1: %{\mbox{\boldmath ${\bf \sigma}$}}
2: %\documentclass[aps,prl,twocolumn,groupedaddress]{revtex4}
3: %\documentclass[prb,aps,twocolumn,showpacs,superscriptaddress,
4: %floatfix]{revtex4}
5: \documentclass[prb,aps,showpacs,superscriptaddress,floatfix]{revtex4}
6: \usepackage{graphicx}
7: \usepackage{bm}
8: \usepackage{amsmath}
9: \begin{document}
10: %%%%%%%%%%%%%%%%%%%%
11: % Title
12: %%%%%%%%%%%%%%%%%%%%
13: \title{
14: Mechanism of $d_{x^2-y^2}$-wave superconductivity
15: based on doped hole induced spin texture
16: in high $T_c$ cuprates
17: }
18: \author{T. Morinari}
19: \affiliation{Yukawa Institute for Theoretical Physics, Kyoto
20: University, Kyoto 606-8502, Japan}
21: \date{\today}
22: \begin{abstract}
23: A mechanism of $d_{x^2-y^2}$-wave superconductivity is proposed for
24: the high-$T_c$ cuprates based on a spin texture with non-zero
25: topological density induced by doped holes through Zhang-Rice singlet
26: formation.
27: The pairing interaction arises from
28: the magnetic Lorentz force like interaction between the holes and the
29: spin textures.
30: The stability of the pairing state against
31: the vortex-vortex interaction and the Coulomb repulsion
32: is examined.
33: The mechanism suggests appearance of a p-wave pairing component
34: by introducing anisotropy in the CuO$_2$ plane.
35: \end{abstract}
36: %\pacs{ Valid PACS appear here.
37: %{\tt$\backslash$\string pacs\{\}} should always be input, even if
38: %empty.}
39: \pacs{74.20.-z,74.72,-h,71.10.-w%75.25.+z,74.72.-h
40: }
41:
42: %\begin{multicols}{2}
43:
44: %\narrowtext
45:
46: \maketitle
47:
48: \section{Introduction}
49: High-temperature superconductivity occurs in a moderately hole doped
50: Mott insulator.\cite{ANDERSON}
51: The symmetry of the Cooper pairs has been established to be
52: $d_{x^2-y^2}$-wave.\cite{Harlingen95}
53: To explain the mechanism of $d_{x^2-y^2}$-wave superconductivity,
54: several mechanisms have been proposed.
55: Among others, the antiferromagnetic spin fluctuation theory
56: suggest a $d_{x^2-y^2}$-wave pairing mechanism\cite{Pines}
57: between d-orbital electrons at copper sites.
58: Another $d_{x^2-y^2}$-wave pairing mechanism is proposed in the
59: resonating valence bond (RVB) theory.\cite{ANDERSON}
60: In the RVB theory, electrons are assumed to be spin-charge separated.
61: In the slave-boson approach, \cite{LNW04}
62: spinons, which carry the spin degrees of
63: freedom, form a $d_{x^2-y^2}$-wave pairing state.
64: This pairing state is associated with the short-range
65: antiferromagnetic correlations.\cite{ABZH87}
66: Combined with Bose-Einstein condensation of holons, which carry the
67: electric charge, the pairing state of the spinons
68: leads to a $d_{x^2-y^2}$-wave superconducting state.
69:
70:
71: Both of these theories are based on a pairing state between
72: copper site electrons which produce antiferromagnetic correlations.
73: However, from the sign of the Hall coefficient\cite{Takagi89}
74: the charge carriers seem to be doped holes.
75: Optical conductivity measurements also support that the charge
76: carriers are doped holes because
77: the Drude weight in the optical conductivity\cite{Uchida91}
78: is proportional to the doped hole concentration.
79: If we assume that the charge carriers are doped holes,
80: then it is natural to expect that high-temperature
81: superconductivity is based on a $d_{x^2-y^2}$-wave pairing mechanism
82: between doped holes.
83: In this paper, we propose such a mechanism
84: based on a picture for the doped holes.
85: The picture is that each hole induces a spin texture
86: that is characterized by a non-vanishing topological charge density
87: in the spin system through the suppression of the antiferromagnetic
88: correlations.
89:
90:
91: In Ref.\cite{TM_PRB05}, a half-skyrmion spin texture
92: \cite{SD02,BASKARAN}
93: formation is discussed in the single-hole-doped cuprate
94: within the non-linear sigma model(NLSM) description\cite{CHN} of the
95: Heisenberg antiferromagnet.
96: The formation of the half-skyrmion spin texture can be understood
97: by using an analogy to a superfluid system:
98: Let us consider a two-dimensional boson system which exhibits
99: superfluidity in the ground state.
100: The Bose-Einstein condensate is described by a Gross-Pitaevskii
101: equation.\cite{FW}
102: If we suppress the condensate at some point $P$, then we see that
103: a vortex is formed around $P$ from the analysis of the
104: Gross-Pitaevskii equation.
105: Note that the field described by the Gross-Pitaevskii equation
106: is a classical field.
107: However,
108: the non-condensed component near the vortex core is described by
109: quantum bosons.
110: In case of a charged boson system, the condensate is locally
111: suppressed by a magnetic flux, or a vortex.
112:
113:
114: We can apply this result to the two-dimensional quantum
115: Heisenberg antiferromagnet.
116: In the Schwinger boson mean field theory,\cite{AA88}
117: the N{\' e}el ordering is described by Bose-Einstein condensation of
118: the Schwinger bosons.\cite{SB_BEC}
119: If we assume that the doped hole forms a spin singlet with a
120: localized spin at a copper site, then the condensate is suppressed
121: at that site.
122: Forming a Zhang-Rice singlet in the antiferromagnet plays a similar
123: role of introducing a magnetic flux in a charged boson condensate.
124: As in the case of a superfluid system, a vortex appears around the
125: site.
126: In terms of the spins, the vortex is a spin texture with the
127: non-vanishing topological charge density
128: \cite{skyrmions}
129: that is given by
130: \begin{equation}
131: q({\bf r})=\frac{1}{4\pi} {\bf n}({\bf r}) \cdot \left[
132: \partial_x {\bf n} ({\bf r})\times
133: \partial_y {\bf n} ({\bf r})\right],
134: \end{equation}
135: where the unit vector ${\bf n}({\bf r})$ describes the staggered
136: component of the spins.
137: Note that the spin at the core of the spin texture is a quantum spin
138: analogous to the normal component near a vortex core in a superfluid,
139: and so the spin can form a singlet state with the doped hole spin,
140: which is the Zhang-Rice singlet.\cite{ZHANG_RICE}
141: Based on the Schwinger boson mean field theory, we find that
142: the topological charge, which is obtained by integrating $q({\bf r})$
143: over a whole region, is quantized
144: to $\pm 1/2$.\cite{TM_PRB05,WENG_ETAL}
145: Because of this quantized value, the spin texture is called
146: half-skyrmion.
147:
148:
149:
150: As argued in Ref.\cite{TM_PRB05}, the dispersion of the half-skyrmion
151: is in good agreement with the results of the angle-resolved
152: photoemission spectroscopy (ARPES) in the undoped
153: compounds.\cite{WELLS_ETAL}
154: The dispersion of the half-skyrmion excitation is the same as
155: that of the quasiparticles in the $\pi$-flux phase with a dynamically
156: induced mass.\cite{MARSTON,KIM_LEE}
157: In fact, the effective theory of the half-skyrmion is similar to that
158: of the $\pi$-flux phase:
159: the half-skyrmions are described by Dirac fermions with a U(1) gauge
160: field interaction.
161: Anomalously broad peaks observed in ARPES are associated with the
162: self-trapping of the half-skyrmions due to the coupling to the
163: longitudinal gauge field.\cite{TM_BROAD}
164:
165:
166: In Ref.\cite{TM_PRB05}, we assume that there is N{\' e}el ordering.
167: The presense of the N{\' e}el ordering is used for obtaining the quantized
168: topological charge.
169: In the absence of the N{\' e}el ordering, the topological charge is no more
170: quantized because the antiferromagnetic correlations
171: decay exponentially with a length scale given by
172: the antiferromagnetic correlation length $\xi_{AF}$.
173: However, the system preserves
174: the local relationship between the hole density and the topological
175: charge density.
176: We consider spin textures with such topological charge density and
177: hereafter we call such a spin texture staggered-spin-vortex.
178:
179:
180:
181: The rest of the paper is organized as follows:
182: In Sec.\ref{sec_model}, we describe the model for
183: the staggered-spin-vortex formation.
184: The pairing interaction is discussed in Sec.\ref{sec_pairing}.
185: We solve the gap equation in Sec.\ref{sec_gap} and show that the
186: stable pairing state has $d_{x^2-y^2}$-wave symmetry.
187: Section \ref{sec_conclusion} is devoted to the conclusion.
188:
189:
190: \section{Model}
191: \label{sec_model}
192: We assume that the holes and the localized spins are independent
193: degrees of freedom.
194: Such a model is, for instance, the spin-fermion model, which is
195: derived from the d-p model by applying a canonical transformation.
196: In such a model, the spin system is described by the antiferromagnetic
197: Heisenberg model,
198: \begin{equation}
199: H_s=J\sum_{\langle i,j \rangle} {\bf S}_i \cdot {\bf S}_j.
200: \end{equation}
201: Here the summation is taken over the nearest neighbor sites and ${\bf
202: S}_i$ denote the spin moment at copper sites.
203: As an effective theory of the spin system, we consider the CP$^1$
204: model,\cite{RAJARAMAN}
205: which is derived from the NLSM by representing ${\bf n}(x)$
206: in terms of the Schwinger bosons:
207: ${\bf n} = \sum_{\sigma,\sigma'}\overline{z}_{\sigma}
208: {\mbox{\boldmath ${\bf \sigma}$}}_{\sigma,\sigma'} z_{\sigma}$:
209: \begin{equation}
210: S_{CP^1} = \frac{2}{g}
211: \int d^3 x \sum_{\sigma=\uparrow,\downarrow}
212: \left[
213: \left| \left( \partial_{\mu}-i\alpha_{\mu} \right)
214: z_{\sigma} (x) \right|^2
215: + \frac{\Delta_{\rm sw}^2}{c_{\rm sw}^2}
216: |z_{\sigma} (x) |^2
217: \right],
218: \label{eq_CP1}
219: \end{equation}
220: where $\Delta_{\rm sw}$ is the antiferromagnetic spin wave excitation
221: gap and $c_{\rm sw}$ is the spin wave velocity.
222: The coupling $g$ is given by $g=c_{\rm sw}/\rho_s$ with
223: $\rho_s=Z_{\rho} J/4$ the spin stiffness constant.
224: $Z_{\rho}$ is the renormalization factor due to quantum phase
225: fluctuations.
226: In the spin disordered phase, $\Delta_{\rm sw}\neq 0$,
227: and there is no antiferromagnetic long-range order.
228: The antiferromagnetic correlations are characterized by
229: the correlation length
230: $\xi_{AF} = c_{\rm sw}/\Delta_{\rm sw}$.
231: For the Schwinger bosons,
232: the presence of the antiferromagnetic correlations is associated with
233: the phase coherence among the Schwinger bosons.
234:
235:
236: For the kinetic energy of the holes, we assume
237: \begin{equation}
238: H_h = \sum_{k\sigma} \epsilon_k c_{k\sigma}^{\dagger} c_{k\sigma},
239: \label{eq_ke}
240: \end{equation}
241: where $\epsilon_k = k^2/2m$ with $m$ the effective mass.
242: This is an approximate form in the continuum.
243:
244:
245:
246:
247: The coupling between the holes and the localized spins ${\bf S}_j$
248: has the form of the antiferromagnetic Kondo coupling:
249: \begin{equation}
250: H_K = J_K \sum_j \left(c_j^{\dagger} \sigma c_j \right)\cdot
251: {\bf S}_j.
252: \end{equation}
253: The parameter $J_K$ is on the order of $1\sim 3$eV.
254: Because of this strong Kondo coupling, holes suppress
255: phase coherence of the Schwinger bosons
256: locally through the Zhang-Rice singlet formation.
257: The suppression of the phase coherence leads to a
258: staggered-spin-vortex:
259: \begin{equation}
260: \nabla \times
261: {\mbox{\boldmath ${\bf \alpha}$}}
262: =
263: \pi \sum_{s=\pm} s \psi^{\dagger}_s ({\bf r}) \psi_s ({\bf r}),
264: \end{equation}
265: where
266: $\nabla \times
267: {\mbox{\boldmath ${\bf \alpha}$}}
268: = \partial_x \alpha_y - \partial_y \alpha_x$
269: and
270: $\psi^{\dagger}_{+(-)}$ is the creation operator of the
271: staggered-spin-vortex (anti-staggered-spin-vortex).
272: The coefficient is determined by considering the limit of
273: $\xi_{\rm AF} \rightarrow \infty$.
274: In this limit, the spin texture is the half-skyrmion.
275: The gauge flux $\nabla \times
276: {\mbox{\boldmath ${\bf \alpha}$}}$ is
277: associated with the spin chirality.\cite{LEE_NAGAOSA02,WWZ}
278: Note that $\langle \nabla \times
279: {\mbox{\boldmath ${\bf \alpha}$}}
280: \rangle =0$,
281: because in the equilibrium there is an equal number of staggered spin
282: vortices and anti-staggered spin vortices, and so
283: $\langle \sum_s s \psi^{\dagger}_s ({\bf r})
284: \psi_s ({\bf r}) \rangle=0$.
285:
286:
287: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
288: By integrating out the Schwinger bosons, we obtain the effective
289: action of the gauge field $\alpha_{\mu}$.
290: Since $\Delta_{\rm sw}\neq 0$, the action takes the Maxwellian form.
291: That is, the gauge field is massless.
292: However, in the spin disordered phase, the antiferromagnetic
293: correlation length $\xi_{AF}$ is finite.
294: Therefore, the gauge field propagator decays
295: with a length scale of $\xi_{AF}$.
296: To include this feature, we perform a duality mapping.
297: We write $z_{\sigma}(x)$ as
298: \begin{equation}
299: z_{\sigma}(x)=\overline{\rho}^{1/2}_{\sigma} \exp(i\theta_0 +
300: i\theta_v).
301: \end{equation}
302: Here the phase $\theta_0$ is associated with the coherent motion of
303: the Schwinger bosons and the phase $\theta_v$ is associated with the
304: staggered-spin-vortices.
305: Note that this form is applicable for the description outside the
306: vortex cores.
307: After performing a Stratonovich-Hubbard transformation in
308: Eq.(\ref{eq_CP1}), we obtain the following Lagrangian density,
309: \begin{equation}
310: {\cal L} = \frac{g}{8\overline{\rho}} J_{\mu}^2 +iJ_{\mu}
311: (\partial_{\mu} \theta_0 + \partial_{\mu} \theta_v - \alpha_{\mu}).
312: \end{equation}
313: We find $\partial_{\mu} J_{\mu}=0$ by integrating out $\theta_0$.
314: This constraint is satisfied by introducing a gauge field:
315: \begin{equation}
316: J_{\mu}=\frac{1}{2\pi} \epsilon_{\mu\nu\lambda} \partial_{\nu}
317: A_{\lambda}.
318: \end{equation}
319: We define the vortex current by
320: \begin{equation}
321: j_{\mu}^v = \frac{1}{2\pi} \epsilon_{\mu\nu\lambda} \partial_{\nu}
322: \partial_{\lambda} \theta_v.
323: \end{equation}
324: In terms of the staggered-spin-vortex
325: and anti-staggered-spin-vortex fields,
326: the vortex density is given by
327: \begin{equation}
328: \rho^v ({\bf r}) = \sum_s s \psi^{\dagger}_s ({\bf r})
329: \psi_s ({\bf r}).
330: \end{equation}
331: From the equation of the continuity, $\partial_t \rho^v + \nabla \cdot
332: {\bf j}^v =0$,
333: we find
334: \begin{equation}
335: {\bf j}^v =
336: \sum_s
337: \frac{s}{2mi} \left[ \psi_s^{\dagger} \nabla \psi_s
338: - (\nabla \psi_s^{\dagger}) \psi_s
339: \right],
340: \end{equation}
341: where we have used the fact that the kinetic energy of the holes is
342: described by Eq.(\ref{eq_ke}).
343: The coupling between the gauge field $A_{\mu}$ and the vortex current
344: $j_{\mu}^v$ is of the minimal coupling:
345: \begin{equation}
346: {\cal L}_{int} = -ij_{\mu}^v A_{\mu}.
347: \label{eq_coupling}
348: \end{equation}
349:
350:
351: Since $J_0=\nabla \times {\bf A}/2\pi$ and $J_0$ describes phase
352: fluctuations of the Schwinger boson density,
353: nonvanishing gauge flux $\nabla \times {\bf A}$
354: describes the suppression of
355: the phase coherence.
356: Contrary to the gauge flux
357: $\nabla \times
358: {\mbox{\boldmath ${\bf \alpha}$}}$,
359: the flux quantum for the Meissner phase of ${\bf A}$ is
360: $2\pi$ because staggered-spin-vortex carry a unit gauge charge.
361: Therefore, the relation between the vortex density and the
362: gauge flux is
363: \begin{equation}
364: \sum_s \psi^{\dagger}_s ({\bf r})
365: \psi_s ({\bf r})
366: =-\frac{1}{2\pi} \nabla \times {\bf A}.
367: \label{eq_flux}
368: \end{equation}
369: The gauge field $A_{\mu}$ is equivalent to the gauge field in the
370: slave-boson mean field theory. \cite{LEE_NAGAOSA02}
371:
372:
373: \section{Pairing interaction}
374: \label{sec_pairing}
375: The pairing interaction is obtained from
376: Eqs.(\ref{eq_flux}) and (\ref{eq_coupling})
377: by eliminating the gauge field $A_{\mu}$.
378: We take the Coulomb gauge:
379: \begin{equation}
380: \nabla \cdot {\bf A} =0.
381: \end{equation}
382: In the momentum space,
383: \begin{equation}
384: A_{qx} = \frac{iq_y}{q^2} A_q,
385: \hspace{2em}
386: A_{qy} = -\frac{iq_x}{q^2} A_q.
387: \end{equation}
388: From Eq.(\ref{eq_flux}) we obtain
389: \begin{equation}
390: A_q = -2\pi \sum_{ks} \psi^{\dagger}_{ks} \psi_{k+q,s}.
391: \end{equation}
392: The staggered-spin-vortex current operator is
393: \begin{equation}
394: {\bf j}_{\bf q}^v =
395: \sum_{{\bf k},s}
396: \frac{s}{m} \left( {\bf k}+\frac{{\bf q}}{2} \right)
397: \psi^{\dagger}_{ks} \psi_{k+q,s}.
398: \end{equation}
399: Substituting these equations into Eq.(\ref{eq_coupling}), we obtain
400: \begin{equation}
401: H_{\rm int}=-\frac{2\pi i}{m\Omega}
402: \sum_{q ,s,s'}
403: \frac{{\bf q}\times {\bf k}'}{q^2}s'
404: \psi^{\dagger}_{k+q,s} \psi_{k's'}^{\dagger}
405: \psi_{k'+q,s'}\psi_{ks},
406: \end{equation}
407: where $\Omega$ is the area of the system.
408: Since we are interested in the Cooper pairing, we set ${\bf k}+{\bf
409: k}'+{\bf q}=0$.
410: Thus, we obtain
411: \begin{equation}
412: H_{\rm int}=-\frac{2\pi i}{m\Omega}
413: \sum_{{\bf k}\neq{\bf k}',s,s'}
414: \frac{{\bf k}\times {\bf k}'}{|{\bf k}-{\bf k}'|^2}
415: s'
416: \psi^{\dagger}_{k's} \psi_{-k',s'}^{\dagger}
417: \psi_{-k,s'}\psi_{ks}.
418: \end{equation}
419: By making replacement of $s\rightarrow s'$,
420: $s'\rightarrow s$, ${\bf k}'\rightarrow -{\bf k}'$,
421: and ${\bf k}\rightarrow -{\bf k}$,
422: we obtain
423: \begin{equation}
424: H_{\rm int}=-\frac{2\pi i}{m\Omega}
425: \sum_{{\bf k}\neq{\bf k}',s,s'}
426: \frac{{\bf k}\times {\bf k}'}{|{\bf k}-{\bf k}'|^2} s
427: \psi^{\dagger}_{k's} \psi_{-k',s'}^{\dagger}
428: \psi_{-k,s'}\psi_{ks}.
429: \end{equation}
430: Therefore, we may write
431: \begin{equation}
432: H_{\rm int}=- \frac{i\pi}{m\Omega}
433: \sum_{{\bf k}\neq{\bf k}',s,s'}
434: \frac{{\bf k}\times {\bf k}'}{|{\bf k}-{\bf k}'|^2}(s+s')
435: \psi^{\dagger}_{k's} \psi_{-k',s'}^{\dagger}
436: \psi_{-k,s'}\psi_{ks}.
437: \label{eq_pairing}
438: \end{equation}
439: From this form, it is apparent that the interaction exists for either
440: staggered-spin-vortex pairs or anti-staggered-spin-vortex pairs.
441: The Hamiltonian is given by
442: \begin{equation}
443: H=\sum_s H^{(s)},
444: \label{eq_h}
445: \end{equation}
446: \begin{equation}
447: H^{(s)} = \sum_{\bf k} \epsilon_k \psi_{ks}^{\dagger}
448: \psi_{ks}
449: - \frac{2\pi i}{m\Omega} \sum_{{\bf k}\neq {\bf k}'}
450: \frac{{\bf k}\times {\bf k}'}{|{\bf k}-{\bf k}'|^2}
451: \psi^{\dagger}_{k's} \psi_{-k',s}^{\dagger}
452: \psi_{-k,s}\psi_{ks}.
453: \label{eq_hs}
454: \end{equation}
455:
456:
457: A similar pairing interaction was discussed in the composite fermion
458: system at half-filled Landau levels.\cite{GWW,TM_PRB00}.
459: In that system, the composite fermions are spinless fermion and there
460: is no other index associated with internal symmetry.
461: It was shown \cite{GWW2,TM_PRB00} that the pairing interaction leads
462: to $p_x \pm ip_y$-wave pairing state.
463: The sign is determined by the direction of the applied magnetic field
464: perpendicular to the system.
465: This pairing state is consistent with numerical
466: simulations.\cite{CF_NUM}
467:
468:
469: Before going into the analysis of the gap equation derived from the
470: Hamiltonian (\ref{eq_h}) and (\ref{eq_hs}),
471: we show an intuitive view about the origin of the attractive
472: interaction.
473: According to Eq.~(\ref{eq_flux}), the gauge flux $\nabla \times {\bf
474: A}$
475: is produced by the doped holes.
476: Suppose a hole passes with the Fermi velocity the region of the gauge
477: flux created by another hole.
478: Then, as schematically shown in Fig.~\ref{fig_p}
479: the motion of the former hole is equivalent to a charged particle
480: motion under a magnetic field.
481: Magnetic Lorentz force like interaction is induced between the two
482: holes.
483: Such an interaction leads to a chiral pairing state.
484: In this pairing mechanism, the gap is scaled by the Fermi energy.
485: Since holes carry either positive or negative gauge charge,
486: we expect there are two chiral pairing states with opposite
487: chiralities.
488: The stable pairing state in the bulk turns out to be
489: $d_{x^2-y^2}$-wave pairing state as we shall show in the next section.
490: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
491: % Figure fig_p.eps
492: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
493: \begin{figure}[htbp]
494: \includegraphics[scale=0.3]{fig1.eps}
495: \caption{Pairing interaction between the doped holes.
496: Shaded region represents the gauge flux $\nabla \times {\bf A}$
497: created by the hole sitting at the center.
498: If another hole passes this region with the Fermi velocity as shown in
499: the figure, a magnetic Lorentz force like interaction is induced
500: between the two holes.
501: }
502: \label{fig_p}
503: \end{figure}
504:
505:
506:
507:
508: \section{Gap equation}
509: \label{sec_gap}
510: Now we apply the BCS mean field theory to Eq.(\ref{eq_h}).
511: First we consider staggered-spin-vortex pairs described by $H^{(+)}$.
512: In the following, we set $\chi_k = \psi_{k+}$ and
513: $\chi_k^{\dagger}=\psi_{k+}^{\dagger}$,
514: to simplify notations.
515: We consider the grand canonical ensemble and
516: define the following mean field:
517: \begin{equation}
518: \Delta_k^{(+)} = -\frac{1}{\Omega}
519: \sum_{{\bf k}'(\neq {\bf k})}
520: V_{{\bf k}{\bf k}'}
521: \langle \chi_{-{\bf k}'} \chi_{{\bf k}'}
522: \rangle,
523: \end{equation}
524: where
525: \begin{equation}
526: V_{{\bf k}{\bf k}'} = -\frac{4\pi i}{m}
527: \frac{{\bf k}\times {\bf k}'}{|{\bf k}-{\bf k}'|^2}.
528: \end{equation}
529: The mean field Hamiltonian is
530: \begin{equation}
531: H^{(+)}_{MF} = {\sum_{\bf k}}' \left[
532: \xi_k \chi_{k}^{\dagger} \chi_{k}
533: -\xi_k \chi_{-k} \chi_{-k}^{\dagger}
534: +(\Delta_k^{(+)})^* \chi_{-k} \chi_{k}
535: +\Delta_k^{(+)} \chi^{\dagger}_{k} \chi^{\dagger}_{-k}
536: \right],
537: \end{equation}
538: where $\xi_k = \epsilon_k-\mu$ with $\mu$ the chemical potential and
539: the summation in the momentum space is taken over half of the
540: Brillouin zone.
541: The gap equation is given by
542: \begin{equation}
543: \Delta_{\bf k}^{(+)} = -\frac{1}{2\Omega}
544: \sum_{{\bf k}'(\neq {\bf k})}
545: V_{{\bf k}{\bf k}'}
546: \frac{\Delta_{{\bf k}'}^{(+)}}{E_{{\bf k}'}}
547: \tanh \frac{\beta E_{{\bf k}'}}{2},
548: \label{eq_gap}
549: \end{equation}
550: with $E_{\bf k}=\sqrt{\xi_k^2+|\Delta_{\bf k}^{(+)}|^2}$.
551:
552:
553: A similar gap equation is analyzed in Ref.\cite{GWW} in the context of
554: the composite fermion pairing.
555: We apply the analysis presented in Ref.\cite{GWW}.
556: In order to solve the gap equation, we introduce the following
557: Ansatz\cite{GWW2}:
558: \begin{equation}
559: \Delta_{\bf k}^{(+)} = \Delta_k
560: \exp \left( -i\ell \theta_{\bf k} \right),
561: \end{equation}
562: where $\ell$ is an integer.
563: In Eq.~(\ref{eq_gap})
564: the integral with respect to $\theta_{{\bf k}'}$ is reduced to
565: \begin{equation}
566: I_{\ell}(\eta)=\int_0^{2\pi} d\theta \frac{\sin \theta}{\cos \theta -\eta}
567: \exp (i\ell \theta),
568: \end{equation}
569: with $\eta=(k^2+k'^2)/(2kk')$.
570: The function $I_{\ell} (\eta)$ is exactly calculated by setting
571: $z=\exp (i\theta)$
572: and applying a contour integral in the complex plane.
573: From Eq.~(\ref{eq_gap}), we see that if the interaction is attractive
574: (repulsive) then the sign of $I_{\ell}(\eta)$ is positive (negative).
575: For $\ell >0$ case, $I_{\ell}(\eta)>0$ while $\ell <0$ case,
576: $I_{\ell}(\eta)<0$, and $I_{\ell=0}(\eta)=0$.
577: Therefore, the gap equation has solutions only for $\ell>0$.
578: Thus, we obtain
579: \begin{equation}
580: \Delta_k = \frac{1}{2m}
581: \left[
582: \int_0^k dk'
583: \frac{k'\Delta_{k'}}{E_{k'}}
584: \left( \frac{k'}{k} \right)^{\ell}
585: +
586: \int_k^{\infty} dk' \frac{k'\Delta_{k'}}{E_{k'}}
587: \left( \frac{k}{k'} \right)^{\ell} \right].
588: \end{equation}
589: From the asymptotic forms in $k\rightarrow \infty$ and
590: $k\rightarrow 0$, we take the following approximate form:\cite{GWW}
591: \begin{equation}
592: \Delta_k =
593: \left\{
594: \begin{array}{ccc}
595: \Delta \epsilon_F (k/k_F)^{\ell} & {\rm for} & k<k_F \\
596: \Delta \epsilon_F (k_F/k)^{\ell} & {\rm for} & k>k_F
597: \end{array}
598: \right.,
599: \end{equation}
600: where $\epsilon_F$ is the Fermi energy of the holes.
601: The gap $\Delta$ is obtained from the following equation:
602: \begin{equation}
603: \int_0^1 dx \frac{x^{2\ell+1}}{\sqrt{(x^2-1)^2+\Delta^2 x^{2\ell}}}
604: + \int_1^{\infty} dx
605: \frac{x^{1-2\ell}}{\sqrt{(x^2-1)^2+\Delta^2 x^{-2\ell}}}
606: =1.
607: \end{equation}
608: The largest gap $\Delta = 0.916$ is obtained for the case of
609: $p$-wave ($\ell=1$).
610: The second largest gap $\Delta =0.406$ is obtained for the case of
611: $d$-wave ($\ell=2$).
612: The third largest gap is $\Delta=0.264$ for $\ell=3$.
613:
614:
615:
616: For anti-staggered-spin-vortex pairs, we may carry out the same
617: analysis.
618: Since the interaction term has opposite sign compared to the
619: staggered-spin-vortex
620: pair case, the gap function $\Delta_{\bf k}^{(-)}$ has the
621: following form:
622: \begin{equation}
623: \Delta_{\bf k}^{(-)} = \Delta_k
624: \exp \left( i\ell \theta_{\bf k} \right),
625: \end{equation}
626: with $\ell>0$.
627:
628:
629: From the above analyses, the linear combination of the gap functions
630: is
631: \begin{equation}
632: \Delta_{\bf k}=(\Delta_{\bf k}^{(+)}+\Delta_{\bf k}^{(-)})/2=
633: \Delta_k \cos \left( \ell \theta_{\bf k} \right).
634: \end{equation}
635: The relative phase between $\Delta_{\bf k}^{(+)}$
636: and $\Delta_{\bf k}^{(-)}$ is set to be zero.
637: (A non-zero phase can appear if there is a magnetic field.)
638: For the case of $\ell=1$, the symmetry of the Cooper pair is $p_x$.
639: Such a state can be stablized only at the boundary of the sample in
640: the square lattice.
641: Meanwhile, $\ell=2$ case, that is, $d_{x^2-y^2}$-wave pairing state is
642: stable in the bulk.
643: %Although the possibility of $d_{xy}$-symmetry is not ruled out,
644: %such a state has a smaller gap than that of $d_{x^2-y^2}$ symmetry on
645: %the square lattice.
646: Thus, we may conclude that the above pairing interaction leads to
647: $d_{x^2-y^2}$-wave superconductivity.
648:
649:
650: %spin
651: The above analysis is carried out for the staggered-spin-vortices.
652: In terms of those fields, the hole spin states are implicit.
653: For spin singlet pairing states, the spin states should be
654: \begin{equation}
655: \langle \chi_{-k}^{\downarrow} \chi_k^{\uparrow}
656: \rangle,
657: \end{equation}
658: and
659: \begin{equation}
660: \langle \chi_{-k}^{\uparrow} \chi_k^{\downarrow}
661: \rangle,
662: \end{equation}
663: where $\sigma$ in $\chi_k^{\sigma}$ denotes the hole spin state.
664: The relative phase, which is arbitrary in the above analysis,
665: is $\pi$ for the spin-singlet pairing state:
666: \begin{equation}
667: \langle \chi_{-k}^{\downarrow} \chi_k^{\uparrow}
668: \rangle=
669: -\langle \chi_{-k}^{\uparrow} \chi_k^{\downarrow}
670: \rangle.
671: \end{equation}
672:
673:
674:
675: Now we consider the effect of the staggered-spin vortex-vortex
676: interaction and the Coulomb interaction.
677: Those repulsive interactions reduce the gap value.
678: In the following we consider those effects separately.
679: %vortex-vortex interaction
680: In order to evaluate the vortex-vortex interaction effect,
681: we consider a two-vortex solution as follows:
682: \begin{equation}
683: \theta_v = \tan^{-1} \frac{y}{x} + \tan^{-1} \frac{y}{x-d}.
684: \end{equation}
685: The interaction energy between the vortices is
686: \begin{equation}
687: V_{vv} (d) = \frac{J}{2}Z_{\rho} \overline{\rho}
688: \int d^2 {\bf r}
689: \frac{{\hat z}\times {\bf r}}{r^2}
690: \cdot
691: \frac{{\hat z}\times \left( x-d,y \right)}{(x-d)^2+y^2}.
692: \end{equation}
693: Taking into account the fact that the range of the interaction is
694: limited by the antiferromagnetic correlation length $\xi_{AF}$, we
695: obtain
696: \begin{equation}
697: V_{vv}(d) \simeq \pi J Z_{\rho} \overline{\rho}
698: \ln \frac{\xi_{AF}}{d}.
699: \end{equation}
700: In the momentum space,
701: \begin{equation}
702: V_{vv} (q) = \frac{2\pi\xi_{AF}^2}{q^2} J Z_{\rho} \overline{\rho},
703: \end{equation}
704: with the constraint, $q>2\pi/\xi_{AF}$.
705:
706:
707: In the presence of the vortex-vortex interaction $V_{vv}(q)$, the gap
708: equation is given by
709: \begin{eqnarray}
710: 1&=&\int_0^1 dx \frac{x^{2\ell+1}}{\sqrt{(x^2-1)^2+\Delta^2 x^{2\ell}}}
711: + \int_1^{\infty} dx
712: \frac{x^{1-2\ell}}{\sqrt{(x^2-1)^2+\Delta^2 x^{-2\ell}}}
713: \nonumber \\
714: & & - C_v
715: \left[
716: \int_0^{1-\zeta} dx
717: \frac{x^{2\ell}}{\sqrt{(x^2-1)^2+\Delta^2 x^{2\ell}}}
718: \frac{x}{1-x^2}
719: + \int_{1+\zeta}^{\infty} dx
720: \frac{x^{-2\ell}}{\sqrt{(x^2-1)^2+\Delta^2/x^{2\ell}}}
721: \frac{x}{x^2-1}
722: \right],
723: \end{eqnarray}
724: where $\zeta=2\pi/(k_F \xi_{AF})$ and the constant $C_v$ is
725: \begin{equation}
726: C_v = \frac{\pi J}{8\epsilon_F} Z_{\rho} \overline{\rho}\xi_{AF}^2.
727: \end{equation}
728: We take $C_v \simeq 0.5$ as a reasonable value at $x=0.10$
729: by setting
730: $\overline{\rho} \xi_{AF}^2 \sim 1$,
731: $J/\epsilon_F \sim 2$.
732: For $Z_{\rho}$, the value of $Z_{\rho}=0.72$ is used which is
733: estimated from quantum Monte Carlo simulations.\cite{QMC}
734: Figure \ref{fig_vv} shows $k_F \xi_{AF}$ dependence of the gap
735: parameters for each $\ell$.
736: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
737: % Figure fig_vv.eps
738: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
739: \begin{figure}[htbp]
740: \includegraphics[scale=0.5]{fig2.eps}
741: \caption{The gap parameter $\Delta$ versus
742: $\xi_{AF}k_F/(2\pi)$.}
743: \label{fig_vv}
744: \end{figure}
745: The effect of the vortex-vortex interaction is large
746: for strong antiferromagnetic correlations.
747: The gap values are somewhat reduced by the vortex-vortex interaction.
748: While the relevant parameter region would be
749: $\xi_{AF} k_F/(2\pi) = 1\sim 2$,
750: around $\xi_{AF} k_F/(2\pi)\sim 3$,
751: the f-wave gap becomes larger than the d-wave gap.
752: Note that as an approximation,
753: $C_v$ is fixed in the above calculation for simplicity.
754: For more precise calculations,
755: we require detail analysis of $\xi_{AF}
756: k_F$ dependence of $\overline{\rho}$ and $Z_{\rho}$.
757: %vortex-anti-vortex
758: %If we consider the vortex-anti-vortex interaction, then there is
759: %another attractive interaction.
760: %However, the gap values associated with such an interaction are
761: %much smaller than the above gap values.
762:
763:
764: Another pairing mechanism based on merons, which presumably correspond
765: to the spin texture with non-zero topological density,
766: was discussed by Berciu and John.\cite{BJ99,BJ00}
767: From the analysis of an extended Hubbard model, where a
768: nearest-neighbor Coulomb repulsion is included, it was suggested that
769: meron-antimeron pairs lead to d-wave superconductivity.
770: In Ref.\cite{BJ99,BJ00}, the pairing interaction between merons and
771: antimerons comes from the vortex-antivortex interaction,
772: and the pairing interaction (\ref{eq_pairing}) is not considered.
773: Although the vortex-antivortex interaction is attractive interaction,
774: the analysis of the gap equation for the vortex-antivortex interaction
775: within our Ansatz shows that there is no pairing state for
776: $\xi_{AF} k_F/(2\pi) < 13$.
777: In this analysis, the coefficient $C_v$ is fixed.
778: More precise analysis requires $\xi_{AF}$ dependence of $C_v$.
779: However, $C_v$ is expected to be an increasing function with respect
780: to $\xi_{AF}$, and
781: the increase of $C_v$ leads to reduction of the gap values for large
782: $\xi_{AF}$.
783:
784:
785:
786:
787: Now we consider the effect of the Coulomb interaction.
788: The Coulomb interaction between the holes is
789: \begin{equation}
790: V^C_q = \frac{2\pi e^2}{\epsilon q},
791: \end{equation}
792: with $\epsilon$ the dielectric constant.
793: The gap equation with the Coulomb interaction is
794: \begin{eqnarray}
795: 1&=&\int_0^1 dx \frac{x^{2\ell+1}}{\sqrt{(x^2-1)^2+\Delta^2 x^{2\ell}}}
796: + \int_1^{\infty} dx
797: \frac{x^{1-2\ell}}{\sqrt{(x^2-1)^2+\Delta^2 x^{-2\ell}}}
798: \nonumber \\
799: & & - C_e
800: \left[
801: \int_0^{1} dx
802: \frac{x^{2\ell+1/2}}{\sqrt{(x^2-1)^2+\Delta^2 x^{2\ell}}}
803: J_{\ell} \left(\frac{x^2+1}{2x} \right)
804: + \int_1^{\infty} dx
805: \frac{x^{1/2-2\ell}}{\sqrt{(x^2-1)^2+\Delta^2/x^{2\ell}}}
806: J_{\ell} \left(\frac{x^2+1}{2x} \right)
807: \right],
808: \end{eqnarray}
809: where
810: \begin{equation}
811: J_{\ell} (\eta)=\int_0^{2\pi} d\theta
812: \frac{\cos (\ell \theta)}{\sqrt{\eta-\cos \theta}},
813: \end{equation}
814: \begin{equation}
815: C_e = \frac{e^2 k_F}{8\sqrt{2}\pi \epsilon}.
816: \end{equation}
817: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
818: % Figure fig_c.eps
819: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
820: \begin{figure}[htbp]
821: \includegraphics[scale=0.5]{fig3.eps}
822: \caption{The gap $\Delta$ versus $1/\epsilon$.}
823: \label{fig_c}
824: \end{figure}
825: Figure \ref{fig_c} shows $1/\epsilon$ dependence of the gap $\Delta$
826: for each $\ell$.
827:
828:
829:
830:
831: \section{Conclusion}
832: \label{sec_conclusion}
833: In this paper, we have proposed a mechanism of $d_{x^2-y^2}$-wave
834: superconductivity based on a spin texture with non-zero topological
835: charge density.
836: The spin texture formation is based on the Zhang-Rice singlet
837: formation in the background of the antiferromagnetic correlations.
838: In terms of a gauge field that describes antiferromagnetic spin
839: correlations, the spin texture is described by a gauge flux.
840: The interaction between the flux and the hole current induces
841: the magnetic Lorentz force like interaction between the holes.
842: Such an interaction leads to chiral pairing states with opposite
843: chiralities.
844: It turns out that the stable pairing state in the bulk is the
845: $d_{x^2-y^2}$-wave pairing state.
846: The stability of the pairing state is examined against the
847: vortex-vortex interaction and the Coulomb interaction.
848: %The appearance of the antiferromagnetic correlations is associated
849: %with the higher pseudogap temperature $T_0$.\cite{TIMUSK_STATT}
850: %While the formation of the Zhang-Rice singlet occurs at the Kondo
851: %temperature, $T_K(x)$.
852: %In the N{\' e}el ordering phase, $T_K(x)$ would be
853: %$k_B T_K(x) \sim 1{\rm eV}$.
854: %However, in the spin disordered phase, $T_K(x)$ could be on the order
855: %of $100K$ because the magnitude of the localized moment at copper
856: %sites is significantly reduced.
857: %It would be interesting to consider the relationship between
858: %$T_K(x)$ and the lower pseudogap temperature
859: %$T^*$.\cite{TIMUSK_STATT}
860:
861:
862: In our pairing mechanism, the $p$-wave state is unstable in the bulk.
863: However, this state can be stabilized in the presence of anisotropy.
864: Since the $p$-wave gap is much larger than the d-wave gap, we expect
865: enhancement of the superconducting transition temperature if we can
866: induce that component.
867: If there is the $p$-wave component, then the parity is broken locally.
868: While the time-reversal symmetry is not broken unless one chiral
869: pairing state is suppressed.
870:
871:
872:
873:
874:
875:
876:
877: \acknowledgments
878: I would like to thank Professor G. Baskaran for discussions and his
879: kind hospitality at Institute for Mathematical Sciences where part of
880: this work was done.
881: I also thank Dr. N. Nakai for discussions.
882: This work was supported by Grant-in-Aid for Young Scientists(B)
883: (17740253) and the 21st Century COE "Center for Diversity and
884: Universality in Physics" from the Ministry of Education, Culture,
885: Sports, Science and Technology (MEXT) of Japan.
886:
887:
888:
889:
890: %\section{Duality mapping approach}
891: %\label{sec_duality}
892:
893: \begin{thebibliography}{99}
894: %\begin{references}
895: %\begin{bibliography}
896: \bibitem{ANDERSON}P.~W.~Anderson, Science {\bf 235}, 1196 (1987).
897: \bibitem{Harlingen95} See, for example, D.~J.~Van Harlingen,
898: Rev. Mod. Phys. {\bf 67}, 515 (1995).
899: \bibitem{Pines} P.~Monthoux, A.~V.~Balatsky, and D.~Pines,
900: Phys. Rev. B {\bf 46}, 14803 (1992);
901: T.~Moriya, Y.~Takahashi, and K.~Ueda,
902: J. Phys. Soc. Jpn. {\bf 59}, 2905 (1990).
903: \bibitem{LNW04} See, for a review, P.~A.~Lee, N.~Nagaosa, and
904: X.-G.~Wen, cond-mat/0410445.
905: \bibitem{ABZH87} P.~W.~Anderson, G.~Baskaran, Z.~Zou, and T.~Hsu,
906: Phys. Rev. Lett. {\bf 58}, 2790 (1987).
907: \bibitem{Takagi89}H.~Takagi, T.~Ido, S.~Ishibashi, M.~Uota, S.~Uchida,
908: and Y.~Tokura, Phys. Rev. B {\bf 40}, 2254 (1989).
909: \bibitem{Uchida91}S.~Uchida, T.~Ido, H.~Takagi, T.~Arima,
910: Y.~Tokura, and S.~Tajima, Phys. Rev. B {\bf 43}, 7942 (1991).
911: \bibitem{TM_PRB05}T.~Morinari, to be published in Phys. Rev. B.
912: \bibitem{SD02} A.~Saxena and R.~Dandoloff,
913: Phys.~Rev.~B {\bf 66}, 104414 (2002).
914: \bibitem{BASKARAN}G.~Baskaran, Phys. Rev. B {\bf 68}, 212409 (2003).
915: \bibitem{CHN}S.~Chakravarty, B.~I.~Halperin, and D.~R.~Nelson,
916: Phys.~Rev.~B {\bf 39}, 2344 (1989).
917: \bibitem{FW} See, for example, A.~L.~Fetter and J.~D.~Walecka,
918: Quantum Theory of Many-Particle Systems (McGraw-Hill, 1971).
919: \bibitem{AA88} D.~P.~Arovas and A.~Auerbach,
920: Phys.~Rev.~B {\bf 38}, 316 (1988).
921: \bibitem{SB_BEC}J.~E.~Hirsch and S.~Tang, Phys.~Rev.~B {\bf 39}, 2850
922: (1989); D.~Yoshioka, J.~Phys.~Soc.~Jpn. {\bf 58}, 3733 (1989);
923: M.~Raykin and A.~Auerbach, Phys.~Rev.~Lett. {\bf 70}, 3808(1993).
924: \bibitem{skyrmions}
925: B.~I.~Shraiman and E.~D.~Siggia,
926: Phys.~Rev.~Lett. {\bf 61}, 467 (1988);
927: R.~J.~Gooding, Phys. Rev. Lett. {\bf 66}, 2266
928: (1991);
929: J.~P.~Rodriguez, Phys.~Rev.~B {\bf 39}, R2906 (1989);
930: A.~Auerbach, B.~E.~Larson, and G.~N.~Murthy, Phys.~Rev.~B {\bf 43},
931: R11515 (1991);
932: S.~Haas, F.-C.~Zhang, F.~Mila, and T.~M.~Rice,
933: Phys.~Rev.~Lett. {\bf 77}, 3021 (1996);
934: E.~C.~Marino, Phys.~Rev.~B {\bf 61}, 1588 (2000).
935: \bibitem{ZHANG_RICE}F.~C.~Zhang and T.~M.~Rice, Phys. Rev. B {\bf 37},
936: R3759 (1988).
937: \bibitem{WENG_ETAL} The same result of half of a topological charge
938: was obtained by S.-P. Kou, X.-L Qi, and Z.-Y. Weng,
939: Phys. Rev. B {\bf 71}, 235102 (2005)
940: in a different formalism.
941: \bibitem{WELLS_ETAL}B.~O.~Wells, Z.-X.~Shen,
942: A.~Matsuura, D.~M.~King, M.~A.~Kastner, M.~Greven,
943: and R.~J.~Birgeneau,
944: Phys. Rev. Lett. {\bf 74}, 964 (1995).
945: \bibitem{MARSTON}J.~B.~Marston, Phys.~Rev.~Lett. {\bf 64}, 1166
946: (1990).
947: \bibitem{KIM_LEE}D.~H.~Kim and P.~A.~Lee, Ann. Phys. {\bf 272}, 130
948: (1999).
949: \bibitem{TM_BROAD}T.~Morinari, preprint, cond-mat/0508251.
950: \bibitem{RAJARAMAN}See, for example,
951: R.~Rajaraman, Solitons and Instantons (North-Holland, 1987).
952: \bibitem{LEE_NAGAOSA02} P.~A.~Lee and N.~Nagaosa,
953: Phys. Rev. B {\bf 46}, 5621 (1992);
954: N.~Nagaosa and P.~A.~Lee, Phys. Rev. Lett. {\bf 64}, 2450 (1990).
955: \bibitem{WWZ} X.-G.~Wen, F.~Wilczek, and A.~Zee,
956: Phys. Rev. B {\bf 39}, 11413 (1989).
957: \bibitem{GWW}M.~Greiter, X.~G.~Wen, and F.~Wilczek,
958: Nucl. Phys. B {\bf 374}, 567 (1992).
959: \bibitem{TM_PRB00} T.~Morinari, Phys. Rev. B {\bf 62}, 15903 (2000);
960: Phys. Rev. Lett. {\bf 81}, 3741 (1998).
961: \bibitem{GWW2} See Sec.3.1 in Ref.\cite{GWW}.
962: \bibitem{CF_NUM} R.~H.~Morf, Phys. Rev. Lett. {\bf 80}, 1505 (1998);
963: K.~Park, V.~Melik-Alaverdian, N.~E.~Bonesteel, and J.~K.~Jain,
964: Phys. Rev. B {\bf 58}, R10167 (1998);
965: E.~H.~Rezayi and F.~D.~M.~Haldane,
966: Phys. Rev. Lett. {\bf 84}, 4685 (2000).
967: \bibitem{QMC} B.~B.~Beard, R.~J.~Birgeneau, M.~Greven, and
968: U.-J.~Wiese, Phys. Rev. Lett. {\bf 80}, 1742 (1998);
969: J.-K.~Kim and M.~Troyer, Phys. Rev. Lett. {\bf 80}, 2705 (1998).
970: \bibitem{BJ99}M.~Berciu and S.~John, Phys. Rev. B {\bf 59}, 15 143
971: (1999).
972: \bibitem{BJ00}M.~Berciu and S.~John, Phys. Rev. B {\bf 61}, 16 454
973: (2000).
974: %\bibitem{TIMUSK_STATT}T.~Timusk and B.~Statt,
975: %Rep. Prog. Phys. {\bf 62}, 61 (1999).
976: %\bibitem{Fetter72}A.~L.~Fetter, Ann. Phys. {\bf 70}, 67 (1972).
977: %\bibitem{}
978: %\end{references}
979: %\end{bibliography}
980: \end{thebibliography}
981: \end{document}
982:
983: %\widetext
984:
985: %\end{multicols}
986:
987: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
988: % Figure
989: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
990: \begin{figure}[htbp]
991: \includegraphics[scale=0.5]{*.eps}
992: \caption{}
993: \label{fig_*}
994: \end{figure}
995:
996:
997:
998:
999: