cond-mat0509257/pra.tex
1:                                                
2: %\documentclass[12pt,a4paper]{iopart} %\usepackage{amssymb,graphicx,cite}
3: %\eqnobysec %\begin{document}
4: 
5:  
6: %LATEX FILE OF MANUSCRIPT
7: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
8:  
9:  
10:  
11: %LATEX file of the manuscript
12:  
13: %\documentclass[preprint,eqsecnum,aps,epsf]{revtex4} % PH. REV.
14: \documentclass[eqsecnum,aps,twocolumn,epsf,showpacs]{revtex4} % PH. REV.
15: \usepackage{graphicx}
16: %\documentclass[eqsecnum,aps,twocolumn]{revtex4} % PH. REV.
17:  
18: %\renewcommand{\baselinestretch}{.89}
19:  
20:  
21: %\documentstyle[aps,epsf]{revtex}
22: %\documentstyle[eqsecnum,aps,epsf]{revtex}
23: %%% <<< epsf commands in the next two lines >>>
24: %\newcommand{\postscript}[2] {\setlength{\epsfxsize}{#2\hsize}
25: %\centerline{\epsfbox{#1}}}
26:  
27:  
28:  
29: %\documentstyle[eqsecnum,aps,epsf]{revtex} % PH. REV. FINAL FORMAT STYLE
30: %\documentstyle[aps,epsf]{revtex} % PH. REV. FINAL FORMAT STYLE
31: %%% <<< epsf commands in the next two lines >>>
32: %\newcommand{\postscript}[2] {\setlength{\epsfxsize}{#2\hsize}
33: %\centerline{\epsfbox{#1}}}
34:  
35: %\documentstyle[preprint,aps]{revtex}
36: %\documentstyle[eqsecnum,aps]{revtex}
37: %\documentstyle[aps]{revtex}
38: %\renewcommand{\baselinestretch}{.945}
39:  
40: \begin{document}
41:    
42: %\twocolumn[\hsize\textwidth\columnwidth\hsize\csname@twocolumnfalse\endcsname
43:  
44: %\title[Jet formation in a collapsing Bose-Einstein condensate]{Mean-field
45: %model of jet formation in a collapsing Bose-Einstein condensate}
46: 
47: \title{Fermionic bright  soliton  in
48: a  boson-fermion mixture}
49:  
50: \author{Sadhan K. Adhikari\footnote{Electronic
51: address: adhikari@ift.unesp.br; \\
52: URL: http://www.ift.unesp.br/users/adhikari/}}
53: \affiliation
54: %\address
55: {Instituto de F\'{\i}sica Te\'orica, Universidade Estadual
56: Paulista, 01.405-900 S\~ao Paulo, S\~ao Paulo, Brazil\\}
57: {Accepted:  Physical Review A} 
58:  
59: \date{\today}
60:  
61: %\maketitle
62:  
63: \begin{abstract}
64: 
65: We use a time-dependent dynamical mean-field-hydrodynamic model to study
66: the formation of fermionic bright 
67: solitons in a trapped 
68: degenerate Fermi gas mixed with a Bose-Einstein
69: condensate in a quasi-one-dimensional cigar-shaped geometry. Due to a
70: strong
71: Pauli-blocking repulsion among 
72: spin-polarized fermions at short distances 
73: there cannot be bright fermionic
74: solitons in the case of repulsive boson-fermion interactions. 
75: However,  we demonstrate that stable bright fermionic
76: solitons can be formed for a sufficiently attractive boson-fermion
77: interaction in a boson-fermion mixture. We also consider the formation of
78: fermionic solitons in the
79: presence of a periodic  axial optical-lattice potential.
80: These solitons can be formed and studied in the laboratory
81: with present
82: technology.
83: 
84: 
85: 
86: 
87: 
88:  
89: \pacs{03.75.Lm, 03.75.Ss}
90: \end{abstract}
91: 
92: \maketitle
93:  
94:  
95:  
96: \section{Introduction}
97: 
98: There cannot be an  effective evaporative
99: cooling  leading to a trapped quantum degenerate  Fermi gas (DFG) 
100: due to a strong repulsive Pauli-blocking interaction at low
101: temperature  among spin-polarized fermions \cite{exp1}. 
102: However, it has been possible to achieve a DFG by sympathetic cooling in the presence of a 
103: second
104: boson or fermion
105: component.  
106: Recently, there have been  successful observation
107: \cite{exp1,exp2,exp3,exp4} and associated  experimental
108: \cite{exp5,exp5x,exp6} and theoretical \cite{yyy1,yyy,zzz,capu,ska} 
109: studies of mixtures of a trapped DFG and a Bose-Einstein condensate
110: (BEC)   by different experimental groups
111: \cite{exp1,exp2,exp3,exp4} in 
112: the following systems: $^{6,7}$Li \cite{exp3}, $^{23}$Na-$^6$Li
113: \cite{exp4} and 
114: $^{87}$Rb-$^{40}$K \cite{exp5,exp5x}. 
115: Also,  there have been 
116: studies of mixtures of two-component trapped DFGs in 
117: $^{40}$K \cite{exp1} and  
118: $^6$Li \cite{exp2} atoms.
119: 
120: The formation and collapse 
121: of a DFG in a boson-fermion mixture
122: $^{87}$Rb-$^{40}$K 
123: have been
124: observed and studied by Modugno {\it et al.}
125: \cite{exp5,zzz,ska}. Although, the fermion-fermion interaction at short
126: distances 
127: is
128: repulsive due to strong Pauli blocking 
129: and hence incapable of leading to collapse,  
130: a sufficiently  
131: attractive boson-fermion interaction could overcome the Pauli repulsion
132: and could result in a collapse of a DGF.
133: Bright solitons in a BEC are formed due to an
134: attractive nonlinear atomic interaction. 
135: As the interaction in a pure
136: DFG at short distances is repulsive, 
137: there cannot be
138: bright solitons in 
139: a DFG. 
140: 
141: In this paper 
142: we study the possibility of the formation of stable fermionic bright
143: solitons  
144: in a mixture of a DFG with a BEC   in the presence of a sufficiently attractive
145: boson-fermion interaction which can overcome the Pauli repulsion among 
146: fermions. 
147: The formation of a fermionic  soliton is related to the fact 
148: that the system can lower its energy by forming high density regions
149: (the solitons) when the   attraction between the bosons and fermions is
150: large enough to overcome the Pauli repulsion in the DFG and any possible 
151: repulsion in the BEC. 
152: In particular we consider 
153: the formation of fermionic bright  solitons, which can freely move in the
154: axial direction,  in
155: such a mixture for  a quasi-one-dimensional cigar-shaped 
156: geometry
157: using a coupled time-dependent mean-field-hydrodynamic
158: model where the  bosonic component is treated by the mean-field
159: Gross-Pitaevskii 
160: equation \cite{11} and the fermionic component is treated by a
161: hydrodynamic
162: model \cite{capu}. 
163: This time-dependent mean-field-hydrodynamic model was
164: suggested
165: recently by the present author \cite{ska} to study the collapse 
166: dynamics
167: of fermions and is a
168: time-dependent extension of a time-independent model used for the
169: stationary  states by Capuzzi {\it et al.} \cite{capu}. 
170: 
171: 
172: 
173: 
174: Bright solitons are really eigenfunction of the one-dimensional nonlinear
175: Schr\"odinger equation. However, the experimental realization of bright
176: solitons in trapped attractive cigar-shaped BECs has been possible under 
177: strong transverse binding which, in the case of weak or no axial binding,  
178: simulates the ideal one-dimensional situation for the formation of bright
179: solitons. 
180: The 
181: dimensionless nonlinear Schr\"odinger 
182: (NLS) equation in the attractive
183: or
184: self-focusing case  \cite{1}
185: \begin{equation}\label{nls}
186: i u_t+u_{xx}+  |u|^2u=0.
187: \end{equation}
188: sustains the following bright
189:  soliton \cite{1}:
190: \begin{eqnarray}\label{DS}
191: u(x,t)&=& \sqrt{2 B}\hskip 3pt \mbox{sech}
192: [\sqrt{B}(x-\delta+2v t)] \nonumber \\ &\times& 
193: \exp[-iv(x-\delta) +i(B -v v)t+i\sigma], 
194: \end{eqnarray}
195: with four parameters. The parameter  $B$ represents the amplitude as well
196: as pulse width, $v$ represents velocity, the parameters $\delta$ and 
197: $\sigma$ are phase constants. The bright soliton 
198: profile  
199: is easily recognized for $v=\delta=0$ as  
200: $|u(x,t)|=\sqrt{2B}\hskip 3pt  \mbox{sech} [x\sqrt{B}]$.  
201: There have been  experimental \cite{exdks} and theoretical \cite{thdks}
202: studies of the formation of  bright  solitons in a
203: BEC. In view of this, here we study for the first time  
204: the possibility of the
205: formation of a stable fermionic bright  soliton in a  boson-fermion
206: mixture.  
207: 
208: 
209: In recent times there have been routine experimental studies on the
210: formation of BEC in the presence of a periodic axial optical-lattice
211: potential \cite{oplat}. These leads to a different condition of
212: trapping from the harmonic trap and generates a BEC of distinct
213: modulation. Hence we
214: also consider in this paper the modulations of the fermionic bright
215: solitons in the presence of an optical-lattice potential.
216: 
217: 
218: 
219:  
220: In Sec. II we present  the time-dependent mean-field model
221: consisting of a set of coupled partial differential equations involving
222: the bosonic and fermionic probability densities. 
223:  In the case
224: of a cigar-shaped geometry with stronger  radial trapping, 
225: the above model is reduced to an effective
226: one-dimensional form appropriate for the study of bright solitons. 
227: In Sec.  
228: III we present our results for stationary axially-free fermionic bright
229: solitons as well as those formed on an axial  periodic optical-lattice
230: potential.
231: We also demonstrate the stability of the bright solitons after a
232: perturbation  is applied. The bright solitons are found to execute
233: stable breathing oscillation upon perturbation. 
234: Finally,
235: a summary of our findings  is given in Sec. IV.
236:  
237: \section{Nonlinear mean-field-hydrodynamic model}
238:  
239: 
240: 
241: The time-dependent Bose-Einstein condensate wave
242: function $\Psi({\bf r},t)$ at position ${\bf r}$ and time $t $
243: may
244: be described by the following  mean-field nonlinear Gross-Pitaevskii 
245: equation
246: \cite{11}
247: \begin{eqnarray}\label{a} \biggr[- i\hbar\frac{\partial
248: }{\partial t}
249: -\frac{\hbar^2\nabla_{\bf r}^2   }{2m_{{B}}}
250: + V_{{B}}({\bf r})
251: + g_{{BB}} n_B
252:  \biggr]\Psi_{{B}}({\bf r},t)=0, 
253: \end{eqnarray}
254: with normalization $ \int d{\bf r} |\Psi_B({\bf r},t)|^2 = N_B. $ 
255: Here $m_{{B}}$
256: is
257: the mass and  $N_{{B}}$ the number of bosonic atoms in the
258: condensate, $n_B\equiv  |\Psi_{{B}}({\bf r},t)|^2$ is the boson 
259: probability density,
260:  $g_{{BB}}=4\pi \hbar^2 a_{{BB}}/m_{{B}} $ the strength of
261: inter-atomic interaction, with
262: $a_{{BB}}$ the boson-boson scattering length. 
263: The trap potential with axial symmetry may be written as  $
264: V_{{B}}({\bf
265: r}) =\frac{1}{2}m_B \omega ^2 (\rho^2+\nu^2 z^2)$ where
266:  $\omega$ and $\nu \omega$ are the angular frequencies in the radial
267: ($\rho$) and axial ($z$) directions with $\nu$ the anisotropy parameter.
268: The probability density of an isolated 
269: DFG  in the Thomas-Fermi approximation  is given by
270: \cite{zzz}
271: \begin{eqnarray}\label{b}
272: n_F= \frac{[\mbox{max}(0,\{\epsilon_F-V_F({\bf r})\})]^{3/2}}{A^{3/2}},
273: \end{eqnarray}
274: where  $A=\hbar^2 (6
275: \pi^2
276: )^{2/3}/ (2m_F)$, $\epsilon_F$ is the Fermi energy, 
277: $m_F$ is the
278: fermionic mass, and the function $\mbox{max}$ denotes the larger of
279: the arguments. The confining trap potential
280: $V_F({\bf
281: r})= \frac{1}{2}m_F
282: \omega_F^2(\rho^2+\nu^2 z^2)$ has axial symmetry as the
283: bosonic potential $V_B({\bf r}),$ where $\omega_F$ is the radial
284: frequency. The
285: anisotropy parameter $\nu$ will be taken to be zero for axially-free
286: solitons in the following. The
287: number of fermionic atoms $N_F$
288: is given by the normalization $\int d{\bf r} n_F({\bf r})=N_F$.
289: 
290: 
291: We developed a set of practical time-dependent mean-field-hydrodynamic
292: equations for the interacting boson-fermion mixture starting from the
293: following Lagrangian density \cite{ska} \begin{eqnarray}\label{yy} {\cal
294: L}&=& \frac{i}{2}\hbar \left[ \Psi_B\frac{\partial \Psi_B^*}{\partial t} -
295: \Psi_B^* \frac{\partial \Psi_B}{\partial t} \right] \nonumber \\ &+&
296: \frac{i}{2}\hbar \left[ \sqrt{n_F}\frac{\partial {\sqrt n_F} ^*}{\partial
297: t} - {\sqrt n_F}^* \frac{\partial \sqrt{n_F}}{\partial t} \right]
298: \nonumber \\ &+& \left(\frac{\hbar^2|\nabla_{\bf r} \Psi_B|^2
299: }{2m_B}+V_B|\Psi_B|^2+\frac{1}{2}g_{BB} |\Psi_B|^4\right)\nonumber \\ &+&
300: \left(\frac{\hbar^2 |\nabla_{\bf r} \sqrt{n_F}|^2 }{6m_F}+
301: V_F|n_F|+\frac{3}{5} A |n_F|^{5/3}\right)\nonumber \\ &+& g_{BF} n_F
302: |\Psi_B|^2, \end{eqnarray} where $g_{BF}=2\pi \hbar^2 a_{BF}/m_R$ with the
303: boson-fermion reduced mass $m_R=m_Bm_F/(m_B+m_F),$ where $ a_{BF}$ is the
304: boson-fermion scattering length.
305: 
306: 
307: The terms in the first round bracket   on the right-hand side of
308: Eq. (\ref{yy}) are the standard 
309: Gross-Pitaevskii terms  for the bosons and are related to a
310: Schr\"odinger-like equation \cite{11}. However, 
311: terms in the second  round bracket
312: are derived from the hydrodynamic equation of motion of the 
313: fermions
314: \cite{capu}. Hence,
315: the second kinetic energy term 
316: has a different mass factor $6m_F$ and not the
317: conventional  Schr\"odinger mass factor $2m_B$ as in the first integral.
318: Finally, the 
319: last term in this equation 
320:  corresponds to an interaction between bosons
321: and fermions. 
322: The interaction between fermions in
323: spin polarized state is highly suppressed due 
324: to Pauli blocking 
325: and has been neglected in Eq. (\ref{yy})  and will be
326: neglected throughout this paper.
327: 
328:  
329:  
330: Recently, Jezek {\it et al.} \cite{jz} used the Thomas-Fermi-Weizs\"acker
331: kinetic energy term $T_F$ of fermions in their formulation which, in our
332: notation, will correspond to a fermionic kinetic energy of
333: $\hbar^2|\nabla_{\bf r}\sqrt{n_F}|^2/(9m_F) $ in Eq. (\ref{yy}) in place
334: of the present term $\hbar^2|\nabla_{\bf r}\sqrt{n_F}|^2/(6m_F) $.  This
335: kinetic energy term contributes little to this problem compared to the
336: dominating $3A|n_F|^{5/3}/5$ term in Eq. (\ref{yy}) and is usually
337: neglected in the Thomas-Fermi approximation.  However, its inclusion leads
338: to an analytic solution for the probability density everywhere
339: \cite{jz}. For a
340: discussion of these two fermionic kinetic energy terms we refer the reader
341: to Refs. \cite{capu,jz,pi}.
342: 
343: 
344: With the Lagrangian density (\ref{yy}), the Euler-Lagrange equations of
345: motion become \cite{ska}:  \begin{eqnarray}\label{e} \biggr[ &-&
346: i\hbar\frac{\partial }{\partial t} -\frac{\hbar^2\nabla_{\bf
347: r}^2}{2m_{{B}}} + V_{{B}}({\bf r}) + g_{{BB}}n_B \nonumber \\ &+& g_{{BF}}
348: n_F
349:  \biggr]\Psi_{{B}}({\bf r},t)=0,
350: \end{eqnarray}
351: \begin{eqnarray}\label{f} \biggr[& -& i\hbar\frac{\partial
352: }{\partial t}
353: -\frac{\hbar^2\nabla_{\bf r}^2}{6m_{{F}}}
354: + V_{{F}}({\bf r})
355: + A |n_F|^{2/3} \nonumber \\
356: &+& g_{{BF}} n_B
357:  \biggr]\sqrt{n_{{F}}({\bf r},t)}=0.
358: \end{eqnarray}
359: 
360: 
361:  
362: When the nonlinearity in Eq.  (\ref{f})
363: is  very large,
364: the kinetic energy term in this equation can be neglected and the
365: time-independent stationary form of this equation becomes
366: \begin{equation}\label{mod}
367: n_F= \frac{[\mbox{max}(0,\{\epsilon_F-V_F({\bf
368: r})-g_{BF}n_B\})]^{3/2}}{A^{3/2}},
369: \end{equation}
370: which is the generalization of Eq. (\ref{b}) in the presence of
371: boson-fermion coupling. Equation (\ref{mod}) has been used by 
372: Modugno
373: {\it et al.} \cite{zzz} for an analysis of a DFG-BEC mixture. 
374: In actual experimental
375: condition the nonlinearity in Eq.  (\ref{f})    is quite large and  Eq.
376: (\ref{mod}) is a good approximation. 
377: 
378: 
379: 
380: 
381: 
382: 
383: For the study of bright  solitons
384: we shall reduce Eqs. (\ref{e}) and  (\ref{f}) to the minimal 
385:  one-dimensional form under the action of stronger radial trapping. 
386: The one-dimensional form is appropriate for     studying bright 
387: solitons in the so-called cigar-shaped quasi-one-dimensional geometry
388: where 
389: $\nu << 1$. For radially-bound and axially-free solitons we eventually set 
390: $\nu =0$.
391: In this case the dynamical equations can be reduced to strict
392: one-dimensional coupled NLS equations  without any trap. 
393: We perform this
394: reduction below where we 
395: take $V_B({\bf r})=V_F({\bf r})= \frac{1}{2}m_B\omega^2(
396: \rho^2+\nu^2 z^2)$ which corresponds to a
397: reduction of $\omega_F$ and
398: $\nu  \omega_F$ in $V_F({\bf r})$
399: by a factor
400: $\sqrt{m_B/m_F}$ as in the study by 
401: Modugno {\it et al.} \cite{zzz} and Jezek  {\it et al.} \cite{jz}.
402: 
403:  
404: 
405: 
406: 
407: 
408: 
409: For  $\nu =0$, Eqs. (\ref{e})
410: and (\ref{f}) can be reduced to an effective 
411: one-dimensional form by considering 
412: solutions of the type 
413: $\Psi_B({\bf r},t)=  \phi_B(z,t)\psi_B^{(0)}( \rho)$ and
414: $\sqrt{n_F({\bf r},t)}=  \phi_F(z,t)\psi_F^{(0)}( \rho),$
415: where
416: \begin{eqnarray}\label{wfx}
417: |\psi_i^{(0)}(\rho)|^2&\equiv&
418: {\frac{M_i\omega}{\pi\hbar}}\exp\left(-\frac{M_i
419: \omega
420: \rho^2}{\hbar}\right),
421: \end{eqnarray}
422: and  $i=B,F$ represents bosons and fermions and $M_B=m_B$ and
423: $M_F=\sqrt{3m_Bm_F}. $ The expression (\ref{wfx})
424: corresponds to the respective
425: ground state wave function in the absence of nonlinear interactions and
426: satisfies
427: \begin{eqnarray}
428: -\frac{\hbar^2}{2m_B}\nabla_\rho ^2\psi_B^{(0)}
429: +
430: \frac{1}{2}m_B\omega^2\rho^2
431: \psi_B^{(0)}&=&\hbar\omega
432: \psi_B^{(0)},\\
433: -\frac{\hbar^2}{6m_F}\nabla_\rho^2\psi_F^{(0)}+
434: \frac{1}{2}m_B\omega^2\rho
435: ^2\psi_F^{(0)}&=&\sqrt{\frac{m_B}{3m_F}}\hbar\omega
436: \psi_F^{(0)},\nonumber \\
437: \end{eqnarray}
438: with normalization 
439: $2\pi \int_{0}^\infty |\psi_i^{(0)}(\rho)|^2 \rho d\rho=1.$
440: Now the dynamics is carried by $ \phi_i(z,t)$ and the radial dependence is
441: frozen in the ground state $\psi_i^{(0)}(\rho)$.
442: The factorization of $\Psi_B$ and $\sqrt{n_F}$ above follows from the
443: structure of the mathematical equations (\ref{e}) and (\ref{f}).
444: Although $n_F$ gives the probability density of DFG it may not be to the
445: point to
446: associate $\psi_F^{0}$ and $\phi_F$ to physical fermionic one-particle
447: wave functions. The true fermionic wave function has the form of a
448: many-particle 
449: Slater determinant. Nevertheless,  
450: the functions $\psi_F^{0}$ and $\phi_F$ 
451: could be regarded as mathematical functions related to
452: fermionic density \cite{yyy1}. In the quasi-one-dimensional cigar-shaped
453: geometry the
454: linear fermionic and bosonic probability densities are given by
455: $|\phi_F(z,t)|^2$ and
456: $|\phi_B(z,t)|^2,$ respectively.
457: 
458: Averaging over the radial mode $\psi_i^{(0)}(\rho)$, 
459: i.e., multiplying
460: Eqs. (\ref{e}) and (\ref{f})
461: by  $\psi_i^{(0)*}(\rho)$
462: and integrating over $\rho$, we obtain the following one-dimensional 
463: dynamical equations \cite{abdul}:
464: \begin{eqnarray}\label{i} \biggr[ &-& i\hbar\frac{\partial
465: }{\partial t}
466: -\frac{\hbar^2}{2m_{{B}}}\frac{\partial^2}{\partial z^2}
467:  \nonumber \\  &+&F_{BB}|
468: \phi_B|^2
469: + F_{BF}| \phi_F|^2
470:  \biggr] \phi_{{B}}(z,t)=0, 
471: \end{eqnarray}
472: \begin{eqnarray}\label{j} 
473: \biggr[& -& i\hbar\frac{\partial
474: }{\partial t}
475: -\frac{\hbar^2}{6m_F}\frac{\partial^2}{\partial z^2}
476: \nonumber \\  
477: &+&F_{FF}|
478: \phi_F|^{4/3}  
479: + F_{BF}| \phi_B|^2
480:  \biggr] \phi_{{F}}(z,t)=0, 
481: \end{eqnarray}
482: where 
483: \begin{eqnarray}
484:  F_{BB}=g_{BB}\frac{\int_0^\infty|\psi_B^{(0)}|^4\rho d\rho}
485: {\int_0^\infty|\psi_B^{(0)}|^2\rho d\rho}=
486: g_{BB}{\frac{m_B\omega}{2\pi\hbar}},
487: \end{eqnarray}
488: \begin{eqnarray}
489: F_{BF}=g_{BF}\frac{\int_0^\infty|\psi_F^{(0)}|^2|\psi_B^{(0)}|^2\rho 
490: d\rho}{\int_0^\infty|\psi_B^{(0)}|^2\rho d\rho}
491: =g_{BF}{\frac{M_{BF}\omega}{\pi\hbar}},
492: \end{eqnarray}
493: \begin{eqnarray}
494: F_{FF}=A
495: \frac{\int_0^\infty|\psi_F^{(0)}|^{2+4/3}\rho
496: d\rho}{\int_0^\infty|\psi_B^{(0)}|^2\rho
497: d\rho} =
498: {\frac{3A}{5}}\left[
499: \frac{M_F\omega}{\pi \hbar}    \right]^{2/3}.
500: \end{eqnarray}
501: In Eqs. (\ref{i}) and (\ref{j}) 
502: the normalization 
503: is given by $\int_{-\infty}^\infty |\phi_i(z,t)|^2
504: dz = N_i$. In these equations we have set the anisotropy parameter 
505: $\nu=0$
506:  to remove the axial trap and thus to generate axially-free
507: quasi-one-dimensional solitons.
508: 
509: 
510: 
511: 
512: For calculational purpose it is convenient to reduce 
513: the set  (\ref{i}) and (\ref{j})  to
514: dimensionless form 
515: by introducing convenient  dimensionless variables. Although the algebra is
516: quite straightforward, the expressions become messy with  different
517: factors
518: of masses. 
519: As we shall not be interested in a particular
520: boson-fermion 
521: system in this
522: paper, but will be concerned with the formation of fermionic bright 
523: solitons 
524: in general,  we take in the rest of this paper $m_B=3 m_F=
525: m({^{87}\mbox{Rb}})$, where $m({^{87}\mbox{Rb}})$ is the mass of
526: $^{87}$Rb, and 
527: whence $m_R=3m_F/4, M_B=M_F=m_B,$ and
528: $M_{BF}=m_B/2. $
529: In
530: the two experimental
531: situations of Refs. \cite{exp4,exp5} 
532: $m_B \approx 
533: 3m_F$. 
534: 
535: 
536: 
537: 
538: In Eqs. (\ref{i}) and (\ref{j}),   
539: we consider the dimensionless variables 
540: $\tau=t \omega/2$,
541: $y=z /l$,
542: ${\chi}_i=
543: \sqrt{(l/N_i)} \phi_i$, with $l=\sqrt{\hbar/( \omega m_B)}$, 
544: so that 
545: \begin{eqnarray}\label{m} \biggr[& - &i\frac{\partial
546: }{\partial \tau}
547: -\frac{d^2}{dy^2} 
548: +    
549:  N_{BB}
550: \left|{{\chi}_B}\right|^2\nonumber \\        
551: &+&N_{BF}
552:   \left|{{\chi}_F}\right|^2                  
553:  \biggr]{\chi}_{{B}}({y},\tau)=0,         
554: \end{eqnarray}
555: \begin{eqnarray}\label{n} \biggr[& - & i\frac{\partial
556: }{\partial \tau}-\frac{d^2}{dy^2} 
557: +
558: N_{FB}
559:   \left|{{\chi}_B} \right|^2
560:   \nonumber \\
561: &+&
562: N_{FF}
563:   \left|{{\chi}_F}
564: \right|^{4/3}
565:  \biggr]{\chi}_{{F}}(y,\tau)=0,
566: \end{eqnarray}
567: where
568: $N_{BB}=4a_{BB}N_B/l,$
569: $N_{BF}=8a_{BF}N_F/l,$ 
570: $N_{FB}=8a_{BF}N_B/l,$ and
571: $N_{FF}=9(6\pi N_F)^{2/3}/5. $
572:  In Eqs.
573: (\ref{m}) and (\ref{n}),
574: the normalization condition  is given by 
575: $\int_{-\infty}^\infty |\chi_i(y,\tau)|^2 dy =1 .$
576: Equations (\ref{m}) and (\ref{n}) are the coupled one-dimensional 
577: NLS equations describing the formation of solitons in the DFG-BEC mixture 
578: in a cigar-shaped quasi-one-dimensional geometry.
579: 
580:  
581: 
582: In Eqs. (\ref{m}) and (\ref{n}) the term $N_{FF}|\chi_F|^{4/3}$ represents
583: a very strong Pauli repulsion which increases with the fermion number
584: $N_F$. The purpose of this study is to show that a sufficiently strong  
585: attractive boson-fermion coupling
586: term $N_{FB}|\chi_B|^2$ can  overcome
587: this repulsion and  form the bright solitons.    
588: 
589:   
590: 
591:  
592:  
593:  
594: \section{Numerical Result}
595:  
596:  
597:  
598: We solve the coupled mean-field-hydrodynamic  equations 
599:  (\ref{m}) and
600: (\ref{n}) for bright  solitons
601: numerically using a time-iteration
602: method based on the Crank-Nicholson discretization scheme
603: elaborated in Ref. \cite{sk1}.  
604: We
605: discretize the mean-field-hydrodynamic  equation
606: using time step $0.0005$ and space step $0.025$.
607: 
608: 
609: 
610: We performed  the time evolution of the set of equations (\ref{m}) and 
611: (\ref{n}) introducing harmonic oscillator potentials $y^2$ in these
612: equations and setting the nonlinear terms to
613: zero: $N_{BB}=N_{BF}=N_{FB}=N_{FF}=0$ and  
614: starting with the eigenfunctions of the linear harmonic
615: oscillator problem, e.g., with $\chi_B(y,\tau)=\chi_F(y,\tau)=
616: \pi^{-1/4}\exp(-y^2/2)\exp(-i\tau).$  The introduction of the extra
617: harmonic oscillator potential in these equations 
618: only aids in starting the time evolution
619: with an exact analytic
620: solution. In the end the harmonic
621: oscillator potentials will be set equal to zero  and will have no effect
622: on
623: the final wave function for solitons. 
624: During the course of time evolution the nonlinear
625: terms are  switched on very slowly and the resultant solutions iterated 
626: until convergence was obtained. Then the time evolution is continued
627: and the harmonic oscillator potential
628: terms in both bosonic and fermionic equations are  slowly switched off by
629: reducing the $y^2$ term to zero in 10000 steps of time evolution.
630: Then 
631: the resultant solutions are iterated about 50000 times 
632: for
633: convergence without any harmonic oscillator potential. 
634: If converged
635: solutions are obtained, they correspond to the
636: required axially-free bright solutions in the absence of any axial
637: potential.   
638: In our numerical investigation 
639: as in the theoretical study of Refs. \cite{jz,ska} 
640: we use  $\omega = 2\pi
641: \times
642: 100$ Hz, and
643: take $m_B$ as the mass of $^{87}$Rb. Consequently, the unit of
644: length $l\approx1$ $\mu$m and unit of time $2/\omega \approx 3$ ms. 
645: 
646: 
647: \begin{figure}%[!ht]
648:  
649: \begin{center}
650: \includegraphics[width=1.\linewidth]{fig1.ps}
651: \end{center}
652: 
653: \caption{(Color online)  The stationary 
654:  function $|\phi_i(z,t)|$
655:  for axially-free bosonic (dotted line) and
656: fermionic (full line) bright solitons  vs. $z$ for 
657:  $N_F=N_B=1000$, $a_{BB}=5 $ nm,  $a_{BF}=-20 $ nm,  harmonic
658: oscillator length $l\approx 1$ $\mu$m and $\nu =0$.  
659: The arrows in dotted
660: and full lines indicate the bosonic and fermionic axes, respectively. 
661: The nonlinearity
662: parameters are
663: $N_{BB}=20 $, $N_{BF}= -160$, $N_{FB}= -160$, and $N_{FF}=274.6$.
664: }
665: \end{figure}
666: 
667: 
668: 
669: 
670: 
671: 
672: \begin{figure}%[!ht]
673:  
674: \begin{center}
675: \includegraphics[width=1.\linewidth]{fig2.ps}
676: \end{center}
677: 
678: \caption{(Color online)  The stationary 
679: function $|\phi_i(z,t)|$
680:  for axially-free bosonic (dotted line) and
681: fermionic (solid line) bright solitons  vs. $z$ for 
682:  $N_F=1000$, $N_B=10000$, $a_{BB}=-1 $ nm,  $a_{BF}=-1.875 $ nm, harmonic
683: oscillator length $l\approx 1$ $\mu$m  and $\nu =0$. The arrows in dotted
684: and full lines indicate the bosonic and fermionic axes, respectively. 
685:  The nonlinearity
686: parameters are
687: $N_{BB}=-40 $, $N_{BF}= -15$, $N_{FB}= -150$, and $N_{FF}=274.6$.
688: }
689: \end{figure}
690: 
691: 
692: 
693: First we solve Eqs. (\ref{m}) and (\ref{n}) with  
694: $N_F=N_B= 1000$, $a_{BB}=5$ nm  and
695: $a_{BF}=-20$ 
696: nm. 
697: This value of  $a_{BB}$ is the
698: experimental scattering length of Rb atoms \cite{11}, and  $a_{BF}=-20$
699: nm is the experimental 
700: scattering
701: length of the Rb-K system \cite{exp5,exp5x}. With these
702: parameters the nonlinearities in Eqs. (\ref{m}) and (\ref{n}) are
703: $N_{BB}=20 $, $N_{BF}= -160$, $N_{FB}= -160$, and $N_{FF}=274.6$.
704: The converged  bright solitons are plotted in Fig. 1. In this
705: case the
706: fermionic and bosonic functions, $\phi_F$ and $\phi_B$, respectively,   
707: have similar spatial extentions.
708: It is possible to have solitons with different 
709: extensions in space by varying the parameters of the system. 
710: We took the experimental values for the scattering lengths in
711: Fig. 1. However,  the scattering length can be manipulated in the
712: boson-fermion
713: $^6$Li-$^{23}$Na and $^{40}$K-$^{87}$Rb systems
714: near the recently
715: discovered Feshbach resonances in them \cite{fesh} by varying a background
716: magnetic field. Thus by varying the scattering length and the number of
717: atoms we could arrive at different values of nonlinearity than in
718: Fig. 1.  
719: 
720: 
721: 
722: 
723: 
724: To simulate a different situation of nonlinearity parameters next we take
725: $a_{BB}=-1$ nm, $a_{BF}=-1.875$ nm, $N_F=1000$, and $N_B=10000$. so that
726: $N_{BB}=-40 $, $N_{BF}= -15$, $N_{FB}= -150$, and $N_{FF}=274.6$. In this
727: case the profiles of the bright solitons shown in Fig. 2 are very
728: different from those in Fig. 1. In Fig. 1 both the solitons are localized
729: to a small region in space, whereas in Fig. 2 only the bosonic soliton is
730: localized to a small region in space whereas the fermionic soliton extends
731: to a very large region of space. 
732: 
733: 
734:  
735: 
736: 
737: \begin{figure}%[!ht]
738:  
739: \begin{center}
740: \includegraphics[width=1.\linewidth]{fig3.ps}
741: \end{center}
742: 
743: \caption{(Color online)  
744: The stationary 
745:  function $|\phi_i(z,t)|$
746:  for axially-free bosonic (dotted line) and
747: fermionic (solid line) bright solitons  vs. $z$ for 
748:  $N_F=1000$, $N_B=10000$, $a_{BB}=0.5 $ nm,  $a_{BF}=-3.75 $ nm,  $\nu
749: =0$
750: and
751: harmonic
752: oscillator length $l\approx 1$ $\mu$m.  The arrows in dotted
753: and full lines indicate the bosonic and fermionic axes, respectively. 
754: The nonlinearity parameters are
755: $N_{BB}=20 $, $N_{BF}= -30$, $N_{FB}= -300$, and $N_{FF}=274.6$.
756: }
757: \end{figure}
758: 
759: \begin{figure}%[!ht]
760:  
761: \begin{center}
762: \includegraphics[width=\linewidth]{fig4a.ps}
763: \includegraphics[width=\linewidth]{fig4b.ps}
764: \end{center}
765: 
766: \caption{(Color online)  The 
767:  function $|\phi_i(z,t)|$
768:  for (a) bosonic and
769: (b) fermionic bright solitons  vs. $z$  and $t$ for
770: the solitons of Fig. 3.
771: At $t=0$ 
772:  $N_F=1000$, $N_B=10000$, $a_{BB}=0.5 $ nm,  $a_{BF}=-3.75 $ nm,  $\nu
773: =0$
774: and
775: harmonic
776: oscillator length $l\approx 1$ $\mu$m.  The nonlinearity parameters at
777: $t=0$ are
778: $N_{BB}=20 $, $N_{BF}= -30$, $N_{FB}= -300$, and $N_{FF}=274.6$. At
779: $t=100$ ms (marked by arrows) the bright solitons are set into small
780: breathing oscillations by suddenly jumping the nonlinearities $N_{BF}$ and
781: $N_{FB}$ to $-33$ and $-330$, respectively. } \end{figure}
782: 
783: Next we consider  $N_F=1000$, $N_B=10000$, $a_{BB}=0.5 $ nm
784: and
785: $a_{BF}=-3.75 $ nm, so that the nonlinearity parameters are 
786: $N_{BB}=20 $, $N_{BF}= -30$, $N_{FB}= -300$, and $N_{FF}=274.6$.
787: The profiles of the solitons in Fig. 3 are different from those in Figs. 1
788: and 2. In this case the bosonic  function extends over  a longer 
789: region in space than the fermionic  function. In Figs. 1 and 2 it was
790: the fermionic  function that extends over a longer region in space. 
791: 
792: 
793: In Fig. 2 the nonlinearity $N_{FB}$
794: appearing in the fermion equation is less attractive compared to that in
795: Figs. 1 and 3. Hence the resultant  nonlinear interaction in the fermion
796: component 
797: is less
798: attractive and hence the fermionic soliton extends to a large distance in
799: space. If the attraction in  $N_{FB}$   is further reduced the fermionic
800: soliton ceases to bind. 
801: 
802: 
803: 
804: Hence by manipulating the parameters one could have different situations 
805: of localization of the solitons. One could either have both the solitons
806: extending up to similar distance in space as in Fig. 1 or one of the
807: solitons extending over a longer region in space as in Figs. 2 and 3. It
808: is worth emphasizing that in the fermionic equation (\ref{n}) the diagonal
809: nonlinearity $N_{FF}$ is repulsive, hence the binding solely comes from
810: the attractive off-diagonal nonlinearity $N_{FB}$  corresponding to an
811: attractive boson-fermion interaction. Hence for a fermionic soliton to
812: appear  $N_{FB}$ and $a_{BF}$ are always taken to be negative or
813: attractive. Consequently,   $N_{BF}$ is also negative. Finally, terms
814: $N_{BB}$ and $a_{BB}$ could be  either  positive or negative. When these
815: terms are positive or repulsive the bosonic solitons are formed due to an
816: attractive or negative $N_{BF}$, as in Fig. 3.
817: 
818: 
819: 
820: 
821: 
822: Next we study the stability of the bright solitons numerically. We
823: consider the soliton of Fig. 3 and during time evolution we suddenly jump
824: at $t=100$ ms the nonlinearity $N_{BF}$ from $-30$ to $-33$ and the
825: nonlinearity $N_{FB}$ from $-300$ to $-330$. This can be achieved by
826: manipulating a
827: background magnetic field near 
828: a
829: Feshbach resonance \cite{fesh} in the boson-fermion interaction and
830: thus varying the boson-fermion scattering length by 10$\%$. Due to
831: the sudden change in the nonlinearity the bosonic and  fermionic bright
832: solitons are set into stable breathing oscillation.  The evolution of the
833: wave function profile of the two solitons are shown in Figs. 4. 
834: The solitons are found to execute stable non-periodic breathing
835: oscillation. In Fig. 5 
836: we plot the root-mean-square size $\langle z \rangle_{rms}$ of the bosonic
837: and fermionic
838: solitons  of Fig. 4 as a function of time. The breathing
839: oscillation of the two solitons after the perturbation is applied results
840: in the stable non-periodic oscillation of the root-mean-square sizes
841: illustrated in Fig. 5. As, after applying the perturbation, the 
842: boson-fermion attraction has been increased this corresponds to a 
843: reduction in the root-mean-square size $\langle z \rangle_{rms}$ as we 
844: find in Fig. 5.  The steady propagation of the solitons in Figs. 4 and the 
845: stable oscillation of their root-mean-square sizes in Fig. 5
846: after the
847: perturbation is
848: applied demonstrate the stability of the solitons. 
849: 
850: 
851: 
852: \begin{figure}%[!ht]
853:  
854: \begin{center}
855: \includegraphics[width=\linewidth]{fig5.ps}
856: \end{center}
857: 
858: \caption{(Color online)  The root-mean-square size $\langle z
859: \rangle_{rms}$ of the bosonic (dotted line)
860: and fermionic (full line)
861: solitons  of Fig. 4 vs. time. } \end{figure}
862: 
863: 
864: Finally, we consider the fermionic bright solitons formed on a periodic
865: optical-lattice potential. For that purpose we include in Eqs. (\ref{m})
866: and (\ref{n}) the following optical-lattice potential formed by a
867: standing-wave laser beam \cite{oplat} 
868: \begin{equation}
869: V_{\mbox{\small{OP}}} = V_0 \sin^2(2\pi  y/\lambda),
870: \end{equation}
871: where $V_0$ is the strength, and $\lambda$ is the wave length of the
872: laser. In our calculation we take $V_0=100$ and $\lambda = \pi/2$. To
873: solve Eqs. (\ref{m}) and (\ref{n}) with this optical-lattice potential and
874: desired nonlinearities, we again start the time evolution with the
875: solution of the linear oscillator problem. In the course of time evolution
876: the nonlinearities and the optical-lattice potential are slowly
877: introduced and eventually the harmonic oscillator potential is slowly
878: removed.  Then the final solutions are iterated for convergence. 
879:  The
880: resultant soliton wave functions are plotted in Fig. 6 for $N_F= 1000$,
881: $N_B=10000$, $a_{BB}=0.3$ nm, $a_{BF}=-2.375$ nm.  
882: The optical-lattice potential introduces modulations in the solitonic 
883: wave function. For
884: the parameters of Fig. 6 the modulations are more prominent on the
885: fermionic soliton than the bosonic one. By changing the
886: parameters it is
887: possible to have modulations on the bosonic soliton as well. 
888: 
889: \begin{figure}[!ht]
890:  
891: \begin{center}
892: \includegraphics[width=\linewidth]{fig6.ps}
893: \end{center}
894: 
895: \caption{(Color online)  
896: The stationary 
897:  function $|\phi_i(z,t)|$
898:  for bosonic (dotted line) and
899: fermionic (full line) bright solitons  vs. $z$ for 
900:  $N_F=1000$, $N_B=10000$, $a_{BB}=0.3 $ nm,  $a_{BF}=-2.375 $ nm,  $\nu
901: =0$
902: and
903: harmonic
904: oscillator length $l\approx 1$ $\mu$m in the presence of the
905: optical-lattice potential $V(y)=100 \sin^2 (4 y)$. The arrows in dotted
906: and full lines indicate the bosonic and fermionic axes, respectively. 
907:  The nonlinearity
908: parameters
909: are $N_{BB}=12 $, $N_{BF}= -19$, $N_{FB}= -190$, and $N_{FF}=274.6$.
910: } \end{figure}
911: 
912: 
913: \section{Summary}
914:  
915: We use a coupled set of time-dependent mean-field-hydrodynamic equations
916: for a boson-fermion mixture to study the formation of fermionic bright
917: soliton in a DFG as a stationary state. In this study we take the
918: boson-boson interaction to be both attractive and repulsive and the
919: boson-fermion interaction to be attractive. An attractive boson-fermion
920: interaction is necessary for the formation of a fermionic bright soliton
921: as the diagonal nonlinearity $N_{FF}$ in the fermion-fermion system is
922: always repulsive. 
923: 
924: 
925: In the present study we demonstrate that stable solitons can be formed 
926: in coupled NLS equations for the boson-fermion mixture
927: with repulsive diagonal nonlinearities  and
928: attractive off-diagonal nonlinearities above some cut-off values. In
929: another study \cite{pla} we showed the possibility of the formation of
930: bright 
931: solitons in coupled bosonic condensates with intraspecies repulsion
932: supported by interspecies
933: attraction. 
934: The stability of the present fermionic and bosonic solitons is
935: demonstrated through
936: their sustained breathing oscillation  initiated by a sudden jump
937: in the boson-fermion scattering length. 
938: Bright solitons have been created experimentally
939: in attractive
940: BECs in three dimensions in the presence of  radial trapping only without
941: any
942: axial trapping \cite{exdks}. 
943: In view of this fermionic bright solitons can be observed in the
944: laboratory in the presence of  radial trapping only in a mixture of a DFG 
945: and BEC. We also suggest the possibility of the formation of fermionic  
946: solitons
947: on a periodic optical-lattice potential.  
948: In the present investigation  we used a set of mean-field equations for 
949: the DFG-BEC mixture. A proper treatment of the DFG should be performed 
950: using a fully antisymmetrized many-body Slater determinant wave
951: function. However, we do not believe that the present conclusion about the
952: existence of robust fermionic solitons in a DFG-BEC mixture to be so
953: peculiar as to have no general validity. 
954: 
955:  
956: \acknowledgments
957: 
958: The work is supported in part by the CNPq 
959: of Brazil.
960:  
961:   
962: %\section*{References}
963:  
964: \begin{thebibliography}{99}
965:  
966: 
967: 
968: \bibitem{exp1}B. DeMarco and D. S. Jin, Science {\bf 285}, 1703
969: (1999).
970: 
971: \bibitem{exp2} K. M. O'Hara, S. L. Hemmer, M. E. Gehm,  S. R. Granade, and
972: J. E. Thomas, Science {\bf 298}, 2179 (2002).
973: 
974: 
975: \bibitem{exp3}F. Schreck, L. Khaykovich, K. L. Corwin, G. Ferrari,
976: T. Bourdel, J.
977: Cubizolles,  and C. Salomon,
978: Phys. Rev. Lett. {\bf 87}, 080403 (2001);
979: A. G. Truscott, K. E. Strecker, W. I. McAlexander,
980: G. B. Partridge, and R. G. Hulet, Science {\bf 291}, 2570
981: (2001). 
982: 
983: 
984: \bibitem{exp4} Z. Hadzibabic, C. A. Stan,
985: K. Dieckmann, S. Gupta, M. W. Zwierlein, A. Gorlitz, and W. Ketterle,
986: Phys. Rev. Lett. {\bf 88}, 160401 (2002).
987: 
988: 
989: \bibitem{exp5}G. Modugno, G. Roati, F. Riboli, F. Ferlaino, R. J. Brecha,
990: and M. Inguscio, Science {\bf 297}, 2240 (2002). 
991: 
992: \bibitem{exp5x} G. Roati,  F. Riboli, G. Modugno, and M. Inguscio,
993: Phys. Rev. Lett. {\bf 89}, 150403 (2002).
994: 
995: 
996: 
997: 
998: \bibitem{exp6}K. E. Strecker, G. B. Partridge, and R. G. Hulet,
999: Phys. Rev. Lett. {\bf 91}, 080406 (2003);
1000:  Z. Hadzibabic,
1001: S. Gupta, C. A. Stan,       C. H. Schunck,  M. W. Zwierlein,
1002:  K. Dieckmann, and W. Ketterle, {\it ibid.} {\bf 91}, 160401 (2003).
1003: 
1004: 
1005: \bibitem{yyy1} K. Molmer, Phys. Rev. Lett.
1006: {\bf 80}, 1804 (1998).
1007: 
1008: \bibitem{yyy}R. Roth, Phys. Rev. A {\bf 66}, 013614 (2002); R. Roth and 
1009: H. Feldmeier,  {\it ibid.}  {\bf 65}, 021603(R) (2002); T. Miyakawa,
1010: T. Suzuki, and H. Yabu,  {\it ibid.}  {\bf 64}, 033611 (2001);
1011: X.-J. Liu, M. Modugno, and H. Hu,
1012:  {\it ibid.}  {\bf 68}, 053605 (2003);  
1013: X.-J. Liu  and H. Hu,
1014:  {\it ibid.}  {\bf 67}, 023613 (2003);  L. Vichi, M. Amoruso, A. Minguzzi, S. Stringari,
1015: and M. Tosi, Eur. Phys. J. D {\bf 11}, 335 (2000).
1016: 
1017: \bibitem{zzz}M. Modugno, F. Ferlaino, F. Riboli, G. Roati, G. Modugno, and
1018: M. Inguscio, 
1019: Phys. Rev. A {\bf 68}, 043626 (2003).           
1020: 
1021:  \bibitem{capu}P. Capuzzi, A. Minguzzi, and M. P. Tosi, 
1022:  Phys. Rev. A {\bf 69},
1023: 053615 (2004); {\bf 68}, 033605
1024: (2003);
1025: {\bf 67}, 053605 (2003). 
1026: 
1027: 
1028: \bibitem{ska}S. K. Adhikari, Phys. Rev. A {\bf 70}, 043617 (2004).
1029: 
1030: 
1031: \bibitem{11}F.  Dalfovo, S. Giorgini, L. P. Pitaevskii, and S.  Stringari,
1032: Rev. Mod. Phys.  {\bf 71}, 463        (1999);  L. P. Pitaevskii and
1033: S.  Stringari, {\it Bose-Einstein Condensation} (Clarendon Press, Oxford,
1034: 2003).
1035: 
1036: 
1037: 
1038: 
1039: 
1040: 
1041: 
1042: 
1043: \bibitem{1}Y. S.  Kivshar  and  G. P. Agrawal,  {\it Optical
1044: Solitons
1045: - From
1046: Fibers to Photonic Crystals} (Academic Press, 
1047: San Diego,  2003); A. Hasegawa and Y. Kodama,
1048: {\it Solitons in Optical Communications}
1049: (Clarendon Press, Oxford, 1995); G. P. Agrawal, {\it Nonlinear Fiber 
1050: Optics, Second  Edition} (Academic Press, San Diego, 1995).
1051: 
1052:  
1053: \bibitem{exdks}K. E. Strecker, G. B. Partridge, A. G. Truscott, and R. G.  
1054: Hulet, Nature (London), {\bf 417}, 150 (2002); L. Khaykovich, F. Schreck,
1055: G.
1056: Ferrari, T. Bourdel, J. Cubizolles, L. D. Carr, Y. Castin, and C. Salomon,
1057: Science, {\bf 296}, 1290 (2002).
1058: 
1059: 
1060: 
1061: \bibitem{thdks} 
1062: V. M. Perez-Garcia, H. Michinel, and H.
1063: Herrero, Phys. Rev. A {\bf 57}, 3837 (1998);  U. Al Khawaja,
1064: H. T. C. Stoof, R. G.
1065: Hulet, K. E. Strecker, and G. B. Partridge, Phys. Rev. Lett. {\bf 89},
1066: 200404  (2002); S. K. Adhikari, New J. Phys. {\bf 5}, 137 (2003). 
1067: 
1068: \bibitem{oplat} F. S. Cataliotti, S. Burger, C. Fort, P. Maddaloni,
1069: F. Minardi, A. Trombettoni, A. Smerzi, and M. Inguscio,
1070: Science
1071: {\bf 293}, 843 (2001); M. Greiner, O. Mandel, T. Esslinger,
1072: T. W. H\"ansch,
1073: and
1074: I. Bloch, Nature (London) {\bf 415}, 39 (2002);
1075: O. Morsch, J. H. Muller, M. Cristiani, D. Ciampini,  and E. Arimondo,
1076: Phys. Rev. Lett. {\bf 87}, 140402 (2001).
1077: 
1078: 
1079: \bibitem{jz} D. M. Jezek, M. Barranco, M. Guilleumas, R. Mayol, and M. Pi, 
1080:  Phys. Rev. A {\bf 70},
1081: 043630 (2004).
1082: 
1083: 
1084: \bibitem{pi} M. Pi, X. Vi\~nas, F. Garcias, and M. Barranco, Phys. Lett. B
1085: {\bf 215}, 5 (1988). 
1086: 
1087: 
1088: 
1089: 
1090: 
1091: 
1092: 
1093: 
1094: \bibitem{abdul} F. K. Abdullaev and R. Galimzyanov, J. Phys. B {\bf 36},
1095: 1099 (2003).  
1096:  
1097:  
1098:  
1099:  
1100:  
1101:  
1102: 
1103: 
1104: 
1105:  
1106:  
1107: 
1108: 
1109: 
1110: 
1111: 
1112: 
1113: 
1114: 
1115:  
1116: \bibitem{sk1}S. K.  Adhikari  and P. Muruganandam, J. Phys. B  
1117: {\bf 35}, 2831 (2002); P.
1118:  Muruganandam  and S. K.  Adhikari,  {\it ibid.}    {\bf 36}, 2501 
1119: (2003). 
1120:  
1121: 
1122: 
1123: 
1124: \bibitem{fesh}C. A. Stan, M. W. Zwierlein, C. H. Schunck,
1125: S. M. F. Raupach, and W. Ketterle, Phys. Rev. Lett. {\bf 93}, 143001
1126: (2004);
1127: S. Inouye, J. Goldwin, M. L. Olsen, C. Ticknor, J. L. Bohn, and D. S. Jin,
1128: {\it ibid.}
1129: {\bf 93}, 183201 (2004).
1130: 
1131: 
1132: \bibitem{pla}S. K. Adhikari, Phys. Lett. A     (2005) (in press). 
1133: 
1134: \end{thebibliography} 
1135:  
1136: \end{document}
1137: 
1138: 
1139: