1: 

2: \documentclass[aps,prl,showpacs,twocolumn]{revtex4}

3: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

4: \usepackage{amssymb}

5: \usepackage{amsmath}

6: \usepackage{graphicx}

7: \usepackage{epsfig}

8: \usepackage{amsfonts}

9: 

10: \setcounter{MaxMatrixCols}{10}

11: %TCIDATA{OutputFilter=Latex.dll}

12: %TCIDATA{Version=5.00.0.2552}

13: %TCIDATA{<META NAME="SaveForMode" CONTENT="1">}

14: %TCIDATA{LastRevised=Monday, February 12, 2007 11:16:38}

15: %TCIDATA{<META NAME="GraphicsSave" CONTENT="32">}

16: 

17: \providecommand{\U}[1]{\protect\rule{.1in}{.1in}}

18: 

19: 

20: \begin{document}

21: 

22: \title{Atomic Entanglement vs Photonic Visibility \\

23: for Quantum Criticality of Hybrid System}

24: \author{M.X. Huo$^{1}$}

25: \author{Ying Li$^{1}$}

26: \author{Z. Song$^{1}$}

27: \email{songtc@nankai.edu.cn}

28: \author{C.P. Sun$^{2}$}

29: \email{suncp@itp.ac.cn}

30: \homepage{http://www.itp.ac.cn/~suncp}

31: \affiliation{$^{1}$Department of Physics, Nankai University, Tianjin 300071, China}

32: \affiliation{$^{2}$Institute of Theoretical Physics, Chinese Academy of Sciences,

33: Beijing, 100080, China}

34: 

35: \begin{abstract}

36: To characterize the novel quantum phase transition for a hybrid system

37: consisting of an array of coupled cavities and two-level atoms doped in each

38: cavity, we study the atomic entanglement and photonic visibility in

39: comparison with the quantum fluctuation of total excitations. Analytical and

40: numerical simulation results show the happen of quantum critical phenomenon

41: similar to the Mott insulator to superfluid transition. Here, the contour

42: lines respectively representing the atomic entanglement, photonic visibility

43: and excitation variance in the phase diagram are consistent in the vicinity

44: of the non-analytic locus of atomic concurrences.

45: \end{abstract}

46: 

47: \pacs{03.65.Ud, 42.50.Dv, 73.43.Nq, 03.67.-a}

48: \maketitle

49: 

50: \textit{Introduction:} It is crucial in the modern theory of second order

51: phase transitions to introduce `order parameter', whose non-vanishing

52: average value characterizes one or more phases and usually breaks a symmetry

53: of the Hamiltonian. But for `quantum' phase transitions on the behavior of

54: matter near zero temperature \cite{qpt}, it is more subtle to use the

55: traditional Ginzburg-Landau-Wilson paradigm, since in some cases the natural

56: description of the quantum criticality is not based on the order parameter

57: \cite{sci}.

58: 

59: Actually, for some systems with complicated structures it might be difficult

60: to chose an appropriate order parameter to correctly characterize the

61: emergent phenomena. The purpose of this paper with a specific example is to

62: demonstrate that, though we can not make sure what is the appropriate order

63: parameter for a photon-atom hybrid system, some physical observable

64: quantities can be used to witness its quantum critical phenomenon.

65: 

66: The hybrid system we consider is a coupled waveguide resonator array (CWRA)

67: where each cavity is doped with a two-level atom (see Fig. 1). This hybrid

68: architecture was suggested as a quantum coherent device to transfer and

69: store quantum information as well as to create the laser-like output \cite%

70: {zhou1, zhou2, hu}. As for the quantum phase transition, it is observed that

71: such a doped CWRA can simulate the Mott like transition of light from

72: \textquotedblleft the Mott insulator (MI) to superfluid (SF)" \cite%

73: {greentree} since a doped atom can induce the effective photon-photon

74: interaction in each cavity. Together with the inter-cavity hopping of

75: localized phonons, this nonlinear photon-photon coupling can result in the

76: Bose-Hubbard model for Mott phase transition \cite{fisher1}. Recent

77: experiments \cite{Greiner} using cold atoms in an optical lattice have

78: clearly demonstrated the quantum phase transition predicted by the

79: Bose-Hubbard model. Actually the Bose-Hubbard theory of Mott phase

80: transition for cold atoms \cite{zoller1} is also based on the assumption of

81: the order parameter, the average of the annihilation operator of boson in

82: each site. In mean field approach, the average of the annihilation operator

83: of boson is usually employed, while \textquotedblleft number variance" is

84: used in many other methods to discriminate Mott insulating and superfluid

85: phases \cite{Hartmann, Angelakis}.

86: 

87: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

88: 

89: \begin{figure}[ptb]

90: \includegraphics[bb=34 357 563 640, width=6 cm]{fig1.eps}

91: \caption{\emph{Schematic setup of a cavity array with each one containing a

92: two-level atom. Photons of mode $\protect\omega_{a}$\ can tunnel between

93: adjacent cavities with hopping integral }$\emph{t}$\emph{\ and couple to the

94: atoms with strength $g$. This atom-photon lattice is expected to simulate

95: Mott insulator and superfluid transition.}}

96: \label{fig1}

97: \end{figure}

98: 

99: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

100: 

101: For the hybrid system, the study of quantum phase transition of light in

102: Ref. \cite{greentree} still assumes the same order parameter in terms of

103: photons, while more strict assumption of order parameter \cite{Hartmann,

104: Angelakis} was implicitly made in terms of the number of the polariton,

105: which is a mixture of photon and atom. Here, we will not adopt such local

106: order parameters of quasi-particle as direct characterizations of quantum

107: phase transition, but pay attention to some observable quantities to

108: characterize the critical phenomenon of a hybrid system. To this end we make

109: use of atomic entanglement from the point view of quantum information, as

110: well as photon visibility in terms of quantum optical theory. We remark

111: that, as a quantum nonlocal property, the quantum entanglement plays an

112: important role in the study of quantum phase transitions \cite{Osterloh02,

113: Osborne02, Gu03, WangZD, Qian}. We examine the signature of the Mott

114: insulator to superfluid transition, the excitation number variance, and

115: other two observable quantities, concurrence between two atoms and the

116: visibility of photons, in lattice atom-photon hybrid systems of small size

117: by analytical and numerical methods respectively. Our results reveal

118: nontrivial connections among the three quantities in such an intriguing way:

119: contour lines of three quantities in the phase diagram are approximately

120: consistent with each other when the non-analyticity of concurrences occurs.

121: It firmly shows that such three quantities are signatures for the MI to SF

122: transition in such a atom-photon hybrid system.

123: 

124: \textit{Model setup and the quasi-excitation fluctuation.} We consider an

125: array of $N$ coupled cavities with each one containing a single two-level

126: atom \cite{Angelakis, zhou1, zhou2, hu}. The photon transmission of

127: cavity-to-cavity occurs as a hopping mechanism if there were the interaction

128: between atom and cavity mode. Such hybrid system can be implemented with the

129: defect array in photonic crystal \cite{fan} or Josephson junction array in

130: cavity \cite{zhou1}. The Hamiltonian of a hybrid system, or a lattice

131: atom-photon system, $H=H_{free}+H_{int}+H_{hop}$ is decomposed as three

132: parts, free Hamiltonians of light and atom,%

133: \begin{equation}

134: H_{free}=\omega _{a}\sum_{i=1}^{N}a_{i}^{\dag }a_{i}+\omega

135: _{b}\sum_{i=1}^{N}\left\vert e\right\rangle _{i}\left\langle e\right\vert ,

136: \end{equation}%

137: the cavity-mode-atom interaction in the $i$th defect

138: \begin{equation}

139: H_{int}=g\sum_{i=1}^{N}\left( a_{i}^{\dag }\left\vert g\right\rangle

140: _{i}\left\langle e\right\vert +\text{H.c.}\right) ,  \label{H2}

141: \end{equation}%

142: with strength $g$ and the photon hopping between NN defects

143: \begin{equation}

144: H_{hop}=-t\sum_{i=1}^{N}\left( a_{i}^{\dag }a_{i+1}+\text{H.c.}\right) ,

145: \end{equation}%

146: with hopping integral constant $t$ for the tunneling between adjacent

147: cavities. Here, $\left\vert g\right\rangle _{i}$ ($\left\vert e\right\rangle

148: _{i}$) denotes the ground (excited) state of the atom placed at $i$th

149: cavity; $a_{i}^{\dag }$ and$\ a_{i}$ are the creation and annihilation

150: operators of a photon at defect $i$. Obviously the total excitation number

151: \begin{equation}

152: \mathcal{\hat{N}}\mathcal{=}\sum_{i=1}\mathcal{\hat{N}}_{i}=\sum_{i=1}\left(

153: a_{i}^{\dag }a_{i}+S_{i}^{z}+\frac{1}{2}\right)

154: \end{equation}%

155: is conserved quantity for the Hamiltonian $H$, i.e., $[H,\mathcal{\hat{N}}%

156: ]=0 $, where $2S_{i}^{z}\left\vert e\right\rangle _{i}=\left\vert

157: e\right\rangle _{i}$ and $2S_{i}^{z}\left\vert g\right\rangle

158: _{i}=-\left\vert g\right\rangle _{i}$.

159: 

160: It can be seen that $\mathcal{\hat{N}}$ is just the single excitation number

161: of the polaritons. It is well known that the conventional Mott insulator to

162: superfluid phase transition occurs in a Bose-Hubbard model. Here, when the

163: repulsive interaction between bosons is large enough in the Mott phase, the

164: number fluctuation would become energetically unfavorable, forcing the

165: system into a number state and exhibiting vanishing particle number

166: fluctuation. In the superfluid regime, atoms are delocalized with

167: non-vanishing particle number fluctuation. As for the present hybrid system,

168: the fundamental excitations are polaritons \cite{Angelakis} and the

169: mechanism of the Mott transition is due to the effect of photon blockade.

170: Since the photon number is not conserved in such system, the photon number

171: fluctuation $\Delta n_{i}=\Delta (a_{i}^{\dag }a_{i})$ is not appropriate to

172: characterize the superfluid phase as that for a pure Bose-Hubbard model.

173: This is because $\Delta n_{i}$\ does not vanish even in the Mott insulator

174: regime due to the couplings between photons and atoms. Hereafter, we define

175: the variance $\Delta A$ by $(\Delta A)^{2}=\left\langle (A)^{2}\right\rangle

176: -\left\langle A\right\rangle ^{2}$. Therefore, one can take the excitation

177: number fluctuation per site $\Delta \mathcal{N}_{i}$ as an\ order parameter

178: to characterize the Mott transition. In the large detuning limit $\delta

179: =\omega _{a}-\omega _{b}\gg 0$, all atoms are in excited states, which is

180: perfectly number squeezed states, i.e., $\Delta \mathcal{N}_{i}=0$ for all

181: sites. In the other limit $\delta \ll 0$, all atoms are in ground states.

182: Obviously two-atom concurrence vanishes and the density fluctuation becomes $%

183: \Delta \mathcal{N}_{i}=\sqrt{\langle a_{i}^{\dag }a_{i}^{\dag

184: }a_{i}a_{i}\rangle }=\sqrt{(N-1)/N}\simeq 1$ since $\mathcal{N}=N$ in this

185: case.

186: 

187: \textit{Atomic entanglement characterized by concurrence. }Intuitively, two

188: atoms in two adjacent cavities should entangle with each other due to the

189: hopping of phonon from one cavity to another. Now we try to describe this

190: kind of atomic entanglement induced by coupled photons. Obviously, if the

191: photon is in quantum phase transition, the critical photon induced atomic

192: entanglers can characterize this critical behavior.

193: 

194: We express the concurrence characterizing quantum entanglement in terms of

195: observable quantities such as correlation functions. The complete basis

196: vectors of the total system are denoted by

197: \begin{equation}

198: |\{n_{j},s_{j}\}\rangle =|n_{1},..,n_{N};s_{1},..,s_{N}\rangle

199: =\prod\limits_{j=1}^{N}|n_{j}\rangle \otimes |s_{j}\rangle

200: \end{equation}%

201: where $|n_{j}\rangle $ is the Fock state of photon and $|s_{j}\rangle

202: =\left\vert g\right\rangle _{i}$, $\left\vert e\right\rangle _{i}$ for $%

203: s_{j}=0,1$ respectively. The fact that $\mathcal{\hat{N}}$ is conserved can

204: be reflected by the matrix element vanishing of the density operator $\rho

205: =\rho (H)$ on the above basis for any state of the hybrid system, that is,

206: \begin{equation}

207: \rho _{\{n_{j},s_{j}\}}^{\{n_{j}^{\prime },s_{j}^{\prime }\}}=\rho

208: _{\{n_{j},s_{j}\}}^{\{n_{j},s_{j}\}}\delta \lbrack \sum

209: (n_{j}+s_{j}-n_{j}^{\prime }-s_{j}^{\prime })]

210: \end{equation}%

211: The functional $\rho (H)$ of the Hamiltonian may be a ground state or

212: thermal equilibrium states. The reduced density matrix $\rho

213: ^{(12)}=Tr_{p}Tr_{3..N}^{s}[\rho (H)]$ for two atomic quasi-spins, e.g., $%

214: s_{1}$ and $s_{2}$ are obtained as

215: \begin{align}

216: \lbrack \rho ^{(12)}]_{s_{1}^{\prime }s_{2}^{\prime }},_{s_{1}s_{2}}&

217: =\sum_{[n_{j};s_{3}...s_{N}]}\rho

218: _{n_{j},s_{1}s_{2}s_{3}...s_{N}}^{n_{j},s_{1}^{\prime }s_{2}^{\prime

219: }s_{3}..s_{N}}\delta \lbrack \sum (s_{j}-s_{j}^{\prime })]  \notag \\

220: & =\delta (s_{1}+s_{2}-s_{1}^{\prime }-s_{2}^{\prime

221: })\sum_{[n_{j};s_{3}...s_{N}]}\rho

222: _{n_{j},s_{1}s_{2}s_{3}..s_{N}}^{n_{j},s_{1}^{\prime }s_{2}^{\prime

223: }s_{3}..s_{N}}

224: \end{align}%

225: by tracing over all photon variables (with $Tr_{p}$) and atomic variables

226: except for $s_{1}$ and $s_{2}$. The corresponding reduced density matrix for

227: two atoms $i$ and $j$ is of the form

228: \begin{equation}

229: \rho ^{(ij)}=\left(

230: \begin{array}{cccc}

231: u_{ij}^{+} & 0 & 0 & 0 \\

232: 0 & w_{ij}^{1} & z_{ij}^{\ast } & 0 \\

233: 0 & z_{ij} & w_{ij}^{2} & 0 \\

234: 0 & 0 & 0 & u_{ij}^{-}%

235: \end{array}%

236: \right) .  \label{reduced}

237: \end{equation}%

238: According to Refs. \cite{Wootters, Wang}, the concurrence $C_{ij}=\max

239: \left\{ 0,\lambda _{1}-\lambda _{2}-\lambda _{3}-\lambda _{4}\right\} $

240: shared between two atoms $i$ and $j$ is obtained in terms of the the square

241: roots $\left\{ \lambda _{i}\right\} $ ($\lambda _{1}=\max \left\{ \lambda

242: _{i}\right\} $) of eigenvalues of the non-Hermitian matrix $\rho \widetilde{%

243: \rho }$. Here $\widetilde{\rho }=\left( \sigma _{y}\otimes \sigma

244: _{y}\right) \rho ^{\ast }\left( \sigma _{y}\otimes \sigma _{y}\right) $.

245: Using the observable quantities, the quantum correlation $%

246: z_{ij}=\left\langle \psi \right\vert S_{i}^{+}S_{j}^{-}\left\vert \psi

247: \right\rangle $, $u_{ij}^{\pm }=\left\langle \psi \right\vert \left( 1/2\pm

248: S_{i}^{z}\right) \left( 1/2\pm S_{j}^{z}\right) \left\vert \psi

249: \right\rangle $, the\ concurrence is rewritten as a computable form

250: \begin{equation}

251: C_{ij}=2\max (0,\left\vert z_{ij}\right\vert -\sqrt{u_{ij}^{+}u_{ij}^{-}}).

252: \label{c2}

253: \end{equation}%

254: We note that this formula for the concurrence of two quasi-spin in a hybrid

255: system is the same as that for pure spin-1/2 system \cite{Wootters, Wang}.

256: The non-analyticity of concurrence arises from the abrupt switch of the sign

257: of quantity $\left\vert z_{ij}\right\vert -\sqrt{u_{ij}^{+}u_{ij}^{-}}$ and

258: can be used to determine quantum phase transitions.

259: 

260: \textit{Photon visibility in hybrid system. }Similar to the transition of

261: superfluid to Mott insulator in Bose-Hubbard model \cite{Gerbier1}, two

262: phases of the atom-photon hybrid system can also be delimited through the

263: quantum coherence of the ground state. In Mott insulating phase, the quantum

264: coherence of photons is completely destroyed due to the photon blockade. In

265: superfluid phase limit, the quantum coherence of photons gets its maximum.

266: Therefore the quantum coherence of photons can be employed to indicate

267: phases, which is characterized by a observable quantity, the visibility of

268: `interference fringes'

269: \begin{equation}

270: \mathcal{V=}\frac{V_{\max }-V_{\min }}{V_{\max }+V_{\min }}.  \label{v}

271: \end{equation}%

272: Here, $V_{\max }$\ and $V_{\min }$\ are the maximum and minimum of the

273: photon number distribution of ground state in the $k$ space

274: \begin{equation}

275: V(k)=\frac{1}{N}\sum_{j,l}e^{ik(j-l)}\left\langle a_{j}^{\dag

276: }a_{l}\right\rangle .

277: \end{equation}%

278: In the strong photon blockade limit, $\mathcal{V}=0$\ while in the

279: superfluid limit, $\mathcal{V}=1$. Comparing with the local quantity, the

280: photon number fluctuation, the visibility is more appropriate to

281: discriminate two phases.

282: 

283: \textit{Characterizing quantum criticality. }For the lattice atom-photon

284: system, we now consider the connections among three quantities $\Delta

285: \mathcal{N}_{i}$, $C_{ij}$, and $\mathcal{V}$\ around the critical point.

286: 

287: We start with an extreme case $g=0$. At zero temperature, the physics of the

288: lattice atom-photon model can be described in two regimes separated by the

289: boundary $\delta =2t$. In the region $\delta >2t$ ($\delta <2t$), the ground

290: state is in typical Mott insulating (superfluid) phase with $\Delta \mathcal{%

291: N}_{i}=0$ ($\sqrt{(N-1)/N}$), $\mathcal{V}=0$ ($1$), and $C_{ij}=0$ ($0$),

292: respectively. At line $\delta =2t$, the model admits multi-fold degenerate

293: ground states with energy $\varepsilon ^{(0)}=N\omega _{b}$, and the

294: excitation number fluctuation and visibility experience a big jump, while

295: the concurrence between two atoms is `uncertainty' due to the energy-level

296: crossing.

297: 

298: When the atom-photon interaction $g$\ is switched on, it becomes avoided

299: level crossing. This fact will result in the quantum fluctuation driving the

300: transition from Mott insulator to superfluid phase, which corresponds to the

301: non-vanishing concurrence between atoms. To illustrate this mechanism

302: quantitatively, we just switch on atom-photon couplings in cavities $i$ and $%

303: j$ and leave all other coupling to be zero. For very small $g$, the

304: un-perturbable ground states

305: \begin{align}

306: \left\vert \phi _{1}\right\rangle & =|N\rangle

307: _{k=0}\prod\limits_{l}\left\vert g\right\rangle _{l},\left\vert \phi

308: _{2}\right\rangle =|N-1\rangle _{k=0}\left\vert e\right\rangle

309: _{i}|G_{i}\rangle ,  \notag \\

310: \left\vert \phi _{3}\right\rangle & =|N-1\rangle _{k=0}\left\vert

311: e\right\rangle _{j}|G_{j}\rangle , \\

312: \left\vert \phi _{4}\right\rangle & =|N-2\rangle _{k=0}\left\vert

313: e\right\rangle _{i}\left\vert e\right\rangle _{j}|G_{i,j}\rangle ,  \notag

314: \end{align}%

315: are degenerate, where\ $|n\rangle _{k}$ denotes the photon Fock state in $k$

316: space and\ $|G_{i,j,..}\rangle =\prod_{l\neq i,j,..}\left\vert

317: g\right\rangle _{l}$ denotes the atomic state of all atoms except $l=i,j,...$%

318: . Up to the first order perturbation with energy correction $\varepsilon

319: ^{(1)}=-g\sqrt{2N(1+\beta ^{2})}$ $=-g\sqrt{N}\eta ^{-1}$ where $\beta =%

320: \sqrt{(N-1)/N}$, the perturbed ground state is

321: 

322: \begin{equation}

323: \left\vert \psi _{g}\right\rangle =\eta (\left\vert \phi _{1}\right\rangle

324: +\beta \left\vert \phi _{4}\right\rangle )-\frac{1}{2}(\left\vert \phi

325: _{2}\right\rangle +\left\vert \phi _{3}\right\rangle ).

326: \end{equation}%

327: The corresponding concurrence can be calculated as

328: \begin{equation}

329: C_{ij}=\frac{(1-\beta )^{2}}{2(1+\beta ^{2})}.  \label{c_app}

330: \end{equation}

331: As $\delta $ being apart from the degenerate point, the concurrence $C_{ij}$%

332: \ decreases due to the energy competition of two phases. Therefore, this

333: heuristic analysis has shown the simple relation among concurrence,

334: visibility, and excitation number fluctuation around quantum phase

335: transition critical point: the excitation number fluctuation and visibility

336: both exhibit an abrupt drop while the concurrence has a sharp maximum. It

337: can be predicted that as $g$ increases, changes of the three quantities will

338: be slow due to the strongly coupling between atoms and photons. In the

339: following, it will be investigated for small system in wide range of

340: parameters by numerical simulations.

341: 

342: We investigate three quantities in a small size system by exact

343: diagonalization method. For open chain cavity array system, the visibility $%

344: \mathcal{V}$\ can be calculated by\ $S(k)=2/(N+1)\sum_{i,j}\sin (ki)\sin

345: (kj)\left\langle a_{i}^{\dagger }a_{j}\right\rangle $, where $k=n\pi /\left(

346: N+1\right) $, $n\in \lbrack 1,N]$, while the concurrence and excitation

347: number fluctuation can be characterized as average concurrence $\overline{C}%

348: =(1/N)\sum_{i<j}C_{ij}$ and average excitation number fluctuation $\overline{%

349: \Delta \mathcal{N}}=1/N\sum_{i}\Delta \mathcal{N}_{i}$. In Fig. 2, contours

350: of three quantities obtained by exact diagonalization are plotted in the $%

351: \delta /g$-$t/g$ plane for 2- (Fig. 2(a, b)), 4- (Fig. 2(c, d)) cavity

352: systems. Contours of excitation number fluctuation $\overline{\Delta

353: \mathcal{N}}$ (dark lines in Fig. 2(a, c)) and visibility of photons $%

354: \mathcal{V}$ (dark lines in Fig. 2(b, d)) are compared with the concurrence $%

355: \overline{C}$\ (color maps in Fig. 2(a-d)) as functions of the scaled

356: detuning $\delta /g$\ and photon hopping integral $t/g$. We see that contour

357: lines of three quantities are consistent in the vicinity of the locus at

358: which the non-analyticity of concurrence occurs. The non-analytic locus in $%

359: \delta /g$-$t/g$\ plane is defined by the equation $\left\vert

360: z_{ij}\right\vert -\sqrt{u_{ij}^{+}u_{ij}^{-}}=0$. Red lines in Fig. 2(a-d)

361: denote closer contour lines of $\mathcal{V}$ and $\overline{\Delta \mathcal{N%

362: }}$ to the non-analytical curve of $\overline{C}$. It also shows that the

363: visibility and excitation number fluctuation start to drop at the

364: non-analytic locus of concurrence. There is a slight difference between

365: profiles of 2 and 4-cavity systems. The red contour line of $\mathcal{V}$ in

366: 4-cavity system is closer to the non-analytic locus of $\overline{C}$ than

367: that in 2-cavity system. It indicates that contour lines of\ three

368: quantities will cover at the vicinity the non-analytic locus of $\overline{C}

369: $ in thermodynamics limit.

370: 

371: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

372: 

373: \begin{figure}[ptb]

374: \includegraphics[bb=19 258 585 764, width=8.5 cm]{fig2.eps}

375: \caption{\textit{(Color online) Contours of three quantities, }$\overline{%

376: \Delta\mathcal{N}}$ \textit{(dark lines in (a,c))}, $\mathcal{V}$ \textit{%

377: (dark lines in (b,d))}, and $\overline{C}$\textit{\ (color maps in (a-d)

378: obtained by exact diagonalization for 2- (a,b), 4- (c,d) cavity systems. Red

379: lines in (a-d) denote closer contour lines of }$\mathcal{V}$\textit{\ and }$%

380: \overline {\Delta\mathcal{N}}$\textit{\ to the non-analytical curve of }$%

381: \overline{C}$\textit{. It shows that contour lines of three

382: quantities are consistent in the vicinity of the non-analytic

383: locus, at which the non-analyticity of concurrence occurs.}}

384: \label{fig2}

385: \end{figure}

386: 

387: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

388: 

389: \textit{Summary: }In summary, we have investigated the Mott

390: insular to superfluid transition in a hybrid system consisting of

391: an array of coupled cavities doped with two level atoms. We

392: investigate two non-local observable quantities, the concurrence

393: between atoms and the visibility of photons, in comparison with

394: the local order parameter, excitation number fluctuation, for the

395: Mott insulator to superfluid transition. It can be observed form

396: analytical and numerical simulation results that the atomic

397: entanglement and photonic visibility in the phase diagram indeed

398: can reflect the quantum critical phenomenon signatured by the

399: total excitation variance. In principle, such non-local observable

400: quantities of the hybrid system can be used to detect the critical

401: point in experiment.

402: 

403: This work is supported by the NSFC with grant Nos. 90203018, 10474104 and

404: 60433050, and NFRPC with Nos. 2006CB921206 and 2005CB724508.

405: 

406: \begin{thebibliography}{99}

407: \bibitem{qpt} S. Sachdev, Quantum Phase Transitions (Cambridge University

408: Press, Cambridge, England, 1999).

409: 

410: \bibitem{sci} T. Senthil, A. Vishwanath, L. Balents, S. Sachdev, M. P. A.

411: Fisher Science \textbf{303}, 1490 (2004).

412: 

413: \bibitem{zhou1} L. Zhou, Y.B. Gao, Z. Song, and C.P. Sun, cond-mat/0608577,

414: submitted to Phys. Rev. Lett..

415: 

416: \bibitem{hu} F.M. Hu, L. Zhou, T. Shi, C.P. Sun, quant-ph/0610250(2006)

417: submitted to Phys. Rev. E.

418: 

419: \bibitem{zhou2} L. Zhou, J. Lu, C.P. Sun, quant-ph/0611159, submitted to

420: Phys. Rev. A.

421: 

422: \bibitem{greentree} A.D. Greentree, C. Tahan, J.H. Cole, and L.C.L.

423: Hollenberg, Nature Phys. \textbf{2}, 856 (2006).

424: 

425: \bibitem{fisher1} M.P.A. Fisher, P.B. Weichman, G. Grinstein, and D.S.

426: Fisher, Phys. Rev. B \textbf{40}, 546 (1989).

427: 

428: \bibitem{Greiner} M. Greiner, O. Mandel, T. Esslinger, T.W. H$\ddot{a}$nsch,

429: and I. Bloch, Nature (London) \textbf{415}, 39 (2002).

430: 

431: \bibitem{zoller1} D. Jaksch, C. Bruder, J.I. Cirac, C.W. Gardiner, and P.

432: Zoller, Phys. Rev. Lett. \textbf{81}, 3108 (1998).

433: 

434: \bibitem{Hartmann} M.J. Hartmann, F.G.S.L. Brandao and M.B. Plenio, Nature

435: Phys. \textbf{2}, 849 (2006).

436: 

437: \bibitem{Angelakis} D.G. Angelakis, M.F. Santos, and S. Bose,

438: quant-ph/0606159.

439: 

440: \bibitem{Osterloh02} {A. Osterloh, L. Amico, G. Falci, and R. Fazio}, Nature

441: \textbf{416}, 608 (2002).

442: 

443: \bibitem{Osborne02} {T.J. Osborne and M. A. Nielsen}, Phys. Rev. A \textbf{66%

444: }, 032110 (2002).

445: 

446: \bibitem{Gu03} {S.J. Gu, H.Q. Lin, and Y.Q. Li}, Phys. Rev. A \textbf{68},

447: 042330 (2003).

448: 

449: \bibitem{WangZD} Y. Chen, Z.D. Wang and F.C. Zhang, Phys. Rev. B \textbf{73}%

450: , 224414 (2006).

451: 

452: \bibitem{Qian} X.F. Qian, Ying Li, Yong Li, Z. Song, and C.P. Sun, Phys.

453: Rev. A \textbf{72}, 062329 (2005).

454: 

455: \bibitem{fan} M.F. Yanik, S. Fan, Phys. Rev. Lett. \textbf{92}, 083901

456: (2004); Q. Xu, S. Sandhu, M.L. Povinelli, J. Shakya, S. Fan, and M. Lipson,

457: Phys. Rev. Lett. \textbf{96}, 123901 (2006).

458: 

459: \bibitem{Wootters} W.K. Wootters, Phys. Rev. Lett. \textbf{80}, 2245 (1998).

460: 

461: \bibitem{Wang} X. Wang and P. Zanardi, Phys. Lett. A \textbf{301}, 1 (2002);

462: X. Wang, Phys. Rev. A \textbf{66}, 034302 (2002).

463: 

464: \bibitem{Gerbier1} F. Gerbier, A. Widera, S. F$\ddot{o}$lling, O. Mandel, T

465: Gericke, and I. Bloch, Phys. Rev. Lett. \textbf{95}, 050404 (2005); Phys.

466: Rev. A \textbf{72}, 053606 (2005).

467: \end{thebibliography}

468: 

469: \end{document}

470: