1:
2: % Group addresses by affiliation; use superscriptaddress for long
3: % author lists, or if there are many overlapping affiliations.
4: % For Phys. Rev. appearance, change preprint to twocolumn.
5: % Choose pra, prb, prc, prd, pre, prl, prstab, or rmp for journal
6: % Add 'draft' option to mark overfull boxes with black boxes
7: % Add 'showpacs' option to make PACS codes appear
8: % Add 'showkeys' option to make keywords appear
9: %\documentclass[aps,prb,preprint,groupedaddress,showpacs]{revtex4}
10: %\documentclass[aps,pre,preprint,superscriptaddress,showpacs]{revtex4}
11: \documentclass[aps,pra,twocolumn,superscriptaddress,showpacs]{revtex4}
12: \usepackage{tipa}
13: \usepackage{pifont}
14: \usepackage{txfonts}
15: \usepackage{amssymb}
16: \usepackage{graphicx}% Include figure files
17:
18: \begin{document}
19:
20: \title{Partial-state fidelity and quantum phase transitions
21: induced by continuous level crossing}
22:
23:
24: \author{Ho-Man Kwok}
25: \affiliation{Department of Physics and ITP, The Chinese University of Hong
26: Kong, Shatin, Hong Kong, China}
27:
28: \author{Chun-Sing Ho}
29: \affiliation{Department of Physics and ITP, The Chinese University of Hong
30: Kong, Shatin, Hong Kong, China}
31:
32: \author{Shi-Jian Gu}
33: \email{sjgu@phy.cuhk.edu.hk}
34: \affiliation{Department of Physics and ITP, The Chinese University of Hong
35: Kong, Shatin, Hong Kong, China}
36:
37: \date{\today}
38:
39: \begin{abstract}
40: The global-state fidelity cannot characterize those quantum phase
41: transitions (QPTs) induced by continuous level crossing due to its
42: collapse around each crossing point. In this paper, we take the
43: isotropic Lipkin-Meshkov-Glick (LMG) model and the
44: antiferromagnetic one-dimensional Heisenberg model as examples to
45: show that the partial-state fidelity can signal such
46: level-crossing QPTs. Extending to the thermodynamic limit we
47: introduce the partial-state fidelity susceptibility and study its
48: scaling behavior. The maximum of the partial-state fidelity
49: susceptibility goes like $N$ for the LMG model and $N^3$ for the
50: Heisenberg model.
51: \end{abstract}
52:
53:
54: \pacs{64.60.-i, 05.70.Fh, 75.10.-b}
55:
56:
57: %05.70.Fh phase transitions: general studies
58: %03.67.Mn entanglement production, characterization, and manipulation
59: %03.75.Gg entanglement and decoherence Bose-Einstein condensates
60: %03.75.Hh static properties of condensates, thermodynamical statistical
61: % and structural properties
62: %75.10.jm quantized spin models
63: %71.10.Fd Lattice fermion models (Hubbard model, etc.)
64: %75.30.kz magnetic phase boundaries (including magnetic transitions,
65: % metamagnetism, etc)
66: %03.67.-a quantum information
67: %64.60.-i General studies of phase transitions (see also 63.70.+h
68: %Statistical mechanics of lattice vibrations and displacive phase transitions;
69: %for critical phenomena in solid surfaces and interfaces, and in
70: %magnetism, see 68.35.Rh, and 75.40.-s, respectively)
71: %75.10.-b General theory and models of magnetic ordering (see also
72: %05.50.+q Lattice theory and statistics)
73:
74:
75: %05.70.Fh, 03.67.Mn, 03.75.Gg, 03.75.Hh
76:
77: % PACS Number
78:
79:
80: \maketitle
81:
82:
83: It has been an interesting issue for merging quantum phase transitions (QPTs)
84: and fidelity. The former one is noticed by the observation of quantities
85: undergoing structural changes around some critical points \cite{sachdev}, while
86: the latter one measures the amount of relevance for two quantum ground states
87: of the system differed by some parameters \cite{nilesen}. A physical phenomenon
88: is then associated with a pure quantum informational consideration
89: \cite{HTQuan06,PZanardi06,HQZhou07,HQZhou071}. The fidelity approach brings
90: advantages to the characterization of QPTs, because by comparing the states, no
91: \textit{a priori} knowledge to the order parameter, symmetry, and type of QPTs
92: of the system are required. The fidelity approach has been examined in various
93: models and proved its ability in characterizing QPTs \cite{PZanardiStat,
94: Buonsante07, MCozzini07, MCozzini07b, Chen07}. This suggests experimental
95: measurements of the quantum state itself which is a rather challenging task.
96: Leading order of the fidelity has been suggested as well \cite{PZanardi07FS,
97: WLYou07FS}, for its critical exponents and divergence helps classification of
98: the universality of the system \cite{LCVenuti07FS, SJGu07FS, MFYang07FS,
99: MFYang08FS, HMKwok08FS}.
100:
101: However, there are still limitations to the fidelity approach. It is useful to
102: study ground states of some continuous variables, but unable to describe
103: discrete global ground states, i.e. states with fixed quantum numbers within a
104: certain continuous range of parameters. Especially when the driving Hamiltonian
105: commutes with the whole Hamiltonian, the leading order of the fidelity is not
106: well-defined. To tackle this problem, the partial-state fidelity was introduced
107: \cite{HQZhou071,NPaunkovic08}. It concerns the quantum relevance of part of a
108: system with respect to a global change of the parameter. It has been
109: investigated in characterizing the QPTs in the XY model, a three-body
110: interacting model \cite{HQZhou071} and a conventional BCS superconductor with
111: an inserted magnetic impurity system \cite{NPaunkovic08}. In addition, the
112: operator fidelity susceptibility was also introduced and was shown that it can
113: signal QPTs regardless of the degeneracy of the system \cite{XWang08}. However,
114: attention to those phase transitions of continuous level crossing were not paid
115: under their definitions.
116:
117: In this paper, we put our attention on thermodynamic systems in
118: which continuous level crossing occurs. One is the isotropic
119: Lipkin-Meshkov-Glick (LMG) model introduced in nuclear physics
120: \cite{LMG}, which is related to Bose-Einstein condensation and
121: Josephson junctions. We make use of its exact spectrum to obtain
122: the partial-state fidelity. The other one is the one-dimensional
123: Heisenberg model, where we adopt the Bethe-Ansatz method to
124: compute the ground state energy, as well as the required reduced
125: density matrix. We show that the partial-state fidelity can be
126: used to locate the critical point for these two models. We
127: defined the corresponding fidelity susceptibility and perform
128: scaling analysis.
129:
130: Let a system be parameterized by $h$, with its density operator
131: $\hat{\rho}(h) = \left|\Psi (h) \right\rangle \left\langle \Psi
132: (h)\right|$ corresponds to the ground state $\left|\Psi
133: (h)\right\rangle$. When $h$ is displaced by $\delta h$ such that
134: $\tilde{h} = h+\delta h$, the density operator becomes
135: $\hat{\rho}(\tilde{h}) = \left|\Psi (\tilde{h})\right\rangle
136: \left\langle \Psi(\tilde{h})\right|$, the fidelity is defined
137: according to their respective density operator
138: \begin{eqnarray}
139: F(h,\tilde{h}) &=&
140: \textrm{Tr}\sqrt{\sqrt{\hat{\rho}(h)}\hat{\rho}(\tilde{h})
141: \sqrt{\hat{\rho}(h)}} \nonumber \\
142: &=& |\langle\Psi(h)|\Psi(\tilde{h})\rangle|.
143: \end{eqnarray}
144: If the system is divided into two subsystems $A$ and $B$, the
145: reduced density operator $\hat{\rho}_A(h) = \textrm{Tr}_B
146: \hat{\rho}(h)$ contributes to the partial-state fidelity
147: \begin{eqnarray}
148: F_A\left(h,\tilde{h}\right) =
149: \textrm{Tr}\sqrt{\sqrt{\hat{\rho}_A(h)}\hat{\rho}_A(\tilde{h})
150: \sqrt{\hat{\rho}_A(h)}}. \label{eq:PSFdef}
151: \end{eqnarray}
152:
153: The partial state of subsystem $A$ can be a single-site state or a
154: two-site state, or even a larger subsystem state. For convenience
155: in this paper we consider tracing out all particles but one. So
156: for a system with definite magnetization $M$, one can make use of
157: the on-site average magnetization basis $\langle \sigma^z \rangle
158: = 2M/N$, where $N$ is the number of spins, to trace out the
159: density operator. This left us the diagonal reduced density
160: matrix $\rho_A(h)$
161: \begin{eqnarray}
162: \rho_A(h) = \frac{1}{2}\left( \begin{array}{cc}
163: 1+\langle\sigma^z\rangle & 0
164: \\ 0 & 1 - \langle\sigma^z\rangle
165: \end{array} \right). \label{eq:RDMsigma}
166: \end{eqnarray}
167:
168: Consider a system of size $N$ with a set of discrete ground state
169: level-crossing points $\{h_j\}$, where $j = 0,1,2...$ and $h_j >
170: h_{j+1}$. Let the partial state within a range $h \in R_j^{(N)} =
171: (h_{j}, h_{j-1})$ with an average magnetization $\langle \sigma_z
172: \rangle^j$, the partial-state fidelity at $h_j$ is defined by
173:
174: \newpage
175: \begin{eqnarray}
176: &&F_A (h_j) = \textrm{Tr}\nonumber \\
177: &&\sqrt{\frac{1}{4}\left( \begin{array}{cc}
178: 1+\langle\sigma^z\rangle^j & 0
179: \\ 0 & 1 - \langle\sigma^z\rangle^j
180: \end{array} \right)\left( \begin{array}{cc}
181: 1+\langle\sigma^z\rangle^{j+1} & 0
182: \\ 0 & 1 - \langle\sigma^z\rangle^{j+1}
183: \end{array} \right)} \nonumber \\
184: &=& \frac{1}{2}\sqrt{\left(1+\langle \sigma_z \rangle^j\right)\left(1+\langle
185: \sigma_z \rangle^{j+1}\right)}
186: \nonumber \\
187: && + \frac{1}{2}\sqrt{\left(1-\langle \sigma_z
188: \rangle^{j}\right)\left(1-\langle \sigma_z \rangle^{j+1}\right)}.
189: \label{eq:PSF}
190: \end{eqnarray}
191: It is the trace of the reduced density matrices at two sides. The
192: non-unity of Eq. (\ref{eq:PSF}) signals the level crossing when
193: $\langle\sigma^z\rangle$ changes.
194:
195:
196: {\it The isotropic LMG model:} The model reads
197: \begin{eqnarray}
198: H_{\rm LMG} &=& - \frac{1}{N}\sum\limits_{i < j} {\left( {\sigma _x^i \sigma _x^j
199: + \sigma _y^i \sigma _y^j } \right)} - h\sum\limits_i {\sigma _z^i}
200: \nonumber \\
201: &=& - \frac{2}{N}\left( {S_x^2 + S_y^2 } \right) - 2hS_z + \frac{1}{2}
202: \nonumber \\
203: &=& - \frac{2}{N}\left( {\textbf{S}^2 - S_z^2 - N/2} \right) -
204: 2hS_z. \label{eq:LMGmodel}
205: \end{eqnarray}
206: For $\sigma _{\kappa}\, (\kappa=x,y,z)$ are the Pauli matrices,
207: and $S_\kappa = \sum _i \sigma _{\alpha}^i/2$. It describes a
208: system of mutually interacting spins subjected to a transverse
209: external field of strength $h$. The ground state lies in the
210: maximum spin sector $S = N/2$. For it is diagonal in the basis
211: $\left|N/2,M\rangle\right.$, its eigenenergies are given by
212: \begin{eqnarray}
213: E\left( {M,h} \right) = \frac{2}{N}\left( {M - \frac{hN}{2}}
214: \right)^2 - \frac{N}{2}\left( {1 + h^2 } \right).
215: \end{eqnarray}
216: The ground state is determined by the minimum of the square,
217: \begin{eqnarray}
218: M_0 = \left\{ \begin{array}{cc}
219: \Large{\mbox{$\frac{N}{2}$}} & h \ge 1 \\
220: \\
221: I\Large{\mbox{$\left( \frac{hN}{2} \right)$}} & 0 \le h < 1 \\
222: \end{array} \right.,
223: \end{eqnarray}
224: where $I(x)$ gives the integer part of $x$. It can be shown that
225: ground state level crossing occurs at some $h_j \equiv 1-(2j+1)/N$
226: and when $h \in (h_j, h_{j-1})$, the ground state is $M = N/2 -
227: j$ \cite{HMKwok08FS, JVidal05}. In the thermodynamic limit $h =
228: 1$ is the critical point. Since the model [Eq.
229: (\ref{eq:LMGmodel})] is infinitely coordinated, we consider the
230: partial state as a single particle state with density matrix still
231: follows Eq. (\ref{eq:RDMsigma}). The partial-state fidelity at
232: $h_j$ has the exact form according to Eq. (\ref{eq:PSF})
233: \begin{eqnarray}
234: F_A = \frac{1}{N}\left(\sqrt{(N-j)(N-j-1)}+\sqrt{j(j+1)}\right).
235: \end{eqnarray}
236:
237: Fig. \ref{fig:FLMG} shows the plot based on the above formula.
238: $F_A$ drops to a minimum at $h = h_0$, the level-crossing point
239: closest to the critical point $h_c = 1$. Since there are no
240: further level-crossing points for $h > 1$, the partial-state
241: fidelity maintains the value one. As system size increases, the
242: minimum of $F_A$ gets closer to one. It is because unlike
243: ordinary treatment in the fidelity where $\delta h$ is fixed, we
244: calculate the partial-state fidelity obtained by two nearest
245: level-crossing points. When $\delta h$ becomes smaller, the
246: similarity between states becomes higher and is reflected by the
247: close-to-unity behavior, similar to that in global fidelity.
248:
249: We emphasis the comparison between neighboring partial states, as
250: a realization to continuous level crossing when $N \to \infty$.
251: The partial-state fidelity helps extrapolating discrete level
252: crossing to continuous level crossing.
253:
254: \begin{figure}
255: \includegraphics[width=\linewidth]{FLMG}
256: \caption{Partial-state fidelity as a function of external field
257: strength $h$ of the LMG model with different system sizes.}
258: \label{fig:FLMG}
259: \end{figure}
260:
261: {\it The one dimensional Heisenberg model:} The isotropic LMG
262: model provides us an analytic form of the partial-state fidelity.
263: Next we try to examine the one dimensional Heisenberg model,
264: which is another system that exhibits ground state level crossing.
265: The Hamiltonian reads
266: \begin{eqnarray}
267: H_{\rm Heisenberg} &=& \sum_i^N S_i^x S_{i+1}^x + S_i^y S_{i+1}^y
268: + S_i^z S_{i+1}^z - 2hS_i^z,
269: \end{eqnarray}
270: where $S_i^\kappa$ is the spin-1/2 operator at site $i$. With a ring geometry
271: $S_{N+1}^\kappa = S_1^\kappa$ imposed, one can solve the spectrum by the
272: Bethe-Ansatz method \cite{HABethe31}
273: \begin{eqnarray}
274: E_0=\frac{N}{4}-Nh+\sum_{j=1}^{N_\downarrow}
275: \left(2h-\frac{2}{x_j^2 +1}\right) \label{eq:BAGE}
276: \end{eqnarray}
277: where $N_\downarrow$ is the number of down spins, and $x_j$ are
278: spin rapidities. They satisfy the Bethe ansatz equations
279: \cite{HABethe31}
280: \begin{eqnarray}
281: 2N\tan^{-1} x_j=2\pi I_j
282: +2\sum_{l=1}^{N_\downarrow}\tan^{-1}\frac{x_j-x_l}{2},
283: \label{eq:rapidities}
284: \end{eqnarray}
285: where $I_j$ are quantum numbers and take values of
286: $-(N_\downarrow -1)/2, \cdots, (N_\downarrow -1)/2$ for the
287: ground state. By solving the Bethe-Ansatz equations numerically,
288: the eigenenergies are obtained and the ground state is determined
289: by the minimum eigenenergy. So are the level-crossing points. For
290: the single-site subsystem $A$ can also be characterized by the
291: on-site magnetization $\langle \sigma^z \rangle$, according to
292: Eq. (\ref{eq:PSF}) again we compute the partial-state fidelity at
293: the level-crossing points by comparing the nearest partial states.
294: The numerical result is shown in Fig. \ref{fig:FHei}. We find
295: similar features as in the LMG model such as a drop of $F_A$ at
296: the critical point, the drop becomes sharper and the minimum gets
297: closer to one as system size increases.
298:
299: \begin{figure}
300: \includegraphics[width=\linewidth]{FHei}
301: \caption{Partial-state fidelity as a function of external field
302: strength $h$ of the 1D Heisenberg model.} \label{fig:FHei}
303: \end{figure}
304:
305: {\it Partial-state fidelity susceptibility:} Interested in the
306: continuous level crossing that corresponds to the thermodynamic
307: limit, we introduce the concept of partial-state fidelity
308: susceptibility. It is because in such case an infinitesimal
309: change of the parameter is sufficiently responsible for an
310: obvious change of the fidelity. The partial-state fidelity
311: susceptibility is defined in a similar manner as \cite{WLYou07FS}:
312: \begin{eqnarray}
313: \chi_{_F}^{(A)} = \lim_{\delta h \to 0}
314: \frac{-2\textrm{ln}F_A}{(\delta h)^2}. \label{eq:PSFS}
315: \end{eqnarray}
316: The above formula combines the ability of partial-state fidelity
317: in observing level-crossing transitions and the idea of
318: global-state fidelity susceptibility that measures the leading
319: response of fidelity to infinitesimal change of parameter.
320:
321: In finite systems, we compute $\chi_{_F}^{(A)}$ by taking the
322: natural log of the partial-state fidelity at $h_j$, and the divide
323: it by the square of the modulus of the range $R_{j+1}^N$, i.e.
324: $\delta h = h_j - h_{j+1}$. With this notion we arrive the
325: analytic form of $\chi_{_F}^{(A)}$ of the LMG model as $\delta h
326: = 2/N$:
327: \begin{eqnarray}
328: \chi_{_F}^{(A)} =
329: -\frac{N^2}{2}\textrm{ln}\left[\sqrt{\left(1-\frac{j}{N}\right)
330: \left(1-\frac{j+1}{N}\right)}+\sqrt{\frac{j(j+1)}{N^2}}\right].
331: \label{eq:FSLMG}
332: \end{eqnarray}
333: \begin{figure}
334: \includegraphics[width=\linewidth]{FSLMG}
335: \caption{Partial-state fidelity susceptibility as a function of
336: external field strength $h$ of the LMG model with different
337: system sizes.} \label{fig:FSLMG}
338: \end{figure}
339:
340: The plot in fig. \ref{fig:FSLMG} shows $\chi_{_F}^{(A)}$ grows
341: with system size, and arrives its maximum at $h_0$, the
342: level-crossing point closest to the critical point. The response
343: near the maximum becomes sharper for larger systems, indicating a
344: divergence in the thermodynamic limit. It suggests
345: $\chi_{_F}^{(A)}$ as a smooth function of $h$ in the
346: thermodynamic limit except at the critical point. It diverges at
347: $h = h_c$ and drops to zero when $h > h_c$. The divergence of the
348: maximum goes like $N$, since at $j = 0$, from Eq. (\ref{eq:FSLMG})
349: \begin{eqnarray}
350: -\frac{N^2}{2}\textrm{ln}\sqrt{1-\frac{1}{N}} = \frac{N}{4}
351: \end{eqnarray}
352: for large $N$.
353:
354: We compute $\chi_{_F}^{(A)}$ for the Heisenberg model and the
355: result is shown in Fig. \ref{fig:FSHei}. The divergence is even
356: sharper. Although the full analytic form of the $\chi_{_F}^{(A)}$
357: is inaccessible as the spin rapidities $x_j$ form a set of
358: transcendental equations, the critical exponent can be estimated
359: by obtaining $h_0 - h_1$.
360:
361: For $N_\downarrow = 1$, from Eq. (\ref{eq:rapidities}), we have
362: \begin{eqnarray}
363: 2N\tan^{-1} x_1 = 0, \;\,\,\,\, x_1 = 0.
364: \end{eqnarray}
365: The ground state energy for $N_\downarrow = 0$ is simply
366: $\frac{N}{4} - Nh$ and that of $N_\downarrow = 1$ is calculated by
367: Eq. (\ref{eq:BAGE})
368: \begin{eqnarray}
369: \frac{N}{4} - Nh + 2(h-1), \label{eq:GENdown1}
370: \end{eqnarray}
371: in which $h_0 = 1$ is determined.
372:
373: For $N_\downarrow = 2$, Eq. (\ref{eq:rapidities}) consists of two
374: equations
375: \begin{eqnarray}
376: 2N\tan^{-1}x_1 &=& -\pi + 2\tan^{-1}\left(\frac{x_1 -x_2}{2}\right)
377: \nonumber \\
378: 2N\tan^{-1}x_2 &=& \pi + 2\tan^{-1}\left(\frac{x_2 -
379: x_1}{2}\right).
380: \end{eqnarray}
381: The value of $x_1$ and $x_2$ can be found, since by symmetry $x_1
382: = -x_2$, the above two equations become one
383: \begin{eqnarray}
384: 2N\tan^{-1}x_2 = \pi + 2\tan^{-1}x_2, \;\,\,\,\, x_2 =
385: \tan\frac{\pi}{2(N-1)}.
386: \end{eqnarray}
387: The ground state energy for $N_\downarrow = 2$ is
388: \begin{eqnarray}
389: \frac{N}{4} - Nh + 2\left(2h -
390: \frac{2}{\left[\tan\frac{\pi}{2(N-1)}\right]^2+1} \right).
391: \label{eq:GENdown2}
392: \end{eqnarray}
393: Then $h_1$ is determined when Eq. (\ref{eq:GENdown1}) equals to
394: Eq. (\ref{eq:GENdown2}), that is
395: \begin{eqnarray}
396: h_1 = -1 + \frac{2}{\left[\tan\frac{\pi}{2(N-1)}\right]^2+1}.
397: \end{eqnarray}
398: Expanding for large $N$ limit, for $\tan y \simeq y$ for small
399: $y$, we have $h_1 \simeq 1 - \frac{\pi^2}{2(N-1)^2}$ and thus
400: \begin{eqnarray}
401: \delta h = h_0 - h_1 \simeq \frac{\pi^2}{2(N-1)^2}.
402: \end{eqnarray}
403: The partial-state fidelity susceptibility scales like
404: \begin{eqnarray}
405: \chi_{_F}^{(A)} =
406: -\frac{8(N-1)^4}{\pi^4}\textrm{ln}\sqrt{1-\frac{1}{N}} \propto
407: N^3,
408: \end{eqnarray}
409: which is apparently different from that of the LMG model.
410:
411: \begin{figure}
412: \includegraphics[width=\linewidth]{FSHei}
413: \caption{Partial-state fidelity susceptibility as a function of
414: external field strength $h$ of the 1D Heisenberg
415: model.}\label{fig:FSHei}
416: \end{figure}
417:
418: Let us make a remark. In many times, it is often to consider some
419: averaged physical quantities to understand the intrinsic response
420: to the driving agent. Fidelity susceptibility is one of them
421: \cite{LCVenuti07FS, SJGu07FS}. But for the partial-state fidelity
422: susceptibility, we have already focused on a part of the system.
423: Such a local response to the global driving has already played a
424: role as a certain averaged quantity. So we believe, supported by
425: the two distinct models above, the divergence of the partial-state
426: fidelity in continuous level crossing could be a general feature.
427: Its divergence in the isotropic LMG model may be related to the
428: $\gamma \to 1$ limit of the averaged fidelity susceptibility
429: driven by external field in \cite{HMKwok08FS}.
430:
431:
432: We introduced the partial-state fidelity susceptibility formalism
433: derived from the partial-state fidelity and showed it is a
434: suitable candidate to describe quantum phase transitions induced
435: by continuous level crossing, which global-state fidelity cannot
436: provide information to. Focusing on a subsystem, the
437: partial-state fidelity susceptibility still diverges in the
438: thermodynamic limit. The sudden drop-to-zero indicates the
439: critical point of the system.
440:
441: By examining two models, the isotropic LMG model and the one
442: dimensional Heisenberg model, we started from the discrete level
443: crossing and extrapolated to the thermodynamic limit which
444: corresponds to continuous level crossing. We find the maximum of
445: the partial-state fidelity susceptibility goes like $N$ for the
446: LMG model and $N^3$ for the Heisenberg model, indicating they
447: belong to different universality classes. The former one could be
448: treated as a complement to the fidelity susceptibility analysis
449: in the LMG model \cite{HMKwok08FS}.
450:
451: We demonstrated the calculation for a single-site partial state.
452: However, the partial-state fidelity susceptibility shall not be
453: limited to (sub)systems with definite magnetization, because it
454: can still be well-defined for two-particle or many-particle
455: partial states, yet not for all-particle (global) states. The
456: partial-state fidelity is a new approach to tackle QPTs, studying
457: its leading order which is independent of the small change of the
458: driving parameter helps understanding the continuous level
459: crossing QPTs as well as to determine the critical points. We
460: hope this encourages discussions on the related topics.
461:
462: {\it Note added:} After finishing this work, we received a preprint from XG
463: Wang, in which the fidelity and its susceptibility of two-site partial state in
464: the LMG model are studied \cite{XGWang}.
465:
466: We thank J. Vidal for comments on our work. This work is supported
467: by the Direct grant of CUHK (A/C 2060344)
468:
469: \begin{thebibliography}{99}
470:
471: \bibitem{sachdev} S. Sachdev, \emph{Quantum Phase Transitions} (Cambridge
472: University Press, Cambridge, England, 1999).
473:
474: \bibitem{nilesen} M. A. Nilesen and I. L. Chuang, \emph{Quantum Computation
475: and Quantum Information} (Cambridge University Press, Cambridge,
476: England, 2000)
477:
478: \bibitem{HTQuan06}
479: H. T. Quan, Z. Song, X. F. Liu, P. Zanardi, and C. P. Sun, Phys.
480: Rev. Lett. \textbf{96}, 140604 (2006).
481:
482: \bibitem{PZanardi06}
483: P. Zanardi and N. Paunkovic, Phys. Rev. E \textbf{74}, 031123
484: (2006).
485:
486: \bibitem{HQZhou07}
487: H. Q. Zhou and J. P. Barjaktarevic, arXiv: cond-mat/0701608; H. Q. Zhou, J. H.
488: Zhao, and B. Li, arXiv:0704.2940.
489:
490: \bibitem{HQZhou071}
491: H. Q. Zhou, arXiv:0704.2945.
492:
493:
494: \bibitem{PZanardiStat}
495: P. Zanardi, M. Cozzini, and P. Giorda, J. Stat. Mech. \textbf{2},
496: L02002 (2007).
497:
498: \bibitem{Buonsante07}
499: P. Buonsante and A. Vezzani, Phys. Rev. Lett. \textbf{98}, 110601
500: (2007).
501:
502: \bibitem{MCozzini07}
503: M. Cozzini, P. Giorda, and P. Zanardi, Phys. Rev. B \textbf{75},
504: 014439 (2007).
505:
506: \bibitem{MCozzini07b}
507: M. Cozzini, R. Ionicioiu, and P. Zanardi, Phys. Rev. B
508: \textbf{76}, 104420 (2007).
509:
510:
511: \bibitem{Chen07}
512: S. Chen, L. Wang, S. J. Gu, and Y. Wang, Phys. Rev. E \textbf{76}, 061108
513: (2007); S. Chen, L. Wang, Y. Hao, and Y. Wang, Phys. Rev. A {\bf 77}, 032111
514: (2008).
515:
516: \bibitem{PZanardi07FS}
517: P. Zanardi, P. Giorda, and M. Cozzini, Phys. Rev. Lett.
518: \textbf{99}, 100603 (2007).
519:
520: \bibitem{WLYou07FS}
521: W. L. You, Y. W. Li, and S. J. Gu, Phys. Rev. E \textbf{76},
522: 022101 (2007)
523:
524: \bibitem{LCVenuti07FS}
525: L. Campos Venuti and P. Zanardi, Phys. Rev. Lett. \textbf{99},
526: 095701 (2007).
527:
528: \bibitem{SJGu07FS}
529: S. J. Gu, H. M. Kwok, W. Q. Ning, and H. Q. Lin, arXiv:0706.2945 [Phys. Rev. B,
530: to appear].
531:
532: \bibitem{MFYang07FS}
533: M. F. Yang, Phys. Rev. B \textbf{76}, 180403 (R) (2007).
534:
535: \bibitem{MFYang08FS}
536: Yu-Chin Tzeng and M. F. Yang, Phys. Rev. A \textbf{77}, 012311
537: (2008).
538:
539: \bibitem{HMKwok08FS}
540: H. M. Kwok, W. Q. Ning, S. J. Gu, and H. Q. Lin, arXiv:0710.2581.
541:
542: \bibitem{NPaunkovic08}
543: N. Paunkovic \textit{et al}, Phys. Rev. A \textbf{77}, 052302
544: (2008).
545:
546: \bibitem{XWang08}
547: X. Wang, Z. Sun, and Z. D. Wang, arXiv:0803.2940.
548:
549: \bibitem{LMG}
550: H. J. Lipkin, N. Meshkov, and A. J. Glick, Nucl. Phys. \textbf{62}, 188 (1965).
551:
552: \bibitem{JVidal05}
553: S. Dusuel and J. Vidal, Phys. Rev. B \textbf{71}, 224420 (2005).
554:
555: \bibitem{HABethe31}
556: H. A. Bethe, Z. Physik {\bf 71}, 205 (1931).
557:
558: \bibitem{XGWang}
559: J. Ma, L. Xu, and X. G. Wang, arXiv:0805.4062.
560:
561: \end{thebibliography}
562:
563: \end{document}
564: