1: \documentclass[pra,twocolumn,showpacs]{revtex4}
2: \usepackage{graphicx}
3: \usepackage{amsmath}
4: \usepackage[abs]{overpic}
5: \usepackage{rotating}
6: \usepackage{hyperref}
7: \newcommand{\ket}[1]{|#1\rangle}
8: \newcommand{\bra}[1]{\langle #1|}
9:
10: \begin{document}
11: \title{Geometric Phase in Entangled Bipartite Systems}
12: \author{H. T. Cui}\email{cuiht@student.dlut.edu.cn}
13: \author{L. C. Wang, X. X. Yi} \homepage{http://qpi.dlut.edu.cn}
14: \affiliation{Department of Physics, Dalian University of Technology,
15: Dalian 116024, China}
16: %\date{}
17:
18: \date{\today}
19: \begin{abstract}
20: The geometric phase (GP) for bipartite systems in transverse
21: external magnetic fields is investigated in this paper. Two
22: different situations have been studied. We first consider two
23: non-interacting particles. The results show that because of
24: entanglement, the geometric phase is very different from that of the
25: non-entangled case. When the initial state is a Werner state, the
26: geometric phase is, in general, zero and moreover the singularity of
27: the geometric phase may appear with a proper evolution time. We next
28: study the geometric phase when intra-couplings appear and choose
29: Werner states as the initial states to entail this discussion. The
30: results show that unlike our first case, the absolute value of the
31: GP is not zero, and attains its maximum when the rescaled coupling
32: constant $J$ is less than 1. The effect of inhomogeneity of the
33: magnetic field is also discussed.
34: \end{abstract}
35: \pacs{03.65.Vf, 03.65.Ud}
36: \maketitle
37:
38: \section{introduction}
39: A system can retain the information of its motion when it undergoes
40: a cyclic evolution, in the form of a geometric phase (GP), which was
41: first put forward by Pancharatnam in optics \cite{p} and later
42: studied explicitly by Berry in a general quantal
43: system\cite{berry}. Great progress has been made in this novel
44: region \cite{shapere}; The original adiabatic condition in Berry's
45: work has been removed by Aharonov and Anandan \cite{aharonov}, and
46: Samuel and Bhandari have generalized the geometric phase by
47: extending to noncyclic evolution and sequential measurements
48: \cite{samuel}. At the same time a kinematic approach to the theory
49: of the geometric phase has also been developed by Mukunda and Simon
50: \cite{mukunda1}. Recently the generalization to mixed states was
51: conducted first by Uhlmann in the mathematical context of
52: purification \cite{uhlmann}, and then by Sj\"{o}qvist {\it et al.}
53: based on the Mach-Zender interferometer \cite{erik}. Consequently
54: the mixed-state geometric phase has been experimentally verified
55: using both NMR interferometry \cite{du} and single photon
56: interferometry \cite{ericsson}. Recently the geometric phase for
57: mixed state was put forward by Singh \textit{et al.} for the case
58: of degenerate density operators \cite{singh} and a general formula
59: for the parallel transporting was also provided. Despite the great
60: progress in this field, Bhandari recently raised the criticism that
61: the definition for geometric phase in mixed state fails when the
62: interference fringes disappear \cite{bhandari}. This can be
63: explained as the disappearance (appearance) of the geometric phase
64: (off-diagonal geometric phase). The definition of off-diagonal
65: geometric phase(OP) was first given by Manini \textit{et al.} for
66: pure states in adiabatic evolutions\cite{manini}, and then was
67: generalized to non-adiabatic situation\cite{mukunda2} and in
68: mixed-states\cite{filipp}. Moreover the off-diagonal geometric phase
69: was studied in degenerate case \cite{tong1}and in bipartite
70: systems\cite{yi1}. Recently the effect of entanglement on the
71: off-diagonal geometric phase has also discussed \cite{cui}.
72:
73:
74: The quantum computation scheme for the geometric phase has been
75: proposed based on the Abelian \cite{jones} or non-Abelian geometric
76: phase \cite{zanardi}, in which geometric phase has been shown to be
77: intrinsic against faults in the presence of some kind of external
78: noise due to the geometric nature of Berry phase. Consequently the
79: quantum gates based on geometric phase have also been proposed in
80: different systems \cite{zheng}, where the interactions play an
81: important role for the realization of some specific operations.
82:
83: Bipartite system is of great importance in quantum computation, such
84: as the transfer of quantum information, the construction of
85: entanglement as well as the realizations of logic operations.
86: Furthermore, it was found that GPs may be used to design quantum
87: logic gates. These facts together give rise to a question what are
88: the geometric phase and its motion in bipartite systems. Recently
89: some papers have addressed this issue \cite{erik2, yi2}, where the
90: discussions respectively focus on the entanglement dependence of the
91: geometric phase for subsystem and the coupling effect on the GP for
92: subsystem under adiabatical evolution. However, another question
93: remains open; how the entanglement or interaction affects the
94: geometric phase for the whole system. This consideration is not
95: trivial since any quantum information procession cannot implemented
96: by only one qubit. Moreover some interesting results may be
97: found with the increment of the size of the system. In this paper,
98: focusing on the entanglement and inter-subsystem couplings, we
99: present explicit discussion on the geometric phases in bipartite
100: systems.
101:
102: For this purpose, our discussion is divided into several sections.
103: In Sec. II, we describe the model to be discussed in this paper and
104: some formulas are also present for the calculation of the
105: geometric phase. Then in Sec. III we study the geometric phase for
106: two noninteracting particles and some interesting results can be
107: found in this section. The intra-subsystem coupling effect on the
108: geometric phase is studied in Sec. IV. Finally, conclusions are
109: given at the end of this paper.
110:
111:
112: \section{model}
113: In order to highlight the effect of the entanglement, we choose the
114: initial state as
115: \begin{equation}\label{rho}
116: \rho(0)=\frac{1-r}{4}I+r\ket{\Phi}\bra{\Phi},
117: \end{equation}
118: where $r\in(0,1]$ determines the mixing of this state and $I$ is the
119: unit matrix in the $2\times2$ Hilbert space. Notice that for $r=0$
120: Eq. \eqref{rho} is the unit matrix and its geometric phase is
121: trivial (always be zero), our discussion excludes this case. The
122: state $\ket{\Phi}$ may be either of the following states,
123: \begin{eqnarray}\label{}
124: \ket{\varphi}&=\sin\theta\ket{11}_{1,2}
125: +\cos\theta\ket{00}_{1,2} \nonumber\\
126: \ket{\psi}&=\sin\theta\ket{10}_{1,2}
127: +\cos\theta\ket{01}_{1,2}
128: \end{eqnarray}
129: where $\theta$ determines the degree of entanglement and
130: $\ket{1(0)}_{i}(i=1,2)$ is the eigenstate of Pauli operator
131: $\sigma_{z}$. One should note that when $\theta=(3)\pi/4$, the
132: equations above are Bell states and Eq.(\ref{rho}) are the Werner
133: states \cite{werner}, which plays an important role in quantum
134: information processings, especially in quantum communication via
135: noisy channels \cite{bennett1} and quantum distillation
136: \cite{bennett2}. In general the mixed state Eq. (\ref{rho}), which
137: first was introduced by Wootters \cite{wootters}, may be entangled;
138: that can be judged by the Peres-Horodecki condition \cite{peres}.
139: Eq. (\ref{rho}) includes all possible cases, such as pure or mixed
140: state and maximal or non-maximal entangled states. One should note
141: that Eq. (\ref{rho}) is triplet-degenerate for $r\neq1$.
142:
143: We should point out that the initial state Eq. \eqref{rho} is not a
144: trivial generalizing from the pure case. The first term in Eq.
145: \eqref{rho} can be regarded as the noise and the mixing coefficient
146: $r$ properly describe the intensity of noise. Recently the
147: one-to-one correspondence between $r$ of Werner state and the
148: temperature $T$ of the one-dimensional Heisenberg two-spin chain
149: with a magnetic field $B$ along the $z$-axis, has been established
150: \cite{batle}. This connection gives us the strong physical support
151: for the initial state Eq. \eqref{rho}.
152:
153: We choose the system composed of two spin-1/2 particles undergoing
154: spin procession in an external time independent magnetic field in
155: the $z$ direction. Then the Hamiltonian is
156: \begin{equation}\label{h}
157: H=H_{0}+H_{I},
158: \end{equation}
159: in which the free Hamiltonian $H_0$ and the XX interaction $H_I$ are respectively
160: \begin{eqnarray}
161: H_{0}&=&\omega_{1}S_1^z+\omega_{2}S_2^z\nonumber \\
162: H_{I}&=&g(S^{\dag}_{1} S^{-}_2 + S^{-}_1 S^{\dag}_2),
163: \end{eqnarray}
164: where $g>0$ is the antiferromagnetism coupling constant and
165: $S_i^z(i=1,2)$ is the $z$ component of spin operator respectively.
166: $S^{\pm}_{i}=S_{i}^x+iS_{i}^y(i=1,2)$ are the raising and decreasing
167: operators of the $z$ component of spin-1/2. Actually the Hamiltonian
168: Eq. (\ref{h}) is two-qubit XX model, which is of fundamental
169: importance to understand the relation between the entanglement and
170: quantum correlation in interacting many-body systems. In general we
171: suppose that $\omega_1$ may not be equal to $\omega_2$ because of
172: the inhomogeneity of the external magnetic field. One will find in
173: the following discussion that the inhomogeneity of the external
174: magnetic field has nontrivial effect on GP.
175:
176: Some formulas should be addressed for the calculation of GP in
177: this model. Recently GP for the mixed state has been discussed by
178: Sj\"oqvist \textit{et al.}in \cite{erik}, based on the Mach-Zender
179: interferometer and a formula was provided for calculation of GP for
180: a mixed state,
181: \begin{equation}\label{}
182: \gamma_{g}=\textrm{Arg}\textrm{Tr}[U^{\|}(t)\rho(0)],
183: \end{equation}
184: in which $U^{\|}(t)$ was defined as the parallel transportation
185: operator. Based on this work, Singh, \textit{et al.,} \cite{singh}
186: studied GP for non-degenerate and degenerate mixed states and
187: provided a general method for the construction of $U^{\|}(t)$ by
188: imposing
189: \begin{equation}\label{}
190: U^{\|}(t)=U(t)V(t)
191: \end{equation}
192: in which $U(t)$ is the unitary evolution operator and is equal to $
193: e^{- iHt}$ in our model and $V(t)$ is a blocked matrix, of which the
194: elements are determined by
195: \begin{eqnarray}
196: V_{\mu\nu}&=&\bra{\mu}e^{it\sum_{\mu',\nu'} \bra{\mu '}H\ket{\nu
197: '}\ket{\mu
198: '}\bra{\nu '}}\ket{\nu},\nonumber\\
199: &&\ket{\mu}, \ket{\nu} , \ket{\mu '}, \ket{\nu '}\in\{degenerate\ \ subspace\} \nonumber\\
200: V_{kk} &=& e^{i \bra{k}H\ket{k}t}, \ket{k}\in\{the\ remaining \
201: space\},
202: \end{eqnarray}
203: where $\ket{\mu}, \ket{\nu}, \ket{\mu '}, \ket{\nu '}, \ket{k}$ are
204: the eigenstates of $\rho(0)$, and the interference terms between the
205: degenerate space and the other space are set to be zero in order to
206: keep the parallel transporting in the degenerate space. One notes
207: that for different mixed states one has different $U^{\|}(t)$. Based
208: on these formulas, one can calculate GP. In the following
209: calculations, we label the initial states as
210: \begin{eqnarray}
211: \rho_1&=&\frac{1-r}{4}I_{4}+r\ket{\varphi}\bra{\varphi} \nonumber \\
212: \rho_2&=&\frac{1-r}{4}I_{4}+r\ket{\psi}\bra{\psi}.
213: \end{eqnarray}
214: and the geometric phases for these states are respectively
215: calculated in the following parts.
216:
217:
218: \section{ $g=0$ case}
219: We first study the geometric phase without interaction. It should
220: point out that this situation still has interest in quantum
221: information, such as the quantum teleportation and communication in
222: which the nonlocal correlation, i. e. the entanglement between two
223: space-liked particles plays a crucial role. The geometric phases for
224: $\rho_{i}$(i=1,2)in this case can be easily obtained from the
225: formulas in the former section,
226: \begin{widetext}
227: \begin{eqnarray}\label{g1}
228: \gamma_{g1}&=&\arctan\frac{-r(\sin[\cos2\theta\frac{\omega_1+\omega_2}{2}
229: t]\cos\frac{\omega_1+\omega_2}{2}t - \cos2\theta
230: \cos[\cos2\theta\frac{\omega_1+\omega_2}{2}
231: t]\sin\frac{\omega_1+\omega_2}{2}t)}
232: {\frac{1-r}{2}+\frac{1+r}{2}(\cos[\cos2\theta\frac{\omega_1+\omega_2}{2}
233: t]\cos\frac{\omega_1+\omega_2}{2}t + \cos2\theta
234: \sin[\cos2\theta\frac{\omega_1+\omega_2}{2}
235: t]\sin\frac{\omega_1+\omega_2}{2}t)}\nonumber\\
236: \gamma_{g2}&=&\arctan\frac{-r(\sin[\cos2\theta\frac{\omega_1-\omega_2}{2}
237: t]\cos\frac{\omega_1-\omega_2}{2}t - \cos2\theta
238: \cos[\cos2\theta\frac{\omega_1-\omega_2}{2}
239: t]\sin\frac{\omega_1-\omega_2}{2}t)}
240: {\frac{1-r}{2}+\frac{1+r}{2}(\cos[\cos2\theta\frac{\omega_1-\omega_2}{2}
241: t]\cos\frac{\omega_1-\omega_2}{2}t + \cos2\theta
242: \sin[\cos2\theta\frac{\omega_1-\omega_2}{2}
243: t]\sin\frac{\omega_1-\omega_2}{2}t)},
244: \end{eqnarray}
245: \end{widetext}
246:
247:
248: \begin{figure}[tbp]
249: \begin{center}
250: \begin{overpic}[width=4cm]{f1a}
251: \put(40,10){$\theta$} \put(20,80){$r$} \put(-8,60){$\gamma_{g1}$}
252: \put(0,75){(a)}
253: \end{overpic}
254: \begin{overpic}[width=4cm]{f1b}
255: \put(40,10){$\theta$} \put(20,70){$r$} \put(0,60){$\gamma_{g2}$}
256: \put(0,75){(b)}
257: \end{overpic}
258: \caption{ \label{1} The geometric phases $\gamma_{gj}(j=1,2)$[Arc]
259: versus $r, \theta$. For $\gamma_{g1}$ (a), we have chosen
260: $\omega_{1}= \omega_{2}$ and $\omega_{1}t=\pi/2$; whereas for
261: $\gamma_{g2}$ (b), $\omega_{1}= 2 \omega_{2}$ and
262: $\omega_{1}t=\pi$.}
263: \end{center}
264: \end{figure}
265:
266: \begin{figure}[t]
267: \begin{center}
268: \begin{overpic}[width=4cm]{f2a}
269: \put(-5,32){\begin{rotate}{90} $\gamma_{g1}$
270: \end{rotate}}
271: \put(55,-5){$\theta$} \put(15, 55){(a)}
272: \end{overpic}
273: \begin{overpic}[width=4cm]{f2b}
274: \put(3,32){\begin{rotate}{90}$\gamma_{g2}$\end{rotate}}
275: \put(55,-5){$\theta$} \put(15, 55){(b)}
276: \end{overpic}
277: \\[0.5cm]
278: \begin{overpic}[width=4cm]{f2c}
279: \put(-5,32){\begin{rotate}{90} $\gamma_{g1}$
280: \end{rotate}}
281: \put(55,-5){$\theta$} \put(18, 53){(c)}
282: \end{overpic}
283: \begin{overpic}[width=4cm]{f2d}
284: \put(3,32){\begin{rotate}{90} $\gamma_{g2}$
285: \end{rotate}}
286: \put(55,-5){$\theta$} \put(18, 53){(d)}
287: \end{overpic}
288: \caption{\label{2} The geometric phase [Arc] vs the inhomogeneity of
289: magnetic fiedl $n=\omega_2/\omega_1$ with $r=1$. Since the figures
290: is symmetrical with the $\theta=\pi/2$, we only draw for
291: $\theta\in[0, \pi/2]$. (a) corresponds to $\rho_{1}$ and the other
292: parameters are the same as that of Fig. \ref{1}(a). The dotted,
293: dashed, longer-dashed and solid lines correspond respectively to
294: $n=0.99, 0.9, 0.5, 0$. When $n=1$, the geometric phase is not
295: discontinued at the point $\theta=\pi/4$. (b) is for $\rho_{2}$ and
296: the other parameters are the same as that of Fig. \ref{1}(b). The
297: dotted, dashed and solid lines correspond to $n=0.5, 0.1, 0.01$.
298: When $n=0$, the geometric phase is not discontinued in the point
299: $\theta=\pi/4$. (c) and (d) demonstrate the case that there is no
300: the singularity ( We have chosen $\omega_1t=\pi/4$ for $\rho_1$ and
301: $\omega_1t=\pi/2$ for $\rho_2$ respectively).}
302: \end{center}
303: \end{figure}
304:
305:
306: It is interesting to note that when the initial states are Werner
307: states, the geometric phase is zero when the denominators in
308: Eqs.\eqref{g1} are not vanishing simultaneously. Furthermore when
309: $\omega_1=\omega_2$, the geometric phase for $\rho_2$ is zero since
310: $\rho_2$ is commutative with the Hamiltonian and it cannot pick up
311: any geometric phase. A detailed demonstration for
312: $\gamma_{gj}(j=1,2)$ with the parameters $r, \theta$ is shown in
313: Fig. \ref{1}. From the figures it is obvious that the absolute
314: values of $\gamma_{g1(2)}$ attain the maximum when $\theta\neq\pi/4,
315: 3\pi/4$ and because of the mixing, scaled by $r$, the absolute value
316: of geometric phase is compressed and tends to be zero with
317: $r\rightarrow 0$. We also note that a singularity about
318: $\gamma_{g1}$ appears when $\theta=(3)\pi/4$ and $r=1$, as displayed
319: in Fig. \ref{1}(a). At this point, the numerator and the dominator
320: in the expression of $\gamma_{g1}$ in Eq. \eqref{g1} are
321: simultaneously zero and the geometric phase is undefined. One has
322: to calculate the so-called off-diagonal geometric phase to retain
323: the information of the evolution \cite{manini}. Furthermore, our
324: calculation shows that the singularity depends not only on the
325: entanglement, but on how the system evolves.
326:
327:
328: With the consideration of the two noninteracting particles, it is of
329: interest to discuss the effect of the inhomogeneity of the external
330: fields. The results have been illustrated in Fig. \ref{2}. It
331: clearly shows that when there is a singular point, for $\rho_{1}$,
332: the absolute value of the geometric phase with $n\rightarrow1$
333: attains the maximum closed to this point. Whereas, for $\rho_{2}$,
334: this happens for $n\rightarrow0$. We should emphasize that this
335: phenomenon appears only when there exists the singularity, which is
336: induced by the entanglement of the initial state. If there is no
337: possibility of the appearance of the singularity, our calculation
338: shows that the points where the absolute value of the geometric
339: phase attains the maximum is independent of the inhomogeneity of the
340: external field,just shown as Fig.\ref{2}(c)and (d) and always is
341: zero for the Werner state. Physically it is thus possible for this
342: phenomenon to act as the signature of the singularity of the
343: geometric phase, which is usually originated from the degenerace
344: \cite{manini}.
345:
346: Another interesting consideration is that one supposes
347: $\omega_{2}=0$, which corresponds to case that particle A is
348: processing with frequency $\omega_{1}$ while particle B keep
349: dynamically fixed. In this case, the affect from the dynamics of the
350: second particle is excluded and one can more clearly check the
351: effect of the entanglement on the geometric phase. The geometric
352: phase for this special situation can be obtained by setting
353: $\omega_2=0$ in Eq. \eqref{g1}. One can find that because of the
354: entanglement, the geometric phase for particle 1 is zero too for
355: Werner state. and it is also possible for the singularity to appear,
356: for example when $r=1$ and $\omega_1t=\pi$.
357:
358:
359: \section{$g\neq0$ case}
360: \begin{figure}[tbp]
361: \begin{overpic}[width=4cm]{f3a}
362: \put(80,10){$J$} \put(10,15){$r$} %\put(100,60){$\gamma_{g1}$}
363: \put(0,75){(a)}
364: \end{overpic}
365: \begin{overpic}[width=4cm]{f3b}
366: \put(40,5){$J$} \put(20,70){$r$} \put(-5,60){$\gamma_{g2}$}
367: \put(0,75){(b)}
368: \end{overpic}
369: \caption{\label{3} GP for $\gamma_{g2}$[Arc] versus $r$ and rescaled
370: coupling constant $J$. We have chosen $\theta=\pi/4$ (a) and
371: $\theta=3\pi/4$ (b) for the highlight of the couplings effect on the
372: geometric phase, which corresponds to the Werner state as the
373: initial state, and set $\omega_{1}t =2\omega_{2}t =\pi$.}
374: \end{figure}
375:
376: Because of the intra-subsystem coupling, the evolution of the system
377: can be very different from the free case. So in this section we
378: focus on the effect of coupling on the geometric phase. Obviously
379: one note that since $[H_I, \rho_1]=0$, $H_I$ has trivial effect on
380: $\gamma_{g1}$ and thus this state have been excluded in this
381: section. Based on the formulas in Sec. II, GP are for $\rho_2$,
382: \begin{eqnarray}
383: \gamma_{g2}= \arctan\frac{-
384: r(\sin\lambda_1t\cos\lambda_2t-\frac{\lambda_2}{\lambda_1}\cos\lambda_1t\sin\lambda_1t)}
385: {\frac{1-r}{2} +
386: \frac{1+r}{2}(\cos\lambda_1t\cos\lambda_2t+\frac{\lambda_2}{\lambda_1}\sin\lambda_1t\sin\lambda_1t)}
387: \end{eqnarray}
388: in which
389: \begin{eqnarray}
390: \lambda_1=&[(\omega_1-\omega_2)\cos2\theta-2g\sin2\theta]/2 \nonumber\\
391: \lambda_2=&\sqrt{(\omega_1-\omega_2)^2+4g^2}/2.\nonumber
392: \end{eqnarray}
393:
394:
395:
396: The results have been illustrated in Fig. \ref{3} with rescaled
397: coupling constant $J=\frac{g}{\omega_1}$. Compared with the $g=0$
398: case, we have chosen the Werner states ($\theta=(3)\pi/4$) as the
399: initial states and $\omega_1 t=2\omega_2 t=\pi$ with consideration
400: of the inhomogeneity of external magnetic field. Because of the
401: coupling, the geometric phase for the Werner state is obviously very
402: different from the free case. From these figures we also note that
403: GP has a maximal or mininmal values when the rescaled coupling
404: constant is less than 1. Furthermore, when $J>1$ the absolute value
405: of GP decrease with the increment of $J$. One also notes that
406: because of the mixing, which is scaled by the parameter $r$, the
407: absolute values of GP is compressed, which is same as the free case.
408: The effect of the inhomogeneity of magnetic field also discussed.
409: Our calculation shows that with $\omega_2/\omega_1 \rightarrow 1$,
410: the point that GP has maximal or minimal value is infinitely closed
411: to $J=0.5$ (Fig. \ref{4}).
412:
413:
414: \begin{figure}
415: \begin{center}
416: \begin{overpic}[width=8cm]{f5.eps}
417: \put(-15,80){$\gamma_{g2}$} \put(145,-5){$J$}
418: \put(40,18){n$\rightarrow 0$} \put(40, 50){n=0.1}
419: \put(55,110){n=0.5} \put(69,125){n=0.8}
420: \put(100,20){\includegraphics[width=4cm]{f5in.eps}}
421: \end{overpic}
422: \caption{\label{4} GP[Arc] vs. the inhomogeneity of magnetic field
423: $n=\omega_2/\omega_1$ . We have choose the Werner state
424: ($\theta=\pi/4$)as the initial state with $r=1, \omega_1t=\pi$.The
425: inset is for $n=0.8$. Similar behavior can be found when
426: $\theta=3\pi/4$.}
427: \end{center}
428: \end{figure}
429:
430:
431: \begin{figure}[tbp]
432: \begin{overpic}[width=4cm]{f4a}
433: \put(40,5){$J$} \put(30,75){$r$} \put(0,60){$\gamma_{g2}$}
434: \put(0,75){(a)}
435: \end{overpic}
436: \begin{overpic}[width=4cm]{f4b}
437: \put(70,5){$J$} \put(10,15){$r$} \put(100,60){$\gamma_{g2}$}
438: \put(0,75){(b)}
439: \end{overpic}
440: \caption{ \label{5} The geometric phase [Arc]for $\rho_2$ when
441: $\omega_2=0$ versus $r, J$. Except of $\omega_2=0$, all other
442: settings are the same as in Fig. \ref{3}. }
443: \end{figure}
444:
445: We also investigate the GP when one particle in zero field. This
446: case is nontrivial since any system cannot avoid the affect from the
447: other party. Also it can be used to manipulate the behavior of one
448: particle by changing of the coupling strength. To this end, we set
449: $\omega_2 =0$ and the geometric phase have been illustrated in Fig.
450: \ref{5}. Different from Fig. \ref{3}, the absolute value of GP
451: attains the maximal value when $J\rightarrow1$.
452:
453:
454: The interesting extension to this discussion above is in the limit
455: of large $J=\frac{g}{\omega_1}$. In this case GP has a novel
456: character, which have been displayed in Fig. \ref{6}. We note that
457: GP tends to be zero in the limit of large $J$. This can be
458: understood as that in the limit of large $J$ the interaction $H_I$
459: is dominant in Eq. (\ref{h})and the Hamiltonian is inclined to be
460: commutative with the Werner state. Then the entanglement of the
461: initial state tends to be unchanged.
462:
463: Based on the analysis above, we can conclude that because of the
464: inter-subsystem couplings, the geometric phase shows two different
465: characters; in the weak coupling limit, the interaction obviously
466: benefits the geometric phase, displayed in Fig. \ref{3} and Fig.
467: \ref{5}. However, with further increment of the couplings, the GP
468: decreases and tends to be zero infinitely. This phenomena can be
469: explained easily only if one note that in the infinite coupling
470: limit, $H_I$ is dominant and its eigenstates are just two of the
471: Bell states. The entanglement is destroyed in the weak couplings
472: limit and then revived by the interaction in the infinite limit. The
473: interesting relation between entanglement and interaction between
474: two parties is an important aspect of quantum information.
475: \begin{figure}[tbp]
476: \begin{overpic}[width=4cm]{f6a}
477: \put(15, 55){(a)}\put(-5,35){\begin{rotate}{90}$\gamma_{g2}$
478: \end{rotate}} \put(60, -5){$J$}
479: \end{overpic}
480: \hspace{1em}
481: \begin{overpic}[width=4cm]{f6b}
482: \put(15,55){(b)}\put(-5,35){\begin{rotate}{90}$\gamma_{g2}$\end{rotate}}
483: \put(60, -5){$J$}
484: \end{overpic}
485: \caption{\label{6} $\gamma_{g2}$[Arc] in the limit of large $J$. The
486: parameters have same values to that of Fig. \ref{3} and $r=1$ is
487: chosen.}
488: \end{figure}
489:
490: \section{conclusions}
491: In conclusion, we have discussed the geometric phase for entangled
492: mixed state Eq. (\ref{rho}) in an external magnetic field. For the
493: free case ($g=0$), our studies show that because of the
494: entanglement, the geometric phase for the system displays two
495: different types of character. The first is that if there was no
496: singularity, the absolute value of the geometric phase attains the
497: maximum when the initial state is not the Werner state, independent
498: of the inhomogeneity of the external field (see Fig. \ref{2}(c) and
499: (d)). Because of the mixing, scaled by $r$, it is compressed.
500: Furthermore the geometric phase is always zero for the Werner state.
501: The second behavior occurs when the singularity appears, induced by
502: the entanglement of the initial states under the proper evolution.
503: In this case the geometric phase reaches the maximum with $\theta$
504: tending to the singular point (see Fig. \ref{1}(a) and Fig.
505: \ref{2}(a) and (b)). We also discuss the situation that one particle
506: is processing and another keeps dynamically fixed, and similar
507: results can be found.
508:
509: For the case $g\neq0$, we choose the Werner states as the initial
510: condition for facilitating our discussion. The results show that GP
511: is completely determined by the rescaled coupling constant $J$ and
512: attain the maximal or minimal value when $J<1$. Similar to the free
513: case, we also discussed the effect of the inhomogeneity of the
514: external magnetic field. The results show that for Werner state GP
515: is always zero when system is in a homogeneous external magnetic
516: field , independent of the interaction. Furthermore with
517: $\omega_2/\omega_1\rightarrow1$, the absolute value of GP attains a
518: maximum at the point $J\rightarrow0.5$. The geometric phase for
519: $\omega_2=0$ is also studied. Further study in the limit of large
520: $J$ displays a novel phenomena that GP tends to be zero and this
521: result have been explained as the revival of entanglement in this
522: limit.
523:
524: This work was supported by NSF of China under grant 10305002 and
525: 60578014.
526:
527: \begin{thebibliography}{99}
528: \bibitem{p} S. Pancharatnam, Proc. Indian Acad. Sci. A {\bf 44}, 247 (1956).
529:
530: \bibitem{berry} M. V. Berry, Proc. R. Soc. London A {\bf 392}, 45(1984).
531:
532: \bibitem{shapere} A. Shapere and F. Wilczek, \textit{Geometric Phase in Physics} (World Scientific, 1989);
533: A. Bohm, A. Mostafazadeh, H. Koizumi, Q. Niu, J. Zwanziger, \emph{The Geometric Phase in Quantum
534: System}, Springer(2003).
535:
536: \bibitem{aharonov} Y. Aharonov, J. Anandan, Phys. Rev. Lett. {\bf 58}, 1593(1988);
537: J. Anandan, Y. Aharonov, Phys. Rev. D {\bf 38}, 1863(1988).
538:
539: \bibitem{samuel} J. Samuel, R. Bhandari, Phys. Rev. Lett. {\bf 60}, 2339(1988);
540: A. K. Pati, Phys. Rev. A {\bf 52}, 2576(1995).
541:
542: \bibitem{mukunda1} N. Mukunda, R. Simon, Ann. Phy. (N. Y.) {\bf 228}, 205(1995).
543:
544: \bibitem{uhlmann} A. Uhlmann, Rep. Math. Phys. {\bf 24}, 229(1986); Lett. Math. Phys. {\bf 21}, 229(1991).
545:
546: \bibitem{erik} E. Sj\"{o}qvist, \textit{et. al.}, Phys. Rev. Lett.{\bf 85}, 2845(2000).
547:
548: \bibitem{du}Jiangfeng Du, Ping Zou, \textit{et al.}, Phys. Rev. Lett. {\bf 91}, 100403
549: (2003),J. Klepp, \textit{et. al.}, Phys. Lett. A {\bf 342}, 48
550: (2005); A. Ghosh and A. Kumar, Phys. Lett. A {\bf 349}, 27 (2005).
551:
552: \bibitem{ericsson} M. Ericsson, \textit{et al.}, Phys. Rev. Lett. {\bf 94}, 050401 (2005).
553:
554: \bibitem{singh} K. Singh, \textit{et al.}, Phys. Rev. A {\bf 67}, 032016 (2003); D. M. Tong, \textit{et al}.
555: Phys. Rev. Lett. {\bf93}, 080405 (2004).
556:
557: \bibitem{bhandari} R. Bhandari, Phys. Rev. Lett. {\bf 89}, 268901(2002); J. Anandan, \textsl{et al}., Phys.
558: Rev. Lett. {\bf 89}, 268902 (2002).
559:
560: \bibitem{manini} N. Manini, F. Pistolesi, Phys. Rev. Lett. {\bf 85}, 3067(2000).
561:
562: \bibitem{mukunda2} N. Munkunda, \textit{et al.}, Phys. Rev. A., {\bf 67}, 042114(2003).
563:
564: \bibitem{filipp} S. Filipp, Erik Sj\"{o}qvist, Phys. Rev. Lett. {\bf 90}, 050403(2003);
565: Phys. Rev. A {\bf 68}, 042112(2003).
566:
567: \bibitem{tong1} D. M. Tong, \textit{et al.}, Phys. Rev. A {\bf 71}, 032106(2005).
568:
569: \bibitem{yi1} X. X. Yi, J. L. Chang, Phys. Rev. A {\bf 70}, 012108 (2004).
570:
571: \bibitem{cui} H. T. Cui, L. C. Wang, X. X. Yi, Europhys. Lett. {\bf
572: 74}, 757(2006)
573:
574: \bibitem{jones} J. Jones, V. Vedral, A. K. Ekert, C. Castagnoli, Nature(London), {\bf 403}, 869(2000);
575: L. M. Duan, J. I. Cirac, P. Zoller, Science {\bf 292}, 1695(2001).
576:
577: \bibitem{zanardi} P. Zanardi, M. Rasetti, Phys. Lett. A {\bf 264}, 94(1999).
578:
579: \bibitem{zheng} A. Nazir, T. P. Spiller, W. J. Munro, Phys. Rev. A {\bf 65}, 042303(2002);
580: S.-L. Zhu and Z. D. Wang, Phys. Rev. Lett. {\bf89}, 097902(2002);{\bf 91}, 187902 (2003);
581: S. B. Zheng, Phys. Rev. A {\bf70}, 052320(2004); R. G. Unanyan , M. Fleischhauer
582: Phys. Rev. A {\bf 69}, 050302(2004).
583:
584: \bibitem{erik2} E. Sj\"{o}qvist, Phys. Rev. A {\bf 62}, 022109(2001); D. M. Tong, {\it et al.}, Phys. Rev. A {\bf 68}, 022106(2003),
585: J. Phys. A: Math. Gen. {\bf 36}, 1149(2003);R. A. Bertlmann, {\it et al.}, Phys. Rev. A {\bf 69}, 032112(2004).
586: \bibitem{yi2} X. X. Yi, L. C. Wang, and T. Y. Zheng, Phys. Rev. Lett. {\bf92}, 150406(2004);
587: X. X. Yi and E. Sj\"oqvist, Phys. Rev. A {\bf70}, 042104(2004);
588: X. X. Yi, \textit{et al.}, Phys. Rev. A. {\bf71}, 022107(2005).
589:
590:
591: \bibitem{werner} R. F. Werner, Phys. Rev. A {\bf 40}, 4277(1989).
592:
593: \bibitem{bennett1} C. H. Bennett, \textit{et al.}, Phys. Rev. Lett. {\bf 76}, 722(1996).
594:
595: \bibitem{bennett2} C. H. Bennett, P. P. DiVincenzo, {\it et al}., Phys. Rev. A {\bf54}, 3824(1996);
596: D. Deutsch, A. Ekert, {\it et al}., Phys. Rev. Lett. {\bf 77}, 2818(1996);
597: N. Gisin, Phys. Rev. A {\bf 210}, 151(1996).
598:
599: \bibitem{wootters} W. K. Wootters, Phys. Rev. Lett. {\bf 80}, 2245(1998).
600:
601: \bibitem{peres} A. Peres, Phys. Rev. Lett. {\bf 77}, 1413(1996);
602: M. Horodecki, P. Horodecki, and R. Horodecki, Phys. Lett. A {\bf
603: 223}, 1 (1996).
604:
605: \bibitem{batle}J. Batle, M. Casas, A. Plastino, A. R. Plastino, axXiv:
606: quant-ph/0603055.
607: \end{thebibliography}
608: \end{document}
609: