1: %\documentclass[aps,showpacs,twocolumn]{revtex4}

2: 

3: 

4: \documentclass[aps,twocolumn,pra,superscriptaddress,showpacs,tightenlines]{revtex4}

5: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

6: \usepackage{amssymb}

7: \usepackage{amsmath}

8: \usepackage{graphicx}

9: \usepackage{epsfig}

10: 

11: \setcounter{MaxMatrixCols}{10}

12: %TCIDATA{OutputFilter=Latex.dll}

13: %TCIDATA{Version=5.00.0.2552}

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

15: %TCIDATA{LastRevised=Saturday, February 25, 2006 11:51:01}

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

17: 

18: 

19: \begin{document}

20: 

21: \title{Quantum dynamics of magnetically controlled network for Bloch

22: electrons }

23: \author{S. Yang$^{1}$, Z. Song}

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

25: \affiliation{Department of Physics, Nankai University, Tianjin 300071, China}

26: \author{C.P. \surname{Sun}}

27: \email{suncp@itp.edu.cn} \homepage{http://www.itp.ac.cn/~suncp}

28: \affiliation{Institute of Theoretical Physics, Chinese Academy of

29: Sciences, Beijing 100080, China}

30: 

31: \begin{abstract}

32: We study quantum dynamics of wave packet motion of Bloch electrons

33: in quantum networks with the tight-binding approach for different

34: types of nearest-neighbor interactions. For various geometrical

35: configurations, these networks can function as some optical

36: devices, such as beam splitters and interferometers. When the

37: Bloch electrons with the Gaussian wave packets input these

38: devices, various quantum coherence phenomena can be observed,

39: e.g., the perfect quantum state transfer without reflection in a

40: Y-shaped beam, the multi- mode entanglers of electron wave by star

41: shaped network and Bloch electron interferometer with the lattice

42: Aharonov-Bohm effects. Behind these conceptual quantum devices are

43: the physical mechanism that, for hopping parameters with some

44: specific values, a connected quantum networks can be reduced into

45: a virtual network, which is a direct sum of some irreducible

46: subnetworks. Thus, the perfect quantum state transfer in each

47: subnetwork in this virtual network can be regarded as a coherent

48: beam splitting process. Analytical and numerical investigations

49: show the controllability of wave packet motion in these quantum

50: networks by the magnetic flux through some loops of these

51: networks, or by adjusting the couplings on nodes. We find the

52: essential differences in these quantum coherence effects when the

53: different wave packets enter these quantum networks initially.

54: With these quantum coherent features, they are expected to be used

55: as quantum information processors for the fermion system based on

56: the possible engineered solid state systems, such as the array of

57: quantum dots that can be implemented experimentally.

58: \end{abstract}

59: 

60: \pacs{03.65.Ud, 75.10.Jm, 03.67.Lx}

61: \maketitle

62: 

63: \section{Introduction}

64: 

65: Quantum information processing (QIP) has been a very active area of research

66: in the past few years \cite{QIP1,QIP2}. The current challenge for QIP is to

67: coherently integrate a sufficiently large and complex controllable system

68: and then requires the ability to transfer quantum information between

69: spatially separated quantum bits. Since then numerous approaches for this

70: purpose have been proposed, ranging from linear and nonlinear quantum

71: optical devices to various interacting quantum systems. Among them, many

72: studies proposed using the internal dynamics of coupled spins for the

73: transfer of quantum information \cite{QST1,QST2,QST3,QST4,QST5,LY,ST,SZ,QST6}%

74: . However, the basic and necessary \textquotedblleft optical

75: devices\textquotedblright\ (for the electron wave or the spin wave) in a

76: solid system are scarce due to the technology at hand. Therefore, it is

77: significant to probe the possibilities to construct the artificial

78: \textquotedblleft optical devices\textquotedblright\ and then build the

79: corresponding electronic networks for the matter wave of electron within a

80: solid state system \cite{quant-network,spin-network,Experiment}. Here, we

81: notice that, for the boson system, Plenio, Hartley and Eisert \cite{Plenio}\

82: have studied the quantum network dynamics of a system consisting of a large

83: number of coupled harmonic oscillators in various geometric configurations

84: for the similar purpose.

85: 

86: This paper will pay attentions to the fermion systems where the Bloch

87: electrons move along the quantum lattice network. We consider various

88: geometrical configurations of tight-binding networks that are analogous to

89: quantum optical devices, such as beam splitters and interferometers. We then

90: consider the functions of these tight-binding networks in details when

91: initially Gaussian wave packets are entering these devices. Analytical and

92: numerical investigations show that these devices are controllable by the

93: magnetic flux through the network. Characteristic parameters of devices can

94: be adjusted by changing the flux or the interactions on nodes. The relevant

95: quantum phenomena, such as generation of entanglement and the Aharonov-Bohm

96: (AB) effects in the solid state based devices are also discussed

97: systematically.

98: 

99: This paper is organized as follows. In section II, we present the

100: Hamiltonians of the simplest tight-binding lattice systems with and without

101: magnetic field as building blocks to construct various networks, which can

102: be formed topologically by the linear and the various connections between

103: the ends of them. In section III, we theoretically design and analytically

104: study the basic properties of a star-shaped TBN, then also explore the

105: further dynamic property of $Y$-shaped network. Surprisingly, for

106: appropriate joint hopping integrals, the complicated network can be reduced

107: into an imaginary linear chain with homogeneous NN hopping terms plus a

108: smaller complicated network. It is known that such TBNs are analogous to

109: quantum optical devices such as beam-splitters, entangler and

110: interferometers. In section IV, we investigate the dynamic properties of a

111: Bloch electron model on a $Q$-shaped lattice, which consists of a terminal

112: chain and a ring threaded by a magnetic flux. The appropriate flux through

113: the network can reduce the network to the linear virtual chain, which

114: indicates that the flux can control the propagation of GWP in the network.

115: In section V, the interferometer network, a mimic of the AB effect

116: experiment, is also studied in the similar way. In addition, in the whole

117: paper, a moving Gauss wave packet (GWP) localized in a linear dispersion

118: regime is a good example to illustrate the properties of the above Bloch

119: electron networks. In section VI, we extend the results of the TBNs to the

120: spin network (SN) for the dynamics of the single magnon. In section VII, we

121: summarize the results of this paper and suggest the possible applications of

122: these TBNs.

123: 

124: \section{Basic setup}

125: 

126: In this section, we introduce the systems under consideration, namely the

127: tight-binding Bloch electron systems and the Hamiltonians of the building

128: blocks to construct various networks. Without loss of the generality, we

129: concentrate our attention on the simplest tight-binding systems, in which

130: only the hopping term or kinetic energy is considered.

131: 

132: A general tight-binding network (TBN) is constructed topologically by the

133: linear tight-binding chains and the various connections between the ends of

134: them. An important element in the system is the Aharonov-Bohm flux through

135: some loops of the TBN. Here, we consider the simplest tight-binding model,

136: in which only the nearest neighbor (NN) hopping terms are taken into

137: account. The Hamiltonian of a tight-binding linear chain of $N_{l}$ sites

138: reads as

139: 

140: \begin{equation}

141: H_{l}=H_{l}(N_{l})\equiv

142: -\sum_{j=1}^{N_{l}-1}(t_{j}^{[l]}a_{l,j}^{\dag }a_{l,j+1}+H.c.).

143: \label{h}

144: \end{equation}%

145: Here, the label $l$ denotes the chain with the distribution of the hopping

146: integrals $\{t_{1}^{[l]},t_{2}^{[l]},...,t_{N_{l}-1}^{[l]}\}$ and $%

147: a_{l,j}^{\dag }$ is the fermion creation operator at $j$th site of the chain

148: $l$. The hopping integral could be the complex number due the presence of

149: the external magnetic field. In this paper, we restrict our study to a

150: simplest case described as following. When the field is absent, the hopping

151: integral for a homogeneous chain%

152: \begin{equation}

153: t_{j}^{[l]}=te^{i\Phi _{l,j+1}}  \label{ct}

154: \end{equation}%

155: in a chain is real and identical (i.e., independent of sites) while in the

156: presence of a vector potential the hopping integral is modified by a phase

157: factor. Here, we defined the link phase

158: \begin{equation}

159: \Phi _{l,j+1}=\frac{2\pi }{\phi _{0}}\int_{j}^{j+1}\mathbf{A}\cdot d\mathbf{l%

160: }  \label{phase}

161: \end{equation}%

162: with flux quanta $\phi _{0}=hc/e$ as an integral of the vector potential $%

163: \mathbf{A}$ along the link between the sites $j$ and $j+1$ in the $l$th

164: chain. The above observation about the phase modification of hopping

165: integral can be found in many modern literatures \cite%

166: {gauge-trans,tight-binding} but the proof can be cast back to Peierls {\cite%

167: {Peierls}}.

168: 

169: Another important portion of the quantum networks is the joints or nodes

170: between two linear chains, of which the Hamiltonian can be presented in the

171: form

172: \begin{equation}

173: H_{joint}\equiv -(t_{ji}^{[lm]}a_{l,j}^{\dag }a_{m,i}+H.c.),

174: \label{joint}

175: \end{equation}%

176: where $t_{ji}^{[lm]}$\ denotes the hopping integral over the $j$th site of

177: chain $l$ and the $i$th site of chain $m$. Here, only one joint term

178: connecting two chains is listed in $H_{joint}$ as an illustration. In a

179: general TBN, the joint Hamiltonian $H_{joint}$ should contain many

180: connection terms.

181: 

182: In remaining parts of this paper, we will show that, under certain

183: conditions, a complex TBNs can be decomposed as a simple sum of several

184: independent imaginary chains. In order to avoid confusion, we describe one

185: of the imaginary chains of $N_{l}$ sites by a Hamiltonian%

186: \begin{equation}

187: \widetilde{H}_{l}=\widetilde{H}_{l}(N_{l})\equiv -t\sum_{j=1}^{N_{l}-1}(%

188: \widetilde{a}_{l,j}^{\dag }\widetilde{a}_{l,j+1}+H.c.).

189: \label{ih}

190: \end{equation}%

191: Here, $\widetilde{a}_{l,j}^{\dag }$ denoted by tilde are the fermion

192: creation operators for $j$th site of the imaginary chain $l$, \ which are

193: linear combinations of $a_{m,i}$. Namely, there exists a mapping $R$ between

194: the two sets of fermion operators, $\{a_{l,j}^{\dag }\}\overset{R}{%

195: \longrightarrow }\{\widetilde{a}_{m,i}^{\dag }\}$ or by a transformation $R$:

196: 

197: \begin{equation}

198: \widetilde{a}_{m,i}^{\dag }=\sum_{l,j}R_{l,j,m,i}a_{l,j}^{\dag }

199: \end{equation}%

200: We will investigate several types of networks based on the notations

201: introduced above. As an example to demonstrate the application of the

202: notation, we can express the main conclusion of this paper by using the

203: above well-defined notations as

204: \begin{equation}

205: \sum\limits_{l}H_{l}+H_{joint}\overset{R}{\longrightarrow }\sum\limits_{m}%

206: \widetilde{H}_{m},

207: \end{equation}%

208: i.e., a network can be equivalent to the simple sum of several independent

209: imaginary chains with the aid of the transformation $R$.

210: 

211: There is a remark to be made here: In usual, the role of the potential $%

212: \mathbf{A}$ shows as the AB effect of Bloch electron in a close chain (or

213: called a tight-binding ring). Here, the local magnetic filed strength for

214: Bloch electron may vanish, but the loop integral of $\mathbf{A}$--the\textbf{%

215: \ }magnetic flux does not. Due to the AB effect, the magnetic flux $\phi $\

216: can be used to control the single-particle spectrum of a homogeneous chain, $%

217: \varepsilon _{k}=-2t\cos (k+2\pi \phi /N)$. When the flux $\phi \sim \phi

218: _{n}$ $\equiv (n/2$ $+1/4)N$, the lower spectrum becomes a linear dispersion

219: approximately, i.e., $\varepsilon _{k}\sim k$. For the wave packets as a

220: superposition of those eigenstates with small $k$, the\ linearization of

221: Hamiltonian lead to the transfer of the wave packet without spreading \cite%

222: {YS1}.

223: 

224: \section{The basic blocks of quantum network: star- and Y-shaped beams}

225: 

226: In this section, we use analytical method and numerical simulation to study

227: the basic blocks of quantum network, the star- and Y-shaped beams, which is

228: constructed by connecting one ends of several chains to the end (a node) of

229: a single chain. Beam splitters are the elementary optical devices frequently

230: used in classical and quantum optics \cite{QOP}, which can even work well in

231: the level of single photon quanta \cite{S-photon} and are applied to

232: generate quantum entanglement \cite{Entng}. For matter wave, an early beam

233: splitter can be referred to the experiments of neutron interference based on

234: a perfect crystal interferometer with wavefront and amplitude division \cite%

235: {N-interfer}. Moreover, for cold atoms, a beam splitter have been

236: experimentally implemented on the atom chip \cite{C-atom}. The theoretical

237: method has been suggested to realize the beam splitter for the Bose-Einstein

238: condensate \cite{BEC}.

239: 

240: In the following, we begin with the basic properties of the corresponding

241: networks for fermions. We will show that, for appropriate joint hopping

242: integrals, the complicated network can be reduced into an imaginary linear

243: chain with homogeneous NN hopping terms plus a smaller complicated network.

244: We also further study the dynamic property of such kind of network by taking

245: the $Y$-shaped network as an example of star shape beams, in which only

246: three chains are involved. We investigate various aspects that are analogous

247: to quantum optical devices, such as beam-splitters, entangler, and

248: interferometers.

249: 

250: \subsection{Star-shaped beam splitter and its reduction}

251: 

252: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

253: \begin{figure}[tbp]

254: \includegraphics[bb=45 125 560 710, width=7 cm,clip]{star.eps}

255: \caption{\textit{(Color on line) (a) The star-shaped Bloch electron network

256: with an input chain $A$ and $m$ output chains in the real space. (b) When

257: the joint hopping constants satisfy $t_{n}=t/\protect\sqrt{m}$, the STBN can

258: be reduced into one homogeneous tight-binding chain $a$ with length $M+N$

259: and $m-1$ virtual chains with length $N$ in the virtual space.} }

260: \label{star}

261: \end{figure}

262: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

263: We consider a simple TBN, a star shape (we also call the star-shaped

264: tight-binding network (STBN)) as shown in Fig. \ref{star}(a). The STBN is

265: constructed by linking the $m$ output chains $B_{p}$ $(p=$ $1,2,$ $...,m)$

266: to the one end of the input chain\ $A$ by the hopping integrals $%

267: t_{N_{A}1}^{[AB_{p}]}$. The Hamiltonian of an STBN consists of the linear

268: chain part (\ref{h}) and the joint part (\ref{joint}) around the last $N_{A}

269: $th site of the input chain $A$. Obviously, since there is no vector

270: potential acting on the network, all the hopping integrals are real.

271: 

272: We will show that, under some restriction for the intrachain hopping

273: constants $t$\ and interchain hopping constants $t_{N_{A}1}^{[AB_{p}]}$, an

274: STBN can be reduced into an imaginary linear tight-binding chain with

275: homogeneous hopping constants plus a smaller complicated network. The fact

276: that the input chain $A$ is a part of this virtual linear chain implies that

277: the Bloch electron can perfectly propagate in this virtual linear chain

278: without the reflection by the node. This also indicates that there is a

279: coherent split of the input electronic wave because the wave function in

280: this virtual chain actually is just a superposition of wave functions in the

281: $m$ chains.

282: 

283: To sketch our central idea, we first consider a special STBN, which consists

284: of $m$ identical \textquotedblleft output\textquotedblright\ chains $%

285: B_{1},B_{2},\cdots ,$ $B_{m}$ with homogeneous intrachain hopping constants$%

286: \ t$, interchain hopping constant$\ t_{N_{A}1}^{[AB_{p}]}=t_{n}$ and the

287: same length $N$, while the length of chain $A$ is $M$. The Hamiltonian of

288: the network%

289: \begin{equation}

290: H=\sum_{p=1}^{m}H_{B_{p}}+H_{A}+H_{joint}

291: \end{equation}%

292: is now explicitly written in terms of the chain Hamiltonians $H_{B_{p}}$ and

293: $H_{A}$ defined by Eq. (\ref{h}). Here, the basic parameters for the network

294: are $t_{j}^{[B_{p}]}=t_{j}^{[A]}=t$, $N_{B_{p}}=N$, and $N_{A}=M$, where $p=$

295: $1,2,$ $...,m$. The joint Hamiltonian is

296: \begin{equation}

297: H_{joint}=-t_{n}(a_{A,M}^{\dag }\sum_{p=1}^{m}a_{B_{p},1}+H.c.).

298: \end{equation}%

299: Note that we only consider the case that all the hopping integrals over the

300: joints are identical for the convenience of illustration. In the next

301: section, the different joint hopping integrals will be taken into account

302: for a simplest case of $m=2$.\

303: 

304: Now we construct the new fermion operators denoted by the tilde notation, $%

305: \widetilde{a}_{a,j}^{\dag }$\ of virtual tight-binding chain $a$ of length $%

306: M+N$ as

307: \begin{align}

308: \widetilde{a}_{a,j}^{\dag }& =a_{A,j}^{\dag },  \notag \\

309: \widetilde{a}_{a,M+l}^{\dag }& =\frac{1}{\sqrt{m}}%

310: \sum_{p=1}^{m}a_{B_{p},l}^{\dag },  \label{EYa}

311: \end{align}%

312: where $j\in \lbrack 1,M]$ and $l\in \lbrack 1,N]$. There exist $m-1$

313: complementary tight-binding chains with the collective operators

314: \begin{equation}

315: \widetilde{a}_{b_{q},j}^{\dag }=\frac{1}{\sqrt{m}}\sum_{p=1}^{m}\exp (-i2\pi

316: pq/m)a_{B_{p},j}^{\dag },  \label{EYb}

317: \end{equation}%

318: where $j\in \lbrack 1,N]$, $p=1,2,$ $\cdots ,m,$ and $q=1,2,\cdots ,$ $m-1$.

319: It can be checked that, all the tilde operators are also the standard

320: fermion operators, which satisfy the anticommutation relation%

321: \begin{equation}

322: \left\{ \widetilde{a}_{\alpha ,i},\widetilde{a}_{\beta ,j}^{\dag }\right\}

323: =\delta _{\alpha \beta }\delta _{ij},

324: \end{equation}%

325: where $\alpha ,\beta \in (a,b_{1},b_{2},\cdots ,b_{m-1})$ denote the labels

326: of the virtual chains. By inverting Eqs. (\ref{EYa}) and (\ref{EYb}) we have%

327: \begin{align*}

328: a_{A,j}^{\dag }& =\widetilde{a}_{a,j}^{\dag },\text{ }(j=1,2,\cdots ,M) \\

329: a_{B_{p},j}^{\dag }& =\frac{1}{\sqrt{m}}\sum_{q=1}^{m-1}e^{i\frac{2\pi pq}{m}%

330: }\widetilde{a}_{b_{q},j}^{\dag }+\frac{1}{\sqrt{m}}\widetilde{a}%

331: _{a,M+j}^{\dag },

332: \end{align*}%

333: where $p\in \lbrack 1,m]$, $q\in \lbrack 1,m-1]$, and $j\in \lbrack 1,N]$.\

334: These establish the mapping $R$ between the two sets of fermion operators, $%

335: \{a_{l,j}^{\dag }\}\overset{R}{\longrightarrow }\{\widetilde{a}_{m,i}^{\dag

336: }\}$. Therefore, we have%

337: \begin{eqnarray}

338: H &=&-t\sum_{j=1}^{M-1}\widetilde{a}_{a,j}^{\dag }\widetilde{a}%

339: _{a,j+1}-t\sum_{j=1}^{N-1}(\sum_{q=1}^{m-1}\widetilde{a}_{b_{q},j}^{\dag }%

340: \widetilde{a}_{b_{q},j+1} \\

341: &&+\widetilde{a}_{a,M+j}^{\dag }\widetilde{a}_{a,M+j+1})-\sqrt{m}t_{n}%

342: \widetilde{a}_{a,M}^{\dag }\widetilde{a}_{a,M+1}+H.c..  \notag

343: \end{eqnarray}

344: 

345: The above Hamiltonian depicts a TBN with different geometry. It is easy to

346: observe that only when the matching condition of the joint hopping constants

347: \begin{equation}

348: t_{N_{A}1}^{[AB_{p}]}=t_{n}=\frac{t}{\sqrt{m}}  \label{march1}

349: \end{equation}%

350: is satisfied, we have $H=\widetilde{H}_{a}+\sum_{q=1}^{m-1}\widetilde{H}%

351: _{b_{q}}:$%

352: \begin{align}

353: \widetilde{H}_{a}& =-t\sum_{j=1}^{M+N-1}\widetilde{a}_{a,j}^{\dag }%

354: \widetilde{a}_{a,j+1}  \notag \\

355: \widetilde{H}_{b_{q}}& =-t\sum_{j=1}^{N-1}\widetilde{a}_{b_{q},j}^{\dag }%

356: \widetilde{a}_{b_{q},j+1}+H.c.  \label{reduction}

357: \end{align}%

358: where $N_{a}=M+N$ and $N_{b_{q}}=N$. The tilde Hamiltonians are also

359: illustrated in Fig. \ref{star}(b). Interestingly, all the sub-Hamiltonians $%

360: \widetilde{H}_{a}$, $\widetilde{H}_{b_{q}}$\ commutate with each other, i.e.,

361: 

362: \begin{equation}

363: \left[ \widetilde{H}_{\alpha },\widetilde{H}_{\beta }\right] =0,

364: \end{equation}%

365: where $\alpha ,\beta \in (a,b_{1},b_{2},\cdots ,b_{m-1})$. This fact means

366: that the virtual chain described by $\widetilde{H}_{a}$ is just a standard

367: tight-binding chain of length $N_{a}$ with uniform NN couplings. For an

368: arbitrary initial state localized within the chain $H_{a}$, it will evolve

369: driven by the virtual chain of length $M+N$. If the local state moves out of

370: chain $H_{a}$, an ideal beam splitter can be realized since there is no

371: reflection occurs at the node.

372: 

373: To demonstrate it, we take an example with the initial state as the Gaussian

374: wave packet (GWP). The GWP with momentum $\pi /2$ has the form

375: \begin{equation}

376: \left\vert \psi (N_{0})\right\rangle =\frac{1}{\sqrt{\Omega }}%

377: \sum_{j=1}^{M}e^{-\frac{^{\alpha ^{2}}}{2}(j-N_{0})^{2}}e^{i\frac{\pi }{2}%

378: j}\left\vert j\right\rangle .  \label{GWP}

379: \end{equation}%

380: Here, $\Omega =\sum_{j=1}^{N_{A}}\exp [-\alpha ^{2}(j-N_{0})^{2}]$ is the

381: normalization factor and $N_{0}\in \lbrack 1,N_{A}]$ is the initial central

382: position of the GWP at the input chain $A$, while the factor $\alpha $ is

383: large enough to guarantee the locality of the state in the chain $A$.

384: Accordingly, the basis $\left\vert j\right\rangle $ is defined as $%

385: \left\vert j\right\rangle =\widetilde{a}_{a,j}^{\dag }\left\vert

386: 0\right\rangle $ for $j\in \lbrack 1,M+N]$. In the previous work \cite{YS1},

387: it has been shown that such GWP can approximately propagate along a

388: homogenous chain without spreading. Actually, at a certain time $\tau $,

389: such GWP evolves into%

390: \begin{eqnarray}

391: \left\vert \Psi (\tau )\right\rangle &=&e^{-i\widetilde{H}_{a}\tau

392: }\left\vert \psi (N_{0})\right\rangle \simeq \left\vert \psi (N_{0}+2t\tau

393: )\right\rangle  \notag \\

394: &=&\frac{1}{\sqrt{\Omega }}\sum_{j=M+1}^{M+N}e^{-\frac{^{\alpha ^{2}}}{2}%

395: (j-N_{0}-2t\tau )^{2}}e^{i\frac{\pi }{2}j}\left\vert j\right\rangle

396: \end{eqnarray}%

397: in the virtual space. From the mapping of the operators (\ref{EYa}), we have

398: the final state as%

399: \begin{equation}

400: \left\vert \Psi (\tau )\right\rangle =\frac{1}{\sqrt{m}}\sum_{p=1}^{m}\left%

401: \vert \phi _{p}(N_{\tau })\right\rangle .

402: \end{equation}%

403: Here, the state

404: 

405: \begin{equation}

406: \left\vert \phi _{p}(N_{\tau })\right\rangle =\frac{1}{\sqrt{\Omega }}%

407: \sum_{j=1}^{N}e^{-\frac{^{\alpha ^{2}}}{2}(j-N_{\tau })^{2}}e^{i\frac{\pi }{2%

408: }j}a_{B_{p},j}^{\dag }\left\vert 0\right\rangle ,

409: \end{equation}%

410: is the clone of the initial GWP with the center at $N_{\tau }=N_{0}+2t\tau

411: -M $ in the chain $B_{p}$. Then we conclude that the beam splitter split the

412: single-particle GWP into $m$ cloned GWPs without any reflection.

413: Furthermore, it is obvious that the function of the splitter originates from

414: the reduction (\ref{reduction}) of the Hamiltonian, which is available for

415: every invariant subspaces of fixed particle number. Therefore, such splitter

416: can be applied to the many-particle system.

417: 

418: The above discussion is limited to the simplest case of identical joint

419: hopping integrals. We would like to say that the marching condition (\ref%

420: {march1}) is not unique for constructing independent virtual chain. We will

421: demonstrate this for the case $m=2$ in the next section.

422: 

423: \subsection{Y-shaped beam splitter}

424: 

425: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

426: \begin{figure}[tbp]

427: \includegraphics[bb=50 210 550 660, width=7 cm,clip]{Y.eps}

428: \caption{\textit{(Color on line) (a) $Y$-shaped TBN or called $Y$-beam, a

429: special STBN with different joint hopping integrals $t_{nB}$ and $t_{nC}$.

430: (b) Reduction of $Y$-shaped TBN under the matching condition. It shows that

431: if $t_{nB}^{2}+t_{nC}^{2}=t^{2}$, the $Y$-shaped TBN in real space is mapped

432: into virtual space as two decoupled virtual chain $a$ and $b$ with length $%

433: M+N$ and $N$ respectively.}}

434: \label{Y}

435: \end{figure}

436: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

437: The reduction for star-shaped network was demonstrated under the

438: restriction\ (\ref{march1}). This kind of reduction can be also performed

439: with different joint hopping integrals. If the splitter is applied to the

440: local GWP, the node interactions of the two output legs are not necessary to

441: be identical. In the following, we will investigate this issue by

442: considering the simplest configuration with $m=2$, which is called $Y$-beam.

443: The asymmetric $Y$-beam consists of three legs $A$, $B$ and $C$ with the

444: intrachain hopping integrals $t$ for $F=A,B,C$ and the joint ones $t_{nF}$

445: for $F=B,C$ (see Fig. \ref{Y}(a)). The total Hamiltonian reads

446: \begin{equation}

447: H=\sum_{F=A,B,C}H_{F}-\sum_{F=B,C}(t_{nF}a_{A,M}^{\dag

448: }a_{F,1}+H.c.)

449: \end{equation}%

450: where $t_{j}^{[A]}=t_{j}^{[B]}$ $=t_{j}^{[C]}=t$, $N_{A}=M,$ and $%

451: N_{B}=N_{C}=N$.

452: 

453: In order to decouple this $Y$-beam\emph{\ }as two virtual linear

454: tight-binding chains, we need to optimize the asymmetric couplings so that

455: the perfect transmission can occur in the decoupled linear tight-binding

456: chains. For this purpose, we introduce the tilde operators of fermion%

457: \begin{align}

458: \widetilde{a}_{a,j}^{\dag }& =a_{A,j}^{\dag },  \notag \\

459: \widetilde{a}_{a,M+l}^{\dag }& =\cos \theta a_{B,l}^{\dag }+\sin \theta

460: a_{C,l}^{\dag },  \notag \\

461: \widetilde{a}_{b,m}^{\dag }& =\sin \theta a_{B,m}^{\dag }-\cos \theta

462: a_{C,m}^{\dag },  \label{tilde_Y}

463: \end{align}%

464: where $j\in \lbrack 1,M]$, $l\in \lbrack 1,N]$,\ and $m\in \lbrack 1,N]$.

465: Here, the mixing angle $\theta $ is to be determined as follows by the

466: optimization for quantum state transmission. In comparison with the optical

467: beam splitter, the above equations (\ref{tilde_Y}) can be regarded as a

468: fundamental issue for the electronic wave beam splitter.

469: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

470: \begin{figure}[tbp]

471: \includegraphics[bb=30 240 500 650, width=7 cm,clip]{fslt.eps} %

472: \includegraphics[bb=20 315 490 785, width=6.5 cm,clip]{fspm.eps}

473: \caption{\textit{(Color on line) (a) The contour map of the reflection

474: factor $R(t_{nC},t_{nB})$ as a function of $t_{nC},t_{nB}$ for the GWP with $%

475: \protect\alpha =0.3$ and momentum $\protect\pi /2$ in a finite system with $%

476: N_{A}=$$N_{B}=$$N_{C}=50$. It shows that around the matching condition, i.e,

477: the circle $t_{nC}^{2}+t_{nB}^{2}=t^{2}$, the reflection factor approaches

478: zero. (b) The profile of $R(t_{nC},t_{nB})$ along $t_{nC}=t_{nB}$. }}

479: \label{R}

480: \end{figure}

481: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

482: 

483: Together with the original creation operator $\widetilde{a}_{a,j}^{\dag

484: }=a_{A,j}^{\dag }$ for the input leg, the set $\{\widetilde{a}_{a,M+j}^{\dag

485: }\mid j\in \lbrack 1,N]\}$ defines a new linear chain $a$ with the effective

486: hopping integrals

487: \begin{eqnarray}

488: t_{aj} &=&t_{a,M+l}=t,  \notag \\

489: t_{aM} &=&t_{nB}\cos \theta +t_{nC}\sin \theta ,

490: \end{eqnarray}%

491: where $j\in \lbrack 1,M-1]$\ and $l\in \lbrack 1,N-1]$.\ Another virtual

492: linear chain $b$ is defined by $\widetilde{a}_{b,j}^{\dag }$ with

493: homogeneous hopping integral $t_{bj}=t$,$\ j\in \lbrack 1,N-1]$.

494: 

495: In general, these two linear chains are dependent, since there is a

496: connection interaction around the node

497: \begin{equation}

498: H_{joint}=-t_{AB}(\widetilde{a}_{a,M}^{\dag

499: }\widetilde{a}_{b,1}+H.c.)

500: \end{equation}%

501: where

502: \begin{equation}

503: t_{AB}=t_{nB}\sin \theta -t_{nC}\cos \theta .

504: \end{equation}%

505: Fortunately, the two virtual chains decouple with each other when the mixing

506: angle $\theta $ and the intrachain connections are optimized by setting $%

507: \tan \theta =t_{nC}/t_{nB}$. And if we take $t_{AB}=t$, the virtual chain $a$%

508: \ becomes a completely homogeneous chain of length $M+N$ as illustrated in

509: Fig. \ref{Y}(b).\ Thus, these lead to the matching\ condition%

510: \begin{equation}

511: \sqrt{t_{nC}^{2}+t_{nB}^{2}}=t  \label{matching}

512: \end{equation}%

513: for $Y$-beam network. It can be employed to transfer the quantum state

514: without reflection on the node in the transformed picture. Transforming back

515: to the original picture, we can see that such network behaves as a perfect

516: beam splitter.

517: 

518: In the point of view of linear optics, such beam splitting process can

519: generate the\ mode entanglement between the separated waves in chains $B$

520: and $C$, and the measure of such mode entanglement is determined by the

521: values of $t_{nB}$ and $t_{nC}$. We will show that the strength of $t_{nB}$

522: and $t_{nC}$ can be used to control the amplitudes of the evolving Bloch

523: electron wave packets on legs $B$ and $C$.

524: 

525: Now we apply the beam splitter to a special Bloch electron wave packet, a

526: GWP with momentum $\pi /2$, which has the form (\ref{GWP}) at $\tau =0$. It

527: is known from the previous work \cite{YS1} that such GWP can approximately

528: propagate along a homogenous chain without spreading. Then at a certain time

529: $\tau $, such GWP evolves into

530: \begin{equation}

531: \left\vert \Psi (\tau )\right\rangle =\cos \theta \left\vert \psi

532: _{B}(N_{\tau })\right\rangle +\sin \theta \left\vert \psi _{C}(N_{\tau

533: })\right\rangle  \label{split}

534: \end{equation}%

535: where

536: \begin{equation}

537: N_{\tau }=N_{0}+2t\tau -M,

538: \end{equation}%

539: i.e., the beam splitter can split the GWP into two cloned GWPs completely.

540: The possibility of GWPs in the arms $B$ and $C$ is determined by the mixed

541: angle $\theta $.

542: 

543: In order to verify the above analysis, a numerical simulation is performed

544: for a GWP with $\alpha =0.3$ in a finite system with $N_{A}=N_{B}$ $=$$%

545: N_{C}=50$. Let $\left\vert \Phi (0)\right\rangle $ be a normalized initial

546: state. Then the reflection factor at time $\tau $ can be defined as%

547: \begin{equation}

548: R(t_{nC},t_{nB},\tau )=\sum_{j_{A}=1}^{M-1}\left\vert \left\langle

549: j_{A}\right\vert e^{-iH\tau }\left\vert \Phi (0)\right\rangle \right\vert

550: ^{2}

551: \end{equation}%

552: to depict the reflection at the node. At an appropriate instant $\tau _{0}$,

553: \begin{equation}

554: R(t_{nC},t_{nB})=R(t_{nC},t_{nB},\tau _{0})

555: \end{equation}%

556: as a function of $t_{nC}$ and $t_{nB}$ is plotted in Fig. \ref{R}. It shows

557: that around the matching condition (\ref{matching}), the reflection factor

558: vanishes, which is just in agreement with our analytical result.

559: 

560: The conclusion for such GWP comes from the reduction of the $Y$-shaped

561: network, which the equal length of two output arms is crucial. However, for

562: the local wave packet (\ref{GWP}), the local environment around the wave

563: packet only result in its behavior at the next instant. This can be seen

564: from the speed of the GWP (\ref{GWP}). According to the study in Ref. \cite%

565: {YS1}, the speed of the GWP is independent of the size and the boundary

566: condition (ring or open chain). Therefore, for a splitter to GWP, the

567: equality of two output arms is not necessary. This argument will be

568: demonstrated in the following content about quantum interferometer.

569: 

570: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

571: \begin{figure}[tbp]

572: \includegraphics[bb=30 240 500 650, width=7 cm,clip]{jclt.eps} %

573: \includegraphics[bb=15 315 490 785, width=6.5 cm,clip]{jcpm.eps}

574: \caption{\textit{(Color on line) (a) The contour map of maximal concurrence

575: of two GWPs at two legs $A$ and $B,$ $C_{\max }(t_{nC},t_{nB})$ for the same

576: setup as that in Fig. \protect\ref{R}. It is found that two GWPs yield the

577: maximal entanglement at the point $t_{nC}=$$t_{nB}=$$t_{A}/\protect\sqrt{2}$%

578: . (b) The profile of $C_{\max }(t_{nC},t_{nB})$ along $t_{nC}=t_{nB}$.}}

579: \label{C}

580: \end{figure}

581: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

582: 

583: \subsection{Entangler of Bloch electron}

584: 

585: Now we consider how the STBN can behave as an entangler to produce

586: entanglement with the $Y$-beam\textbf{\ }as an illustration. Let the input

587: state $|\phi (0)\rangle $, represents single-particle state located in the

588: arm $A$. It can propagate into the arms $B$ and $C$ through the node with

589: some reflection. On the other hand, the electronic wave can be regarded as

590: being transferred along the virtual legs $a$ and $b$. Once we manipulate the

591: joint hopping integrals to satisfy the matching condition, the electronic

592: wave can only enter the virtual chain $a$ rather than $b$ without any

593: reflection. Then the final state is in the subspace of the virtual chain $a$%

594: . Similar to optical splitter, such $Y$-beam splitting can also be regarded

595: as an entangler of fermion. For instance, consider a state $|D(j)\rangle =%

596: \widetilde{a}_{a,j}^{\dag }|0\rangle $ for $j\in \lbrack M,M+N]$, which is a

597: local state in the view of point of the virtual chain $a$. However, in the

598: real space, this state is nonlocal and possesses mode entanglement, while

599: state $|D(j)\rangle $ for $j\in $ $[1,M]$ is still a non-entangled state.

600: Obviously, the $Y$-beam acts as an entangler similar to that in quantum

601: optical systems.

602: 

603: To quantitatively characterize mode entanglement generated by the splitter

604: on the joint hopping integrals $t_{nC}$, $t_{nB}$, we consider the GWP (\ref%

605: {GWP}) as an initial state. Through the splitter, two separated GWPs are

606: obtained. The total concurrence with respect to the two waves located at the

607: arms $B$ and $C$ can be calculated as

608: \begin{equation}

609: C(\tau )=\sum_{j=1}^{N}\left\vert \left\langle \Psi (\tau )\right\vert

610: (a_{B,j}^{\dag }a_{C,j}+a_{B,j}a_{C,j}^{\dag })\left\vert \Psi (\tau

611: )\right\rangle \right\vert

612: \end{equation}%

613: according to refs.\cite{wang,qian}. When the interchain connections are

614: optimized by setting $t_{nB}=t\cos \theta $, $t_{nc}=t\sin \theta $, the

615: above mode concurrence can be calculated as%

616: \begin{equation}

617: C(\tau )=\sin (2\theta )

618: \end{equation}%

619: from the Eq. (\ref{split}).

620: 

621: It is obvious that if $\cos \theta =$ $\sin \theta =1/\sqrt{2}$, i.e.,

622: \begin{equation}

623: t_{nB}=t_{nC}=\frac{t}{\sqrt{2}},  \label{MaxC}

624: \end{equation}%

625: $C(\tau )$ reaches its maximum value $1$. Numerical simulation for $%

626: t_{nB},t_{nC}\in \lbrack 0,2t]$ is performed for a GWP with $\alpha =0.3$

627: and momentum $\pi /2$ in a finite system with $N_{A}=50,$ $N_{B}=50,$ and $%

628: N_{C}=50$. The concurrence is also the function of time $\tau $\ due to the

629: dynamics of the system. We choose maximal concurrence

630: \begin{equation}

631: C_{\max }(t_{nC},t_{nB})=\max \{C(\tau )\}

632: \end{equation}%

633: as a function of $t_{nC}$ and $t_{nB}$ to depict the property of the

634: splitter. Numerical result is plotted in Fig. \ref{C}. It shows that two

635: split wave packets yield the maximal entanglement just at the matching point

636: (\ref{MaxC}).

637: 

638: \subsection{Quantum interferometer for Bloch electron}

639: 

640: Now we consider in details a more complicated TBN than the $Y$-beam, the

641: quantum interferometer for Bloch electron wave. This setup consists of two $%

642: Y $-beams, which is illustrated schematically in Fig. \ref{inter}(a). It is

643: similar to the optical interferometer, where state $\left\vert

644: a\right\rangle $\ of a single photon is split into two parts $\left\vert

645: b\right\rangle $ and $\left\vert c\right\rangle $ by the splitter and then a

646: new state $\left\vert d\right\rangle =U_{B}\left\vert b\right\rangle

647: +U_{C}\left\vert c\right\rangle $ can be achieved by the unitary

648: transformations $U_{B}$\ and $U_{C}$. In the tight-binding Bloch electron

649: interferometer, the analogue of the import state $\left\vert a\right\rangle $%

650: \ is the moving GWP (\ref{GWP}).

651: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

652: \begin{figure}[tbp]

653: \includegraphics[bb=45 330 520 625, width=7 cm,clip]{gsy.eps} %

654: \includegraphics[bb=40 445 530 690, width=7 cm,clip]{gs.eps}

655: \caption{\textit{(Color on line) (a) The interferometric network with an

656: input chain $A$ and output chain $D$, which consists of two $Y$-beams. $%

657: \Delta $ is the \textquotedblleft optical path difference\textquotedblright\

658: which determines the interference pattern of output spin wave. (b) The

659: interference pattern of output wave in the leg $D$ ($r_{0}=50$, $%

660: t_{0}=100/J_{A}$) for the GWP with $\protect\alpha =0.3$ in the

661: interferometric network with $N_{A}=N_{B}$ $=N_{D}=50$, $N_{C}=N_{B}+\Delta $%

662: . }}

663: \label{inter}

664: \end{figure}

665: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

666: 

667: Firstly, we consider the simplest case with the path difference (defined in

668: Fig. \ref{inter}(a)) $\Delta =0$. It is shown that such network is

669: equivalent to two independent virtual chains with lengths $N_{A}+N_{B}+N_{D}$

670: and $N_{B}$ respectively when the coupling matching condition is satisfied.

671: Then the initial GWP will be transmitted into the arm $D$ without any

672: reflection. This fact can be understood according to the interference of two

673: split GWPs. Actually, from the above analysis about the GWP propagating in

674: the $Y$-beam, we note that the conclusion can be extended to the $Y$-beam

675: consisting of two different length arms $N_{B}\neq N_{C}$. It is due to the

676: locality of the GWP and the fact that the speed of the GWP only depends on

677: the hopping integral. Then the arrival time of the two split GWPs at arm $D$

678: depends on the lengths $N_{B}\ $and $N_{C}$.\ It means that the nonzero $%

679: \Delta $ should affect the shape of the pattern of output wave.

680: 

681: To verify the analysis above, we investigate this problem again numerically.

682: According to quantum mechanics, the interference pattern at site $r_{0}$ and

683: time $\tau _{0}$ in arm $D$ can be presented as

684: \begin{equation}

685: I(r_{0},\tau _{0},\Delta )=\left\vert \left\langle r_{0}\right\vert \exp

686: (-iH\tau _{0})\left\vert \Phi (0)\right\rangle \right\vert ^{2}.  \label{I}

687: \end{equation}%

688: Numerical simulation of $I(r_{0},\tau _{0},\Delta )$ for the input GWP in

689: the interferometric network with $N_{A}=N_{B}$ $=N_{D}=50$, $%

690: N_{C}=N_{B}+\Delta $\ is performed. For $r_{0}=50$, $\tau _{0}=100/t$, a

691: perfect interference phenomenon by $I(r_{0},\tau _{0},\Delta )$ is observed

692: for the range $\Delta \in \lbrack -25,25]$ in Fig. \ref{inter}(b). This

693: observation shows that the quantum interferometer can be realized by the TBN.

694: 

695: \section{$Q$-shaped TBN controlled by flux}

696: 

697: From the above discussion, it can be found that the essence of the reduction

698: for the TBN lies on the interference of the matter wave. On the other hand,

699: the presence of vector potential can induce a phase factor to the wave

700: function and then the magnetic flux can control the coherent reduction to

701: some extent. In this section, we investigate how to control the motion of

702: the Bloch electron along this TBN by an external magnetic field. We will

703: show that the appropriate flux through the network can reduce the network to

704: the linear virtual chain, which indicate that the flux can control the

705: propagation of GWP in the network.

706: 

707: \subsection{Model and Hamiltonian}

708: 

709: Consider a quantum network constructed by connecting the two free ends of

710: the two identical chains in $Y$-beam. Such a network is called $Q$-shaped

711: TBN, or QTBN labelled by $\{A,B,C\}$. As illustrated schematically in Fig. %

712: \ref{Q}(a), QTBN is placed in an external magnetic field. The ring of the

713: model is threaded by a magnetic flux $\phi $ in the unit of flux quanta.

714: Here, we only consider effect of the vector potential $\mathbf{A}$\ without

715: the Zeeman term for simply. The Hamiltonian of our $Q$-shaped lattice model

716: reads

717: \begin{equation}

718: H=H_{A}+H_{B}+H_{C}+H_{joint}  \label{H-Q}

719: \end{equation}%

720: where

721: \begin{eqnarray}

722: H_{joint} &=&-(t_{nB}a_{A,M}^{\dag }a_{B,1}e^{i\Phi

723: _{B,1}}+t_{nC}a_{A,M}^{\dag }a_{C,1}e^{-i\Phi _{C,1}})  \notag \\

724: &&-ta_{B,N}^{\dag }a_{C,N}e^{i\Phi _{BC}}+H.c.

725: \end{eqnarray}%

726: and the parameters $N_{A}=M$, $N_{B}=N_{C}=N$, $t_{j}^{[A]}=t$, $%

727: t_{j}^{[B]}=t\exp (i\Phi _{B,j+1})$, $t_{j}^{[C]}=t\exp (-i\Phi _{C,j+1})$.

728: Here, $\Phi _{B,l}$, $\Phi _{C,l}$, $l\in \lbrack 1,,N]$ denote the phase

729: differences between the neighboring sites $l$ and $l+1$ in the chains $B$

730: and $C$, while $\Phi _{BC}$ is respect to the connection between the two

731: chain. The values of the phase difference is defined as (\ref{phase}) and $%

732: \Phi _{B,l}$, $\Phi _{C,l}$, and $\Phi _{BC}$\ are not required to be

733: identical in order to avoid losing generality.

734: 

735: In the following discussion, what we concern is only the sum of the phase

736: difference along the loop%

737: \begin{equation}

738: \Phi =\sum_{l=1}^{N}(\Phi _{B,l}+\Phi _{C,l})+\Phi _{BC}

739: \end{equation}%

740: corresponding to the flux $\phi =\Phi /2\pi $\ through the loop. We will

741: show that the flux $\phi $\ can control the dynamics of the $Q$-shaped

742: lattice system.

743: 

744: \subsection{Reduction of the $Q$-shaped lattice model}

745: 

746: The QTBN with the flux $\phi $, and the joint hopping strengths $t_{nB}$, $%

747: t_{nC}$, exhibits a rich variety of dynamic behaviors. Fortunately, we find

748: that there exist analytical results in some range of parameters. Together

749: with the numerical simulation, these analytical results are helpful to get a

750: comprehensive understanding of the mechanism. We start with the cases of

751: fixed $\Phi $ but various $t_{nB}$, $t_{nC}$ and then investigate the cases,

752: vice versa.%

753: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

754: \begin{figure}[tbp]

755: \includegraphics[bb=60 100 535 760, width=7 cm,clip]{Q.eps}

756: \caption{\textit{(Color on line) (a) The $Q$-shaped Bloch electron network

757: with an input chain $A$ and a ring ${B,C}$ threaded by a magnetic flux. (b)

758: When $\protect\phi =n/2$ and $t_{nB}=t_{nC}=t/\protect\sqrt{2}$, the $Q$%

759: -shaped Bloch electron network can be decoupled into two virtual homogeneous

760: linear chains $a$ and $b$ with length $M+N$ and $N$ respectively. (c) When $%

761: \protect\phi =n/2+1/4$ and $t_{nB}^{2}+t_{nC}^{2}=t^{2}$, the $Q$-shaped

762: Bloch electron network can be decoupled into a long virtual homogeneous

763: linear chain $a$ with length $M+2N$. (d) If $t_{nB}=t_{nC}=t/\protect\sqrt{2}

764: $, as to an arbitrary $\protect\phi $, the virtual homogeneous linear chains

765: $a$ and $b$ are connected by the hopping integral $\widetilde{t}$. There

766: also exist chemical potentials $-\protect\mu $ and $\protect\mu $ at the

767: ends of the virtual chains $a$ and $b$ respectively.}}

768: \label{Q}

769: \end{figure}

770: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

771: 

772: \subsubsection{Case: $\protect\phi =\frac{n}{2}$}

773: 

774: Our aim is try to decouple this $Q$-shaped model as two virtual linear

775: chains.\ We first introduce two anticommutative sets of fermion operator $\{%

776: \widetilde{a}_{a,M+j}^{\dag }\}$, $\{\widetilde{a}_{b,j}^{\dag }\}$\ defined

777: by%

778: \begin{eqnarray}

779: \widetilde{a}_{a,M+j}^{\dag } &=&\cos \theta e^{-i\varphi

780: _{B}^{j}}a_{B,j}^{\dag }+\sin \theta e^{i\varphi _{C}^{j}}a_{C,j}^{\dag }

781: \notag \\

782: \widetilde{a}_{b,j}^{\dag } &=&\sin \theta e^{-i\varphi

783: _{B}^{j}}a_{B,j}^{\dag }-\cos \theta e^{i\varphi _{C}^{j}}a_{C,j}^{\dag }

784: \label{tildeQ}

785: \end{eqnarray}%

786: for $j\in \lbrack 1,N]$, where $\varphi _{\alpha }^{j}=\sum_{l=1}^{j}\Phi

787: _{\alpha ,l}$, $(\alpha =B,C)$.

788: 

789: We can check that they still satisfy the anticommutation relations $\{%

790: \widetilde{a}_{\alpha ,i},\widetilde{a}_{\beta ,j}^{\dag }\}=\delta _{\alpha

791: \beta }\delta _{ij}$, where $\alpha ,\beta \in (a,b)$. The inverse

792: transformations of Eq. (\ref{tildeQ}), together with the original fermion

793: operator $\widetilde{a}_{a,j}^{\dag }=a_{A,j}^{\dag },$ $j\in \lbrack 1,M]$,

794: define a new linear chain $a$, while another virtual linear chain $b$ is

795: only constructed by $\widetilde{a}_{b,j}^{\dag }$, $j\in \lbrack 1,N]$.

796: Therefore, the parameters are taken as

797: 

798: \begin{equation}

799: \phi =\frac{n}{2};t_{nB}=t_{nC}=\frac{t}{\sqrt{2}},

800: \end{equation}%

801: the Hamiltonian can be reduced as%

802: \begin{eqnarray}

803: H &=&\widetilde{H}_{a}+\widetilde{H}_{b}+\widetilde{H}_{\mu }  \notag \\

804: \widetilde{H}_{\mu } &=&t(-1)^{n}\widetilde{n}_{a,M+N}-t(-1)^{n}\widetilde{n}%

805: _{b,N}

806: \end{eqnarray}%

807: with $N_{a}=M+N$ and $N_{b}=N$.\ Here, $\widetilde{n}_{a,M+N}=\widetilde{a}%

808: _{a,M+N}^{\dag }\widetilde{a}_{a,M+N}$ and $\widetilde{n}_{b,N}=\widetilde{a}%

809: _{b,N}^{\dag }\widetilde{a}_{b,N}$ are the particle number operators. The $%

810: \widetilde{H}_{\mu }$\ represents the chemical potential at the ends of

811: chains $a$ and $b$. For large $N$ system, the effect of the end potentials

812: can be ignored. Thus the $Q$-type lattice can be reduced into two

813: independent virtual linear chains $a$ and $b$ with homogeneous NN hopping

814: integrals, and length $M+N$ and $N$ respectively as illustrated in Fig. \ref%

815: {Q}(b).\

816: 

817: \subsubsection{Case: $\protect\phi =\frac{n}{2}+\frac{1}{4}$}

818: 

819: In this case, we will show that the model can be reduced to a virtual chain

820: with $M+2N$ sites if the joint hopping integrals satisfy $\sqrt{%

821: t_{nC}^{2}+t_{nB}^{2}}=t$. We introduce the fermion operators%

822: \begin{eqnarray}

823: \widetilde{a}_{a,j}^{\dag } &=&a_{A,j}^{\dag },  \notag \\

824: \widetilde{a}_{a,M+l}^{\dag } &=&\cos \theta e^{-i\varphi

825: _{B}^{l}}a_{B,l}^{\dag }+\sin \theta e^{i\varphi _{C}^{l}}a_{C,l}^{\dag },

826: \notag \\

827: \widetilde{a}_{a,M+N+l}^{\dag } &=&i(-1)^{n}[\sin \theta e^{-i\varphi

828: _{B}^{l}}a_{B,l}^{\dag }-\cos \theta e^{i\varphi _{C}^{l}}a_{C,l}^{\dag }],

829: \notag \\

830: &&

831: \end{eqnarray}%

832: where $\ j\in \lbrack 1,M]$, $l\in \lbrack 1,N]$. Similarly, when we set%

833: \begin{equation}

834: \phi =\frac{n}{2}+\frac{1}{4};t_{nB}=t\cos \theta ,t_{nC}=t\sin \theta

835: \end{equation}%

836: the Hamiltonian becomes%

837: \begin{equation}

838: H=\widetilde{H}_{a},

839: \end{equation}%

840: with $N_{a}=M+2N,$\ which is illustrated schematically in Fig. \ref{Q}(c).

841: Then we conclude that, when the flux is $\phi =n/2+1/4$ and the joint

842: hopping integrals satisfy $\sqrt{t_{nC}^{2}+t_{nB}^{2}}=t$, the QTBN is

843: equivalent to a single virtual open chain $a$ with length $M+2N$.

844: 

845: \subsubsection{Case: arbitrary $\protect\phi ,$ $t_{nB}=t_{nC}=\frac{t}{%

846: \protect\sqrt{2}}$}

847: 

848: When we take the interchain hopping integrals as $t_{nB}=t_{nC}=t/\sqrt{2}$,

849: the mapping (\ref{tildeQ}) reduces the network Hamiltonian as

850: 

851: \begin{eqnarray}

852: H &=&\widetilde{H}_{a}+\widetilde{H}_{b}+\widetilde{H}_{ab},  \notag

853: \label{film} \\

854: \widetilde{H}_{ab} &=&-\mu (\widetilde{a}_{a,M+N}^{\dag }\widetilde{a}%

855: _{a,M+N}-\widetilde{a}_{b,N}^{\dag }\widetilde{a}_{b,N})  \notag \\

856: &&-\widetilde{t}(\widetilde{a}_{b,N}^{\dag

857: }\widetilde{a}_{a,M+N}+H.c.).

858: \end{eqnarray}%

859: with $N_{a}=M+N$, $N_{b}=N$. We have replaced $i\widetilde{a}_{b,N}^{\dag }$

860: by $\widetilde{a}_{b,N}^{\dag }$ for $j\in \lbrack 1,N]$ without influence

861: on the physics of dynamical process.\ Here, $\widetilde{H}_{a}$ and $%

862: \widetilde{H}_{b}$ stand for two virtual chains with length $M+N$ and $N$

863: respectively, $\widetilde{H}_{ab}$\ represents the chemical potential at the

864: ends of chains $a$ and $b$ and the connection between the two sites. Note

865: that the end-site chemical potentials possess the same magnitude $\mu =t\cos

866: \Phi $, but of opposite sign and the hopping integral between the two end

867: sites is $\widetilde{t}=t\sin \Phi $. The reduced model is also illustrated

868: in Fig. \ref{Q}(d).

869: 

870: Physically, the chemical potentials $\mu $ and the hopping integral $%

871: \widetilde{t}\ $have the complementary relation $\mu ^{2}+\widetilde{t}%

872: ^{2}=t^{2}$. When $\Phi =(n+1/2)\pi $, the network is equivalent to a linear

873: chain with length $M+2N$;\ while for $\Phi =n\pi $ it corresponds to two

874: independent chains with lengths $M+N$ and $N$. In the next section, we will

875: focus on such system for arbitrary $\Phi $. We will show that such system

876: behaves as an optical device, a transmission-reflection film, while the flux

877: determines the coefficient. In conclusion, the magnetic flux $\phi $ can

878: influence the \textquotedblleft effective length\textquotedblright\ or the

879: connective status of the virtual chains and then can be used to control the

880: dynamics of the network.

881: 

882: \subsection{Transmission-reflection film}

883: 

884: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

885: \begin{figure}[tbp]

886: \includegraphics[bb=30 195 560 700, width=7 cm,clip]{half.eps}

887: \caption{\textit{(Color on line) (a) Schematic illustration for the

888: transmission-reflection film in real space. Two terminal sites of the two

889: TBNs $\{A,B\}$ are connected by the hopping integral $\widetilde{t}=t\sin

890: \Phi $. There also exist chemical potentials $\protect\mu =-t\cos \Phi $ for

891: the terminal site of chain $A$ and $\protect\mu =t\cos \Phi $ for the one of

892: chain $B$. (b) The above TBN can be recomposed as a homogeneous chain $%

893: \{a,b\}$ in virtual space.} }

894: \label{half}

895: \end{figure}

896: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

897: 

898: To make the above observation more clear, we consider two identical

899: tight-binding chains $\{A,B\}$, which consist of $N$ sites respectively.

900: There exists a connection interaction between the two terminal sites of the

901: two virtual chains. The hopping constant is $t\sin \Phi $ and each terminal

902: site also has a chemical potential, $\pm t\cos \Phi $. The Hamiltonian reads

903: 

904: \begin{equation}

905: H=H_{A}+H_{B}+H_{joint}  \label{Htr}

906: \end{equation}%

907: where $N_{A}=N_{B}=N$, $t_{j}^{[A]}=t_{j}^{[B]}=t$, and%

908: \begin{eqnarray}

909: H_{joint} &=&-t\sin \Phi (a_{A,N}^{\dag }a_{B,N}+H.c.)  \notag \\

910: &&-t\cos \Phi a_{A,N}^{\dag }a_{A,N}+t\cos \Phi a_{B,N}^{\dag }a_{B,N}.

911: \notag \\

912: &&

913: \end{eqnarray}%

914: Obviously, it is the simplest case of the system described by Eq. (\ref{film}%

915: ).

916: 

917: In order to study the properties of this Bloch electron model more clearly,

918: we introduce two anticommutative sets of fermion operators%

919: \begin{eqnarray}

920: \widetilde{a}_{a,j}^{\dag } &=&\frac{\sqrt{2}}{2}\left( f_{+}a_{A,j}^{\dag

921: }-f_{-}a_{B,j}^{\dag }\right)  \notag \\

922: \widetilde{a}_{b,j}^{\dag } &=&\frac{\sqrt{2}}{2}\left( f_{-}a_{A,j}^{\dag

923: }+f_{+}a_{B,j}^{\dag }\right)  \label{tran}

924: \end{eqnarray}%

925: where $j\in \lbrack 1,N]$ and%

926: \begin{equation}

927: f_{\pm }=\cos \frac{\Phi }{2}\pm \sin \frac{\Phi }{2}.

928: \end{equation}%

929: The inverse transformation of Eq. (\ref{tran}) results in the reduction of

930: the network described by%

931: \begin{equation}

932: H=\widetilde{H}_{a}+\widetilde{H}_{b}+\widetilde{H}_{joint}  \label{ab}

933: \end{equation}%

934: where $N_{a}=N_{b}=N$, and%

935: \begin{equation}

936: \widetilde{H}_{joint}=-t(\widetilde{a}_{a,N}^{\dag }\widetilde{a}%

937: _{b,N}+H.c.).

938: \end{equation}%

939: Obviously, the Hamiltonian (\ref{ab}) depicts an imaginary linear chain with

940: homogeneous couplings $t$ no matter how much the magnitude of the flux $\Phi

941: $\ is taken. Then, the Hamiltonian describing transmission-reflection is

942: mapped into a chain in virtual space. Interestingly, such mapping is

943: irrelevant to the state concerned.

944: 

945: In order to demonstrate the function of such network, we study the

946: propagation of a moving GWP. To this end, we consider a GWP defined as (\ref%

947: {GWP}) at chain $A$, i.e.,%

948: \begin{equation}

949: \left\vert \psi _{A}(N_{0})\right\rangle =\frac{1}{\sqrt{\Omega _{1}}}%

950: \sum_{j=1}^{N}e^{-\frac{\alpha ^{2}}{2}(j-N_{0})^{2}}e^{i\frac{\pi }{2}%

951: j}a_{A,j}^{\dag }\left\vert 0\right\rangle .

952: \end{equation}%

953: We require this wave packet to satisfy $\left\langle 0\right\vert

954: a_{B,j}\left\vert \psi _{A}(N_{0})\right\rangle \simeq 0$,\ so that it

955: ensures the initial GWP being located in chain $A$. The transformation (\ref%

956: {tran}) means that such GWP corresponds to the combination of two GWPs in

957: virtual space with the centers at $N_{0}$ respectively,%

958: \begin{equation}

959: \left\vert \psi _{a(b)}(N_{0})\right\rangle =\frac{f_{+(-)}}{\sqrt{2\Omega

960: _{1}}}\sum_{j=1}^{N}e^{-\frac{\alpha ^{2}}{2}(j-N_{0})^{2}}e^{i\frac{\pi }{2}%

961: j}a_{a(b),j}^{\dag }\left\vert 0\right\rangle .

962: \end{equation}

963: 

964: Based on the analytical result in Ref. \cite{YS1}, the two GWPs in the

965: virtual chain should travel along the chain defined by Eq. (\ref{ab})\

966: without spreading as time evolution. Then, at a certain time $\tau $, the

967: evolved state $\left\vert \phi _{a(b)}(\tau )\right\rangle =$ $\exp (-i%

968: \widetilde{H}\tau )\left\vert \psi _{a(b)}(N_{0})\right\rangle $, or

969: 

970: \begin{eqnarray}

971: \left\vert \phi _{a(b)}(\tau )\right\rangle &=&\frac{1}{\sqrt{2\Omega _{1}}}%

972: [f_{+(-)}\sum_{j=1}^{N}e^{-\frac{\alpha ^{2}}{2}(j-N_{\tau })^{2}}e^{i\frac{%

973: \pi }{2}j}  \notag \\

974: &&+f_{-(+)}\sum_{j=1}^{N}e^{-\frac{\alpha ^{2}}{2}(Pj-N_{\tau })^{2}}e^{i%

975: \frac{\pi }{2}Pj}]a_{a(b),j}^{\dag }\left\vert 0\right\rangle  \notag \\

976: &&

977: \end{eqnarray}%

978: describes the superposition of two GWPs. Here, $Pj=2N+1-j,$ $N_{\tau

979: }=N_{0}+2t\tau $.

980: 

981: Transforming back to the real space, we rewrite the time evolution by the

982: state

983: 

984: \begin{equation}

985: \left\vert \Psi (\tau )\right\rangle =\cos \Phi \left\vert \phi _{A}(\tau

986: )\right\rangle +\sin \left\vert \phi _{B}(\tau )\right\rangle

987: \end{equation}%

988: in terms of the two components of the wave function%

989: \begin{eqnarray}

990: \left\vert \phi _{A}(\tau )\right\rangle &=&\frac{1}{\sqrt{\Omega _{1}}}%

991: \sum_{j=1}^{N}[e^{-\frac{\alpha ^{2}}{2}(Pj-N_{\tau })^{2}}e^{i\frac{\pi }{2}%

992: Pj}  \notag \\

993: &&+\frac{1}{\cos \Phi }e^{-\frac{\alpha ^{2}}{2}(j-N_{\tau })^{2}}e^{i\frac{%

994: \pi }{2}j}]a_{A,j}^{\dag }\left\vert 0\right\rangle  \notag \\

995: &\approx &\frac{1}{\sqrt{\Omega _{1}}}\sum_{j=1}^{N}e^{-\frac{\alpha ^{2}}{2}%

996: (Pj-N_{\tau })^{2}}e^{i\frac{\pi }{2}Pj}a_{A,j}^{\dag }\left\vert

997: 0\right\rangle  \notag \\

998: &&  \label{phia}

999: \end{eqnarray}%

1000: and

1001: 

1002: \begin{equation}

1003: \left\vert \phi _{B}(\tau )\right\rangle =\frac{1}{\sqrt{\Omega _{1}}}%

1004: \sum_{j=1}^{N}e^{-\frac{\alpha ^{2}}{2}(Pj-N_{\tau })^{2}}e^{i\frac{\pi }{2}%

1005: Pj}a_{B,j}^{\dag }\left\vert 0\right\rangle .

1006: \end{equation}%

1007: Here, in Eq. (\ref{phia}), the second term is ignored in the case $%

1008: \left\vert j-N_{\tau }\right\vert \gg 1$.

1009: 

1010: Obviously, the central positions of the final sub-GWPs $2N+1-N_{\tau }$

1011: decrease with time $\tau $. This observation indicates that, the beam

1012: splitter can split the GWP into two cloned GWPs with opposite moving

1013: directions along with $AB$ chain. Therefore, state $\left\vert \phi

1014: _{A}(\tau )\right\rangle $ represents the reflection component with

1015: probability $\cos ^{2}\Phi $, while state $\left\vert \phi _{B}(\tau

1016: )\right\rangle $\ is the transmission component through the connection of $%

1017: AB $\ with probability $\sin ^{2}\Phi $. So this Bloch electron network for

1018: a moving GWP behaves like an optical transmission-reflection film for

1019: photons. Interestingly, transmission and reflection coefficients are

1020: governed by the parameter $\Phi $, the flux through the network. This

1021: feature is illustrated in Fig. \ref{half} in details.

1022: 

1023: \subsection{The dynamic properties of the $Q$-shaped Bloch electron model}

1024: 

1025: Now we take the propagation of the GWP $\left\vert \psi

1026: _{A}(N_{0})\right\rangle $ as an example. Its advantage is that the GWP we

1027: often used can move along a homogeneous chain without spreading

1028: approximately. So we can easily see the various characteristics of the model

1029: through the propagation of the GWP.

1030: 

1031: \subsubsection{Case: $\protect\phi =\frac{n}{2}$, $t_{nB}=t_{nC}=\frac{t}{%

1032: \protect\sqrt{2}}$}

1033: 

1034: The initial GWP is moving with speed $2t$ along chain $A$. Before it reaches

1035: the node, it can also be regarded as moving along the virtual chain $a$.

1036: From the above discussion, the virtual chain $a$ is homogeneous with length $%

1037: M+N$ and decoupled with another virtual chain $b$. Thus in the virtual

1038: space, we can see that\ the GWP moves toward the end site of virtual chain $%

1039: a $ and then reflects at the boundaries with \textquotedblleft $\pi $-phase

1040: shift\textquotedblright . It never appears on virtual chain $b$. Notice

1041: that, in this case, $t_{nB}=t_{nC}=t/\sqrt{2}$ must be satisfied, and then

1042: the GWP in virtual space can be mapped into two identical GWPs with half

1043: amplitude of the initial one in the real space. Therefore, the whole

1044: propagation process in the real space is as follows: When the initial GWP

1045: reaches the node, it is divided into the two identical GWPs which also move

1046: with speed $2t$ along the legs $B$ and $C$ respectively without spreading.

1047: Then the two GWPs reflect completely at the opposite site of the node and

1048: come back along the original paths. When they reach the node again, they

1049: merge as a big GWP and get out of the ring.

1050: 

1051: \subsubsection{Case: $\protect\phi =\frac{n}{2}+\frac{1}{4}$, $%

1052: t_{nB}^{2}+t_{nC}^{2}=t^{2}$.}

1053: 

1054: As it is shown in the subsection (2), the reduction of $Q$-shaped Bloch

1055: electron model have two mainly characters. First and foremost, it is

1056: decoupled as a long virtual chain $a$ with length $M+2N$. Secondly, $%

1057: t_{nB}=t\cos \theta $, $t_{nC}=t\sin \theta $. So when the initial GWP

1058: reaches the node for the first time, it is divided into two GWPs with $%

1059: \left( t_{nB}/t\right) ^{2}$ and $\left( t_{nC}/t\right) ^{2}$ amplitude of

1060: the initial one. They move along the ring for one circle and reach the node

1061: again. This time, instead of going out of the ring to the real chain $A$,

1062: they reflect back and continue moving along the ring for another circle

1063: until they meet at the node for the third time. After circumambulating two

1064: circles, they finally merge into a big one and run out of the ring towards

1065: to the input leg.

1066: 

1067: \subsubsection{Case: arbitrary $\protect\phi $, $t_{nB}=t_{nC}=\frac{t}{%

1068: \protect\sqrt{2}}$}

1069: 

1070: For other values of $\phi $ and $t_{nB},t_{nC}$, when the two GWPs reach the

1071: node for the second time, parts of them get out while the rest parts remain

1072: moving in the ring. Especially, when $t_{nB}=t_{nC}=t/\sqrt{2}$, the

1073: coupling constants and the chemical potentials satisfy the relation

1074: discussed in the section \textquotedblleft Transmission-reflection

1075: film\textquotedblright . So when the initial GWP reaches the end of the

1076: virtual chain $a$, some novel phenomena occur. Part of it can move onto the

1077: virtual chain $b$ and forms a new GWP with $\sin ^{2}\Phi $ amplitude of the

1078: initial one. At the same time, the other part is reflected by the joint and

1079: forms a GWP with $\cos ^{2}\Phi $ amplitude of the initial one. On mapping

1080: them to the real space, we can image that when the two sub-GWPs reach the

1081: node again. Parts of them are merged as a GWP with $\cos ^{2}\Phi $

1082: amplitude getting out of the ring. However, the rest parts move along the

1083: ring for another circle before going out.

1084: 

1085: Therefore, the magnetic flux $\phi $ can control the amplitude of the out

1086: coming GWP. Such an $Q$-shaped Bloch electron model can also be used to test

1087: the flux $\phi $ by measuring the probability of the out coming GWP at some

1088: certain instants.

1089: 

1090: \section{Flux-controlled Interferometer and its reduction}

1091: 

1092: \subsection{Model and Hamiltonian}

1093: 

1094: In this subsection, we consider the interferometer model $\{A,B,C,D\}$ for

1095: Bloch electron, illustrated schematically in Fig. \ref{phai}(a). This

1096: quantum interferometer consists of two chains $A,D$ and one ring $\{B,C\}$

1097: with one end of each chain connecting to two opposite point of the ring. The

1098: ring is threaded by a magnetic flux $\phi $ in the unit of flux quanta. The

1099: Hamiltonian reads

1100: \begin{equation}

1101: H=H_{A}+H_{D}+H_{B}+H_{C}+H_{joint}  \label{Hphi}

1102: \end{equation}%

1103: where $N_{A}=M$, $N_{D}=L$, $N_{B}=$ $N_{C}=N$, $t_{j}^{[A]}=$ $%

1104: t_{j}^{[D]}=t $, $t_{j}^{[B]}=$ $t\exp (i\Phi _{B,j+1})$, and $t_{j}^{[C]}=$

1105: $t\exp (-i\Phi _{C,j+1})$. The connection Hamiltonian reads%

1106: \begin{eqnarray}

1107: H_{joint} &=&-(t_{nAB}a_{A,M}^{\dag }a_{B,1}e^{i\Phi

1108: _{B,1}}+t_{nAC}a_{A,M}^{\dag }a_{C,1}e^{-i\Phi _{C,1}}  \notag \\

1109: &&+t_{nBD}a_{B,N}^{\dag }a_{D,1}e^{i\Phi _{B,N+1}}+  \notag \\

1110: &&t_{nCD}a_{C,N}^{\dag }a_{D,1}e^{-i\Phi _{C,N+1}}+H.c.).

1111: \end{eqnarray}%

1112: Here, $\Phi =$ $\sum_{l=1}^{N+1}\Phi _{B,l}$ $+\sum_{l=1}^{N+1}\Phi _{C,l}$ $%

1113: =2\pi \phi $ is the sum of $\Phi $ along the ring.

1114: 

1115: In this section, we investigate the basic properties of the flux-controlled

1116: interferometer. Similarly, we will find out that there still exist some

1117: analytical results, which reveal the dynamics of such network for the

1118: appropriate parameters.

1119: 

1120: \subsection{Reduction of the interferometer network}

1121: 

1122: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1123: \begin{figure}[tbp]

1124: \includegraphics[bb=55 85 550 780, width=7 cm,clip]{phai.eps}

1125: \caption{\textit{(Color on line) (a) The $\protect\phi $-shaped Bloch

1126: electron network with an input chain $A$, an output chain $D$ and a ring ${%

1127: B,C}$ threaded by a magnetic flux. (b) When $t_{nAB}=t_{nCD}=t\cos \protect%

1128: \theta $, $t_{nAC}=t_{nBD}=t\sin \protect\theta $ and $\protect\phi =n+1/2$,

1129: the $\protect\phi $-shaped Bloch electron network can be decoupled into two

1130: virtual homogeneous linear chains $a$ and $b$ with length $M+N$ and $N+L$

1131: respectively. (c) When $t_{nAB}=t_{nBD}=t\cos \protect\theta $, $%

1132: t_{nAC}=t_{nCD}=t\sin \protect\theta $ and $\protect\phi =n$, the $\protect%

1133: \phi $-shaped Bloch electron network can be decoupled into a long virtual

1134: homogeneous linear chain $a$ with length $M+N+L$ and a short virtual

1135: homogeneous linear chain $b$ with length $N$. (d) If $%

1136: t_{nAB}=t_{nAC}=t_{nBD}=t_{nCD}=t/\protect\sqrt{2}$, for an arbitrary $%

1137: \protect\phi $, the network can be decoupled into three virtual homogeneous

1138: linear chains $a$, $b$ and $c$. They connect with each other by the hopping

1139: integrals $\widetilde{t}_{ac}$ and $\widetilde{t}_{bc}$. }}

1140: \label{phai}

1141: \end{figure}

1142: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1143: To reduce the network of interferometers, the four sets of new fermion

1144: operator%

1145: \begin{eqnarray}

1146: \widetilde{a}_{a,j}^{\dag } &=&a_{A,j}^{\dag },  \notag \\

1147: \widetilde{a}_{a,M+l}^{\dag } &=&\cos \theta e^{-i\varphi

1148: _{B}^{l}}a_{B,l}^{\dag }+\sin \theta e^{i\varphi _{C}^{l}}a_{C,l}^{\dag },

1149: \notag \\

1150: \widetilde{a}_{b,l}^{\dag } &=&\sin \theta e^{-i\varphi

1151: _{B}^{l}}a_{B,l}^{\dag }-\cos \theta e^{i\varphi _{C}^{l}}a_{C,l}^{\dag },

1152: \notag \\

1153: \widetilde{a}_{c,s}^{\dag } &=&e^{i\varphi _{C}^{N+1}}a_{D,s}^{\dag },

1154: \label{Tran2}

1155: \end{eqnarray}%

1156: for $j\in \lbrack 1,M]$,\ $l\in \lbrack 1,N]$,\ and $s\in \lbrack 1,L],$\

1157: are introduced to satisfy%

1158: \begin{equation}

1159: \left\{ \widetilde{a}_{a,M+j},\widetilde{a}_{b,j}^{\dag }\right\} =0.

1160: \end{equation}%

1161: Here,

1162: \begin{equation}

1163: \varphi _{\alpha }^{j}=\sum_{l=1}^{j}\Phi _{\alpha ,l},(\alpha =B,C),

1164: \label{sumphase}

1165: \end{equation}%

1166: is the sum of the phase.

1167: 

1168: The inverse transformation of the above Eqs. (\ref{Tran2}) reduces the

1169: Hamiltonian (\ref{Hphi}) into%

1170: \begin{eqnarray}

1171: H &=&\widetilde{H}_{a}+\widetilde{H}_{b}+\widetilde{H}_{d}  \notag \\

1172: &&-t\sum_{j=1}^{N-1}(\widetilde{a}_{a,M+j}^{\dag }\widetilde{a}%

1173: _{a,M+j+1}+H.c.)  \notag \\

1174: &&-[(t_{nAB}\cos \theta +t_{nAC}\sin \theta )\widetilde{a}_{a,M}^{\dag }%

1175: \widetilde{a}_{a,M+1}  \notag \\

1176: &&+(-t_{nAB}\sin \theta +t_{nAC}\cos \theta )\widetilde{a}_{a,M}^{\dag }%

1177: \widetilde{a}_{b,1}  \notag \\

1178: &&+(t_{nBD}\cos \theta e^{i\Phi }+t_{nCD}\sin \theta )\widetilde{a}%

1179: _{a,M+N}^{\dag }\widetilde{a}_{c,1}  \notag \\

1180: &&+(-t_{nBD}\sin \theta e^{i\Phi }+t_{nCD}\cos \theta )\widetilde{a}%

1181: _{b,N}^{\dag }\widetilde{a}_{c,1}+H.c.]  \notag \\

1182: &&

1183: \end{eqnarray}%

1184: where $N_{a}=M$, $N_{b}=N$, and $N_{c}=L$. Now we concentrate on two special

1185: cases with different $\phi $ and other parameters:

1186: 

1187: \subsubsection{Case: $\protect\phi =n+\frac{1}{2}$, $t_{nAB}=t_{nCD}=t\cos

1188: \protect\theta $, $t_{nAC}=t_{nBD}=t\sin \protect\theta $}

1189: 

1190: It is obvious that $\exp (i\Phi )=-1$. The Hamiltonian can be rewritten as%

1191: \begin{eqnarray}

1192: H &=&-t(\sum_{j=1}^{M+N-1}\widetilde{a}_{a,j}^{\dag }\widetilde{a}%

1193: _{a,j+1}+\sum_{j=1}^{L-1}\widetilde{a}_{c,j}^{\dag }\widetilde{a}_{c,j+1}

1194: \notag \\

1195: &&+\sum_{j=1}^{N-1}\widetilde{a}_{b,j}^{\dag }\widetilde{a}_{b,j+1}+%

1196: \widetilde{a}_{b,N}^{\dag }\widetilde{a}_{c,1}+H.c.).  \notag \\

1197: &&

1198: \end{eqnarray}%

1199: We define the new fermion operator%

1200: \begin{equation}

1201: \widetilde{a}_{b,N+j}^{\dag }=\widetilde{a}_{c,j}^{\dag },(j\in \lbrack 1,L])

1202: \end{equation}%

1203: to extend the virtual chain $b$. Its Hamiltonian

1204: \begin{equation}

1205: H=\widetilde{H}_{a}+\widetilde{H}_{b}  \label{Hphir}

1206: \end{equation}%

1207: is given by the parameters $N_{a}=M+N$ and $N_{b}=N+L$.

1208: 

1209: From the reduced Hamiltonian (\ref{Hphir}), we can see that the

1210: interferometer network is decoupled into two imaginary linear chains with

1211: homogeneous couplings $t$. The set of $\{\widetilde{a}_{a,j}^{\dag }\mid

1212: j\in \lbrack 1,N+M]\}$ defines one of them with length $M+N$ sites, and the

1213: set $\{\widetilde{a}_{a,j}^{\dag }\mid j\in \lbrack 1,N+L]\}$\ defines the

1214: other one with length $N+L$. As illustrated schematically in Fig. \ref{phai}%

1215: (b).

1216: 

1217: \subsubsection{Case: $\protect\phi =n$, $t_{nAB}=t_{nBD}=t\cos \protect%

1218: \theta $, $t_{nAC}=t_{nCD}=t\sin \protect\theta $}

1219: 

1220: In this case, the Hamiltonian becomes%

1221: \begin{eqnarray}

1222: H &=&-t(\sum_{j=1}^{M+N-1}\widetilde{a}_{a,j}^{\dag }\widetilde{a}%

1223: _{a,j+1}+\sum_{j=1}^{L-1}\widetilde{a}_{c,j}^{\dag }\widetilde{a}_{c,j+1}

1224: \notag \\

1225: &&+\sum_{j=1}^{N-1}\widetilde{a}_{b,j}^{\dag }\widetilde{a}_{b,j+1}+%

1226: \widetilde{a}_{a,M+N}^{\dag }\widetilde{a}_{c,1}+H.c.)  \notag \\

1227: &&

1228: \end{eqnarray}%

1229: with the newly defined operators%

1230: \begin{equation}

1231: \widetilde{a}_{a,M+N+j}^{\dag }=\widetilde{a}_{c,j}^{\dag },(j\in \lbrack

1232: 1,L])

1233: \end{equation}%

1234: the reduced Hamiltonian

1235: 

1236: \begin{equation}

1237: H=\widetilde{H}_{a}+\widetilde{H}_{b}

1238: \end{equation}%

1239: describe an extended virtual chain of length $N_{a}=M+N+L$ and another of $%

1240: N_{b}=N$.

1241: 

1242: It is clear that, when the conditions

1243: \begin{eqnarray}

1244: t_{nAB} &=&t_{nBD}=t\cos \theta  \notag \\

1245: t_{nAC} &=&t_{nCD}=t\sin \theta  \notag \\

1246: \phi &=&n

1247: \end{eqnarray}%

1248: are satisfied, the interferometer network is decoupled into two imaginary

1249: linear chains with length $M+N+L$ sites and $N$ sites respectively. See also

1250: Fig. \ref{phai}(c).

1251: 

1252: \subsubsection{Case: arbitrary $\protect\phi $, $t_{nAB}=t_{nAC}=$ $%

1253: t_{nBD}=t_{nCD}$ $=\frac{t}{\protect\sqrt{2}}$}

1254: 

1255: Under this condition, the Hamiltonian is reduced as%

1256: \begin{equation}

1257: H=\widetilde{H}_{a}+\widetilde{H}_{b}+\widetilde{H}_{c}+\widetilde{H}_{joint}

1258: \end{equation}%

1259: where $N_{a}=M+N$, $N_{b}=N$, and $N_{c}=L$.

1260: 

1261: Here, the joint Hamiltonian is%

1262: \begin{eqnarray}

1263: \widetilde{H}_{joint} &=&-t[e^{i\frac{\Phi }{2}}\cos \left( \frac{\Phi }{2}%

1264: \right) \widetilde{a}_{a,M+N}^{\dag }\widetilde{a}_{c,1}  \notag \\

1265: &&-ie^{i\frac{\Phi }{2}}\sin \left( \frac{\Phi }{2}\right) \widetilde{a}%

1266: _{b,N}^{\dag }\widetilde{a}_{c,1}+H.c.].

1267: \end{eqnarray}%

1268: while the sub-Hamiltonians $\widetilde{H}_{a}$, $\widetilde{H}_{b}$ and $%

1269: \widetilde{H}_{c}$ present three homogeneous virtual linear chains $\{a,b,c\}

1270: $ with length $M+N$, $N$ and $L$ respectively. In $\widetilde{H}_{joint}$,

1271: there exists a connection interaction $\exp (i\Phi /2)\cos \left( \Phi

1272: /2\right) $ between the two end sites of virtual chain $a$ and $c$.

1273: Meanwhile, there exists another connection interaction $-i\exp (i\Phi

1274: /2)\sin (\Phi /2)$ between the two end sites of virtual chain $b$ and $c$.

1275: The geometry of such network is illustrated in Fig. \ref{phai}(d). Obviously

1276: in the virtual space, such network is equivalent to the $Y$-shaped beam

1277: splitter with different lengths of output arms and complex joint hopping

1278: constants controlled by the flux $\Phi $. From the discussion about $Y$%

1279: -shaped network, we have known that the lengths of output arms do not affect

1280: the feature as beam splitter for local input wave packet. In the following

1281: we will investigate property of such $Y$-shaped network by considering

1282: equi-length case for simplicity.

1283: 

1284: \subsection{Y-shaped Beam splitter controlled by $\Phi $}

1285: 

1286: Now we consider a $Y$-shaped network $\{A,B,C\}$ with complex joint hopping

1287: constants. The model Hamiltonian reads

1288: 

1289: \begin{equation}

1290: H=H_{A}+H_{B}+H_{C}+H_{joint}

1291: \end{equation}%

1292: where $N_{A}=L$, $N_{B}=N_{C}=N$, and $t_{j}^{[A]}=$ $t_{j}^{[B]}=$ $%

1293: t_{j}^{[C]}=t$. The joint Hamiltonian%

1294: \begin{equation}

1295: H_{joint}=-(t_{AB}a_{A,L}^{\dag }a_{B,1}+t_{AC}a_{A,L}^{\dag

1296: }a_{C,1}+H.c.), \notag

1297: \end{equation}%

1298: describes the connections with the complex hopping integrals

1299: 

1300: \begin{equation}

1301: t_{AB}=te^{-i\frac{\Phi }{2}}\cos \left( \frac{\Phi }{2}\right)

1302: ,t_{AC}=ite^{-i\frac{\Phi }{2}}\sin \left( \frac{\Phi }{2}\right) .

1303: \end{equation}

1304: 

1305: Interestingly, if we get rid of the exponential terms in the hopping

1306: integrals, i.e., $t\cos \left( \Phi /2\right) $, $t\sin \left( \Phi

1307: /2\right) $, we recover the matching condition (\ref{matching}) in the

1308: original $Y$-shaped beam splitter $t_{AB}^{2}+t_{AC}^{2}=t^{2}$. In order to

1309: decouple this network, we introduce three communitative sets of fermion

1310: operators%

1311: \begin{eqnarray}

1312: \widetilde{a}_{a,l}^{\dag } &=&a_{A,l}^{\dag },  \notag \\

1313: \widetilde{a}_{a,L+j}^{\dag } &=&e^{i\frac{\Phi }{2}}[\cos \left( \frac{\Phi

1314: }{2}\right) a_{B,j}^{\dag }-i\sin \left( \frac{\Phi }{2}\right)

1315: a_{C,j}^{\dag }],  \notag \\

1316: \widetilde{a}_{b,j}^{\dag } &=&e^{-i\frac{\Phi }{2}}[i\sin \left( \frac{\Phi

1317: }{2}\right) a_{B,j}^{\dag }+\cos \left( \frac{\Phi }{2}\right) a_{C,j}^{\dag

1318: }],  \notag \\

1319: &&  \label{tran3}

1320: \end{eqnarray}%

1321: for $l\in \lbrack 1,L]$\ and $j\in \lbrack 1,N]$.

1322: 

1323: The inverse transformations of Eqs. (\ref{tran3}) result in the reduction of

1324: Hamiltonian in terms of $\widetilde{a}_{a,j}^{\dag }$, $\widetilde{a}%

1325: _{a,L+j}^{\dag }$, and $\widetilde{a}_{b,j}^{\dag }$:%

1326: \begin{equation}

1327: H=\widetilde{H}_{a}+\widetilde{H}_{b}

1328: \end{equation}%

1329: where $N_{a}=L+N$ and $N_{b}=N$.

1330: 

1331: Thus this kind of $Y$-shaped Bloch electron network is also decoupled into

1332: two imaginary chains. Similarly, we apply the beam splitter to the special

1333: Bloch electron GWP $\left\vert \psi _{A}(N_{0})\right\rangle $. At a certain

1334: time $\tau $, such GWP evolves into%

1335: \begin{eqnarray}

1336: \left\vert \Psi (\tau )\right\rangle &\sim &\cos \left( \frac{\Phi }{2}%

1337: \right) \left\vert \psi _{B}(N_{\tau })\right\rangle  \notag \\

1338: &&-i\sin \left( \frac{\Phi }{2}\right) \left\vert \psi _{C}(N_{\tau

1339: })\right\rangle ,

1340: \end{eqnarray}%

1341: where $N_{\tau }=N_{0}+2t\tau -L$. This means that the beam splitter can

1342: split the GWP into two cloned GWPs completely with the probabilities%

1343: \begin{eqnarray}

1344: \left\vert \left\langle j_{B}\right. \left\vert \psi _{B}(N_{\tau

1345: })\right\rangle \right\vert ^{2} &=&\cos ^{2}\frac{\Phi }{2};  \notag \\

1346: \left\vert \left\langle j_{C}\right. \left\vert \psi _{C}(N_{\tau

1347: })\right\rangle \right\vert ^{2} &=&\sin ^{2}\frac{\Phi }{2},

1348: \end{eqnarray}%

1349: which can be controlled by the external flux $\phi $.

1350: 

1351: \subsection{AB effect in a solid system}

1352: 

1353: This virtual model of interferometer network is very similar to the second

1354: type of $Y$-shaped beam splitter we discussed before. The only difference

1355: between them is that in this virtual model the lengths of the two legs are

1356: unequal. Fortunately, by appropriately choosing $\alpha $, the width of the

1357: wave packet, the GWP can be regarded as a classical electron. It not only

1358: propagates along a homogeneous chain without spreading approximately, but

1359: also does not regard the length of the chain. Now we prepare such a moving

1360: GWP at the input leg. When it reaches to the node, it will be split into two

1361: cloned GWPs with the amplitudes of $\cos ^{2}(\Phi /2)$ and $\sin ^{2}(\Phi

1362: /2)$ respectively. According to our discussion above, the sub-GWP with the

1363: amplitude of $\sin ^{2}(\Phi /2)$ will be reflected by the opposite node of

1364: the ring, but the other sub-GWP with the amplitude of $\cos ^{2}(\Phi /2)$

1365: will move onto the output leg directly. So some time later we will receive a

1366: cloned GWP with the probability $\cos ^{2}(\Phi /2)$ at the output leg.

1367: 

1368: The interferometer based on Bloch electron network can be regarded as a

1369: mimic of AB effect \cite{ABeffect} experimental device in a solid system

1370: illustrated in Fig. \ref{AB}(a). Here, the initial GWP%

1371: \begin{equation}

1372: \left\vert \psi _{A}(N_{0})\right\rangle =\frac{1}{\sqrt{\Omega _{1}}}%

1373: \sum_{j=1}^{M}e^{-\frac{\alpha ^{2}}{2}(j-N_{0})^{2}}e^{i\frac{\pi }{2}%

1374: j}\left\vert j\right\rangle

1375: \end{equation}%

1376: is taken as a good example to demonstrate the physical mechanism of such

1377: setup.

1378: 

1379: We first focus on the GWP at the input site $N_{0}=A$ and detected it later

1380: on a distant site $D_{1}$ or $D_{2}$. The maximal probability of the GWP in

1381: some certain site $j$ is

1382: \begin{equation}

1383: |\psi (j,\alpha )|_{\max }^{2}=\max \{\left\vert \left\langle j\right\vert

1384: e^{-iH\tau }\left\vert \psi _{A}(N_{0})\right\rangle \right\vert ^{2}\}

1385: \end{equation}%

1386: Thus we can define the relative probability $Q$ as a function of $\alpha $,

1387: the magnetic flux $\phi $, and the site of the detector $j$,%

1388: \begin{equation}

1389: Q(j,\phi ,\alpha )=\frac{|\psi (j,\alpha )|_{\max }^{2}}{|\psi (A,\alpha

1390: )|_{\max }^{2}}.

1391: \end{equation}

1392: 

1393: Obviously, $Q$ is an observable physical quantity, which describes the AB

1394: effect and the influence of lattice scattering. Numerical simulation of $%

1395: Q(D_{1},\phi ,\alpha )$ and $Q(D_{2},\phi ,\alpha )$ for different initial

1396: GWP with half-width $\Delta =16.65$, $(\alpha =0.1)$, $\Delta =5.55$, $%

1397: (\alpha =0.3)$, and $\Delta =1$, $(\alpha =\infty )$ are plotted in Fig. \ref%

1398: {AB}(b). Here, the optical paths between the input site and the

1399: detector-sites are $L_{1}=200$, $L_{2}=400$. The ring of the system is

1400: threaded by a magnetic flux $\phi \in \lbrack -2,2]$. The numerical results

1401: show that the relative probabilities $Q$\ are periodic in the magnetic flux $%

1402: \phi $ with a period of unit flux quantum $\Phi _{0}=h/e$. This is the so

1403: called AB effect in a solid system. At $\phi =$integer, our previous

1404: discussion shows that the interferometer network of Bloch electron model is

1405: decoupled into one long chain and one short chain. The initial GWP localized

1406: in the input arm can be transmitted to the detector-arm without any

1407: reflection. Consequently, the relative probability $Q$ reaches its maximum

1408: of the curve. On the other hand, when $\phi $ is a half-integer, the initial

1409: GWP cannot be transmitted to the detector-arm. Thus, the corresponding $Q$

1410: equals to its minimum zero.

1411: 

1412: Then we consider the GWPs with different half-width $\Delta =2\sqrt{\ln 2}%

1413: /\alpha $. If $\Delta $ is larger, the GWP is localized in the linear

1414: dispersion regime more exactly. In this case, it can be well transferred

1415: without spreading \cite{YS1}; or in another point of view, it is a free

1416: particle which will not be scattered by the lattice. We can see from the

1417: numerical results, $Q=$ $(1+\cos \Phi )/2$, the maximum $Q$ of $\Delta

1418: =16.65 $, $(\alpha =0.1)$ is approximately equals to $1$. Otherwise, a GWP

1419: with smaller $\Delta $, i.e., $\Delta =5.55$, $(\alpha =0.3)$ is scattered

1420: by the lattice severely, so the relative probability $Q$ of which are

1421: smaller than $Q$ of wider GWP. It is reasonable that when the optical path

1422: is longer, the influence of lattice scattering is larger, the relative

1423: probability $Q$ is smaller. The black dot dash line also shows that in large

1424: $\alpha $ limit, $Q $ are approximately equal to zero. Therefore, a GWP

1425: localized beyond the linear dispersion regime is not suitable for

1426: demonstrating the AB effect experiment in a solid system.

1427: 

1428: To sum up, when the half-width of the initial wave packet is narrower or the

1429: detect-length is longer, the relative probability $Q$ is smaller, the AB

1430: effect is weaker to be observed. From these results, we get two possible

1431: reasons why we cannot observe AB effect in a macroscopically solid system.

1432: Firstly, we do not choose an appropriate wave packet. Secondly, the total

1433: site of the macroscopically solid system is so large that the influence of

1434: lattice scattering cannot be ignored. To solve these problems and to realize

1435: AB effect in a solid system, we should choose a wider GWP mentioned in our

1436: previous work \cite{YS1}. At the same time, we should decrease the optical

1437: paths between the input site and the detector-sites.

1438: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1439: \begin{figure}[tbp]

1440: \includegraphics[bb=25 430 570 670, width=7 cm,clip]{det.eps} %

1441: \includegraphics[bb=18 436 580 720, width=7 cm,clip]{ip.eps}

1442: \caption{\textit{(Color on line) (a) The schematic illustration for the

1443: mimic A-B effect device in a solid system. (b) The comparison of relative

1444: probability $Q$ as a function of the magnetic flux $\protect\phi $. The

1445: half-width of the initial wave packets are 16.65 ($\protect\alpha =0.1$,

1446: blue solid line), 5.55 ($\protect\alpha =0.3$, red dash line), 1 ($\protect%

1447: \alpha =\infty $, black dot dash line). The optical paths between the input

1448: site $A$ and the detector sites $D_{1}$, $D_{2}$ are $L_{1}=200$ (left), $%

1449: L_{2}=400$ (right). It shows that the relative probability $Q$ are periodic

1450: in the magnetic flux $\protect\phi $ with a period of unit flux quantum $%

1451: \Phi _{0}=h/e$. When the half-width of the initial wave packet is narrower

1452: or the optical path is longer, the relative probability $Q$ is smaller, the

1453: A-B effect is weaker.}}

1454: \label{AB}

1455: \end{figure}

1456: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

1457: 

1458: \section{Applications for spin network}

1459: 

1460: In the above discussion, we studied the fermion systems where the Bloch

1461: electrons move along the quantum lattice network. We consider various

1462: geometrical configurations of TBN that are analogous to quantum optical

1463: devices, such as beam splitters and interferometers. In this section, we

1464: will apply the results obtained for TBN to another analogue system, spin

1465: network (SN) where the spin wave acts as the Bloch electron.

1466: 

1467: The basic setup of a SN is constructed topologically by the linear spin

1468: chains and the various connections between the ends of them. Here, we

1469: consider the spin-$1/2$ $XY$ model, in which only the nearest neighbor (NN)

1470: coupling term is taken into account. The Hamiltonian of a SN reads

1471: 

1472: \begin{equation}

1473: H^{s}=\sum\limits_{l}H_{l}^{s}+H_{joint}^{s},  \label{Hs}

1474: \end{equation}%

1475: \ where the Hamiltonians of leg $l$ consisting of $N_{l}$ spins and the

1476: joints are%

1477: \begin{eqnarray}

1478: H_{l}^{s} &=&H_{l}^{s}(J_{l},N_{l})\circeq

1479: \sum_{j=1}^{N_{l}-1}J_{j}^{[l]}(S_{l,j}^{+}S_{l,j+1}^{-}+H.c.),  \notag \\

1480: H_{joint}^{s} &\equiv &J_{ji}^{[lm]}S_{l,j}^{+}S_{m,i}^{-}+H.c..

1481: \end{eqnarray}%

1482: Here, $S_{l,j}^{\pm }$ are the Pauli spin operators acting on the internal

1483: space of electron on the $j$th site of the $l$th leg. Although the SNs and

1484: TBNs have the same structure, the physics should be different due to the

1485: difference of the intrinsic statistical properties. Then the analytical

1486: conclusions for TBN are not available to SN. However, in the context of

1487: state transfer, only the dynamics of the single-magnon is concerned. Notice

1488: that for Hamiltonian (\ref{Hs}), the $z$-component of the total spin $%

1489: S^{z}=\sum\nolimits_{l,i}S_{l,i}^{z}$ is conserved, i.e. $[S^{z},H^{s}]=0$.

1490: Thus in the invariant subspace with $S^{z}=(\sum\nolimits_{l}N_{l}-1)/2$,

1491: this model can be mapped into single-particle TBN.

1492: 

1493: \section{Conclusion and remark}

1494: 

1495: In summary, we studied various geometrical configurations of tight-binding

1496: networks\ for the fermion systems. It is found that the lattice networks for

1497: moving GWPs are analogous to quantum optical devices, such as beam splitters

1498: and interferometers. In practice, our coherent quantum network for

1499: electronic wave can be implemented by an array of quantum dots, Josephson

1500: junctions or other artificial atoms. It will enable an elementary quantum

1501: device for scalable quantum computation, which can coherently transfer

1502: quantum information among the integrated qubits.\textbf{\ }The observable

1503: effects for electronic wave interference may be discovered in the dynamics

1504: of magnetic domain in some artificial quantum material.

1505: 

1506: In the above studies, we only consider the spinless\ Bloch electron.

1507: Actually, all the conclusions we obtained can be extended to the networks of

1508: spin-$1/2$ electrons, if the external magnetic field does not exert any

1509: forces or torques on the magnetic moment of spin, but only a phase on the

1510: wave function of electron. The Hamiltonian of such system has the similar

1511: form with (\ref{h}) and (\ref{joint}) under the transformation $a_{j}^{\dag

1512: }a_{i}\longrightarrow \sum_{\sigma =\pm 1}a_{j,\sigma }^{\dag }a_{i,\sigma }$%

1513: . Note that, for the new Hamiltonian, the spin of electron is a conservative

1514: quantity that cannot be influenced during the propagation \cite{YS1}. The

1515: electronic wave packet with spin polarization is an analogue of photon

1516: \textquotedblleft flying qubit\textquotedblright , where the quantum

1517: information was encoded in its two polarization states. Thus, these networks

1518: can function as some optical devices, such as beam splitters and

1519: interferometers. These are expected to be used as quantum information

1520: processors for the fermion system based on the possible engineered solid

1521: state systems, such as the array of quantum dots, Josephson junctions or

1522: other artificial atoms that can be implemented experimentally.

1523: 

1524: This work is supported by the NSFC with grant Nos. 90203018, 10474104 and

1525: 60433050; and by the National Fundamental Research Program of China with

1526: Nos. 2001CB309310 and 2005CB724508.

1527: 

1528: \begin{thebibliography}{99}

1529: \bibitem{QIP1} D. Bouwnmeester, A. Ekert, A. Zeilinger(Eds.),

1530: \textquotedblleft The Physics of Quantum Information\textquotedblright ,

1531: Springer, Berlin, 2000.

1532: 

1533: \bibitem{QIP2} M.A. Nielsen , I.L. Chuang, \textquotedblright Quantum

1534: Computation and Quantum Information\textquotedblright . Cambridge University

1535: Press, Cambridge, U.K. 2000.

1536: 

1537: \bibitem{QST1} S. Bose, Phys. Rev. Lett. \textbf{91}, 207901 (2003); S.

1538: Bose, B-Q. Jin and V. E. Korepin, Phys. Rev. A \textbf{72}, 022345 (2005);

1539: S. Bose, Phys. Rev. Lett. \textbf{91}, 207901 (2003); M-H. Yung and S. Bose,

1540: Phys. Rev. A \textbf{71}, 032310 (2005).

1541: 

1542: \bibitem{QST2} V. Subrahmanyam, Phys. Rev. A \textbf{69}, 034304 (2004).

1543: 

1544: \bibitem{QST3} M. Christandl, N. Datta, A. Ekert and A. J. Landahl, Phys.

1545: Rev. Lett. \textbf{92}, 187902 (2004)

1546: 

1547: \bibitem{QST4} C. Albanese, M. Christandl, N. Datta and A. Ekert, Phys. Rev.

1548: Lett. \textbf{93}, 230502 (2004)

1549: 

1550: \bibitem{QST5} T. J. Osborne and N. Linden, Phys. Rev. A \textbf{69}, 052315

1551: (2004).

1552: 

1553: \bibitem{LY} Y. Li, T. Shi, B. Chen, Z. Song, C.P. Sun, Phys. Rev. A \textbf{%

1554: 71}, 022301 (2005).

1555: 

1556: \bibitem{ST} T. Shi, Y. Li, Z. Song, and C.P. Sun, Phys. Rev. A \textbf{71},

1557: 032309 (2005)

1558: 

1559: \bibitem{SZ} Z. Song, C.P. Sun, Low Temperature Physics \textbf{31}, 686

1560: (2005).

1561: 

1562: \bibitem{QST6} M.B. Plenio, F. L. Semiao, New. J. Phys. \textbf{7}, 73 (2005)

1563: 

1564: \bibitem{quant-network} J. E. Avron; A. Raveh and B. Zur, Rev. Mod. Phys.

1565: \textbf{60}, 873 (1988); C. H. Wu, G. Mahler, Phys. Rev. B \textbf{43}, 5012

1566: (1991); J. Vidal, G. Montambaux and B. Doucot, Phys. Rev. B \textbf{62},

1567: R16294 (2000); P. S. Deo and A. M. Jayannavar, Phys. Rev. B \textbf{50},

1568: 11629 (1994).

1569: 

1570: \bibitem{spin-network} A. Kay and M. Ericsson, New J. Phys. \textbf{7}, 143

1571: (2005); G. D. Chiara, R. Fazio, C. Macchiavello, S. Montangero and G. M.

1572: Palma, Phys. Rev. A \textbf{72}, 012328 (2005); M. Paternostro, G. M. Palma,

1573: M. S. Kim and G. Falci, Phys. Rev. A \textbf{71}, 042311 (2005).

1574: 

1575: \bibitem{Experiment} R. A. Webb, S. Washburn, C. P. Umbach, and R. B.

1576: Laibowitz, Phys. Rev. Lett. \textbf{54}, 2696 (1985); V. Chandrasekhar, M.

1577: J. Rooks, S. Wind, and D. E. Prober, Phys. Rev. Lett. \textbf{55}, 1610

1578: (1985).

1579: 

1580: \bibitem{Plenio} M.B. Plenio, J. Hartley and J. Eisert, New J. Phys. \textbf{%

1581: 6}, 36 (2004); A Perales, M.B. Plenio, J. Opt. B \textbf{7},

1582: S601-S609 (2005).

1583: 

1584: \bibitem{gauge-trans} N. Byers and C.N. Yang, Phys. Rev. Lett. \textbf{7},

1585: 46 (1961);\ H. T. Nieh, G. Su, B-H. Zhao, Phys. Rev. B \textbf{51}, 3760

1586: (1995).

1587: 

1588: \bibitem{tight-binding} D. Langbein, Phys. Rev. \textbf{180}, 633 (1969);

1589: G-Y. Oh, J. Korean Phys. Soc. 42, 714 (2003).

1590: 

1591: \bibitem{Peierls} R. Peierls, Z. Physik \textbf{80}, 763 (1933).

1592: 

1593: \bibitem{YS1} S. Yang, Z. Song, and C.P. Sun, Phys. Rev. A \textbf{73},

1594: 022317 (2006).

1595: 

1596: \bibitem{QOP} R. Loudon, \textit{The quantum theory of light,} (Oxford,

1597: 2000); M.O. Scully and M.S. Zubairy, Quantum Optics, (Oxford, 1997).

1598: 

1599: \bibitem{S-photon} J.D. Franson; Phys. Rev. A \textbf{56}, 1800-1805 (1997);

1600: K. Jacobs and P.L. Knight; Phys. Rev. A \textbf{54}, R3738(1996); T. Wang,

1601: M. Kostrun, and S.F. Yelin; Phys. Rev. A \textbf{70}, 053822 (2004).

1602: 

1603: \bibitem{Entng} M. Zukowski, A. Zeilinger, and M.A. Horne, Phys. Rev. A

1604: \textbf{55}, 2564(1997); J.L. van Velsen, Phys. Rev. A \textbf{72}, 012334

1605: (2005).

1606: 

1607: \bibitem{N-interfer} H. Rauch, W. Treimer, and U. Bonse, Phys. Lett. A

1608: \textbf{47}, 369 (1974).

1609: 

1610: \bibitem{C-atom} D. Cassettari, B. Hessmo, R. Folman, T. Maier, and J.

1611: Schmiedmayer, Phys. Rev. Lett. \textbf{85}, 5483(2000); U. V. Poulsen and K.

1612: Momer, Phys. Rev. A \textbf{65}, 033613 (2002); D. C. E. Bortolotti and J.L.

1613: Bohn; Phys. Rev. A \textbf{69}, 033607 (2004).

1614: 

1615: \bibitem{BEC} F. Burgbacher and J. Audretsch; Phys. Rev. A \textbf{60},

1616: R3385(1999); N.M. Bogoliubov, A. G. Izergin, N.A. Kitanine, A.G.

1617: Pronko, and J. Timonen; Phys. Rev. Lett. \textbf{86}, 4439 (2001)

1618: 

1619: \bibitem{wang} X. Wang and P. Zanardi, Phys. Lett. A \textbf{301}, 1 (2002);

1620: X. Wang, Phys. Rev. A \textbf{66}, 034302 (2002).

1621: 

1622: \bibitem{qian} X-F. Qian, Y. Li, Y. Li, Z. Song, and C.P. Sun, Phys. Rev. A

1623: \textbf{72}, 062329 (2005).

1624: 

1625: \bibitem{ABeffect} Y. Aharonov and D. Bohm, Phys. Rev. \textbf{115}, 485

1626: (1959).

1627: \end{thebibliography}

1628: 

1629: \end{document}

1630: