giving rise to the resonance at
603: $\omega=\mu_{N}$ for temperatures $T \leq T_{K}$. Magnetic field
604: eventually splits this resonance as illustrated in figure
605: \ref{Kondo_peak}.
606:
607: % ---------- Figure ---------- %
608: \begin{figure}
609: {\epsfxsize=9cm \centerline{\epsffile{figure6.eps}}}
610: \caption{Spectral function of the correlated QD obtained for
611: $\varepsilon_{d}=-1.5\Gamma_{S}$, $U\!=\!10\Gamma_{S}$,
612: $\Gamma_{N} \! = \! \Gamma_{S}$ and temperature $T=10^{-3}
613: \Gamma_{S}$ ($<< T_{K}$) in the limit $\Delta \! \rightarrow
614: \! \infty$. Solid line corresponds to $\mu_{B}B\!=\!
615: \Gamma_{S}/3$.}
616: \label{Kondo_peak}
617: \end{figure}
618: % ------------------------------- %
619:
620: Any features present in the QD spectrum are further showing
621: up in the measurable differential conductance. This is also
622: valid for the Kondo resonance. Since it forms near the chemical
623: potential $\mu_{N}$ therefore its signatures appear predominantly
624: in the low voltage current. In fact, it has been shown that Kondo
625: resonance enhances at low temperatures the zero bias Andreev
626: conductance \cite{Fazio-98,Domanski-07}, however its magnitude
627: remains much smaller than the unitary limit value $2e^{2}/h$
628: typical for N-QD-N systems in the Kondo regime. In the present
629: context we emphasize that magnetic field enforces the Zeeman
630: splitting of the zero bias Andreev anomaly in much the same
631: way as it affects the zero bias anomaly for the QD coupled
632: to both metallic leads \cite{magn_N-QD-N,Kastner_etal}.
633:
634: % ---------- Figure ---------- %
635: \begin{figure}
636: {\epsfxsize=9cm \centerline{\epsffile{figure7.eps}}}
637: \caption{Effect of magnetic field on the subgap Andreev
638: conductance $G_{A}(V)$ in the Kondo regime with $\varepsilon_{d}
639: \!=\!-1.5 \Gamma_{N}$, $U\!=\!10\Gamma_{N}$, $\Gamma_{S}\!=\!
640: 5 \Gamma_{N}$ and $T \!=\! 10^{-3} \Gamma_{N} \ll T_{K}$.
641: Notice appearance of: (i) the zero-bias Kondo anomaly
642: (showing the Zeeman splitting for $\mu_{B}B\!=\!\Gamma_{N}/3$,
643: (ii) the quasiparticle peaks at $|eV|\! \simeq \!E_{d}$, and
644: (iii) Coulomb satellite peaks near $|eV| \simeq U$.}
645: \label{anomaly}
646: \end{figure}
647: % ------------------------------- %
648:
649:
650: The zero bias enhancement of the Andreev conductance is a feature
651: whose presence might be difficult to notice \cite{Fazio-98,Clerk-00,
652: Domanski-07} unless some stringent requirements are fulfilled
653: \cite{TDAD-08}. It turns out that optimal conditions for the low
654: temperature enhancement of $G_{A}(V\!\sim\!0)$ take place when
655: $\Gamma_{S}$ is comparable to $\Gamma_{N}$ (see figure
656: \ref{opt_conditions}) and $\varepsilon_{d}$ is located slightly
657: below the energy gap center. For an increasing asymmetry between
658: the hybridizations $\Gamma_{N}$, $\Gamma_{S}$
659: the magnitude of low voltage Andreev conductance diminishes
660: (similarly as we have been shown in section IV upon neglecting
661: the correlations). On the other hand, for $\varepsilon_{d}$ moving
662: far aside from the superconductor's gap center the proximity
663: effect becomes weakened and the overall Andreev conductance
664: is again suppressed.
665:
666: In general it seems that an interplay between the on-dot pairing
667: (absorbed from the superconducting electrode) and the Kondo state
668: (due to screening of QD spin by the metallic lead electrons) has
669: the same character as a competition of superconductivity versus
670: magnetism in the solid state physics. Since this is outside
671: the main scope of the present topic we shall discuss it
672: separately \cite{TDAD-08}. A combination of the Kondo physics,
673: superconductivity and the Zeeman polarization is a complex
674: problem and to our knowledge only few papers have so far
675: attempted to address this challenging issue \cite{Yamada-07}.
676:
677: % ---------- Figure ---------- %
678: \begin{figure}
679: {\epsfxsize=9.5cm \centerline{\epsffile{figure8.eps}}}
680: \caption{The differential Andreev conductance $G_{A}(V)$
681: (in units $4e^{2}/h$) as a function of the bias voltage $V$
682: and the asymmetry ratio $\Gamma_{S}/\Gamma_{N}$. We used
683: the same set of parameters as in figure \ref{anomaly}.}
684: \label{opt_conditions}
685: \end{figure}
686: % ------------------------------- %
687:
688:
689: \section{Summary}
690:
691: We have explored the effect of magnetic field on charge
692: transport through the quantum dot attached to one normal
693: and one superconducting electrode. For a bias voltage
694: $V\!\simeq\!\pm\Delta/e$ we find the Zeeman splitting
695: of the square root singularities in the differential
696: conductance. This resembles the experimental result
697: of Meservey, Tedrow and Fulde observed in the N-I-S
698: junction \cite{Meservey-70} which for the N-QD-S
699: structures it seems rather easy to achieve.
700:
701: We have extended our study also on the in-gap Andreev current.
702: Due to the proximity effect the particles and holes of the
703: quantum dot get mixed and effectively the spectrum acquires
704: the BCS-like structure (\ref{rho_ingap}). Differential
705: conductance $G_{A}(V)$ of the in-gap current indirectly probes
706: such structure of the bound Andreev states. We have shown
707: that magnetic field leads to appearance of four peaks via
708: the combined particle-hole and Zeeman splittings. We hope
709: that this result might stimulate a search for the experimental
710: detection of above mentioned structures.
711:
712: Moreover, we have explored influence of the on-dot Coulomb
713: interactions on the subgap Andreev current assuming the
714: extreme limit $\Delta\!\rightarrow\!\infty$. In general,
715: the on-dot correlations contribute to the QD spectrum:
716: (i) appearance of the Coulomb satellite near $\omega\!=\!
717: \varepsilon_{d,\uparrow}+\varepsilon_{d,\downarrow}+U$
718: (charging effect), and (ii) at sufficiently low temperatures
719: can produce the narrow Kondo resonance at the chemical potential
720: $\mu_{N}$. Magnetic field imposes the hyperfine splitting onto
721: such spectrum in a similar way as has been observed in N-QD-N junctions
722: \cite{Kastner_etal}. The Kondo effect alone is exemplified
723: in the zero bias Andreev conductance where under appropriate
724: conditions \cite{TDAD-08} a low temperature enhancement can
725: be seen if $\Gamma_{N} \sim \Gamma_{S}$ and the gate voltage
726: tunes $\varepsilon_{d}$ nearly to the energy gap center.
727:
728: It would be of interest to use some more sophisticated methods
729: for treating the on-dot interaction $U$ in order to check
730: whether there exist a minimal magnetic field necessary for
731: splitting the Kondo peak (as theoretically predicted for N-QD-N
732: junctions \cite{magn_N-QD-N}) observable in the Andreev conductance.
733: %analyze an interplay between the Zeeman polarization and efficiency
734: %of the QD screening competing with the on-dot pairing.
735: One can also study QD coupled with $d$-wave superconductor,
736: where the square root singularities are replaced by weaker
737: kinks. We think that the Meservey-Tedrow-Fulde effect would
738: be observable there too (but in a less pronounced manner)
739: whereas the subgap conductance might qualitatively change.
740:
741:
742:
743: \vspace{0.5cm}
744: {\em Acknowledgment}
745: We thank A.M.\ Gabovich for discussion leading to this
746: study and P. Fulde for interest and encouragement.
747: This work was partly supported by the Ministry of Science
748: and Education under the grants NN202187833 and NN202373333.
749:
750:
751: \begin{thebibliography}{11}
752: \bibitem{Meservey-70}
753: R.\ Meservey, P.M.\ Tedrow, and P.\ Fulde,
754: Phys.\ Rev.\ Lett.\ {\bf 25}, 1270 (1970); P.\ Fulde,
755: Adv. Phys. {\bf 22} 667 (1973).
756: \bibitem{Alvarez06}
757: G.A. Alvarez, I.\ Iguchi, X.L.\ Wang, S.X.\ Dou, and
758: Q.W. Yao, J.\ Appl.\ Phys.\ {\bf 99}, 08C912 (2006).
759: \bibitem{Blonder}
760: G.E.\ Blonder, M.\ Tinkham, and T.M.\ Klapwijk,
761: Phys.\ Rev.\ B {\bf 25}, 4515 (1982);
762: G.\ Deutscher, Rev.\ Mod.\ Phys.\ {\bf 77},
763: 109 (2005).
764: % ---- tunneling: general formalism ------
765: \bibitem{Beenakker92}
766: C.W.J.\ Beenakker, Phys.\ Rev.\ B {\bf 46}, 12841 (1992).
767: \bibitem{Claughton95}
768: N.R.\ Claughton, M.\ Leadbeater and C.J.\ Lambert,
769: J.\ Phys.: Cond.\ Matter {\bf 7}, 8757 (1995).
770: \bibitem{Yeyati97}
771: A.L.\ Yeyati, J.C.\ Cuevas, A.\ Lopez-Davalos, and A.\ Martin-Rodero,
772: Phys.\ Rev.\ B {\bf 55}, R6137 (1997).
773: \bibitem{Zhao99}
774: H.-K. Zhao, Phys. Lett. A {\bf 264}, 218 (1999);
775: H.-K. Zhao and J. Wang, Eur.\ Phys.\ J.\ B {\bf 44}, 93 (2005).
776: \bibitem{Sun99}
777: Q.-F.\ Sun, J.\ Wang, T.-H.\ Lin,
778: Phys.\ Rev.\ B {\bf 59}, 3831 (1999).
779: %----- correlations -----------
780: \bibitem{Fazio-98}
781: R.\ Fazio and R.\ Raimondi, Phys.\ Rev.\ Lett.\ {\bf 80},
782: 2913 (1998); Phys.\ Rev.\ Lett.\ {\bf 82}, 4950 (1999);
783: P.\ Schwab and R.\ Raimondi, Phys.\ Rev.\ B {\bf 59},
784: 1637 (1999).
785: \bibitem{Clerk-00}
786: A.A.\ Clerk, V.\ Ambegaokar, and S.\ Hershfield,
787: Phys.\ Rev.\ B {\bf 61}, 3555 (2000).
788: \bibitem{Cuevas-01}
789: J.C.\ Cuevas, A.\ Levy Yeyati, and A.\ Martin-Rodero,
790: Phys.\ Rev.\ B {\bf 63}, 094515 (2001).
791: \bibitem{Krawiec-04}
792: M.\ Krawiec and K.I.\ Wysoki\'nski,
793: Supercond.\ Sci.\ Technol.\ {\bf 17}, 103 (2004).
794: \bibitem{Oguri-04}
795: A.\ Oguri, Y.\ Tanaka, and A.C.\ Hewson,
796: J.\ Phys.\ Soc.\ Jpn.\ {\bf 73}, 2494 (2004).
797: \bibitem{Tanaka-07}
798: Y.\ Tanaka, N.\ Kawakami, and A.\ Oguri,
799: J.\ Phys.\ Soc.\ Jpn.\ {\bf 76}, 074701 (2007).
800: \bibitem{Domanski-07}
801: T.\ Doma\'nski, A.\ Donabidowicz, and K.I.\ Wysoki\'nski,
802: Phys.\ Rev.\ B {\bf 76}, 104514 (2007).
803: \bibitem{Taddei-07}
804: J.\ Splettstoesser, M.\ Governale, J.\ K\"onig, F.\ Taddei,
805: and R.\ Fazio, Phys.\ Rev.\ B {\bf 75}, 235302 (2007).
806: \bibitem{Gezzi08}
807: R.\ Gezzi, A.\ Dirks, and Th.\ Pruschke,
808: cond-mat/0707.0289 (preprint).
809: \bibitem{Hecht-08}
810: T.\ Hecht, A.\ Weichselbaum, J.\ von Delft, and R.\ Bulla,
811: J.\ Phys.: Condens.\ Matter {\bf 20}, 275213 (2008).
812: % ------ applications -----------
813: \bibitem{Giazotto06}
814: F.\ Giazotto, F.\ Taddei, M.\ Governale, C.\ Castellana,
815: R.\ Fazio, and F.\ Beltram,
816: Phys. Rev. Lett. {\bf 97}, 197001 (2006).
817: \bibitem{Sadovsky07}
818: I.A.\ Sadovskyy, G.B.\ Lesovik, and G.\ Blatter,
819: JETP Lett.\ {\bf 86}, 210 (2007).
820: % ----- review paper ---
821: \bibitem{Aleiner-02}
822: I.L.\ Aleiner, P.W.\ Brouwer, and L.I.\ Glazman,
823: Phys.\ Rep.\ {\bf 358}, 309 (2002).
824: % ----- reference to EOM
825: \bibitem{EOM}
826: H.\ Haug and A.-P. Jauho, {\em Quantum Kinetics in Transport
827: and Optics of Semiconductors}, Springer Verlag, Berlin (1996).
828: % ---- magnetic filed infl on N-QD-N : theoretical papers
829: \bibitem{magn_N-QD-N}
830: J.\ Hong and W.\ Woo,
831: Phys.\ Rev.\ Lett.\ {\bf 99}, 196801 (2007);
832: A.\ Rosch, J.\ Paaske, J.\ Kroha, and P.\ W\"olfle,
833: Phys.\ Rev.\ Lett.\ {\bf 90}, 076804 (2003);
834: T.A.\ Costi, Phys.\ Rev.\ Lett.\ {\bf 85}, 1504 (2000);
835: % ----- studies of the magnetic field effect on the QD
836: \bibitem{Kastner_etal}
837: S.\ Amasha, I.J.\ Gelfand, M.A.\ Kastner, and A.\ Kogan,
838: Phys.\ Rev.\ B {\bf 72}, 045308 (2005);
839: A.\ Kogan, S.\ Amasha, D.\ Golhaber-Gordon, G.\ Granger,
840: M.A.\ Kastner, and H.\ Shtrikman,
841: Phys.\ Rev.\ Lett.\ {\bf 93}, 166602 (2004).
842: \bibitem{TDAD-08}
843: T.\ Doma\'nski and A.\ Donabidowicz,
844: Phys.\ Rev.\ B {\bf 78}, 073105 (2008).
845: \bibitem{Yamada-07}
846: Y.\ Yamada, Y.\ Tanaka, and N.\ Kawakami,
847: Physica E {\bf 40}, 265 (2007).
848: % ----- experimental data ----------
849: \bibitem{S-QD-S}
850: C.\ Buizert, A.\ Oiwa, K.\ Shibata, K.\ Hirakawa,
851: and S. Tarucha, Phys.\ Rev.\ Lett.\ {\bf 99}, 136806 (2007);
852: T.\ Sand-Jesperson {\em et al},
853: Phys.\ Rev.\ Lett.\ {\bf 99}, 126603 (2007);
854: K.\ Grove-Rasmussen, H.I.\ Jorgensen, and P.E.\ Lindelof,
855: New J.\ Phys.\ {\bf 9}, 124 (2007);
856: A.\ Eichler {\em et al},
857: Phys.\ Rev.\ Lett.\ {\bf 99}, 126602 (2007).
858: \end{thebibliography}
859:
860: \end{document}
861:
862: