1:
2: %\documentclass[aps,twocolumn]{revtex4}
3: \documentclass[aps,prb,twocolumn,superscriptaddress,floatfix,amsmath]{revtex4}
4:
5: \usepackage{graphicx}% Include figure files
6: %\usepackage{dcolumn}% Align table columns on decimal point
7: %\usepackage{bm}% bold math
8: %\usepackage{natbib}
9: %\usepackage{color}
10: \usepackage{amsmath}
11: \usepackage{amssymb}
12: %\usepackage{ulem}
13:
14: %\setlength{\textheight}{28cm}
15:
16: \renewcommand{\vec}[1]{\boldsymbol{#1}}
17: %\newcommand{\strike}[1]{\textcolor{blue}{\sout{#1}}}
18: %\newcommand{\comment}[1]{\textcolor{red}{\texttt{#1}}}
19:
20: \begin{document}
21:
22:
23: \title{Doping and temperature dependence of electron spectrum and
24: quasiparticle dispersion in doped bilayer cuprates}
25:
26:
27: \author{Yu Lan}
28: \affiliation{Department of Physics, Beijing Normal University,
29: Beijing 100875, China}
30:
31: \author{Jihong Qin}
32: \affiliation{Department of Physics, Beijing University of Science
33: and Technology, Beijing 100083, China}
34:
35: \author{Shiping Feng}
36: \affiliation{Department of Physics, Beijing Normal University,
37: Beijing 100875, China}
38:
39: %\date{\today}
40:
41: \begin{abstract}
42: Within the $t$-$t'$-$J$ model, the electron spectrum and
43: quasiparticle dispersion in doped bilayer cuprates in the normal
44: state are discussed by considering the bilayer interaction. It is
45: shown that the bilayer interaction splits the electron spectrum of
46: doped bilayer cuprates into the bonding and antibonding components
47: around the $[\pi,0]$ point. The differentiation between the bonding
48: and antibonding components is essential, which leads to two main
49: flat bands around the $[\pi,0]$ point below the Fermi energy. In
50: analogy to the doped single layer cuprates, the lowest energy states
51: in doped bilayer cuprates are located at the $[\pi/2,\pi/2]$ point.
52: Our results also show that the striking behavior of the electronic
53: structure in doped bilayer cuprates is intriguingly related to the
54: bilayer interaction together with strong coupling between the
55: electron quasiparticles and collective magnetic excitations.
56:
57: \end{abstract}
58:
59: %\pacs{71.20.Eh, 75.30.Ds, 75.50.Ee}
60:
61: \maketitle
62:
63: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
64:
65: \section{Introduction}
66:
67: Cuprate materials are unusual in that the undoped cuprates are
68: insulators with an antiferromagnetic (AF) long-range-order (AFLRO),
69: and changing the carrier concentration by ionic substitution or
70: increase of the oxygen content turns these compounds into correlated
71: metals leaving short-range AF correlations still intact
72: \cite{bedell,kastner,shen,campuzano,shen1}. The single common
73: feature of cuprate superconductors is the presence of the
74: two-dimensional CuO$_{2}$ plane
75: \cite{bedell,kastner,shen,campuzano,shen1}, and it seems evident
76: that the unusual behaviors of doped cuprates are dominated by the
77: CuO$_{2}$ plane \cite{anderson}. This layered crystal structure
78: leads to that cuprate superconductors are highly anisotropic
79: materials, then the electron spectral function $A({\bf k},\omega)$
80: is dependent on the in-plane momentum
81: \cite{bedell,kastner,shen,campuzano,shen1}. After twenty years
82: extensive studies, it has been shown that many of the unusual
83: physical properties have often been attributed to particular
84: characteristics of low energy excitations determined by the
85: electronic structure \cite{shen,campuzano,shen1}.
86:
87: Although the electronic structure of doped cuprates is
88: well-established by now \cite{shen,campuzano,shen1}, its full
89: understanding is still a challenging issue. Experimentally,
90: angle-resolved photoemission spectroscopy (ARPES) experiments have
91: provided rather detailed information on the electronic structure of
92: the doped single layer and bilayer cuprates
93: \cite{shen,campuzano,shen1,shen2,dessau,dfeng,kordyuk,chuang,campuzano1}.
94: An important issue is whether the behavior of the low energy
95: excitations determined by the electronic structure is universal or
96: not. The early ARPES measurements \cite{shen1,shen2,dessau} showed
97: that the charge carriers doped into the parent compounds first enter
98: into the ${\bf k}= [\pi/2,\pi/2]$ (in units of inverse lattice
99: constant) point in the Brillouin zone. Moreover, the electron
100: spectral function $A({\bf k},\omega)$ has a flat band form as a
101: function of energy $\omega$ for momentum ${\bf k}$ in the vicinity
102: of the $[\pi,0]$ point, which leads to the unusual quasiparticle
103: dispersion around the $[\pi,0]$ point with anomalously small changes
104: of electron energy as a function of momentum
105: \cite{shen1,shen2,dessau}. Furthermore, this flat band is just below
106: the Fermi energy. Recently, the improvements in the resolution of
107: the ARPES experiments \cite{shen,campuzano,dfeng,kordyuk,chuang}
108: allowed to resolve additional features in the electron spectral
109: function $A({\bf k},\omega)$. Among these new achievements is the
110: observation of the bilayer splitting (BS) in doped bilayer cuprates
111: in a wide doping range \cite{shen,dfeng,kordyuk,chuang}. In this
112: case, whether the electronic structure of doped cuprates can be
113: influenced by the interaction between CuO$_{2}$ planes has been an
114: interesting issue. The study of the electronic structure is
115: complicated by the BS, that is, the BS of the CuO$_{2}$ plane
116: derived the electronic structure in the bonding and antibonding
117: bands due to the present of CuO$_{2}$ bilayer blocks in the unit
118: cell of doped bilayer cuprates \cite{dfeng,kordyuk,chuang}. The
119: magnitude of the BS is the doping independent, and increases upon
120: approaching the $[\pi,0]$ point, where the BS exhibits the largest
121: value. As a result of the maximal BS at the $[\pi,0]$ point, there
122: are two main flat bands around the $[\pi,0]$ point
123: \cite{dfeng,kordyuk,chuang}. Moreover, it has been shown that a well
124: pronounced peak-dip-hump structure in the electron spectrum of doped
125: bilayer cuprates in the superconducting state is partially caused by
126: the BS \cite{kordyuk1,borisenko}. Theoretically, the
127: electron-removal spectral functions at the $[\pi,0]$ point and
128: quasiparticle dispersion of the multilayer cuprates in the normal
129: state has been discussed \cite{mori} within the two-dimensional
130: multilayer $t$-$t'$-$t''$-$J$ model based on the resonating valence
131: bond wave function and Gutzwiller approximation, in particular,
132: their results show that there are two main flat bands around the
133: $[\pi,0]$ point in the bilayer system. However, the resonating
134: valence bond wave function and Gutzwiller approximation can be only
135: applied to discuss the zero temperature physical properties of doped
136: cuprates. To the best of our knowledge, the doping and temperature
137: dependence of the electron spectrum and quasiparticle dispersion in
138: both doped single layer and bilayer cuprates have not been treated
139: from a unified point of view for the normal state.
140:
141: In our earlier work using the charge-spin separation (CSS)
142: fermion-spin theory \cite{feng}, the electronic structure of the
143: doped single layer cuprates has been calculated within the single
144: layer $t$-$t'$-$J$ model \cite{guo}, and the obtained doping
145: dependence of the electron spectrum and quasiparticle dispersion are
146: consistent with the corresponding ARPES experiments
147: \cite{shen1,shen2,dessau}. In this paper we show explicitly if the
148: bilayer interaction is included, one can reproduce some main
149: features in the normal state observed experimentally on doped
150: bilayer cuprates \cite{shen,dfeng,kordyuk,chuang}. The
151: differentiation between the bonding and antibonding components is
152: essential. Two main flat bands around the $[\pi,0]$ point are rather
153: similar to the scenario argued in Refs. \onlinecite{dfeng},
154: \cite{kordyuk}, and \cite{mori} for the normal state, and the
155: quasiparticle dispersion we derive from the bilayer $t$-$t'$-$J$
156: model (without additional terms and adjustable parameters)
157: demonstrates explicitly this energy band splitting. In comparison
158: with the case of the doped single layer cuprates \cite{guo}, our
159: present results also show that the striking behavior of the
160: electronic structure in doped bilayer cuprates is intriguingly
161: related to the bilayer interaction together with strong coupling
162: between the electron quasiparticles and collective magnetic
163: excitations.
164:
165: The rest of this paper is organized as follows. In Sec. II, we
166: introduce the bilayer $t$-$t'$-$J$ model that include the bilayer
167: hopping and bilayer magnetic exchange interaction. Within this
168: bilayer $t$-$t'$-$J$ model, we calculate explicitly the longitudinal
169: and transverse components of the electron Green's functions based on
170: the CSS fermion-spin theory by considering charge carrier
171: fluctuations, and then obtain the bonding and antibonding electron
172: spectral functions according to these longitudinal and transverse
173: components of the electron Green's functions. The doping and
174: temperature dependence of the electron spectrum and quasiparticle
175: dispersion of doped bilayer cuprates in the normal state are
176: presented in Sec. III. Sec. IV is devoted to a summary and
177: discussions.
178:
179: \section{Theoretical Framework}
180:
181: The basic element of cuprate materials is two-dimensional CuO$_{2}$
182: plane, and it is believed that the unusual behaviors of doped
183: cuprates are closely related to the doped CuO$_{2}$ plane as
184: mentioned above \cite{bedell,kastner,shen,campuzano,shen1}. It has
185: been argued that the essential physics of the doped CuO$_{2}$ plane
186: is contained in the $t$-$t'$-$J$ model on a square lattice
187: \cite{anderson}. In this case, the $t$-$t'$-$J$ model in the bilayer
188: structure is expressed as,
189: \begin{eqnarray}
190: H&=&-t\sum_{i\hat{\eta}a\sigma}C^{\dagger}_{ia\sigma}
191: C_{i+\hat{\eta}a\sigma}+t'\sum_{i\hat{\tau}a\sigma}
192: C^{\dagger}_{ia\sigma}C_{i+\hat{\tau}a\sigma}\nonumber\\
193: &-&\sum_{i\sigma}
194: t_{\perp}(i)(C^{\dagger}_{i1\sigma}C_{i2\sigma}+H.c.)\
195: +\mu\sum_{ia\sigma}C^{\dagger}_{ia\sigma}C_{ia\sigma}\nonumber\\
196: &+&J\sum_{i\hat{\eta}a}{\bf S}_{ia} \cdot {\bf S}_{i+\hat{\eta}a}
197: +J_{\perp}\sum_{i}{\bf S}_{i1} \cdot {\bf S}_{i2},
198: \end{eqnarray}
199: where $\hat{\eta}=\pm\hat{x},\pm\hat{y}$, $\hat{\tau}=\pm\hat{x}
200: \pm\hat{y}$, $a=1,2$ is plane index, $C^{\dagger}_{ia\sigma}$
201: ($C_{ia\sigma}$) is the electron creation (annihilation) operator,
202: ${\bf S}_{ia}= C^{\dagger}_{ia}{\vec\sigma} C_{ia}/2$ is the spin
203: operator with ${\vec\sigma}=(\sigma_{x},\sigma_{y},\sigma_{z})$ as
204: Pauli matrices, $\mu$ is the chemical potential, and the interlayer
205: hopping \cite{chakarvarty},
206: \begin{eqnarray}
207: t_{\perp}({\bf k})={t_{\perp}\over 4}(\cos k_{x} -\cos k_{y})^{2},
208: \end{eqnarray}
209: describes coherent hopping between the CuO$_{2}$ planes. This
210: functional form of the interlayer hopping in Eq. (2) is predicted on
211: the basis of the local density approximation calculations
212: \cite{chakarvarty}, and later the experimental observed BS agrees
213: well with it \cite{shen,dfeng,kordyuk,chuang}. The $t$-$t'$-$J$
214: Hamiltonian (1) is supplemented by the single occupancy local
215: constraint $\sum_\sigma C_{ia\sigma}^{\dagger}C_{ia\sigma}\leq 1$.
216: This local constraint can be treated properly in analytical
217: calculations within the CSS fermion-spin theory \cite{feng}, where
218: the constrained electron operators in the $t$-$J$ type model are
219: decoupled as $C_{ia\uparrow}= h^{\dagger}_{ia\uparrow}S^{-}_{ia}$
220: and $C_{ia\downarrow}= h^{\dagger}_{ia\downarrow}S^{+}_{ia}$, with
221: the spinful fermion operator $h_{ia\sigma}= e^{-i\Phi_{ia\sigma}}
222: h_{ia}$ keeps track of the charge degree of freedom together with
223: some effects of the spin configuration rearrangements due to the
224: presence of the doped hole itself (dressed holon), while the spin
225: operator $S_{ia}$ keeps track of the spin degree of freedom. The
226: advantage of this CSS fermion-spin theory is that the electron local
227: constraint for single occupancy,
228: $\sum_{\sigma}C^{\dagger}_{ia\sigma}
229: C_{ia\sigma}=S^{+}_{ia}h_{ia\uparrow}h^{\dagger}_{ia\uparrow}
230: S^{-}_{ia}+S^{-}_{ia}h_{ia\downarrow}h^{\dagger}_{ia\downarrow}
231: S^{+}_{ia}=h_{ia}h^{\dagger}_{ia}(S^{+}_{ia}S^{-}_{ia}+S^{-}_{ia}
232: S^{+}_{ia})=1-h^{\dagger}_{ia} h_{ia}\leq 1$, is satisfied in
233: analytical calculations, and the double spinful fermion occupancies
234: $h^{\dagger}_{ia\sigma} h^{\dagger}_{ia-\sigma}=
235: e^{i\Phi_{ia\sigma}}h^{\dagger}_{ia} h^{\dagger}_{ia}
236: e^{i\Phi_{ia-\sigma}}=0$ and $h_{ia\sigma} h_{ia-\sigma}=
237: e^{-i\Phi_{ia\sigma}}h_{ia}h_{ia} e^{-i\Phi_{ia-\sigma}}=0$, are
238: ruled out automatically. It has been shown \cite{feng} that these
239: dressed holons and spins are gauge invariant, and in this sense they
240: are real and can be interpreted as the physical excitations
241: \cite{laughlin}. Although in common sense $h_{ia\sigma}$ is not a
242: real spinful fermion, it behaves like a spinful fermion. In this CSS
243: fermion-spin representation, the low-energy behavior of the bilayer
244: $t$-$t'$-$J$ Hamiltonian (1) can be expressed as,
245: \begin{eqnarray}
246: H&=&t\sum_{i\hat{\eta}a}(h^{\dagger}_{i+\hat{\eta}a\uparrow}
247: h_{ia\uparrow}S^{+}_{ia}S^{-}_{i+\hat{\eta}a}+
248: h^{\dagger}_{i+\hat{\eta}a\downarrow}h_{ia\downarrow}S^{-}_{ia}
249: S^{+}_{i+\hat{\eta}a})\nonumber\\
250: &-&t'\sum_{i\hat{\tau}a} (h^{\dagger}_{i+\hat{\tau}a\uparrow}
251: h_{ia\uparrow}S^{+}_{ia} S^{-}_{i+\hat{\tau}a}+
252: h^{\dagger}_{i+\hat{\tau}a\downarrow}h_{ia\downarrow}
253: S^{-}_{ia}S^{+}_{i+\hat{\tau}a})\nonumber\\
254: &+&\sum_{i}t_{\perp}(i)
255: (h^{\dagger}_{i2\uparrow}h_{i1\uparrow}S^{+}_{i1}S^{-}_{i2}+
256: h^{\dagger}_{i1\uparrow}h_{i2\uparrow}S^{+}_{i2}S^{-}_{i1}\nonumber\\
257: &+&h^{\dagger}_{i2\downarrow}h_{i1\downarrow}S^{-}_{i1} S^{+}_{i2}
258: +h^{\dagger}_{i1\downarrow}h_{i2\downarrow}S^{-}_{i2}
259: S^{+}_{i1})-\mu\sum_{ia\sigma}h^{\dagger}_{ia\sigma}h_{ia\sigma}\nonumber\\
260: &+&{J_{\rm eff}}\sum_{i\hat{\eta}a}{\bf S}_{ia}\cdot {\bf
261: S}_{i+\hat{\eta}a}+ {J_{\rm eff\perp}}\sum_{i}{\bf S}_{i1}\cdot {\bf
262: S}_{i2},
263: \end{eqnarray}
264: where $J_{\rm eff}=J(1-\delta)^{2}$, $J_{\rm eff\perp}=J_{\perp}
265: (1-\delta)^{2}$, and $\delta=\langle h^{\dagger}_{ia\sigma}
266: h_{ia\sigma}\rangle=\langle h^{\dagger}_{ia} h_{ia}\rangle$ is the
267: hole doping concentration. As a result, the magnetic energy term in
268: the bilayer $t$-$t'$-$J$ model is only to form an adequate spin
269: configuration \cite{anderson1}, while the kinetic energy part has
270: been expressed as the interaction between the dressed holons and
271: spins, and then this interaction dominates the essential physics of
272: doped cuprates \cite{feng}.
273:
274: Within the CSS fermion-spin theory, the electron spectral function
275: and related quasiparticle dispersion of the doped single layer
276: cuprates in the normal state have been discussed by considering the
277: dressed holon fluctuations around the mean-field (MF) solution
278: \cite{guo}, where the spin part is limited to the first-order (the
279: MF level), while the full dressed holon Green's function is treated
280: self-consistently in terms of the Eliashberg's strong coupling
281: theory \cite{eliashberg}. For doped bilayer cuprates, there are two
282: coupled CuO$_{2}$ planes in the unit cell. This leads to that the
283: energy spectrum has two branches, then the dressed holon and spin
284: Green's functions are matrices, and can be expressed as $g({\bf k},
285: \omega)=g_{L}({\bf k},\omega)+\sigma_{x} g_{T}({\bf k}, \omega)$ and
286: $D({\bf k},\omega)=D_{L}({\bf k}, \omega)+\sigma_{x} D_{T}({\bf k},
287: \omega)$, respectively, where the longitudinal and transverse parts
288: are corresponding to the in-layer and inter-layer Green's functions
289: \cite{feng1}. In this case, we follow the previous discussions for
290: the single layer case \cite{guo}, and obtain explicitly the
291: longitudinal and transverse parts of the full dressed holon Green's
292: functions of the doped bilayer system as [see the Appendix],
293: \begin{subequations}
294: \begin{eqnarray}
295: g_{L}({\bf k},\omega)&=&{1\over 2}\sum_{\nu=1,2}{Z^{(\nu)}_{FA}
296: \over\omega-\bar{\xi}_{\nu{\bf k}}}, \\
297: g_{T}({\bf k},\omega)&=&{1\over 2}\sum_{\nu=1,2}(-1)^{\nu+1}{
298: Z^{(\nu)}_{FA }\over \omega-\bar{\xi}_{\nu{\bf k}}},
299: \end{eqnarray}
300: \end{subequations}
301: with the renormalized dressed holon excitation spectrum
302: ${\bar\xi}_{\nu{\bf k}}=Z^{(\nu)}_{FA}\xi_{\nu{\bf k}}$, where the
303: MF dressed holon excitation spectrum is given as,
304: \begin{eqnarray}
305: \xi_{\nu{\bf k}}=Zt\chi_{1}\gamma_{\bf{k}}-Zt'\chi_{2}
306: \gamma'_{\bf{k}}-\mu+(-1)^{\nu+1}\chi_{\perp}t_{\perp}({\bf{k}}),
307: \end{eqnarray}
308: with the spin correlation functions $\chi_{1}=\langle S_{ia}^{+}
309: S_{i+\hat{\eta}a}^{-} \rangle$, $\chi_{2}=\langle S_{ia}^{+}
310: S_{i+\hat{\tau}a}^{-} \rangle$, $\chi_{\perp}=\langle S^{+}_{i1}
311: S^{-}_{i2}\rangle$, $\gamma_{\bf{k}}=(1/Z)\sum_{\hat{\eta}}
312: {e^{i{\bf{k}}\cdot{\hat{\eta}}}}$, $\gamma'_{\bf{k}}=(1/Z)
313: \sum_{\hat{\tau}}{e^{i{\bf{k}} \cdot{\hat{\tau}}}}$, and $Z$ is the
314: number of the nearest neighbor or second-nearest neighbor sites,
315: while the quasiparticle coherent weights $Z^{(1)-1}_{FA}=
316: Z^{-1}_{F1}- Z^{-1}_{F2}$ and $Z^{(2)-1}_{FA}=Z^{-1}_{F1}+
317: Z^{-1}_{F2}$, with $Z^{-1}_{F1}=1- \Sigma^{(ho)}_{L}({\bf k}_{0},
318: \omega)\mid_{\omega=0}$, $Z^{-1}_{F2}=\Sigma^{(ho)}_{T}({\bf k}_{0},
319: \omega) \mid_{\omega=0}$, and ${\bf k}_{0}=[\pi/2,\pi/2]$, where
320: $\Sigma^{(ho)}_{L}({\bf k},\omega)$ and $\Sigma^{(ho)}_{T}({\bf k},
321: \omega)$ are the corresponding antisymmetric parts of the
322: longitudinal and transverse dressed holon self-energy functions,
323: \begin{subequations}
324: \begin{eqnarray}
325: &\Sigma^{(h)}_{L}&({\bf k},i\omega_{n})={1\over N^{2}}\sum_{\bf
326: p,q}[R^{(1)}_{{\bf{p+q+k}}}{1\over\beta}\sum_{ip_{m}}g_{L}({\bf
327: p+k},ip_{m}+i\omega_{n})\nonumber\\
328: &\times&\Pi_{LL}({\bf p},{\bf q},ip_{m})+R^{(2)}_{{\bf p+q+k}}
329: {1\over\beta}\sum_{ip_{m}}g_{T}({\bf p+k},ip_{m}+i\omega_{n})
330: \nonumber\\
331: &\times& \Pi_{TL}({\bf p},{\bf q},ip_{m})],\\
332: &\Sigma^{(h)}_{T}&({\bf k},i\omega_{n})={1\over N^{2}}\sum_{\bf p,
333: q}[R^{(1)}_{{\bf{p+q+k}}}{1\over\beta} \sum_{ip_{m}}g_{T}({\bf
334: p+k},ip_{m}+i\omega_{n})\nonumber \\
335: &\times&\Pi_{TT}({\bf p},{\bf q},ip_{m})+R^{(2)}_{{\bf p+q+k}}
336: {1\over\beta}\sum_{ip_{m}}g_{L}({\bf p+k}, ip_{m} +i\omega_{n})
337: \nonumber\\
338: &\times&\Pi_{LT}({\bf p},{\bf q},ip_{m})],
339: \end{eqnarray}
340: \end{subequations}
341: with $R^{(1)}_{{\bf k}}=[Z(t\gamma_{\bf k}-t'\gamma'_{\bf k})]^{2}
342: +t_{\perp}^{2}({\bf k})$, $R^{(2)}_{{\bf k}}=2Z(t\gamma_{\bf k}-t'
343: \gamma'_{\bf k})t_{\perp}({\bf k})$, and the spin bubbles
344: $\Pi_{\eta,\eta'}({\bf p},{\bf q},ip_{m})=(1/\beta)\sum_{iq_{m}}
345: D^{(0)}_{\eta}({\bf q},iq_{m}) D^{(0)}_{\eta'}({\bf q+p},
346: iq_{m}+ip_{m})$, where $\eta=L,T$, $\eta'=L,T$, and the longitudinal
347: and transverse parts of the MF spin Green's functions,
348: \begin{subequations}
349: \begin{eqnarray}
350: D^{(0)}_{L}({\bf k},\omega)&=&{1\over 2}\sum_{\nu=1,2}
351: {B_{\nu{\bf k}}\over \omega^{2}-\omega^{2}_{\nu{\bf k}}}, \\
352: D^{(0)}_{T}({\bf k},\omega)&=&{1\over 2}
353: \sum_{\nu=1,2}(-1)^{\nu+1}{B_{\nu{\bf k}}\over \omega^{2}-
354: \omega^{2}_{\nu{\bf k}}},
355: \end{eqnarray}
356: \end{subequations}
357: with $B_{\nu {\bf k}}=\lambda (2\epsilon_{\parallel}\chi_{1}^{z}
358: +\chi_{1})\gamma_{\bf{k}}-2\lambda{'}\chi_{2}^{z}\gamma'_{\bf{k}}
359: -J_{\rm eff \perp}[\chi_{\perp}+2\chi_{\perp}^{z}(-1)^{\nu}]
360: \epsilon_{\perp}({\bf{k}})-\lambda(\epsilon_{\parallel}\chi_{1}+
361: 2\chi_{1}^{z})+\lambda'\chi_{2}-J_{\rm eff\perp}[2\chi_{\perp}^{z}
362: +\chi_{\perp}(-1)^{\nu}]$, $\lambda=2ZJ_{\rm eff}$, $\lambda'=4Z
363: \phi_{2}t'$, $\epsilon_{\parallel}=1+2t\phi_{1}/J_{\rm eff}$,
364: $\epsilon_{\perp}({\bf{k}})=1+4\phi_{\perp}t_{\perp}({\bf{k}})
365: /J_{\rm eff\perp}$, the spin correlation functions
366: $\chi_{1}^{z}=\langle S_{ia}^{z} S_{i+\hat{\eta}a}^{z}\rangle$,
367: $\chi_{2}^{z}=\langle S_{ia}^{z} S_{i+\hat{\tau}a}^{z}\rangle$,
368: $\chi^{z}_{\perp}=\langle S_{i1}^{z} S_{i2}^{z}\rangle$, the dressed
369: holon particle-hole order parameters $\phi_{1}=\langle
370: h^{\dagger}_{ia\sigma} h_{i+\hat{\eta}a\sigma}\rangle$,
371: $\phi_{2}=\langle h^{\dagger}_{ia\sigma}
372: h_{i+\hat{\tau}a\sigma}\rangle$, $\phi_{\perp}=\langle
373: h^{\dagger}_{i1\sigma}h_{i2\sigma}\rangle$, and the MF dressed spin
374: excitation spectrum,
375: \begin{eqnarray}
376: \omega^{2}_{\nu{\bf k}}&=&\lambda^{2}[(A_{2}-\alpha
377: \epsilon_{\parallel}\chi_{1}^{z}\gamma_{\bf k}-{1\over 2Z}\alpha
378: \epsilon_{\parallel}\chi_{1})(1-\epsilon_{\parallel}\gamma_{\bf k})
379: \nonumber\\
380: &+&{1 \over 2}\epsilon_{\parallel}(A_{1}-{2\over Z}\alpha
381: \chi_{1}^{z}-\alpha\chi_{1}\gamma_{\bf k})(\epsilon_{\parallel}
382: -\gamma_{\bf k})]\nonumber\\
383: &+&\lambda'^{2}[\alpha(\chi_{2}^{z}\gamma'_{\bf k}-{Z-1\over 2Z}
384: \chi_{2})\gamma'_{\bf k}+{1 \over 2}(A_{3}-{2 \over Z}\alpha
385: \chi_{2}^{z})]\nonumber\\
386: &+&\lambda\lambda'\alpha[\chi_{1}^{z}(1-
387: \epsilon_{\parallel}\gamma_{\bf k})\gamma'_{\bf k}+{1\over
388: 2}(\chi_{1}\gamma'_{\bf k}-C_{2})(\epsilon_{\parallel} -\gamma_{\bf k})
389: \nonumber\\
390: &+&\gamma'_{\bf k}(C_{2}^{z}- \epsilon_{\parallel}
391: \chi_{2}^{z}\gamma_{\bf k})-{1 \over 2}\epsilon_{\parallel}
392: (C_{2}-\chi_{2}\gamma_{\bf k})] \nonumber\\
393: &+&\lambda J_{\rm eff \perp}\alpha[{1 \over 2}C_{\perp}
394: \epsilon_{\perp}({\bf k})(\epsilon_{\parallel}-\gamma_{\bf k})+{1
395: \over 2}\epsilon_{\parallel}\epsilon_{\perp}({\bf k})(C_{\perp}-
396: \chi_{\perp}\gamma_{\bf k})\nonumber\\
397: &+&C_{\perp}^{z}(1-\epsilon_{\parallel} \gamma_{\bf
398: k})+(C_{\perp}^{z}-\epsilon_{\parallel}\chi_{\perp}^{z} \gamma_{\bf
399: k})\nonumber\\
400: &+&(-1)^{\nu}({1 \over 2}\chi_{1}\epsilon_{\perp} ({\bf
401: k})(\epsilon_{\parallel}-\gamma_{\bf k})+{1 \over 2}
402: \epsilon_{\parallel}(C_{\perp}-\chi_{\perp}\gamma_{\bf k})\nonumber\\
403: &+&\chi_{1}^{z}\epsilon_{\perp}({\bf k})(1-\epsilon_{\parallel}
404: \gamma_{\bf k})+\epsilon_{\perp}({\bf
405: k})(C_{\perp}^{z}-\epsilon_{\parallel} \chi_{\perp}^{z}\gamma_{\bf
406: k}))]\nonumber\\
407: &+&\lambda' J_{\rm eff \perp} \alpha[\epsilon_{\perp}({\bf k})({1
408: \over 2}\chi_{\perp} \gamma'_{\bf
409: k}-C'_{\perp})+(C'^{z}_{\perp}+\chi_{\perp}^{z})
410: \gamma'_{\bf k}\nonumber\\
411: &+&(-1)^{\nu}(\epsilon_{\perp}({\bf k})(\chi_{2}^{z}+
412: \chi_{\perp}^{z})\gamma'_{\bf k}-{1 \over 2}(\chi_{2}
413: \epsilon_{\perp}({\bf k})-\chi_{\perp}\gamma'_{\bf
414: k})\nonumber\\
415: &-&{1\over 2} C'_{\perp})]+{1\over 4}J_{\rm eff \perp}^{2}
416: [\epsilon_{\perp}({\bf k})+ (-1)^{\nu}]^{2},
417: \end{eqnarray}
418: where $A_{1}=\alpha C_{1}+(1-\alpha)/2Z$, $A_{2}=\alpha C_{1}^{z}+
419: (1-\alpha)/4Z$, $A_{3}=\alpha C_{3}+(1-\alpha)/2Z$, and the spin
420: correlation functions $C_{1}=(1/Z^{2})
421: \sum_{\hat{\eta}\hat{\eta'}}\langle S_{i+\hat{\eta}a}^{+}
422: S_{i+\hat{\eta'}a}^{-}\rangle$, $C_{2}=(1/Z^{2})
423: \sum_{\hat{\eta}\hat{\tau}}\langle S_{i+\hat{\eta}a}^{+}
424: S_{i+\hat{\tau}a}^{-}\rangle$, $C_{3}=(1/Z^{2})
425: \sum_{\hat{\tau}\hat{\tau'}}\langle S_{i+\hat{\tau}a}^{+}
426: S_{i+\hat{\tau'}a}^{-}\rangle$, $C_{1}^{z}=(1/Z^{2})
427: \sum_{\hat{\eta}\hat{\eta'}}\langle S_{i+\hat{\eta}a}^{z}
428: S_{i+\hat{\eta'}a}^{z}\rangle$, $C_{2}^{z}=(1/Z^{2})
429: \sum_{\hat{\eta}\hat{\tau}}\langle S_{i+\hat{\eta}a}^{z}
430: S_{i+\hat{\tau}a}^{z}\rangle$, $C_{\perp}=(1/Z)\sum_{\hat{\eta}}
431: \langle S_{i1}^{+} S_{i+\hat{\eta}2}^{-}\rangle$,
432: $C{'}_{\perp}=(1/Z)\sum_{\hat{\tau}} \langle S_{i1}^{+}
433: S_{i+\hat{\tau}2}^{-}\rangle$, $C_{\perp}^{z}=(1/Z)
434: \sum_{\hat{\eta}}\langle S_{i1}^{z}S_{i+ \hat{\eta}2}^{z}\rangle$,
435: and $C{'}_{\perp}^{z}=(1/Z) \sum_{\hat{\tau}}\langle S_{i1}^{z}S_{i+
436: \hat{\tau}2}^{z}\rangle$. In order not to violate the sum rule of
437: the correlation function $\langle S^{+}_{i}S^{-}_{i}\rangle=1/2$ in
438: the case without AFLRO, the important decoupling parameter $\alpha$
439: has been introduced in the above MF calculation, which can be
440: regarded as the vertex correction \cite{kondo}.
441:
442: For discussions of the electronic structure of doped bilayer
443: cuprates in the normal state, we need to calculate the electron
444: Green's function $G(i-j,t-t')=\langle\langle C_{i\sigma}(t);
445: C^{\dagger}_{j\sigma} (t')\rangle\rangle=G_{L}(i-j,t-t')
446: +\sigma_{x}G_{T}(i-j,t-t')$, which is a convolution of the spin
447: Green's function and dressed holon Green's function in the CSS
448: fermion-spin theory, and reflect the charge-spin recombination
449: \cite{anderson1}. In the present bilayer system, the longitudinal
450: and transverse parts of the electron Green's function can be
451: obtained explicitly in terms of the corresponding MF spin Green's
452: functions (7) and full dressed holon Green's functions (4) as,
453: \begin{subequations}
454: \begin{eqnarray}
455: &&G_{L}({\bf k},\omega)={1\over 8N}\sum_{\bf p}\sum_{\mu\nu}
456: Z_{FA}^{(\mu)}{B_{\nu{\bf p}}\over \omega_{\nu{\bf p}}}\nonumber\\
457: &&\times\left ({L^{(1)}_{\mu\nu}({\bf
458: k,p})\over\omega+\bar{\xi}_{\mu{\bf p-k}} -\omega_{\nu{\bf
459: p}}}+{L^{(2)}_{\mu\nu}({\bf k,p})\over\omega+
460: \bar{\xi}_{\mu{\bf p-k}}+\omega_{\nu{\bf p}}}\right ),\\
461: &&G_{T}({\bf k},\omega)={1\over 8N}\sum_{\bf p}\sum_{\mu\nu}
462: (-1)^{\mu+\nu}Z_{FA}^{(\mu)}{B_{\nu{\bf p}}\over\omega_{\nu{\bf p}
463: }}\nonumber\\
464: &&\times\left ({L^{(1)}_{\mu\nu}({\bf k,p})\over\omega+
465: \bar{\xi}_{\mu{\bf p-k}}-\omega_{\nu{\bf p}}}+
466: {L^{(2)}_{\mu\nu}({\bf k,p})\over\omega+ \bar{\xi}_{\mu{\bf p-k}}+
467: \omega_{\nu{\bf p}}}\right ),
468: \end{eqnarray}
469: \end{subequations}
470: where $L^{(1)}_{\mu\nu}({\bf k,p})=n_{F}(\bar{\xi}_{\mu{\bf p-k}})
471: +n_{B}(\omega_{\nu{\bf p}})$, $L^{(2)}_{\mu\nu}({\bf k,p})=1-n_{F}
472: (\bar{\xi}_{\mu{\bf p-k}})+n_{B}(\omega_{\nu{\bf p}})$, and
473: $n_{B}(\omega)$ and $n_{F}(\omega)$ are the boson and fermion
474: distribution functions, respectively, then the longitudinal and
475: transverse electron spectral functions $A_{L}({\bf k},\omega)=-
476: 2{\rm Im}G_{L}({\bf k},\omega)$ and $A_{T}({\bf k}, \omega) =-
477: 2{\rm Im}G_{T}({\bf k},\omega)$ are obtained from the above
478: longitudinal and transverse electron Green's functions as,
479: \begin{subequations}
480: \begin{eqnarray}
481: A_{L}({\bf k},\omega)&=&\pi{1\over 4N}\sum_{\bf p}\sum_{\mu\nu}
482: Z_{FA}^{(\mu)}{B_{\nu{\bf p}}\over\omega_{\nu{\bf p}}}\nonumber\\
483: &\times&[L^{(1)}_{\mu\nu}({\bf k,p})\delta(\omega+\bar{\xi}_{\mu{\bf
484: p-k}} -\omega_{\nu{\bf p}})\nonumber\\
485: &&+L^{(2)}_{\mu\nu}({\bf k,p})\delta(\omega+
486: \bar{\xi}_{\mu{\bf p-k}}+\omega_{\nu{\bf p}})], \\
487: A_{T}({\bf k},\omega)&=&\pi{1\over 4N}\sum_{\bf p}\sum_{\mu\nu}
488: (-1)^{\mu+\nu}Z_{FA}^{(\mu)}{B_{\nu{\bf p}}\over\omega_{\nu{\bf p}}}
489: \nonumber\\
490: &\times&[L^{(1)}_{\mu\nu}({\bf k,p})\delta(\omega+\bar{\xi}_{\mu{\bf
491: p-k}}-\omega_{\nu{\bf p}})\nonumber\\
492: &&+L^{(2)}_{\mu\nu}({\bf k,p})\delta(\omega+ \bar{\xi}_{\mu{\bf
493: p-k}}+\omega_{\nu{\bf p}})].
494: \end{eqnarray}
495: \end{subequations}
496: With the help of the above longitudinal and transverse electron
497: spectral functions, the bonding and antibonding electron spectral
498: functions of doped bilayer cuprates are obtained as,
499: \begin{subequations}
500: \begin{eqnarray}
501: A^{+}({\bf k},\omega)&=&{1\over 2}[A_{L}({\bf k},\omega)+A_{T}
502: ({\bf k},\omega)],\\
503: A^{-}({\bf k},\omega)&=&{1\over 2}[A_{L}({\bf k},\omega)-A_{T} ({\bf
504: k},\omega)],
505: \end{eqnarray}
506: \end{subequations}
507: respectively.
508:
509: \section{Electron spectrum and quasiparticle dispersion of doped
510: bilayer cuprates}
511:
512: We are now ready to discuss the doping and temperature dependence of
513: the electron spectrum and quasiparticle dispersion of bilayer
514: cuprate superconductors in the normal state. We have performed a
515: calculation for the electron spectral functions in Eq. (11), and the
516: results of the bonding (solid line) and antibonding (dashed line)
517: electron spectral functions at (a) the $[\pi,0]$ point and (b)
518: $[\pi/2,\pi/2]$ point for parameters $t/J=2.5$, $t'/t=0.15$, and
519: $t_{\perp}/t=0.3$ with temperature $T=0.1J$ at the doping
520: concentration $\delta=0.15$ are plotted in Fig. 1. Apparently, there
521: is a double-peak structure in the electron spectral function around
522: the $[\pi,0]$ point, i.e., the bonding and antibonding quasiparticle
523: peaks around the $[\pi,0]$ point are located at the different
524: positions, while the bonding and antibonding peaks around the
525: $[\pi/2,\pi/2]$ point are located at the same position, which leads
526: to that the BS appears around the $[\pi,0]$ point, and is absent
527: from the vicinity of the $[\pi/2,\pi/2]$ point. In particular, the
528: positions of the antibonding peaks at the $[\pi,0]$ point are more
529: closer to the Fermi energy than these for the bonding peaks. In this
530: sense, the differentiation between the bonding and antibonding
531: components of the electron spectral function around the $[\pi,0]$
532: point is essential. In analogy to the single layer cuprates
533: \cite{guo}, both positions of the quasiparticle peaks from the
534: bonding and antibonding electron spectral functions at the $[\pi,0]$
535: and $[\pi/2,\pi/2]$ points are below the Fermi energy, but the
536: positions of the peaks at the $[\pi/2,\pi/2]$ point are more closer
537: to the Fermi energy, which indicates that the lowest energy states
538: are located at the $[\pi/2,\pi/2]$ point, in other words, the low
539: energy spectral weight with the majority contribution to the
540: low-energy properties of doped bilayer cuprates comes from the
541: $[\pi/2,\pi/2]$ point, in qualitative agreement with the ARPES
542: experimental data on doped bilayer cuprates
543: \cite{shen,dfeng,kordyuk,chuang}. The double-peak structure in the
544: electron spectral functions around the $[\pi,0]$ point is closely
545: related to the interlayer hopping form in Eq. (2). With decreasing
546: the values of $t_{\perp}$ and $J_{\perp}$, the distance between the
547: bonding and antibonding peaks in the electron spectral functions
548: decreases. When $t_{\perp}=0$ and $J_{\perp}=0$, we find that the
549: transverse part of the dressed holon Green's functions in Eq. (4b)
550: (then the transverse part of the electron Green's functions in Eq.
551: (9b) and transverse part of the electron spectral functions in Eq.
552: (10b)) is equal to the zero. In this case, the bonding electron
553: spectral function in Eq. (11a) is exactly same as the antibonding
554: electron spectral function in Eq. (11b), then the electron spectral
555: functions are reduced to the doped single layer case \cite{guo}.
556:
557: \begin{figure}
558: \includegraphics[scale=0.65]{fig1.eps}
559: \caption{The bonding (solid line) and antibonding (dashed line)
560: electron spectral functions as a function of energy at (a) the
561: $[\pi,0]$ point and (b) $[\pi/2,\pi/2]$ point for $t/J=2.5$,
562: $t'/t=0.15$, and $t_{\perp}/t=0.3$ with $T=0.1J$ at $\delta=0.15$.}
563: \end{figure}
564:
565: For a better understanding of the physical properties of the
566: electron spectrum in doped bilayer cuprates, we have studied the
567: electron spectrum at different doping concentrations, and the result
568: of the electron spectral functions at $[\pi,0]$ point for $t/J=2.5$,
569: $t'/t=0.15$, and $t_{\perp}/t=0.3$ with $T=0.1J$ at $\delta=0.09$
570: (solid line), $\delta=0.12$ (dashed line), and $\delta=0.15$ (dotted
571: line) are plotted in Fig. 2, which indicates that with increasing
572: the doping concentration, both bonding and antibonding quasiparticle
573: peaks become sharper, and the spectral weights of these peaks
574: increase in intensity. Furthermore, we have also discussed the
575: temperature dependence of the electron spectrum, and the result of
576: the electron spectral functions at $[\pi,0]$ point for $t/J=2.5$,
577: $t'/t=0.15$, and $t_{\perp}/t=0.3$ at $\delta=0.15$ with $T=0.1J$
578: (solid line), $T=0.05J$ (dashed line), and $T=0.01J$ (dotted line)
579: are plotted in Fig. 3, it is shown obviously that both bonding and
580: antibonding spectral weights are suppressed with increasing
581: temperatures. Our these results are also qualitatively consistent
582: with the ARPES experimental results on doped bilayer cuprates
583: \cite{shen,dfeng,kordyuk,chuang}.
584:
585: \begin{figure}
586: \includegraphics[scale=0.56]{fig2.eps}
587: \caption{The electron spectral functions at $[\pi,0]$ point for
588: $t/J=2.5$, $t'/t=0.15$, and $t_{\perp}/t=0.3$ with $T=0.1J$ at
589: $\delta=0.09$ (solid line), $\delta=0.12$ (dashed line), and
590: $\delta=0.15$ (dotted line).}
591: \end{figure}
592:
593: \begin{figure}
594: \includegraphics[scale=0.56]{fig3.eps}
595: \caption{The electron spectral functions at $[\pi,0]$ point for
596: $t/J=2.5$, $t'/t=0.15$, and $t_{\perp}/t=0.3$ at $\delta=0.15$ with
597: $T=0.1J$ (solid line), $T=0.05J$ (dashed line), and $T=0.01J$
598: (dotted line).}
599: \end{figure}
600:
601: For considering the quasiparticle dispersion of doped bilayer
602: cuprates, we have made a series of calculations for both bonding and
603: antibonding electron spectral functions at different momenta, and
604: find that the lowest energy peaks are well defined at all momenta.
605: In particular, the positions of both bonding and antibonding
606: quasiparticle peaks as a function of energy $\omega$ for momentum
607: ${\bf k}$ in the vicinity of the $[\pi,0]$ point are almost not
608: changeable, which leads to the unusual quasiparticle dispersion
609: around the $[\pi,0]$ point. To show this broad feature clearly, we
610: plot the positions of the lowest energy quasiparticle peaks in the
611: bonding and antibonding electron spectra as a function of momentum
612: along the high symmetry directions with $T=0.1J$ at $\delta=0.15$
613: for $t/J=2.5$, $t'/t=0.15$, and $t_{\perp}/t=0.3$ in Fig. 4. For
614: comparison, the corresponding result from the tight binding fit to
615: the experimental data of the doped bilayer cuprate
616: Bi$_{2}$Sr$_{2}$CaCu$_{2}$O$_{8+\delta}$ \cite{kordyuk} is also
617: shown in Fig. 2 (inset). Our result shows that in analogy to the
618: doped single layer cuprates \cite{guo}, both electron bonding and
619: antibonding quasiparticles around the $[\pi,0]$ point disperse very
620: weakly with momentum, and then the two main flat bands appear, while
621: the Fermi energy is only slightly above these flat bands. Moreover,
622: this bilayer energy band splitting reaches its maximum at the
623: $[\pi,0]$ point. Our this result shows that the bilayer interaction
624: has significant contributions to the electronic structure of doped
625: bilayer cuprates, and is in qualitative agreement with these
626: obtained from the ARPES experimental measurements on doped bilayer
627: cuprates \cite{shen,dfeng,kordyuk,chuang}. Within the
628: $t$-$t'$-$t''$-$J$ model, this quasiparticle dispersion of doped
629: bilayer cuprates has been studied based on the resonating valence
630: bond wave function and Gutzwiller approximation \cite{mori}, where
631: the quasiparticle dispersion relations are degenerate along the
632: $[0,0]$ to $[\pi,\pi]$ direction, and the splitting becomes maximum
633: at the $[\pi,0]$ point, which is consistent with our present result.
634:
635: \begin{figure}
636: \includegraphics[scale=0.50]{fig4.eps}
637: \caption{The positions of the lowest energy quasiparticle peaks in
638: the bonding (solid line) and antibonding (dashed line) electron
639: spectra as a function of momentum with $T=0.1J$ at $\delta=0.15$ for
640: $t/J=2.5$, $t'/t=0.15$, and $t_{\perp}/t=0.3$. Inset: the
641: corresponding result from the tight binding fit to the experimental
642: data of the doped bilayer cuprate
643: Bi$_{2}$Sr$_{2}$CaCu$_{2}$O$_{8+\delta}$ taken from Ref.
644: \protect\onlinecite{kordyuk}.}
645: \end{figure}
646:
647: The essential physics of the double-peak structure of the electron
648: spectral function around the $[\pi,0]$ point in the present doped
649: bilayer cuprates is dominated by the bilayer interaction. The full
650: electron Green's function in doped bilayer cuprates is divided into
651: the longitudinal and transverse parts, respectively, due to the
652: bilayer interaction, then these longitudinal and transverse Green's
653: function (then the bonding and antibonding electron spectral
654: functions and corresponding quasiparticle dispersions) are obtained
655: beyond the MF approximation by considering the dressed holon
656: fluctuation due to the spin pair bubble, therefore the nature of the
657: bonding and antibonding electron spectral functions of doped bilayer
658: cuprates in the normal state are closely related to the strong
659: interaction between the dressed holons (then electron
660: quasiparticles) and collective magnetic excitations. In this case,
661: the single-particle hoppings in the bilayer $t$-$t'$-$J$ model are
662: strongly renormalized by the magnetic interaction. As a consequence,
663: both bonding and antibonding quasiparticle bandwidths are reduced to
664: the order of (a few) $J$, and then the energy scales of both bonding
665: and antibonding quasiparticle energy bands are controlled by the
666: magnetic interaction. These renormalizations for both bonding and
667: antibonding energy bands are then responsible for the unusual
668: bonding and antibonding electron quasiparticle spectra and
669: production of the two main flat bands around the $[\pi,0]$ point. As
670: in the single layer case, our present results also show that the
671: electron quasiparticle excitations in doped bilayer cuprates
672: originating from the dressed holons and spins are due to the
673: charge-spin recombination, this reflects the composite nature of the
674: electron quasiparticle excitations, and then the unconventional
675: normal-state properties in doped bilayer cuprates are attributed to
676: the presence of the dressed holons, spins, electron quasiparticle
677: excitations, and bilayer coupling.
678:
679: \section{Summary and discussions}
680:
681: In conclusion we have shown that if the strong dressed holon-spin
682: coupling and bilayer interaction are taken into account, the
683: $t$-$t'$-$J$ model per se can correctly reproduce some main features
684: of the ARPES electron spectrum and quasiparticle dispersion of doped
685: bilayer cuprates in the normal state. In our opinion, the
686: differentiation between the bonding and antibonding components is
687: essential, which leads to two main flat bands around the $[\pi,0]$
688: point below the Fermi energy. In analogy to the doped single layer
689: cuprates, the lowest energy states in doped bilayer cuprates are
690: located at the $[\pi/2,\pi/2]$ point. Our results also show that the
691: striking behavior of the electronic structure in doped bilayer
692: cuprates is intriguingly related to the bilayer interaction together
693: with strong coupling between the electron quasiparticles and
694: collective magnetic excitations.
695:
696: The experimental \cite{tanaka} and theoretical
697: \cite{chakarvarty,mori} analysis shows that the physical properties
698: for different families of doped cuprates are strongly correlated
699: with $t'$ and $t_{\perp}$, this means that the different values of
700: parameters $t/J$, $t'/t$, and $t_{\perp}/t$ should be chosen for
701: different families of doped cuprates. For comparison, the values of
702: parameters $t/J$ and $t'/t$ in the present work have been chosen as
703: the same as in the single layer case \cite{guo}. Furthermore, the
704: value of parameter $t_{\perp}/t$ in the present work is
705: qualitatively consistent with these in the local density
706: approximation calculations \cite{chakarvarty} based on the bilayer
707: Hubbard model and the calculations within the bilayer
708: $t$-$t'$-$t''$-$J$ model based on the resonating valence bond wave
709: function and Gutzwiller approximation \cite{mori}. Although the
710: simplest bilayer $t$-$t'$-$J$ model can not be regarded as a
711: complete model for the quantitative comparison with doped bilayer
712: cuprates, our present results for the normal state are in
713: qualitative agreement with the major experimental observations in
714: the normal state of doped bilayer cuprates
715: \cite{shen,dfeng,kordyuk,chuang}.
716:
717:
718: Finally, we have noted that the very recent ARPES experiments have
719: shown that although the largest BS is found in the electron spectral
720: function around the $[\pi,0]$ point, a small BS remains also around
721: the $[\pi/2,\pi/2]$ point in a wide doping range
722: \cite{kordyuk2,shen3}. It has been argued that the observed this
723: small BS around the $[\pi/2,\pi/2]$ point is in good agreement with
724: the local density approximation based band structure calculations,
725: and is caused by the vertical O $2p_{\sigma}$-O $2p_{\sigma}$
726: hopping $t^{pp}_{\perp}$ between two adjacent planes
727: \cite{kordyuk2,shen3}. Therefore the significance of the $p-p$
728: transfer $t^{pp}_{\perp}$ should be taken into account. These and
729: related issues are under investigations now.
730:
731:
732: \acknowledgments The authors would like to thank Dr. H. Guo and Dr.
733: L. Cheng for the helpful discussions. This work was supported by the
734: National Natural Science Foundation of China under Grant Nos.
735: 90403005 and 10547104, and the funds from the Ministry of Science
736: and Technology of China under Grant Nos. 2006CB601002 and
737: 2006CB921300.
738:
739:
740: \appendix
741:
742: \section*{The dressed holon Green's function in the bilayer system}
743:
744: As in the single layer case \cite{guo}, the full dressed holon
745: Green's function in the doped bilayer cuprates satisfies the
746: equation \cite{eliashberg},
747: \begin{eqnarray}
748: g({\bf k},\omega)=g^{(0)}({\bf k},\omega)+g^{(0)}({\bf k},\omega)
749: \Sigma^{(h)}({\bf k},\omega)g({\bf k},\omega)
750: \end{eqnarray}
751: with the MF dressed holon Green's function $g^{(0)}({\bf k},
752: \omega)=g^{(0)}_{L}({\bf k},\omega)+\sigma_{x}g^{(0)}_{T}({\bf k},
753: \omega)$, where the longitudinal and transverse parts are obtained
754: as,
755: \begin{subequations}
756: \begin{eqnarray}
757: g^{(0)}_{L}({\bf k},\omega)&=&{1\over 2}\sum_{\nu=1,2}{1\over
758: \omega-\xi_{\nu{\bf k}}},\\
759: g^{(0)}_{T}({\bf k},\omega)&=&{1\over 2}\sum_{\nu=1,2}(-1)^{\nu+1}
760: {1\over\omega-\xi_{\nu{\bf k}}},
761: \end{eqnarray}
762: \end{subequations}
763: and the dressed holon self-energy function $\Sigma^{(h)}({\bf k},
764: \omega)=\Sigma^{(h)}_{L}({\bf k},\omega)+\sigma_{x}\Sigma^{(h)}_{T}
765: ({\bf k},\omega)$, where $\Sigma^{(h)}_{L}({\bf k},\omega)$ and
766: $\Sigma^{(h)}_{T}({\bf k},\omega)$ are the corresponding
767: longitudinal and transverse parts, and have been given in Eq. (6).
768: This self-energy function $\Sigma^{(h)}({\bf k},\omega)$
769: renormalizes the MF dressed holon spectrum, and thus it describes
770: the quasiparticle coherence. On the other hand, since we only study
771: the low-energy behavior of doped bilayer cuprates, then the
772: quasiparticle coherence can be discussed in the static limit. In
773: this case, we follow the previous discussions for the single layer
774: case \cite{guo}, and obtain explicitly the longitudinal and
775: transverse parts of the full dressed holon Green's functions of the
776: doped bilayer system as,
777: \begin{subequations}
778: \begin{eqnarray}
779: g_{L}({\bf k},\omega)&=&{1\over 2}\sum_{\nu=1,2}{Z^{(\nu)}_{FA}
780: \over\omega-\bar{\xi}_{\nu{\bf k}}}, \\
781: g_{T}({\bf k},\omega)&=&{1\over 2}\sum_{\nu=1,2}(-1)^{\nu+1}{
782: Z^{(\nu)}_{FA }\over \omega-\bar{\xi}_{\nu{\bf k}}},
783: \end{eqnarray}
784: \end{subequations}
785: with the quasiparticle coherent weights satisfy the following
786: equations,
787: \begin{subequations}
788: \begin{eqnarray}
789: {1\over Z_{FA}^{(1)}}&=&1+{1\over 32N^{2}}\sum_{{\bf q,p}}
790: \sum_{\nu,\nu',\nu''}[1+(-1)^{\nu+\nu'+\nu''+1}]\nonumber\\
791: &\times&\Lambda_{\nu\nu'\nu''}({\bf q,p}),\\
792: {1\over Z_{FA}^{(2)}}&=&1+{1\over 32N^{2}}\sum_{{\bf q,p}}
793: \sum_{\nu,\nu',\nu''}[1-(-1)^{\nu+\nu'+\nu''+1}]\nonumber\\
794: &\times&\Lambda_{\nu\nu'\nu''}({\bf q,p}),
795: \end{eqnarray}
796: \end{subequations}
797: where the kernel function $\Lambda_{\nu\nu'\nu''}({\bf q,p})$ can be
798: evaluated as,
799: \begin{eqnarray}
800: \Lambda_{\nu\nu'\nu''}({\bf q,p})&=&C_{\nu\nu''}({\bf p+k_{0}})
801: Z^{(\nu'')}_{FA}{B_{\nu'{\bf p}}B_{\nu{\bf q}}\over \omega_{\nu'{\bf
802: p}} \omega_{\nu{\bf q}}}\nonumber\\
803: &\times&\left( {F^{(1)}_{\nu\nu'\nu''}({\bf q,p})\over
804: [\omega_{\nu'{\bf p} } -\omega_{\nu{\bf q}}+\bar{\xi}_{\nu''{\bf
805: p-q+k_{0}}}]^{2}}\right.\nonumber\\
806: &&\left.+{F^{(2)}_{\nu\nu'\nu''}({\bf q,p})\over [\omega_{\nu'{\bf
807: p}}- \omega_{\nu{\bf q}}-\bar{\xi}_{\nu''{\bf
808: p-q+k_{0}}}]^{2}}\right.\nonumber\\
809: &&\left. +{F^{(3)}_{\nu\nu'\nu''}({\bf q,p})\over [\omega_{\nu'{\bf
810: p}} +\omega_{\nu{\bf q}}+\bar{\xi}_{\nu''{\bf p-q+k_{0}}}]^{2}}\right.
811: \nonumber\\
812: &&\left.+{F^{(4)}_{\nu\nu'\nu''}({\bf q,p})\over [\omega_{\nu'{\bf
813: p}}+\omega_{\nu{\bf q}}-\bar{\xi}_{\nu''{\bf p- q+k_{0}}}]^{2}}
814: \right),~~~~
815: \end{eqnarray}
816: with $C_{\nu\nu''}({\bf p})=[Z(t\gamma_{\bf p}-t'\gamma{'}_{\bf p}
817: )+(-1)^{\nu+\nu''}t_{\perp}({\bf p})]^{2}$,
818: $F^{(1)}_{\nu\nu'\nu''}({\bf q,p})=n_{F}(\bar{\xi}_{\nu''{\bf p-
819: q+k_{0}}}) [n_{B}(\omega_{\nu'{\bf p}})-n_{B}(\omega_{\nu{\bf q}})
820: ]+n_{B}(\omega_{\nu{\bf q}})[1+n_{B}(\omega_{\nu'{\bf p}})]$,
821: $F^{(2)}_{\nu\nu'\nu''}({\bf q,p})=n_{F}(\bar{\xi}_{\nu''{\bf p-
822: q+k_{0}}}) [n_{B}(\omega_{\nu{\bf q}})-n_{B}(\omega_{\nu'{\bf p}})
823: ]+ n_{B}(\omega_{\nu'{\bf p}})[1+n_{B}(\omega_{\nu{\bf q}})]$,
824: $F^{(3)}_{\nu\nu'\nu''}({\bf q,p})=[1-n_{F}(\bar{\xi}_{\nu''{\bf p
825: -q+k_{0}}})][1+n_{B}(\omega_{\nu{\bf q}})+n_{B}(\omega_{\nu'{\bf p
826: }})]+n_{B}(\omega_{\nu{\bf q}})n_{B}(\omega_{\nu'{\bf p}})$,
827: $F^{(4)}_{\nu\nu'\nu''}({\bf q,p})=n_{F}(\bar{\xi}_{\nu''{\bf p-
828: q+k_{0}}}) [1+n_{B}(\omega_{\nu{\bf q}})+n_{B}(\omega_{\nu'{\bf p}
829: })]+ n_{B}(\omega_{\nu{\bf q}})n_{B}(\omega_{\nu'{\bf p}})$. These
830: two equations in Eq. (A4) must be solved together with other
831: self-consistent equations \cite{guo},
832: \begin{subequations}
833: \begin{eqnarray}
834: \delta&=&{1\over 4N}\sum_{\nu,{\bf k}}Z^{(\nu)}_{FA}\left(1-{\rm th}
835: [{1\over 2}\beta \bar{\xi}_{\nu{\bf k}}]\right ), \\
836: \phi_{1}&=&{1\over 4N}\sum_{\nu,{\bf k}}\gamma_{\bf k}Z^{(\nu)}_{FA}
837: \left ( 1-{\rm th}[{1\over 2}\beta \bar{\xi}_{\nu{\bf k}}]\right ),\\
838: \phi_{2}&=&{1\over 4N}\sum_{\nu,{\bf k}}\gamma{'}_{\bf k}
839: Z^{(\nu)}_{FA}\left ( 1-{\rm th}[{1\over 2}\beta
840: \bar{\xi}_{\nu{\bf k}}]\right ), \\
841: \phi_{\perp}&=&{1\over 4N}\sum_{\nu,{\bf k}} (-1)^{\nu+1}
842: Z^{(\nu)}_{FA} \left ( 1-{\rm th}[{1\over 2}\beta
843: \bar{\xi}_{\nu{\bf k}}]\right ), ~~~~~\\
844: 1\over 2&=&{1\over 4N}\sum_{\nu,{\bf k}}{B_{\nu{\bf k}}\over
845: \omega_{\nu{\bf k}}}{\rm coth}[{1\over 2}\beta\omega_{\nu{\bf k}}],\\
846: \chi_{1} &=& {1\over 4N} \sum_{\nu,{\bf k}}\gamma_{\bf k}
847: {B_{\nu{\bf k}}\over\omega_{\nu{\bf k}}}{\rm coth}[{1\over 2}
848: \beta\omega_{\nu{\bf k}}], \\
849: \chi_{2}&=&{1\over 4N}\sum_{\nu,{\bf k}} \gamma{'}_{\bf k}
850: {B_{\nu{\bf k}}\over\omega_{\nu{\bf k}}}{\rm coth}
851: [{1\over 2}\beta\omega_{\nu{\bf k}}], \\
852: C_{1} &=& {1\over 4N} \sum_{\nu,{\bf k}} \gamma_{\bf k}^{2}
853: {B_{\nu{\bf k}}\over\omega_{\nu{\bf k}}}{\rm coth}
854: [{1\over 2}\beta\omega_{\nu{\bf k}}], \\
855: C_{2}&=&{1\over 4N}\sum_{\nu,{\bf k}}\gamma_{\bf k}\gamma{'}_{\bf k}
856: {B_{\nu{\bf k}}\over\omega_{\nu{\bf k}}}{\rm coth}
857: [{1\over 2}\beta\omega_{\nu{\bf k}}], \\
858: C_{3}&=&{1\over 4N}\sum_{\nu,{\bf k}} \gamma{'}_{\bf k}^{2}
859: {B_{\nu{\bf k}}\over\omega_{\nu{\bf k}}}{\rm coth}
860: [{1\over 2}\beta\omega_{\nu{\bf k}}], \\
861: \chi_{1}^{z}&=&{1\over 4N}\sum_{\nu,{\bf k}} \gamma_{\bf k}
862: {B_{z\nu{\bf k}}\over\omega_{z\nu{\bf k}}}{\rm coth}
863: [{1\over 2}\beta\omega_{z\nu{\bf k}}], \\
864: \chi_{2}^{z}&=&{1\over 4N}\sum_{\nu,{\bf k}} \gamma{'}_{\bf k}
865: {B_{z\nu{\bf k}}\over\omega_{z\nu{\bf k}}}{\rm coth}
866: [{1\over 2}\beta\omega_{z\nu{\bf k}}], \\
867: C_{1}^{z}&=&{1\over 4N}\sum_{\nu,{\bf k}} \gamma_{\bf k}^{2}
868: {B_{z\nu{\bf k}}\over\omega_{z\nu{\bf k}}}{\rm coth}
869: [{1\over 2}\beta\omega_{z\nu{\bf k}}], \\
870: C_{2}^{z}&=&{1\over 4N}\sum_{\nu,{\bf k}}\gamma_{\bf k}
871: \gamma{'}_{\bf k} {B_{z\nu{\bf k}} \over \omega_{z\nu{\bf k}}}
872: {\rm coth}[{1\over 2}\beta\omega_{z\nu{\bf k}}], \\
873: \chi_{\perp} &=& {1\over 4N}\sum_{\nu,{\bf k}}(-1)^{\nu+1}
874: {B_{\nu{\bf k}}\over\omega_{\nu{\bf k}}}{\rm coth}[{1\over
875: 2}\beta\omega_{\nu{\bf k}}], \\
876: C_{\perp} &=& {1\over 4N} \sum_{\nu,{\bf k}} (-1)^{\nu+1}
877: \gamma_{\bf k}{B_{\nu{\bf k}}\over\omega_{\nu{\bf k}}}
878: {\rm coth}[{1\over 2}\beta\omega_{\nu{\bf k}}], \\
879: C{'}_{\perp}&=&{1\over 4N}\sum_{\nu,{\bf k}}(-1)^{\nu+1}
880: \gamma{'}_{\bf k}{B_{\nu{\bf k}}\over\omega_{\nu{\bf k}}}
881: {\rm coth}[{1\over 2}\beta\omega_{\nu{\bf k}}], \\
882: \chi^{z}_{\perp}&=&{1\over 4N}\sum_{\nu,{\bf k}}(-1)^{\nu+1}
883: {B_{z\nu{\bf k}}\over\omega_{z\nu{\bf k}}}{\rm coth}[{1\over
884: 2}\beta\omega_{z\nu{\bf k}}],\\
885: C_{\perp}^{z}&=&{1\over 4N}\sum_{\nu,{\bf k}}(-1)^{\nu+1}
886: \gamma_{\bf k} {B_{z\nu{\bf k}} \over \omega_{z\nu{\bf k}}}
887: {\rm coth}[{1\over 2}\beta\omega_{z\nu{\bf k}}],\\
888: C{'}_{\perp}^{z}&=&{1\over 4N}\sum_{\nu,{\bf k}}(-1)^{\nu+1}
889: \gamma{'}_{\bf k} {B_{z\nu{\bf k}} \over \omega_{z\nu{\bf k}}}
890: {\rm coth}[{1\over 2}\beta\omega_{z\nu{\bf k}}],
891: \end{eqnarray}
892: \end{subequations}
893: then all the order parameters, decoupling parameter $\alpha$,
894: chemical potential $\mu$ and the quasiparticle coherent weights are
895: obtained by the self-consistent calculation.
896:
897:
898: \begin{thebibliography}{}
899: %\begin{references}
900:
901: \bibitem{bedell} See, e.g., {\it Proceedings of Los Alamos
902: Symposium}, edited by K.S. Bedell, D. Coffey, D.E. Meltzer, D.
903: Pines, and J.R. Schrieffer (Addison-Wesley, Redwood city,
904: California, 1990).
905:
906: \bibitem{kastner} See, e.g., M.A.Kastner, R.J. Birgeneau, G.
907: Shiran, and Y. Endoh, Rev. Mod. Phys. {\bf 70}, 897 (1998), and
908: referenes therein.
909:
910: \bibitem{shen} See, e.g., A. Damascelli, Z. Hussain, and Z.-X.
911: Shen, Rev. Mod. Phys. {\bf 75}, 473 (2003), and referenes therein.
912:
913: \bibitem{campuzano} See, e.g., J. Campuzano, M. Norman, and M.
914: Randeira, in {\it Physics of Superconductors}, vol. II, edited by K.
915: Bennemann and J. Ketterson (Springer, Berlin Heidelberg New York,
916: 2004), p. 167, and referenes therein.
917:
918: \bibitem{shen1} See, e.g., Z.X. Shen and D.S. Dessau, Phys. Rep.
919: {\bf 70}, 253 (1995), and referenes therein.
920:
921: \bibitem {anderson} P.W. Anderson, in {\it Frontiers and
922: Borderlines in Many Particle Physics}, edited by R.A. Broglia and
923: J.R. Schrieffer (North-Holland, Amsterdam, 1987), p. 1; Science {\bf
924: 235}, 1196 (1987).
925:
926: \bibitem{shen2} Z.X. Shen, W.E. Spicer, D.M. King, D.S. Dessau,
927: and B.O. Wells, Science {\bf 267}, 343 (1995); D.M. King, Z.X. Shen,
928: D.S. Dessau, D.S. Marshall, C.H. Park, W.E. Spicer, J.L. Peng, Z.Y.
929: Li, and R.L. Greene, Phys. Rev. Lett. {\bf 73}, 3298 (1994).
930:
931: \bibitem{dessau} D.S. Dessau, Z.X. Shen, D.M. King, D.S. Marshall,
932: L.W. Lombardo, P.H. Dickinson, A.G. Loeser, J. DiCarlo, C.H. Park,
933: A. Kapitulnik, and W.E. Spicer, Phys. Rev. Lett. {\bf 71}, 2781
934: (1993); D.S. Marshall, D.S. Dessau, A.G. Loeser, C.H. Park, A.
935: Matsuura, J.N. Eckstein, I. Bozovic, P. Fournier, A. Kapitulnik,
936: W.E. Spicer, and Z.X. Shen, Phys. Rev. Lett. {\bf 76}, 4841 (1996).
937:
938: \bibitem{dfeng} D.L. Feng, N.P. Armitage, D.H. Lu, A. Damascelli,
939: J.P. Hu, P. Bogdanov, A. Lanzara, F. Ronning, K.M. Shen, H. Eisaki,
940: C. Kim, Z.-X. Shen, J.-i. Shimoyama, and K. Kishio, Phys. Rev. Lett.
941: {\bf 86}, 5550 (2001).
942:
943: \bibitem{kordyuk} A.A. Kordyuk, S.V. Borisenko, M. Knupfer, and
944: J. Fink, Phys. Rev. B {\bf 67}, 064504 (2003); A.A. Kordyuk and S.V.
945: Borisenko, Low Temp. Phys. {\bf 32}, 298 (2006).
946:
947: \bibitem{chuang} Y.-D. Chuang, A.D. Gromko, A. Fedorov, Y. Aiura,
948: K. Oka, Y. Ando, H. Eisaki, S.I. Uchida, and D.S. Dessau, Phys. Rev.
949: Lett. {\bf 87}, 117002 (2001); P.V. Bogdanov, A. Lanzara, X.J. Zhou,
950: S.A. Kellar, D.L. Feng, E.D. Lu, H. Eisaki, J.-I. Shimoyama, K.
951: Kishio, Z. Hussain, and Z.X. Shen, Phys. Rev. B {\bf 64}, 180505
952: (2001).
953:
954: \bibitem{campuzano1} J.C. Campuzano, H. Ding, M.R. Norman, H.M.
955: Fretwell, M. Randeria, A. Kaminski, J. Mesot, T. Takeuchi, T. Sato,
956: T. Yokoya, T. Takahashi, T. Mochiku, K. Kadowaki, P. Guptasarma,
957: D.G. Hinks, Z. Konstantinovic, Z.Z. Li, and H. Raffy, Phys. Rev.
958: Lett. {\bf 83}, 3709 (1999).
959:
960: \bibitem{kordyuk1} A.A. Kordyuk, S.V. Borisenko, T.K. Kim, K.A.
961: Nenkov, M. Knupfer, J. Fink, M.S. Golden, H. Berger, and R. Follath,
962: Phys. Rev. Lett. {\bf 89}, 077003 (2002).
963:
964: \bibitem{borisenko} S.V. Borisenko, A.A. Kordyuk, T.K. Kim, A.
965: Koitzsch, M. Knupfer, J. Fink, M.S. Golden, M. Eschrig, H. Berger,
966: and R. Follath, Phys. Rev. Lett. {\bf 90}, 207001 (2003); S.V.
967: Borisenko, A.A. Kordyuk, T.K. Kim, S. Legner, K.A. Nenkov, M.
968: Knupfer, M.S. Golden, J. Fink, H. Berger, and R. Follath, Phys. Rev.
969: B {\bf 66}, 140509 (2002).
970:
971: \bibitem{mori} M. Mori, T. Tohyama, and S. Maekawa, Phys. Rev. B
972: {\bf 66}, 064502 (2002).
973:
974: \bibitem{feng} Shiping Feng, Jihong Qin, and Tianxing Ma, J.
975: Phys. Condens. Matter {\bf 16}, 343 (2004); Shiping Feng, Tianxing
976: Ma, and Jihong Qin, Mod. Phys. Lett. B {\bf 17}, 361 (2003).
977:
978: \bibitem{guo} Huaiming Guo and Shiping Feng, Phys. Lett. A
979: {\bf 355}, 473 (2006).
980:
981: \bibitem{chakarvarty} O.K. Anderson, A.I. Liechtenstein, O. Jepsen,
982: and F. Plulsen, J. Phys. Chem. Solids {\bf 56}, 1573 (1995); A.I.
983: Liechtenstein, O.K. Anderson, and R.M. Martin, Phys. Rev. B {\bf
984: 54}, 12505 (1996); S. Chakarvarty, A. Sudbo, P.W. Anderson, and S.
985: Strong, Science {\bf 261}, 337 (1993).
986:
987: \bibitem{laughlin} R.B Laughlin, Phys. Rev. Lett. {\bf 79},
988: 1726 (1997); J. Low. Tem. Phys. {\bf 99}, 443 (1995).
989:
990: \bibitem{anderson1} P.W. Anderson, Phys. Rev. Lett. {\bf 67}, 2092
991: (1991); Science {\bf 288}, 480 (2000).
992:
993: \bibitem{eliashberg} G.M. Eliashberg, Sov. Phys. JETP {\bf 11},
994: 696 (1960); D.J. Scalapino, J.R. Schrieffer, and J.W. Wilkins, Phys.
995: Rev. {\bf 148}, 263 (1966).
996:
997: \bibitem{feng1} Feng Yuan, Jihong Qin, Shiping Feng, and Wei Yeu
998: Chen, Phys. Rev. B {\bf 67}, 134504 (2003).
999:
1000: \bibitem{kondo} J. Kondo and K. Yamaji, Prog. Theor. Phys.
1001: {\bf 47}, 807 (1972); Shiping Feng and Yun Song, Phys. Rev. B {\bf
1002: 55}, 642 (1997).
1003:
1004: \bibitem{tanaka} K. Tanaka, T. Yoshida, A. Fujimori, D.H. Lu, Z.-X.
1005: Shen, X.-J. Zhou, H. Eisaki, Z. Hussain, S. Uchida, Y. Aiura, K.
1006: Ono, T. Sugaya, T. Mizuno, and I. Terasaki, Phys. Rev. B {\bf 70},
1007: 092503 (2004).
1008:
1009: \bibitem{kordyuk2} A.A. Kordyuk, S.V. Borisenko, A.N. Yaresko,
1010: S.-L. Drechsler, H. Rosner, T.K. Kim, A. Koitzsch, K.A. Nenkov, M.
1011: Knupfer, J. Fink, R. Follath, H. Berger, B. Keimer, S. Ono,8 and Y.
1012: Ando, Phys. Rev. B {\bf 70}, 214525 (2004); A.A. Kordyuk, S.V.
1013: Borisenko, A. Koitzsch, J. Fink, M. Knupfer, B. Buechner, H. Berger,
1014: J. Phys. Chem. Solids {\bf 67}, 201 (2006).
1015:
1016: \bibitem{shen3} T. Yamasaki, K. Yamazaki, A. Ino, M. Arita, H.
1017: Namatame, M. Taniguchi, A. Fujimori, Z.-X. Shen, M. Ishikado, S.
1018: Uchida, cond-mat/0603006.
1019:
1020: \end{thebibliography}{}
1021: %\end{references}
1022:
1023:
1024: \end{document}
1025: