1: \documentstyle[aps,multicol,epsfig]{revtex}
2: %\documentstyle[aps,prb,preprint]{revtex}
3: \begin{document}
4: \draft
5: \title{Tunneling effect on composite fermion pairing state in bilayer
6: quantum Hall system}
7: \author{Takao Morinari}
8: \address{Yukawa Institute for Theoretical Physics,
9: Kyoto University, Kyoto 606-8502, Japan}
10:
11: \date{\today}
12: \maketitle
13: \begin{abstract}
14: We discuss the composite fermion pairing state in bilayer quantum Hall
15: systems.
16: After the evaluation of the range of the inter-layer
17: separation in which the quantum Hall state is stabilized,
18: we discuss the effect of inter-layer tunneling on the composite
19: fermion pairing state at $\nu=1/2$.
20: We show that there is a cusp at the transition point between the
21: Halperin $(3,3,1)$ state and the Pfaffian state.
22: \end{abstract}
23:
24: \pacs{73.40.Hm, 71.10.Pm}
25:
26: \begin{multicols}{2}
27: \narrowtext
28: \section{Introduction}
29: As a remarkable example of strongly correlated electron systems,
30: the fractional quantum Hall effect has been studied
31: extensively.\cite{review,review2}
32: In the fractional quantum Hall systems, the Coulomb interaction plays
33: a dominant role.
34: Due to the presence of the strong magnetic field, the Coulomb
35: interaction gives rise to a two-body correlation of non-zero relative
36: angular momentum.
37: The Laughlin wave function,\cite{LAUGHLIN} which captures
38: the essential properties of the system, consists of this two-body
39: correlation only.
40: The role of the Coulomb interaction to generate the two-body
41: correlation becomes clear when it is described in terms of Haldane's
42: pseudopotential.\cite{HALDANE}
43: From the analysis of the pseudopotentials, one can see that
44: the most fundamental contribution comes from the short-range Coulomb
45: interaction.
46:
47: In the single-layer quantum Hall systems, it is hard to change the
48: short-range Coulomb interaction.
49: Whereas in the bilayer quantum Hall systems, we can adjust the
50: inter-layer Coulomb interaction by varying the inter-layer separation
51: $d$.
52: In addition, there is another parameter, the inter-layer tunneling
53: amplitude $t$ to control the system.
54: Since there are such additional parameters to control the system, we
55: can expect rich phase diagram in the bilayer quantum Hall systems.
56:
57: In the absence of inter-layer tunneling, the properties of the system
58: seems to be well described by the Halperin $(m,m,n)$ wave
59: function\cite{HALPERIN_MMN}, which is an extention of the Laughlin
60: wave function:
61: \begin{eqnarray}
62: \lefteqn{\Psi_{mmn} (z_1^{\uparrow},z_2^{\uparrow},\cdots,z_N^{\uparrow};
63: z_1^{\downarrow},z_2^{\downarrow},\cdots,z_N^{\downarrow})}
64: \nonumber \\
65: &=&\prod_{i<j} (z_i^{\uparrow}-z_j^{\uparrow})^m
66: \prod_{i<j} (z_i^{\downarrow}-z_j^{\downarrow})^m
67: \prod_{i,j} (z_i^{\uparrow}-z_j^{\downarrow})^n
68: \nonumber \\ & & \times
69: \exp \left[ - \frac{1}{4\ell_B^2}
70: \sum_j (|z_j^{\uparrow}|^2 + |z_j^{\downarrow}|^2) \right],
71: \label{mmn_wf}
72: \end{eqnarray}
73: where $\uparrow$ and $\downarrow$ are the indices for layers and
74: $z_j^{\sigma} = x_j^{\sigma} - i y_j^{\sigma}$ is the complex
75: coordinate of the $j$-th electron.
76: The integer numbers $m$ and $n$ is associated with
77: the total Landau level filling $\nu$ as $\nu=2/(m+n)$.
78:
79: At $\nu=1/2$, the $(3,3,1)$ wave function has good overlap with the
80: wave function of the finite size system \cite{YOSHIOKA_ETAL}
81: in the absence of the inter-layer tunneling and in the appropriate
82: region of the inter-layer separation.
83: It was pointed out by Ho\cite{HO} that the $(3,3,1)$ state and
84: the Pfaffian state \cite{MOORE_READ}, which
85: may be stabilized in the strong inter-layer tunneling
86: limit,\cite{HALPERIN_BI}
87: are unified to the form of the p-wave pairing state of composite
88: fermions.
89:
90: In this paper, we discuss the stability of the composite fermion
91: paring states in the bilayer quantum Hall systems.
92: First, we discuss the range of the inter-layer separation $d$ in which
93: the $(m,m,n)$ state is stabilized and the appropriate choice of the
94: number of fluxes for composite fermions.
95: After deriving the mean field equations for the composite fermion
96: pairing states,
97: we consider the effect of the inter-layer tunneling $t$ on the p-wave
98: pairing state at $\nu=1/2$.
99: The evolution of the $(3,3,1)$ state to the Pfaffian state is shown
100: being based on the analysis of the gap equation.
101: We also examine the possibility of Haldane-Rezayi
102: state\cite{HALDANE_REZAYI} at $\nu=1/2$.
103:
104: The outline of this paper is as follows.
105: In Sec.~\ref{sec_two}, we determine the appropriate choice of $(m,n)$
106: and we describe the relationship between the Halperin $(m,m,n)$ state
107: and the p-wave pairing state of composite fermions.
108: The formulation for the analysis of the pairing state is given in
109: Sec.~\ref{sec_cfp}.
110: The gap equations for the triplet pairing state and the singlet
111: pairing state are derived in the presence of the inter-layer
112: tunneling.
113: In Sec.~\ref{sec_t}, we discuss the inter-layer tunneling effect.
114: Finally, we summarize the results in Sec.~\ref{sec_summary}.
115:
116: \section{Two-body correlation}
117: \label{sec_two}
118: In this section, we discuss the bilayer quantum Hall systems in the
119: absence of the inter-layer tunneling in order to find the appropriate
120: choice of the number of attached fluxes for composite fermions.
121: For the determination of those numbers, the effect of inter-layer
122: tunneling may be negligible because they are associated with the
123: two-body correlation due to the short-range Coulomb interaction as
124: discussed in Introduction.
125:
126: Since the two-body correlations, which is connected with
127: the numbers $m$ and $n$ in the $(m,m,n)$ wave function
128: (\ref{mmn_wf}), is associated with the short-range Coulomb
129: interaction, $m$ and $n$ are determined by ``high-enrgy'' physics.
130: We compare Haldane's pseudopotentials with various choices of $m$
131: and $n$ \cite{MORINARI_BI}
132: because ``high-energy'' physics is governed by Haldane's
133: pseudopotentials.
134:
135: The basis for the two-body electron correlations is given by the wave
136: function for an electron pair with the relative angular momentum $m$
137: and the angular momentum of the central motion being zero:
138: \begin{eqnarray}
139: \langle z_1,z_2|\psi_m \rangle
140: &=&
141: \frac{1}{\sqrt{4^{m+1}\ell_B^{2m+4}\pi^2 m!}}
142: \left( z_1-z_2 \right)^m
143: \nonumber \\ & & \times
144: \exp \left[ -\frac{1}{4\ell_B^2}
145: \left( |z_1|^2+|z_2|^2 \right) \right],
146: \label{eq_2}
147: \end{eqnarray}
148: where $\ell_B = \sqrt{c\hbar/eB}$ is the magnetic length.
149: Since Haldane's pseudopotential is equivalent to the Coulomb energy
150: estimated in first order, the total energy for
151: ``high-energy'' physics for total $N$ electrons is given by
152: \begin{equation}
153: E_C^{(2)}(m,n)
154: = \frac{N(N-1)}{2} \times \epsilon (m,d=0) \times 2
155: + N^2 \times \epsilon (n,d),
156: \label{pre_Ec2}
157: \end{equation}
158: where
159: \begin{eqnarray}
160: \epsilon (m,d)
161: &=& \langle \psi_m| \frac{e^2}{\epsilon \sqrt{r^2+d^2}}|\psi_m \rangle
162: \nonumber \\ &=&
163: \frac{e^2}{\epsilon \ell_B} \times \frac{1}{m!}
164: \int_0^{\infty} dx \frac{x^{2m+1}}{\sqrt{x^2+\lambda^2}} {\rm e}^{-x^2},
165: \end{eqnarray}
166: with $\epsilon$ the dielectric constant and $\lambda
167: =d/2\ell_B$.
168: In the thermodynamic limit, we obtain
169: \begin{equation}
170: E_C^{(2)}(m,n)/N^2=\epsilon (m,d=0) + \epsilon (n,d).
171: \label{eq_e2}
172: \end{equation}
173: Although the two-body correlation energies are manifestly
174: overestimated in $E_C^{(2)}(m,n)$, the right hand side of
175: Eq.~(\ref{eq_e2}) may be reliable as far as we are concerned with
176: the pseudopotentials.
177: For the choice of $m$ and $n$, we cannot choose arbitrary pair of
178: $m$ and $n$.
179: There is a constraint on the choice of $m$ and $n$.
180: From the Halperin $(m,m,n)$ wave function,
181: one can see that the angular momentum of the electron at the edge of the
182: sample is equal to $(N-1) \times m+N\times n \equiv M$.
183: Since the wave function of this electron is proportional to
184: $z^M {\rm e}^{-r^2/4\ell_B^2}$, the density of it has its maximum
185: at $r=\sqrt{2M} \ell_B \equiv R$.
186: Of course $\pi R^2$ is the area of the system.
187: Taking the thermodynamic limit $N \rightarrow \infty$, we obtain
188: \begin{equation}
189: 2\pi \ell_B^2 \times N \times (m+n) = \Omega.
190: \label{mn_area}
191: \end{equation}
192: We consider the case of symmetric electron density
193: $\rho_{\uparrow}=\rho_{\downarrow}$.
194: In this case, Eq.~(\ref{mn_area}) is reduced to
195: \begin{equation}
196: m + n = 2/\nu.
197: \label{constraint}
198: \end{equation}
199:
200: Now we determine $m$ and $n$ which gives the lowest $E_C^{(2)}(m,n)$
201: under the constraint (\ref{constraint}).
202: For the case of $\nu=1/2$, the constraint (\ref{constraint})
203: is $m+n=4$. Therefore, the possible choice of $(m,n)$ is,
204: $(4,0)$, $(3,1)$ and $(2,2)$.
205: The pair $(m,n)$ with $m<n$ always has larger energy than that with
206: $m\geq n$.
207: Note that in case of $(m,n)=({\rm even},{\rm even})$, the system is
208: not the quantum Hall state.
209: It is compressible liquid of composite fermions and the total wave
210: function can not be determined by the two-body correlations only.
211:
212: In Fig.~\ref{nu_half}, we show the energy $E_C^{(2)} (m,n)/N^2$ for
213: $(m,n)=(4,0),(3,1)$, and $(2,2)$.
214: The region where the choice of $(m,n)=(3,1)$
215: gives the lowest energy $E_C^{(2)}(m,n)$ is $0.789 < d/2\ell_B < 1.480$.
216: The Halperin $(3,3,1)$ state is stabilized in this region.
217:
218: For the case of $\nu=1$, the constraint (\ref{constraint})
219: is $m+n=2$. Therefore, the possible choice for $(m,n)$ is
220: $(2,0)$ and $(1,1)$.
221: In Fig.\ref{nu_one}, we show the energy $E_C^{(2)}(m,n)/N^2$ for
222: $(2,0)$ and $(1,1)$.
223: The region where the choice of $(m,n)=(1,1)$ gives the lowest energy
224: $E_C^{(2)}(m,n)$ is $0 < d/2\ell_B < 0.703(\equiv d_c)$.
225: The Halperin $(1,1,1)$ state is stabilized in this region.
226: Though the above estimation is crude, the critical value $d_c$ for
227: $\nu=1$ is close to $2 \ell_B$ that was obtained by
228: Murphy {\it et al.} experimentally \cite{MURPHY}.
229:
230: In general, the estimation of $E^{(2)}_C (m,n)$ shows that
231: the pair $(m,n)$ giving the lowest $E^{(2)}_C (m,n)$ is $(2/\nu,0)$
232: for $d \gg \ell_B$. As we decrease the value of $d$, it
233: changes as $(2/\nu,0) \rightarrow (2/\nu-1,1) \rightarrow \cdots
234: \rightarrow (1/\nu,1/\nu)$.
235: In this sequence, the quantum Hall state is stable at $(m,n)=({\rm
236: odd},{\rm odd})$, whereas the compressible state of composite fermions
237: is stable at $(m,n)=({\rm even},{\rm even})$.
238:
239: Now we discuss the relationship between the $(m,m,n)$ wave function
240: and the p-wave pairing state of composite fermions.\cite{HO}
241: As shown in Ref.~\cite{HO}, the wave function of the p-wave pairing
242: state of composite fermions at $\nu=1/2$ is, in the second quantized
243: form,
244: \begin{eqnarray}
245: |N,\chi \rangle
246: &=&
247: \int \prod_{j=1}^{N} d^2 z_j
248: \prod_{i<j} (z_i-z_j)^2
249: {\rm e}^{-\frac{1}{4\ell_B^2} \sum_{j=1}^{N} |z_j|^2}
250: \nonumber \\ & & \times
251: \prod_{j=1}^{N} \left[ \frac{\sum_{\sigma \sigma'}
252: \chi_{\sigma \sigma'}
253: \psi_{\sigma}^{\dagger} (z_{2j-1}) \psi_{\sigma'}^{\dagger} (z_{2j})}
254: {z_{2j-1}-z_{2j}} \right]~|0\rangle,
255: \end{eqnarray}
256: where $\psi_{\sigma}^{\dagger} (z)$ is the creation operator of the
257: electron at $z$ with spin $\sigma$.
258: For the $(3,3,1)$ state, $\chi$ is given by
259: $\chi= \left[\begin{array}{cc} 0 & 1 \\ 1 & 0 \end{array} \right]$.
260: Note that in this case the p-wave pairing wave function is the
261: $(1,1,-1)$ state.\cite{HALDANE_REZAYI,HO}
262: Therefore, in general the $(m,m,n)$ state is described by the p-wave
263: pairing state with $\chi= \left[\begin{array}{cc} 0 & 1 \\ 1 & 0
264: \end{array} \right]$ of composite fermions with the number of attached
265: fluxes $(\phi_1,\phi_2)=(m-1,n+1)$.
266: (The even integer $\phi_1$ ($\phi_2$) is for the intra(inter)-layer
267: correlations.)
268: For the Pfaffian state,
269: $\chi$ is given by $\chi=\left[\begin{array}{cc} 1/\sqrt{2} &
270: 1/\sqrt{2} \\
271: 1/\sqrt{2} & 1/\sqrt{2} \end{array} \right]$.
272: If the system is not the quantum Hall state, then the wave function of
273: composite fermions is not the form of the pairing state.
274:
275: \section{Composite fermion pairing}
276: \label{sec_cfp}
277: In the last section, we have discussed the appropritate choice of the
278: number of attached fluxes for composite fermions.
279: Now we introduce composite fermions in the second quantized form:
280: \begin{equation}
281: \left\{
282: \begin{array}{l}
283: \tilde{\psi}_{\alpha} ({\bf r})
284: = {\rm e}^{-iJ_{\alpha} ({\bf r})} \psi_{\alpha} ({\bf r}), \\
285: \tilde{\psi}^{\dagger}_{\alpha} ({\bf r})
286: = \psi_{\alpha}^{\dagger} ({\bf r}) {\rm e}^{iJ_{\alpha}
287: ({\bf r})},
288: \end{array}
289: \right.
290: \label{eq_cf}
291: \end{equation}
292: where the function $J_{\alpha} ({\bf r})$ is given by
293: \begin{equation}
294: J_{\alpha} ({\bf r}) = \sum_{\beta} K_{\alpha \beta}
295: \int d^2 {\bf r}^{\prime} \rho_{\beta} \left( {\bf r}^{\prime} \right)
296: {\rm Im} \ln \left( z - z^{\prime} \right).
297: \label{eq_J}
298: \end{equation}
299: Here $\rho_{\alpha} ({\bf r})
300: = \psi^{\dagger}_{\alpha} ({\bf r}) \psi_{\alpha} ({\bf r})
301: = \tilde{\psi}^{\dagger}_{\alpha} ({\bf r})
302: \tilde{\psi}_{\alpha} ({\bf r})$ and
303: $K=\left( \begin{array}{cc} \phi_1 & \phi_2 \\ \phi_2 & \phi_1
304: \end{array} \right)$ with $\phi_1$ and $\phi_2$ being even integer.
305: In terms of the composite fermion fields $\tilde{\psi}^{\dagger}$
306: and $\tilde{\psi}$, the kinetic energy term of the electrons is
307: rewritten as
308: \begin{eqnarray}
309: H_0
310: &=&
311: \sum_{\alpha} \frac{1}{2m} \int d^2 {\bf r}
312: \psi^{\dagger}_{\alpha} ({\bf r})
313: \left( -i \hbar \nabla + \frac{e}{c} {\bf A} \right)^2
314: \psi_{\alpha} ({\bf r}) \nonumber \\
315: &=&
316: \sum_{\alpha} \frac{1}{2m} \int d^2 {\bf r}
317: \tilde{\psi}^{\dagger}_{\alpha} ({\bf r})
318: \left( -i \hbar \nabla + \frac{e}{c} {\bf A}
319: + \frac{e}{c} {\bf a}_{\alpha}
320: \right)^2
321: \tilde{\psi}_{\alpha} ({\bf r}).\nonumber \\
322: \label{eq_H0}
323: \end{eqnarray}
324: Here ${\bf a}_{\alpha}$ is the Chern-Simons gauge field,
325: \begin{equation}
326: {\bf a}_{\alpha} ({\bf r}) = - \frac{ic\hbar}{e} \nabla
327: J_{\alpha} ({\bf r}).
328: \end{equation}
329: The Chern-Simons gauge field ${\bf a}_{\alpha}$ obeys the constraint
330: \begin{equation}
331: \nabla \times {\bf a}_{\alpha} ({\bf r} )
332: = \phi_0 \sum_{\beta} K_{\alpha \beta} \rho_{\beta} ({\bf r}),
333: \label{eq_flux}
334: \end{equation}
335: where $\phi_0 = ch/|e|$ is the flux quantum.
336:
337: The first order term of Eq.~(\ref{eq_H0}) with respect to the
338: fluctuation of the Chern-Simons gauge field
339: ${\bf A} + {\bf a}_{\alpha}$ yields the minimal coupling term.
340: Eliminating the Chern-Simons gauge field fluctuations upon using the
341: constraint (\ref{eq_flux}), we obtain
342: \begin{eqnarray}
343: V
344: &=& \frac{1}{2\Omega} \sum_{{\bf k}_1 \neq {\bf k}_2, {\bf q}}
345: \sum_{\alpha \beta} K_{\alpha \beta}
346: V_{{\bf k}_1, {\bf k}_2}
347: \tilde{\psi}^{\dagger}_{{\bf k}_1+{\bf q}/2,\alpha}
348: \tilde{\psi}^{\dagger}_{-{\bf k}_1+{\bf q}/2,\beta}
349: \nonumber \\ & & \times
350: \tilde{\psi}_{-{\bf k}_2+{\bf q}/2,\beta}
351: \tilde{\psi}_{{\bf k}_2+{\bf q}/2,\alpha},
352: \label{eq_Vp}
353: \end{eqnarray}
354: where $V_{{\bf k}_1 {\bf k}_2}=
355: \frac{4\pi i}{m} {\tilde \phi} \frac{{\bf
356: k}_1 \times {\bf k}_2}{|{\bf k}_1-{\bf k}_2|^2}$.
357: This interaction gives rise to an attractive interaction that leads to
358: the p-wave pairing state.\cite{GREITER_ETAL}
359: The second order term of Eq.~(\ref{eq_H0}) with respect to the
360: fluctuations of the Chern-Simons gauge field yields the three-body
361: interaction term after eliminating the Chern-Simons gauge field
362: fluctuations.
363: From the analysis of non-unitary transformation, this three-body
364: interaction term turns out to be the counter term to the short-range
365: Coulomb interaction.\cite{MORINARI}
366: However, if we restrict ourselves to the range of the inter-layer
367: separation, $d$ where the states
368: based on composite fermions are stabilized, we may neglect the
369: three-body interaction term.
370: In addition, we neglect the long-range Coulomb
371: interaction, which gives rise to a pair-breaking effect,
372: because the pairing state of compoisite fermions may be
373: stable in the region where the Halperin $(m,m,n)$ state is stable.
374: In the following analysis, we concentrate on the analysis of the
375: pairing interaction and the inter-layer tunneling.
376:
377: Including the inter-layer tunneling effect,
378: $H_t = -t \int d^2 {\bf r} \left[
379: \psi^{\dagger}_{\uparrow} ({\bf r})
380: \psi_{\downarrow} ({\bf r})
381: +
382: \psi^{\dagger}_{\downarrow} ({\bf r})
383: \psi_{\uparrow} ({\bf r})
384: \right]$,
385: the Hamiltonian for
386: composite fermions may be written as
387: \begin{eqnarray}
388: H &=& \sum_{{\bf k} \alpha \beta} \xi^{{\bf k}}_{\alpha \beta}
389: \tilde{\psi}^{\dagger}_{{\bf k} \alpha} \tilde{\psi}_{{\bf k} \beta}
390: \nonumber \\ & &
391: + \frac{1}{2\Omega} \sum_{{\bf k}_1 \neq {\bf k}_2} \sum_{\alpha \beta}
392: V_{{\bf k}_1,{\bf k}_2}^{\alpha \beta}
393: \tilde{\psi}^{\dagger}_{{\bf k}_1 \alpha}
394: \tilde{\psi}^{\dagger}_{-{\bf k}_1 \beta}
395: \tilde{\psi}_{-{\bf k}_2 \beta} \tilde{\psi}_{{\bf k}_2 \alpha},
396: \label{h_bi}
397: \end{eqnarray}
398: where $\xi^{\bf k}_{\uparrow \uparrow} =
399: \xi^{\bf k}_{\downarrow \downarrow} = k^2/2m - \mu$ and
400: $\xi^{\bf k}_{\uparrow \downarrow}
401: = \xi^{\bf k}_{\downarrow \uparrow} = -t$.
402: In the interaction term, we have restricted ourselves to the
403: scattering processes of pairs with zero total momentum.
404: Note that the formulation in this section can be easily extended to
405: the multicomponet systems.
406:
407: From Eq.(\ref{h_bi}), we can define the mean field Hamiltonian as
408: \begin{eqnarray}
409: H_{\rm MF}&=& \frac12
410: \sum_{\bf k}
411: \left(
412: \begin{array}{cccc}
413: \tilde{\psi}_{{\bf k}\uparrow}^{\dagger} &
414: \tilde{\psi}_{{\bf k}\downarrow}^{\dagger} &
415: \tilde{\psi}_{-{\bf k}\uparrow} &
416: \tilde{\psi}_{-{\bf k}\downarrow}
417: \end{array}
418: \right)
419: \nonumber \\ & & \times
420: \left(
421: \begin{array}{cc}
422: \xi^k & \Delta^{\bf k} \\
423: \left( \Delta^{\bf k} \right)^{\dagger} & -\xi^k
424: \end{array}
425: \right)
426: \left(
427: \begin{array}{c}
428: \tilde{\psi}_{{\bf k}\uparrow} \\
429: \tilde{\psi}_{{\bf k}\downarrow} \\
430: \tilde{\psi}_{-{\bf k}\uparrow}^{\dagger} \\
431: \tilde{\psi}_{-{\bf k}\downarrow}^{\dagger}
432: \end{array}
433: \right),
434: \end{eqnarray}
435: where the pairing matrix is defined as
436: \begin{equation}
437: \Delta^{{\bf k}}_{\alpha \beta}
438: = - \frac{1}{2\Omega}
439: \sum_{{\bf k} ( \neq {\bf k}^{\prime} )}
440: V^{\alpha \beta}_{{\bf k} {\bf k}^{\prime}}
441: \langle
442: \tilde{\psi}_{-{\bf k}^{\prime} \beta}
443: \tilde{\psi}_{{\bf k}^{\prime} \alpha}
444: \rangle.
445: \label{eq_Delta}
446: \end{equation}
447:
448: First we consider the triplet pairing case.
449: Since we consider the symmetric bilayer systems, we take the symmetric
450: form of the pairing matrix:
451: $\Delta^{{\bf k}}_{\uparrow \downarrow} =
452: \Delta^{{\bf k}}_{\downarrow \uparrow}$
453: and
454: $\Delta^{{\bf k}}_{\uparrow \uparrow} =
455: \Delta^{{\bf k}}_{\downarrow \downarrow}$.
456: Diagonalization of the mean field Hamiltonian yields the following gap
457: equations at zero temperature:
458: \begin{equation}
459: \Delta^{\bf k}_{\uparrow \uparrow}
460: = - \frac{1}{4\Omega} \sum_{{\bf k}^{\prime} (\neq {\bf k})}
461: V_{{\bf k} {\bf k}^{\prime}}^{\uparrow \uparrow}
462: \left[ \frac{\Delta^{{\bf k}^{\prime}}_{\uparrow \uparrow}
463: + \Delta^{{\bf k}^{\prime}}_{\uparrow \downarrow}}
464: {E_{{\bf k}^{\prime}}^+}
465: +
466: \frac{\Delta^{{\bf k}^{\prime}}_{\uparrow \uparrow}
467: - \Delta^{{\bf k}^{\prime}}_{\uparrow \downarrow}}
468: {E_{{\bf k}^{\prime}}^-} \right],
469: \label{t_gap1}
470: \end{equation}
471: \begin{equation}
472: \Delta^{\bf k}_{\uparrow \downarrow}
473: = - \frac{1}{4\Omega} \sum_{{\bf k}^{\prime} (\neq {\bf k})}
474: V_{{\bf k} {\bf k}^{\prime}}^{\uparrow \downarrow}
475: \left[ \frac{\Delta^{{\bf k}^{\prime}}_{\uparrow \uparrow}
476: + \Delta^{{\bf k}^{\prime}}_{\uparrow \downarrow}}
477: {E_{{\bf k}^{\prime}}^+}
478: -
479: \frac{\Delta^{{\bf k}^{\prime}}_{\uparrow \uparrow}
480: - \Delta^{{\bf k}^{\prime}}_{\uparrow \downarrow}}
481: {E_{{\bf k}^{\prime}}^-} \right],
482: \label{t_gap2}
483: \end{equation}
484: where $E^{\pm}_{\bf k}=\sqrt{(\xi_k-t)^2+|\Delta^{\bf k}_{\uparrow
485: \uparrow} \pm \Delta^{\bf k}_{\uparrow \downarrow}|^2}$.
486: Meanwhile for the singlet pairing state, $\Delta^{\bf k}_{\uparrow
487: \downarrow} = - \Delta^{\bf k}_{\downarrow \uparrow} \equiv
488: \Delta_{\bf k}$,
489: the gap equation is given by
490: \begin{eqnarray}
491: \Delta_{{\bf k}} &=& - \frac{1}{4\Omega}
492: \sum_{{\bf k}^{\prime} (\neq {\bf k})}
493: V_{{\bf k} {\bf k}^{\prime}}^{\uparrow \downarrow}
494: \frac{\Delta_{{\bf k}^{\prime}}}{E_{{\bf k}^{\prime}}}
495: \left[ \tanh \frac{\beta (E_{{\bf k}^{\prime}}-t)}{2}
496: \right. \nonumber \\
497: & & \left. ~~~~~+
498: \tanh \frac{\beta (E_{{\bf k}^{\prime}} + t)}{2} \right],
499: \end{eqnarray}
500: where $E_{\bf k}=\sqrt{\xi_k^2 + |\Delta_{\bf k}|^2}$.
501: At zero temperature, this equation is reduced to
502: \begin{equation}
503: \Delta_{{\bf k}} = - \frac{1}{2\Omega}
504: \sum_{{\bf k}^{\prime} (\neq {\bf k}), E_{{\bf k}^{\prime}}>t}
505: V_{{\bf k}^{\prime} {\bf k}}^{\uparrow \downarrow}
506: \frac{\Delta_{{\bf k}^{\prime}}}{E_{{\bf k}^{\prime}}}.
507: \end{equation}
508: Note that there is the constraint $E_{{\bf k}^{\prime}}>t$
509: in the summation over ${\bf k}'$-space.
510:
511: In the absence of the inter-layer tunneling, we can take
512: $\Delta^{\bf k}_{\uparrow \uparrow} =0$ because a pairing state
513: with $\Delta^{\bf k}_{\uparrow \uparrow} \neq 0$
514: may be stable only in the vicinity of the sample boundary.
515: By this choice of the pairing matrix, the gap equation has the same
516: form both for the triplet pairing state and for the singlet pairing
517: state:
518: \begin{equation}
519: \Delta_{{\bf k}} = - \frac{1}{2\Omega}
520: \sum_{{\bf k}^{\prime} (\neq {\bf k})}
521: V_{{\bf k} {\bf k}^{\prime}}^{\uparrow \downarrow}
522: \frac{\Delta_{{\bf k}^{\prime}}}{E_{{\bf k}^{\prime}}}.
523: \end{equation}
524: We can solve this gap equation by taking the form of the gap as
525: $\Delta_{\bf k}=\Delta \epsilon_F f(k,k_F) \exp (-i \ell \theta_{\bf
526: k})$, where $f(k,k_F)=(k/k_F)^{\ell}$ for $k<k_F$ and
527: $f(k,k_F)=(k_F/k)^{\ell}$ for $k>k_F$ and $\ell$ is an
528: integer.\cite{GREITER_ETAL,MORINARI}
529: From the analysis of this gap equation, we find that the ground state
530: is the p-wave pairing state.
531: Note that the $s$-wave pairing state is excluded
532: because the pairing interaction originates from the Lorentz
533: interaction induced by the Chern-Simons gauge field
534: interaction.\cite{MORINARI}
535: At $\nu=1/2$, the p-wave pairing state corresponds to the $(3,3,1)$
536: state.
537: At $\nu=1$ with the choice of $(\phi_1,\phi_2)=(0,2)$,
538: the p-wave state corresponds to the $(1,1,1)$ state.\cite{MORINARI_BI}
539:
540: \section{Effect of inter-layer tunneling}
541: \label{sec_t}
542: Now let us take into account the inter-layer tunneling effect.
543: We consider the effect of it on the p-wave pairing state at
544: $\nu=1/2$.
545: The possibility of Haldane-Rezayi state is discussed later.
546: As we have discussed in Sec.~\ref{sec_two}, the appropriate choice of
547: $(\phi_1,\phi_2)$ is $(2,2)$.
548: Since the number of attached fluxes $\phi_1$ and $\phi_2$ is
549: symmetric, we may take
550: $\Delta^{\bf k}_{\uparrow \uparrow} = \Delta_{\bf k} a$ and
551: $\Delta^{\bf k}_{\uparrow \downarrow}= \Delta_{\bf k} b$.
552: In order to discuss the evolution of the pairing state,
553: we define an angle as
554: \begin{equation}
555: \theta = \tan^{-1} \frac{a}{b}.
556: \end{equation}
557: This angle $\theta$ characterizes the pairing state.
558: For the case of $\theta=0$, we have the p-wave pairing state that
559: corresponds to the $(3,3,1)$ state.
560: Meanwhile, for the case of $\theta=\pi/4$, we have the Pfaffian state.
561: In Fig.~\ref{tunneling_theta}, we show the inter-layer tunneling
562: dependence of $\theta$.
563: Note that the Pfaffian state is stabilized in the $\tau > 2$ region.
564:
565: In Fig.~\ref{tunneling_gap}, we show the inter-layer tunneling
566: dependence of the gap $\Delta$.
567: Note that there is a cusp at $\tau=2$.
568: Reflecting the presence of the cusp in the gap, the ground state
569: energy also has a cusp at $\tau=2$.
570:
571: Now we discuss the possibility of the Haldane-Rezayi state, or
572: d-wave pairing state.
573: Since the $s$-wave pairing state is excluded as mentioned above, the
574: next leading singlet pairing state is the d-wave pairing state.
575: From the analysis of the gap equation, we find that
576: for the d-wave pairing state $\Delta \sim 1.2$, which is smaller
577: than that for the p-wave pairing state, in the region of
578: $\tau < \tau_c$ where $\tau_c \sim 0.8$ and the d-wave pairing
579: state is not stabilized in $\tau > \tau_c$.
580: However, in the above analysis we have neglected the effect of the
581: long-range Coulomb interaction.
582: Since the effect of it is expected to be larger for the p-wave pairing
583: state than for the d-wave pairing state, it might be possible that the
584: d-wave pairing state becomes stable due to the effect of the
585: long-range Coulomb interaction.
586: In addition, impurites may affect the p-wave pairing state more than
587: the d-wave pairing state.
588:
589: \section{Conclusion}
590: \label{sec_summary}
591: In this paper, we have discussed the region of the inter-layer
592: separation where the p-wave pairing state is stabilized and the
593: effect of the inter-layer tunneling at $\nu=1/2$.
594: The Pfaffian state is stable above the critical tunneling strength and
595: there is a cusp at the transition point between the Pfaffian state and
596: the state continuously connected with the $(3,3,1)$ state.
597:
598: \acknowledgments
599: This work was supported in part by a Grant-in-Aid from the Ministry of
600: Education, Culture, Sports, Science and Technology.
601:
602: \begin{references}
603: \bibitem{review} {\it The Quantum Hall Effect},
604: 2nd ed., edited by R. Prange and S. Girvin
605: (Springer Verlag, New York, 1990)
606: \bibitem{review2} {\it Perspectives in Quantum Hall Effects}, edited
607: by Sankar Das Sarma and Aron Pinczuk (John Wiley \& Sons, New York,
608: 1997).
609: \bibitem{LAUGHLIN} R. B. Laughlin, Phys. Rev. Lett. {\bf 50}, 1395 (1983).
610: \bibitem{HALDANE} F. D. M. Haldane, in Ref.~\cite{review}.
611: \bibitem{HALPERIN_MMN} B. I. Halperin, Helv. Phys. Acta. {\bf 56}, 75
612: (1983).
613: \bibitem{YOSHIOKA_ETAL} D. Yoshioka, A. H. MacDonald and S. M. Girvin,
614: Phys. Rev. B {\bf 39}, 1932 (1989).
615: \bibitem{HO} T. L. Ho, Phys. Rev. Lett. {\bf 75}, 1186 (1995).
616: \bibitem{MOORE_READ} G. Moore and N. Read, Nucl. Phys. B {\bf 360},
617: 362 (1991).
618: \bibitem{HALPERIN_BI} B. I. Halperin, Surf. Sci. {\bf 305}, 1 (1994).
619: \bibitem{HALDANE_REZAYI} F. D. M. Haldane and E. H. Rezayi,
620: Phys. Rev. Lett. {\bf 60}, 956 (1988); {\it ibid}, {\bf 60}, 1886 (1988).
621: \bibitem{MORINARI_BI}T. Morinari, Phys. Rev. B {\bf 59}, 7320 (1999).
622: \bibitem{MURPHY} S. Q. Murphy, J. P. Eisenstein, G. S. Boebinger,
623: L. N. Pfeiffer and K. W. West, Phys. Rev. Lett. {\bf 72}, 728 (1994).
624: \bibitem{GREITER_ETAL} M. Greiter, X. G. Wen and F. Wilczek,
625: Nucl. Phys. B {\bf 374}, 567 (1992).
626: \bibitem{MORINARI} T. Morinari, Phys. Rev. B {\bf 62}, 15903 (2000).
627: \end{references}
628: \end{multicols}
629:
630: \begin{figure}
631: \center
632: \epsfxsize=2.0truein
633: \psfig{file=fig1.ps,width=2.0in,angle=270}
634: \vspace{0.1in}
635: \caption{The two-body correlation energy $E_C^{(2)}(m,n)/N^2$ for
636: $(m,n)=(4,0),(3,1)$, and $(2,2)$ in units of $e^2/\epsilon \ell_B$.}
637: \label{nu_half}
638: \end{figure}
639:
640: \begin{figure}[htbp]
641: \center
642: \epsfxsize=2.0truein
643: \psfig{file=fig2.ps,width=2.0in,angle=270}
644: \vspace{0.1in}
645: \caption{The energy $E_C^{(2)}(m,n)/N^2$ for $(m,n)=(2,0)$, and
646: $(1,1)$
647: in units of $e^2/\epsilon \ell_B$.}
648: \label{nu_one}
649: \end{figure}
650:
651: \begin{figure}[htbp]
652: \center
653: \epsfxsize=2.0truein
654: \psfig{file=fig3.ps,width=2.0in,angle=270}
655: \vspace{0.1in}
656: \caption{The inter-layer tunneling dependence of $\theta$.}
657: \label{tunneling_theta}
658: \end{figure}
659:
660: \begin{figure}[htbp]
661: \center
662: \epsfxsize=2.0truein
663: \psfig{file=fig4.ps,width=2.0in,angle=270}
664: \vspace{0.1in}
665: \caption{The inter-layer tunneling dependence of the gap $\Delta$.}
666: \label{tunneling_gap}
667: \end{figure}
668:
669:
670: \end{document}
671:
672:
673:
674:
675:
676:
677:
678:
679:
680:
681:
682:
683: