1: \documentclass[aps,twocolumn,superscriptaddress,showpacs,amsmath]{revtex4}
2: %\documentclass[aps,twocolumn,showpacs]{revtex4}
3: %\documentstyle[prl,aps,epsf]{revtex}
4: \usepackage{graphicx}% Include figure files
5: \usepackage{bm}% bold math
6: \usepackage{amsmath}
7: \begin{document}
8:
9: \title{Spin-Orbital Entanglement and Phase Diagram of
10: Spin-orbital Chain with $SU(2) \times SU(2)$ Symmetry}
11:
12: \author{Yan Chen}
13: \affiliation{Department of Physics and Center of Theoretical and
14: Computational Physics, The University of Hong Kong, Pokfulam Road,
15: Hong Kong, China}
16: \author{Z. D. Wang}
17: \affiliation{Department of Physics and Center of Theoretical and
18: Computational Physics, The University of Hong Kong, Pokfulam Road,
19: Hong Kong, China} \affiliation{ National Laboratory of Solid State
20: Microstructures, Nanjing University, Nanjing, China}
21: \author{Y. Q. Li}
22: \affiliation{ Department of Physics, Zhejiang University, Hangzhou,
23: China}
24: \author{F. C. Zhang}
25: \affiliation{Department of Physics and Center of Theoretical and
26: Computational Physics, The University of Hong Kong, Pokfulam Road,
27: Hong Kong, China} \affiliation{ Department of Physics, Zhejiang
28: University, Hangzhou, China}
29:
30:
31: \begin{abstract}
32: Spin-orbital entanglement in quantum spin-orbital systems is
33: quantified by a reduced von Neumann entropy, and is calculated for
34: the ground state of a coupled spin-orbital chain with $SU(2)\times
35: SU(2)$ symmetry. By analyzing the discontinuity and local extreme of
36: the reduced entropy as functions of the model parameters, we deduce
37: a rich phase diagram to describe the quantum phase transitions in
38: the model. Our approach provides an efficient and powerful method to
39: identify phase boundaries in a system with complex correlation
40: between multiply degrees of freedom.
41:
42: \end{abstract}
43:
44: %%%%%%%%%PACS # checked ######
45: \pacs{71.70.Ej, 73.43.Nq, 03.67.Mn}
46:
47: \maketitle Exotic states associated with the orbital degrees of
48: freedom in transition-metal oxides have attracted considerable
49: interest recently. Examples of such systems with spin-orbital
50: couplings include spin-gap materials Na$_{2}$Ti$_{2}$Sb$_{2}$O and
51: NaV$_{2}$O$_{5}$, manganites La$_{1-x}$Sr$_x$MnO$_3$, and
52: V$_{2}$O$_{3}$~\cite{Book1,Tokura98,Tokura00,Bao93,Feiner97}.
53: Intriguing physical properties in these systems include the
54: emergence of orbital ordering, the appearance of complex coupled
55: excitations involving both spin and orbital degrees of freedom.
56: Starting from a multi-band Hubbard model at strong coupling limit
57: and at the integer fillings of electrons per unit cell, the charge
58: degree of freedom is frozen and one may derive an effective
59: spin-orbital model~\cite{Kugel73,Cast78}. One of the simplest such
60: systems is the $SU(2) \times SU(2)$ model with $SU(2)$ symmetries
61: for spin-1/2 operator $\mathbf{S}_i$ as well as for pseudospin-1/2
62: operator $\mathbf{T}_i$ representing two degenerate orbitals. There
63: have been a lot of activities recently on the one-dimensional
64: spin-orbital coupled systems ~\cite{gri99,Mila99}, especially on its
65: phase diagram ~\cite{Pati98,Azaria99,yam,Itoi00,Zheng01}. Rich
66: quantum phases include both conventional
67: ferromagnetic/antiferromagnetic gapless phases and symmetry broken
68: gapped states. In the strong coupling regime where the interplay
69: between spin and orbital quantum fluctuations is crucial, the
70: detailed phase diagram still remains controversial and a more
71: comprehensive understanding is awaited.
72:
73: More recently, the investigation of quantum
74: entanglement
75: %~\cite{Nielsen1}
76: from the perspective of quantum information theory has gained much
77: insight for a deeper understanding of quantum many particle physics,
78: especially quantum phase transitions. Many theoretical studies have
79: been devoted to the entanglement in one dimensional spin-1/2
80: systems~\cite{Wootters98,Wang02,AOsterloh2002,GVidal03,Cirac04,chen04}
81: and in interacting fermion and boson systems~\cite{Gu04,chen05}.
82: Quantum entanglement has been quantified in terms of the spin-spin
83: concurrence~\cite{Wootters98}, contiguous block
84: entanglement~\cite{GVidal03}, and sublattice
85: entanglement~\cite{chen04,chen05}. There are evidences to suggest a
86: close connection between quantum phase transition and local extreme
87: or singularity of the quantum entanglement when it is measured
88: appropriately~\cite{chen04}. In the coupled spin-orbital systems,
89: one expects the spin-orbital entanglement (SOE) to be important. In
90: particular, near a quantum transition point, one may naturally
91: expect that the entanglement may manifest itself accordingly. Recent
92: theoretical study has also demonstrated that the SOE could lead to
93: the violation of the Goodenough-Kanamori rules~\cite{Khaliullin06}.
94:
95: In this paper, we propose a reduced von Neumann entropy to quantify
96: the SOE, which measures the interplay between spin and orbital
97: degrees of freedom of the quantum states. We use small size exact
98: numerical method to calculate the reduced entropy of the ground
99: state of the spin-orbital chain given by Hamiltonian Eq. (1) and
100: study its relation with the quantum phase transition of the system.
101: Our results show that this novel measure of the entanglement can
102: reveal faithfully the quantum transition points and phase boundaries
103: of the complex phase diagram of the system. Our results indicate
104: that the strategy of evaluating entanglement measure is powerful and
105: efficient for extracting valuable information of the quantum
106: systems.
107:
108: We consider a one-dimensional spin-orbital Hamiltonian with
109: $SU(2)\times SU(2)$ symmetry,
110: \begin{equation}
111: H=\sum\limits_{i}\left( \mathbf{S}_{i}\cdot\mathbf{S}_{i+1}+x\right)
112: \left( \mathbf{T}_{i}\cdot\mathbf{T}_{i+1}+y\right) ,
113: \end{equation}
114: where $\mathbf{S}_i$ are spin-1/2 operators while $\mathbf{T}_i$
115: denote the orbital pseudo-spin 1/2 operators. $x$ and $y$ are two
116: tuning parameters. At $x=y$, the model has an interchange symmetry
117: between spin and orbital. The model at $x=y=1/4$ is a special case
118: possessing a higher $SU(4)$ symmetry, and there are three gapless
119: modes (spin, orbital and spin-orbital) in the low-lying
120: excitations~\cite{Sutherland75,YQLi98}. It is also known that the
121: model at $x=y=3/4$ has an exact ground state, in which the spin and
122: orbital form dimerized singlets in a staggered pattern, and the
123: doubly degenerate ground states can be expressed as gapped matrix
124: product state~\cite{Kolezhuk98}.
125:
126: In the spin-orbital model, the importance of the SOE has long been
127: recognized~\cite{YQLi98,Khaliullin06}. However, a quantitative
128: measure for the entanglement is still lacking. We propose to measure
129: the SOE by a reduced von Neumann entropy defined as
130: \begin{eqnarray}
131: S^{so} := - \rm{tr}_s \left( \rho_{s} \log_2 \rho_{s} \right),
132: \label{ent-ent}
133: \end{eqnarray}
134: where $\rho_s\equiv \rm{tr}_{o} |{\Psi}\rangle\langle{\Psi}|$ is the
135: reduced density matrix of the spin part in the state
136: $|{\Psi}\rangle$ by integrating out all the orbital degree of freedom.
137: %Alternatively, we could define the density
138: %matrix of the orbital part by integrating out all the spin degrees
139: %of freedom.
140: Obviously, Eq. (2) gives $S^{so}=0$ if spin $\mathbf{S}$ and orbital
141: $\mathbf{T}$ are decoupled. The motivation for such a measure is to
142: better reveal the correlation between two distinctive degrees of
143: freedom. This measure is similar to the recent proposal of the
144: reduced entropy $S_L$ of a block of subsystem in study of the
145: relations between entanglement and quantum phase transition, where
146: $S_L$ is defined by
147: \begin{eqnarray}
148: S_L := - \rm{tr} \left( \rho_{L} \log_2 \rho_{L} \right),
149: \label{ent-ent2}
150: \end{eqnarray}
151: where $\rho_L\equiv \rm{tr}_{L} |{\Psi}\rangle\langle{\Psi}|$ is the
152: reduced density matrix for a block of subsystem $B_L$. The analogy
153: of the SOE with the block subsystem becomes more clear if we map the
154: model of Eq. (1) onto a two-leg "spin" ladder system with one chain
155: described by spin $\mathbf{S}$ and the other chain by orbital
156: $\mathbf{T}$ and the two sites on each leg are coupled by a
157: four-operator interactions~\cite{Nersesyan97}.
158:
159:
160: Let us first examine the SOE defined in Eq. (2) in a few simplest
161: cases. For a single site system, the SOE has a one-to-one
162: correspondence to the two pure spin-1/2 system with one spin for
163: $\mathbf{S}$ and the other for $\mathbf{T}$. For a two-site system,
164: the spin (orbital) states can be either a singlet $|{\Psi}_S^s
165: \rangle$ ( $|{\Psi}_O^s \rangle$) or triplet $|{\Psi}_S^{t} \rangle$
166: ( $|{\Psi}_O^{t} \rangle$). It is easy to check that $S^{so}=0$ for
167: all the spin-orbital decoupled states, and $S^{so}=1$ for the state
168: $1/\sqrt 2 (|\Psi_S^{s} \rangle|\Psi_O^{t} \rangle \pm |\Psi_S^{t}
169: \rangle |\Psi_O^{s} \rangle)$. Next we proceed to the 4-site (1234)
170: cluster, which is the smallest system size to have $SU(4)$ singlet
171: state $|SGL \rangle$. This $|SGL \rangle$ state contains 24 terms,
172: and is rotational invariant under the fifteen $SU(4)$
173: generators~\cite{YQLi98,Bossche00,Footnote1}. After tracing over the
174: orbital degrees of freedom, we find $S^{so}=1$ for this high
175: symmetry state. For the dimerized state at $(x=y=3/4)$, it is known
176: that its ground state is a matrix product state in both spin and
177: orbital part~\cite{Kolezhuk98}. After some algebra, we find its
178: value of entanglement is about 0.40.
179:
180:
181: \begin{figure}[t]
182: \includegraphics[width=8.0cm,height=7.2cm]{Fig1.eps}
183: \caption{\label{Fig1} The rescaled SOE $S^{so}/L$ $(L=8)$ as a
184: function of the $x$ and $y$. The phase boundaries (red lines) are
185: drawn to guide the eyes.}
186: \end{figure}
187:
188: In what follows we will calculate the ground state SOE of the
189: Hamiltonian (1) in a finite size system. We will demonstrate the
190: close connection between SOE and the quantum phase transitions in
191: the model. Since the Hamiltonian has a rotational symmetry around
192: the $z$-axes in $\mathbf{S}$-space as well as in $\mathbf{T}$-space,
193: the exact diagonalization calculations are carried out in an
194: invariant subspace with $S_z=0$ and $T_z=0$ to get the ground state
195: $|\Psi_G\rangle$, from which we construct the density matrix for the
196: whole system. The reduced density matrix $\rho_L$ of spin part is
197: obtained by tracing out the orbital degree of freedom, and compute
198: its reduced entropy. In our calculation, the chain length ranges
199: from 8 to 12.
200:
201:
202: The spatial profiles of SOE $S^{so}/L$ as a function of $x$ and $y$
203: are displayed in Fig. 1. A salient feature shows the existence of
204: zero entanglement regime. When either $\mathbf{S}_i$ or
205: $\mathbf{T}_i$ are aligned ferromagnetically, this model has no
206: frustration and qualitative results can be obtained from physical
207: considerations alone. Intuitively, for $x
208: > -1/4$ ($x < -1/4$) and $y < -1/4$ ($y> -1/4$), the ground states
209: are antiferromagnetic for spin $\mathbf{S}$ and ferromagnetic for
210: orbital $\mathbf{T}$ (and vice versa). In the case of large
211: negative $x$ and $y$, the ground state is the ferromagnetic state
212: with respect to both $\mathbf{S}$ and $\mathbf{T}$. These three
213: conventional states correspond to the decoupling between the spin
214: and orbital degrees of freedom. Therefore, these states have zero
215: SOE. On the other hand, if both $\mathbf{S}$ and $\mathbf{T}$ are
216: coupled antiferromagnetically, the four-operator term may frustrate
217: the system, which may lead to the emergence of various non-trivial
218: ground states with finite values of SOE. As depicted in Fig. 1,
219: there are boundary lines with finite discontinuous jump of SOE
220: between the zero-value and finite-value regions which indicates a
221: first-order phase transition. Ground states of the boundary lines of
222: the fully ferromagnetic spin and orbital phases are highly
223: degenerate. It is obvious that the quantum fluctuation effect pushes
224: the classical phase boundary closer to the symmetric line $x=y$.
225:
226:
227: Two special points manifest themselves clearly: the $SU(4)$
228: symmetric point corresponds a local maximum of $S^{so}/L$ while the
229: dimerized state point corresponds a local minimum. This feature may
230: be simply understood as follows. At the $SU(4)$ point, there is the
231: largest correlation between spin and orbital degrees of freedom,
232: while at the dimer phase point, both the spin $\mathbf{S}$ and
233: orbital $\mathbf{T}$ are weakly coupled so that the entanglement is
234: much suppressed. In addition, the symmetric line $x=y$ is more
235: special and interesting. In Fig. 1, along the line (referred to as
236: the line $A$) connecting the point $(-1/4,-1/4)$ and the $SU(4)$
237: point where both of them have high symmetries, $S^{so}/L$ reaches
238: the local maxima ( ridge-like), while along the line (referred to as
239: the line $B$) connecting the $SU(4)$ point and (0.66,0.66), it
240: behaves as the local minima (valley-like). According to our previous
241: analysis and wisdom~\cite{chen04,chen05}, we know that both the
242: ridges and valleys may correspond to phase boundaries. We conclude
243: here that both lines $A$ and $B$ may serve as phase boundaries. It
244: is worth noting that the $SU(4)$ symmetric point is a multi-critical
245: point. There are four distinct neighboring quantum phases around
246: this point. For example, moving off this symmetric point, one may
247: enter the gapped states to the right upwards or enter the gapless
248: phases to the left downwards. These results are likely supported by
249: other studies. (i) The critical line A is consistent with that of
250: the analysis by Yamashita {\em et al.}~\cite{yam}. (ii) In a recent
251: Schwinger boson mean field study~\cite{Pengli05}, both spin and
252: orbital valence bond states are found in the parameter region
253: separated by the critical line B. Around the dimerized state point,
254: one may observe that there exists a curved phase boundary line
255: dividing the regions where the discontinuity of first derivative of
256: entanglement as a function of parameters occurs. The high
257: temperature series expansion approach suggests the existence of two
258: distinct gapped phases in the parameter region for both positive $x$
259: and $y$~\cite{Zheng01}. Our calculation supports the existence of
260: such gapped phases.
261:
262: For the large $x$ and $y$ region, mean field studies always suggest
263: that the ground state is the antiferromagnetic state with respect to
264: $\mathbf{S}$ and $\mathbf{T}$. In that case, we would expect that
265: its corresponding SOE is equal to zero. However, the SOE in this
266: parameter region shows plateau-like behavior with finite value,
267: which contradicts the conclusion of mean field studies. Since the
268: well-known dimerized state point is located within the large $x$ and
269: $y$ region, we conclude that this phase regime belongs to the gapped
270: dimerized state rather than the gapless antiferromagnetic phase. It
271: is worth noting that the strength of SOE may be regarded as an
272: indicator to discern how good the mean field approximation will be.
273: In the strongly coupled regime, the interplay between spin and
274: orbital quantum fluctuations may be important and leads to some
275: highly nontrivial quantum phases. Therefore it is necessary to
276: consider the effects of quantum fluctuations more seriously beyond
277: the mean field theory. Certainly, we also find that both
278: entanglement measures vanish in the infinitely large limits of $x$
279: and $y$.
280:
281: \begin{figure}[t]
282: \includegraphics[width=8.0cm,height=7.2cm]{Fig2.eps}
283: \caption{\label{Fig2} The rescaled sublattice entanglement
284: $S_{L/2}/L$ $(L=8)$ as a function of the $x$ and $y$. The phase
285: boundaries (red lines) are essentially the same as that of Fig. 1.}
286: \end{figure}
287: To study the additional entanglement between the intercalated
288: sublattices of composite degrees of freedom and to best reveal all
289: possible quantum phase boundaries, we also look into the standard
290: sublattice entanglement~\cite{chen04,chen05}, which is obtained by
291: tracing out both spin and orbital degrees of freedom at even (or
292: odd) sites in the present chain. In Fig. 2, we plot the sublattice
293: entanglement versus the coupling parameters. It is interesting to
294: note that there is roughly one-to-one correspondence of local
295: extreme and discontinuity between these two measures of
296: entanglement. In contrast to that of SOE, the $SU(4)$ point reaches
297: a local minimum of sublattice entanglement while the dimerized state
298: point corresponds to a local maximum. Since the SOE mainly captures
299: the correlation between the spin and orbital degrees of freedom
300: while the sublattice entanglement focuses on the correlation between
301: the intercalated sublattices of composite degrees of freedom, these
302: two measures may provide certain complementary information. In the
303: case of conventional ferromagnetic/antiferromagnetic phases, the SOE
304: vanishes while the sublattice entanglement remains nonzero. Thus the
305: measure of SOE is unlikely to distinguish these conventional phases.
306: Instead, in Fig. 2, these phases are clearly separated. In addition,
307: the enhancement of sublattice entanglement for the gapped dimerized
308: state is clearly observed.
309:
310:
311:
312: \begin{figure}[t]
313: \hspace{5cm}
314: \includegraphics[width=7.6cm,height=6.6cm]{Fig3.eps}
315: \caption{\label{Fig3} Ground state phase diagram of a coupled
316: spin-orbital chain. The dotted point is at (1/4,1/4) while the
317: diamond point is located at (3/4,3/4). Eight distinct phases are
318: identified according to the analysis of entanglement as a function
319: of parameters $x$ and $y$. See text for details.}
320: \end{figure}
321:
322:
323: Quantum phase diagram can be distilled from the analysis of the
324: spatial profiles of entanglement as a function of parameters. In
325: other words, both the ridges and valleys in the three-dimensional
326: plot may correspond to possible phase boundaries. Derived from the
327: numerical results presented in Figs. 1 and 2, we plot the phase
328: boundaries of a coupled spin-orbital chain for $L=8$ in Fig. 3. The
329: results for $L=12$ are essentially the same. There are totally eight
330: distinct quantum phases. Most phase boundaries are in good agreement
331: with previous studies~\cite{Pati98,yam,Itoi00,Zheng01}. Phases I, II
332: and III, are conventional spin and orbital ferromagnetic or
333: antiferromagnetic states. Phase IV and V belong to gapless
334: states.~\cite{yam}. Phase VI and VII may correspond to orbital and
335: spin valence bond phases~\cite{Pengli05}, respectively. In view of
336: the fact that the exact ground state at the point $(3/4,3/4)$
337: belongs to staggered dimerized singlet and this dimerized state
338: point is located within phase VIII, we conclude that the phase VIII
339: is a staggered dimerized singlet state. It is remarkable that the
340: most comprehensive phase diagram is now efficiently and
341: straightforwardly obtained.
342:
343:
344:
345: Since the coupled spin-orbital chain can be regarded equivalently to
346: a two-leg spin ladder with four-spin interactions, we may also
347: employ the measure of concurrence to quantify the bipartite
348: entanglement in terms of spin-spin, orbital-orbital as well as
349: spin-orbital concurrence. Our results show that the concurrence can
350: merely show a few features of phase diagram such as the conventional
351: phases I, II and III, but unfortunately, it fails to identify the
352: detailed phase diagram in the strong coupling regime. Another
353: scenario is to analyze so-called single-site entanglement. In this
354: case, we obtain the reduced density matrix by tracing out all
355: degrees of freedom except for a single site and then get its reduced
356: entropy. However, we are still unable to identify many phase
357: boundaries. In our opinion, the failure of these two measures
358: highlights the importance of the nonlocal many body correlation
359: effect in characterizing some nontrivial quantum phases.
360:
361:
362:
363: %{\em Conclusions}.
364: In conclusion, we present a novel approach to study phase diagram of
365: the coupled spin-orbital chain by coherently examining the
366: entanglement related to two distinctive degrees of freedom. The
367: analysis of the SOE supplemented by the sublattice entanglement
368: scenario enables us to establish an one-to-one link between its
369: local extreme/discontinuity and quantum transition points. The most
370: comprehensive phase diagram has been deduced for the first time
371: based on exact numerical results for a finite lattice system. Our
372: approach presents a superior and efficient way to identify quantum
373: phase transitions in a coupled spin-orbital system. The present work
374: may shed new light on the understanding of the complicated interplay
375: among charge, spin and
376: orbital degrees of freedom in transition-metal oxides %quantum many
377: %particle physics
378: in terms of entanglement.
379:
380:
381:
382:
383: The authors thank P. Zanardi, P. Li, S. Q. Shen and G. M. Zhang for
384: helpful discussions. This work was supported by the RGC grants of
385: Hong Kong, the RGC central allocation grant (HKU-3/05C), and Seed
386: Funding grants of HKU.
387:
388:
389: \begin{thebibliography}{}
390:
391: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% general references of 1D SO model %%%%%%%%%%%%%%%%%%%%
392:
393:
394: \bibitem{Book1} S. Maekawa {\em et al.}, {\it Physics of Transition Metal
395: Oxides} (Springer-Verlag, Heidelberg, 2004).
396:
397: \bibitem{Tokura98} M. Imada, A. Fujimori, and Y. Tokura, Rev. Mod.
398: Phys. \textbf{70}, 1039 (1998).
399:
400: \bibitem{Tokura00} Y. Tokura and N. Nagaosa, Science \textbf{288}, 462 (2000).
401:
402:
403: \bibitem{Bao93} W. Bao, C. Broholm, S. A. Carter, T. F. Rosenbaum, G. Aeppli, S. F.
404: Trevino, P. Metcalf, J. M. Honig, and J. Spalek, Phys. Rev. Lett.
405: \textbf{71}, 766 (1993).
406:
407: %\bibitem{Arovas95} D.P. Arovas and A. Auerbach, Phys. Rev. B \textbf{52}, 10114 (1995).
408:
409: \bibitem{Feiner97} L. F. Feiner, A. M. Oles, and J. Zaanen, Phys. Rev. Lett. \textbf{78}, 2799
410: (1997).
411:
412: \bibitem{Kugel73} I. Kugel and D. I. Khomskii, Sov. Phys. JETP Lett. \textbf{37}, 725 (1973);
413: Sov. Phys. Usp. \textbf{25}, 231 (1982).
414:
415: \bibitem{Cast78} C. Castellani, C. R. Natoli, and J. Ranninger, Phys. Rev. B \textbf{18}, 4945
416: (1978).
417:
418: \bibitem{gri99}B. Frischmuth, F. Mila and M. Troyer, Phys. Rev. Lett. {\bf 82}, 835 (1999).
419:
420: \bibitem{Mila99} F. Mila, B. Frischmuth, A. Deppeler, and M. Troyer Phys. Rev. Lett. \textbf{82}, 3697 (1999).
421:
422: \bibitem{Pati98} S. K. Pati, R. R. P. Singh and D. I. Khomskii, Phys. Rev.
423: Lett. \textbf{81}, 5406 (1998).
424:
425: \bibitem{Azaria99} P. Azaria, A. O. Gogolin, P. Lecheminant,
426: and A. A. Nersesyan, Phys. Rev. Lett. {\bf 83}, 624 (1999).
427:
428:
429: \bibitem{yam}Y. Yamashita, N. Shibata, and K. Ueda, Phys. Rev. B{\bf 58}, 9114(1998);
430: J. Phys. Soc. Japan {\bf 69}, 242(2000).
431:
432: \bibitem{Itoi00}C. Itoi, S. Qin and I. Affleck, Phys. Rev. B {\bf 61}, 6747(2000).
433:
434: \bibitem{Zheng01} W. Zheng and J. Oitmaa, Phys. Rev. B \textbf{64}, 014410 (2001).
435:
436: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% general references of entanglement %%%%%%%%%%%%%%%%%%%%
437: %\bibitem{Nielsen1} M.A. Nielsen and I.L. Chuang, {\it Quantum Computation and
438: %Quantum Information} (Cambridge University Press, Cambridge,
439: %England, 2000).
440:
441: \bibitem{Wootters98} W.K. Wootters, Phys. Rev. Lett. {\bf 80}, 2245 (1998).
442:
443: \bibitem{Wang02} X. Wang, and P. Zanardi, Phys. Lett. A {\bf 301}, 1(2002).
444:
445: \bibitem{AOsterloh2002} A. Osterloh, {\em et al.}, Nature {\bf
446: 416}, 608 (2002).
447:
448: \bibitem{GVidal03} G. Vidal, J.I. Latorre, E. Rico, and A. Kitaev, Phys. Rev. Lett,
449: {\bf 90}, 227902 (2003); V. E. Korepin, {\it ibid}. {\bf 92},
450: 096402(2004).
451:
452: \bibitem{Cirac04} F. Verstraete, M.A. Martin-Delgado, and J.I.
453: Cirac, Phys. Rev. Lett. {\bf 92}, 087201 (2004).
454:
455: \bibitem{chen04} Y. Chen, P. Zanardi, Z.D. Wang and F.C. Zhang, New J. Phys. (in press);
456: quant-ph/0407228.
457:
458: \bibitem{Gu04} S.J. Gu, S.S. Deng, Y.Q. Li, and H.Q. Lin, Phys. Rev. Lett. {\bf 93}, 086402 (2004).
459:
460: \bibitem{chen05} Y. Chen, Z.D. Wang and F.C. Zhang, Phys. Rev. B (in press); quant-ph/0512143.
461:
462: \bibitem{Khaliullin06} A.M. Oles, P. Horsch, L.F. Feiner and G. Khaliullin, Phys. Rev.
463: Lett. \textbf{96}, 147205 (2006).
464:
465: \bibitem{YQLi98} Y.Q. Li, M. Ma, D.N. Shi and F.C. Zhang, Phys. Rev. Lett. {\bf 81}, 3527(1998);
466: Phys. Rev. B{\bf 60}, 12781(1999).
467:
468: \bibitem{Footnote1} The $SU(4)$ singlet state can be most conveniently written as, for the purpose
469: of studying SOE, $|SGL \rangle = \sqrt{2/3} [(12)_S (34)_S (14)_O
470: (23)_O -(14)_S (23)_S (12)_O (34)_O]$ where $(12)_{S(O)}$ stands for
471: the spin (orbital) singlet state of the sites 1 and 2.
472:
473:
474: \bibitem{Sutherland75} B. Sutherland, Phys. Rev. B \textbf{12}, 3795 (1975).
475:
476: \bibitem{Kolezhuk98} A.K. Kolezhuk and H. -J. Mikeska, Phys. Rev. Lett. \textbf{80},
477: 2709 (1998); K. Itoh, J. Phys. Soc. Jpn. \textbf{68}, 322 (1999).
478:
479: \bibitem{Nersesyan97} A.A. Nersesyan and A.M. Tsvelik, Phys. Rev. Lett. {\bf 78}, 3939 (1997).
480:
481:
482: \bibitem{Bossche00} M. Bossche, F.C. Zhang and F. Mila, Eur. Phys. J. B \textbf{17}, 367 (2000).
483:
484: %\bibitem{GMZhang03} G. M. Zhang, H. Hu and L. Yu, Phys. Rev. B \textbf{67}, 064420
485: %(2003).
486:
487: \bibitem{Pengli05} P. Li and S.Q. Shen, Phys. Rev. B \textbf{72}, 214439 (2005).
488:
489:
490:
491: \end{thebibliography}
492:
493: \end{document}
494: