cond-mat0204376/fj1.tex
1: \documentclass[prl,twocolumn,floatfix,superscriptaddress,showpacs]{revtex4}
2: 
3: 
4: \usepackage{graphicx}
5: %\usepackage{dcolumn,float}
6: %\usepackage{bm}
7:   
8: 
9: \begin{document}
10: 
11: \title{Josephson effect in superfluid atomic Fermi-gases}
12: 
13: \date{\today}
14: 
15: \author{Gh.-S.\ Paraoanu}
16: %\email{paraoanu@phys.jyu.fi}
17: \altaffiliation[also at the ]{Department
18:         of Theoretical Physics, National Institute for Physics and Nuclear
19:         Engineering, PO BOX MG-6, R-76900, Bucharest, Romania.}
20: \affiliation{Department of Physics, University of
21: Jyv\"askyl\"a, P.O.Box 35, FIN-40351 Jyv\"askyl\"a, Finland}
22: \affiliation{Department of Physics, Loomis Laboratory, 1110 W.\ Green
23: Street,
24: University of Illinois at Urbana-Champaign, Urbana IL61801, USA}
25: 
26: 
27: \author{M.\ Rodriguez}
28: %\email{mrodrigu@lce.hut.fi}
29: \affiliation{Laboratory of Computational Engineering, P.O.Box 9400, 
30: FIN-02015
31: Helsinki University of Technology, Finland}
32: 
33: 
34: 
35: \author{P.\ T\"orm\"a}
36: %\email{ptorma@phys.jyu.fi}
37: \affiliation{Department of Physics, University of Jyv\"askyl\"a, P.O.Box 
38: 35,
39: FIN-40351 Jyv\"askyl\"a, Finland}
40: 
41: 
42: 
43: 
44: 
45: \begin{abstract}
46: 
47: We consider an analog of the internal Josephson effect in superfluid 
48: atomic Fermi-gases.
49: Four different hyperfine states of the atoms are assumed to be
50: trapped and to form two superfluids {\it via} the BCS-type of pairing.  
51: We show that Josephson oscillations can be realized by coupling
52: the superfluids with two laser fields. Choosing the laser detunings
53: in a suitable way leads to  
54: an asymmetric below-gap tunneling effect
55: for which there exists no analogue in the 
56: context of solid-state superconductivity. 
57: 
58: \end{abstract}
59: 
60: \pacs{05.30.Fk, 32.80.-t, 74.25.-q}
61: 
62: \maketitle
63: 
64: 
65: 
66: 
67: 
68:   Cooling of trapped gases of Fermionic atoms well below the Fermi
69:   temperature \cite{Debbie,Hulet,Salomon,Thomas,KetterleF}  makes it reasonable to
70:   anticipate the achievement of the predicted BCS-transition
71:   \cite{Stoof,Holland,Zoller,Pethick}.  The existence of a gap in the excitation
72:   spectrum of the superfluid Fermi-gas will be the first issue to
73:   address, and several methods for detecting it have already been
74:   proposed \cite{alli,ours}.  Trapped atomic Fermi-gases will allow to
75:   study and test fermion-fermion pairing 
76:   theories in a tunable, controlled manner. For example
77:   the classic problem of the BCS-BEC crossover when the interparticle
78:   attraction varies \cite{bcsbec} could be studied using the
79:   possibility of tuning the interatomic scattering length using 
80: Feshbach resonances \cite{Holland,feshbach}. 
81: Besides the standard superfluid phenomenology, 
82: gases of Cooper-paired atoms are expected to have properties which are
83:   specific to atomic gases only and not present, or not easily realizable,
84:   in metallic superconductors or Helium. For instance, the trapping potential
85: has a major effect on the characteristic lengths of the 
86: superfluid Fermi-gas \cite{vortex}. 
87: 
88: In this paper we propose a way to investigate the
89: Josephson 
90: effect in trapped superfluids of Fermionic atoms. We find a phenomenon 
91:   that is unique to atomic
92:   Fermi-gases, namely an asymmetry in the Josephson currents 
93: corresponding to the "up" and "down" spin states. 
94: We assume that Fermionic atoms in four different hyperfine states (we label
95: them $|g\rangle$, $|g'\rangle$, $|e\rangle$ and $|e'\rangle$) are trapped
96: simultaneously in an optical trap --- recently  all-optical 
97: trapping and cooling below the
98: degeneracy point of the two lowest hyperfine states of $^6$Li has been 
99: demonstrated \cite{Thomas}.
100: The s-wave scattering lengths are assumed to be large 
101: and negative between atoms 
102: in states
103: $|g\rangle$ and $|g'\rangle$, as well as between those in $|e\rangle$ 
104: and $|e'\rangle$,
105: and the chemical potentials 
106: $\mu_g \simeq \mu_{g'}$ and $\mu_e \simeq \mu_{e'}$.
107: For all other combinations of two atoms in different states 
108: the scattering length
109: is assumed to be small and/or the chemical potentials unequal. This leads to
110: the existence of two superfluids, one consisting of Cooper pairs of atoms
111: in the states $|g\rangle$ and $|g'\rangle$, and the other of 
112: $|e\rangle$--$|e'\rangle$
113: pairs. The configuration is experimentally challenging, but
114: in principle possible by the choice of right atoms and hyperfine states,
115: adjusting the number of atoms, and tuning the scattering lengths in 
116: magnetic fields by using Feshbach resonances \cite{Holland,feshbach}. 
117: 
118: The two superfluids are coupled 
119: by driving laser-induced transitions between the states $|g\rangle$ and 
120: $|e\rangle$
121: with the laser Rabi frequency $\Omega$ and detuning $\delta$, and between
122: the states $|g'\rangle$ and $|e'\rangle$ with the Rabi frequency $\Omega'$ 
123: and detuning $\delta'$. For a Raman-process these are effective
124: quantities. If several 
125: lasers are used then in order to be able to see the Josephson
126: oscillations they should maintain their phase coherence for a
127: time much longer than the inverse of the detunings. 
128: In the case of metals, 
129: the two superconductors are spatially separated and connected by a 
130: tunneling junction. In our scheme, 
131: the superfluids share 
132: the same spatial region 
133: and are connected by the laser-coupling of the atoms' internal states; 
134: this resembles the internal Josephson effect 
135: in atomic Bose-Einstein condensates \cite{internal} 
136: or in superfluid $^3$He-A \cite{leggett}.
137: 
138: For metallic 
139: superconductors the a.c. Josephson current is driven by 
140: applying a voltage over the junction -- here the role of the voltage is 
141: played by the laser detunings. The difference is that the detunings 
142: can be different for the 
143: two states forming the pair; in the metallic superconductor analogy 
144: this would mean having a different voltage for the 
145: spin-up and spin-down electrons, a situation which has 
146: not been investigated in the context 
147: of metallic superconductors. There is an interesting connection
148: to recent experiments on superconductor-ferromagnet
149: proximity effects, where the chemical potentials of the spin
150: up and down electrons are slightly different in the ferromagnet 
151: due to the exchange interaction \cite{ferro}.
152: 
153: 
154: %the eponymous effect
155: 
156: 
157: 
158: 
159: 
160: 
161: 
162: 
163: 
164: 
165: 
166: 
167: 
168: 
169: We consider a system described by the standard BCS-theory. The laser
170: interaction is assumed to be a small perturbation and its effect is calculated
171: using linear response theory. The observable of interest is the change in
172: the number of particles in one of the states, say $|e\rangle$ or $|e'\rangle$.
173: 
174: In the rotating wave approximation the interaction of the laser light
175: with the matter fields can be described by a time-independent 
176: Hamiltonian in which the detunings $\delta$ and $\delta'$ play the role of an
177: externally imposed difference in the chemical potential
178: of the two states. The total Hamiltonian becomes then $\hat{H} = 
179: \hat{H_0} +
180: \hat{H}_{T}$, where
181: \begin{widetext}
182: \begin{eqnarray}
183:  \hat{H_0} &=& \hat{H}_{BCS}
184: +\left(\mu_{e} + \frac{\delta}{2}\right)\int 
185: d\vec{r}
186: \hat{\psi}_e^{\dagger}(\vec{r})\hat{\psi}_e
187: (\vec{r})
188: + \left(\mu_{g} - \frac{\delta}{2}\right)\int d\vec{r}
189: \hat{\psi}_g^{\dagger}(\vec{r})\hat{\psi}_g
190: (\vec{r})   \nonumber \\
191: && +\left(\mu_{e} + \frac{\delta'}{2}\right)\int 
192: d\vec{r}
193: \hat{\psi}_{e'}^{\dagger}(\vec{r})\hat{\psi}_{e'}
194: (\vec{r})
195: + \left(\mu_{g} - \frac{\delta'}{2}\right)\int d\vec{r}
196: \hat{\psi}_{g'}^{\dagger}(\vec{r})\hat{\psi}_{g'}
197: (\vec{r}) \nonumber. 
198: \end{eqnarray}
199: \end{widetext}
200: Here $\mu_{e}$ and $\mu_{g}$ are the chemical potentials of the Fermi 
201: gases before 
202: the laser was turned on, $\mu_g =\mu_{g'}$ and $\mu_e = \mu_{e'}$ 
203: in order to allow standard BCS pairing.
204: The Hamiltonian $\hat{H}_{BCS}$ is the BCS-approximation 
205: of the matter Hamiltonian with 
206: the chemical potential included \cite{boogie}. 
207: Figure {\ref{fj_fig1}} presents a schematic view of what happens
208: in this case: the laser detunings shift the chemical potentials of 
209: the four hyperfine states.
210: \begin{figure}
211: \includegraphics[scale=0.65]{fj_fig1.eps}
212: \caption{The effect of the detunings $\delta$ and $\delta '$ of the two 
213: lasers. The initial chemical potentials (in the absence of laser 
214: couplings) were $\mu_{g}=\mu_{g'}$ and $\mu_{e}=\mu_{e'}$. The states
215: $(gg')$ and $(ee')$ are Cooper-paired.}
216: \label{fj_fig1}
217: \end{figure}
218: 
219: The transfer Hamiltonian is given by 
220: \begin{eqnarray}
221: \hat{H}_{T} &=& \int d\vec{r}\Omega (\vec{r})\hat{\psi}_e
222: ^{\dagger}(\vec{r})\hat{\psi}_g(\vec{r}) + \Omega^{*}
223: (\vec{r})\hat{\psi}_g^{\dagger}(\vec{r})
224: \hat{\psi}_e(\vec{r}) \nonumber \\
225: &+& \int d\vec{r}\Omega' (\vec{r})\hat{\psi}_{e'}
226: ^{\dagger}(\vec{r})\hat{\psi}_{g'}(\vec{r}) + \Omega'^{*}
227: (\vec{r})\hat{\psi}_{g'}^{\dagger}(\vec{r})
228: \hat{\psi}_{e'}(\vec{r}),
229: \end{eqnarray}
230: with $\Omega (\vec{r})$ and $\Omega' (\vec{r})$ characterizing 
231: the local strength of the 
232: matter-field interaction.  
233: 
234: The main observable of interest, the 
235: rate of transferred atoms from, say, state $|g\rangle$ to 
236: state $|e\rangle$, is defined by
237: \begin{equation}
238: I_e = \frac{\partial}{\partial t}\int d\vec{r}\langle\Psi 
239: (t)|\hat{\psi}_e^{\dagger}(\vec{r})\hat{\psi}_e(\vec{r})
240: |\Psi (t)\rangle 
241: \end{equation}
242: (the definition for $I_{e'}$ is similar)
243: and can be further evaluated with the help of the Schr\"odinger equation
244: $i\hbar\frac{\partial}{\partial t}|\Psi (t)\rangle = \hat{H} |\Psi (t)
245: \rangle$ as
246: \begin{equation}
247: I_e = i \int d\vec{r}\langle\Psi (t)|
248: \Omega^{*}(\vec{r})\hat{\psi}_g^{\dagger}(\vec{r})
249: \hat{\psi}_e(\vec{r}) - \Omega (\vec{r})
250: \hat{\psi}_e^{\dagger}(\vec{r})
251: \hat{\psi}_g(\vec{r})|\Psi (t)\rangle.
252: \end{equation}
253: In the following we call $I_e$ the current in analogy to
254: metallic superconductors where the flux of electrons out of the
255: superconductor constitutes the electrical current.
256: 
257: We introduce an interaction representation with respect to 
258: $\hat{H_0}$ and use linear response theory with 
259: respect to $\hat{H}_{T}$. 
260: Validity of the linear response theory requires that
261: the laser intensity is small and the transfer of atoms can be
262: treated as a perturbation. 
263: 
264: We split the result for the current $I_e$ into a part which corresponds to
265: the Josephson current $I_{eJ}$ and to the part 
266: which describes normal single-particle current
267: $I_{eS}$, $I_e = I_{eJ} + I_{eS}$. The single-particle current can be 
268: evaluated at finite temperature using the standard techniques of 
269: superconducting Green's functions; we present however only the 
270: result for $T=0$ and positive detunings: 
271: \begin{eqnarray}
272: I_{eS} &=& -2\pi\sum_{n,m}\left|\int d\vec{r}\Omega 
273: (\vec{r})v^e_{n}(\vec{r})u^g_{m}(\vec{r})\right|^{2}\delta (\epsilon^e_{n} + 
274: \epsilon^g_{m} - \tilde{\delta} ) . \nonumber
275: \end{eqnarray}
276: Here the triplet $(u_{n}, v_{n}); \epsilon_{n}$ is a solution of the 
277: (nonuniform) Bogoliubov-de Gennes equations for superconductors 
278: \cite{boogie} and $\tilde{\delta} = \mu_e - \mu_g + \delta$. This is 
279: the 
280: standard Fermi Golden rule result and very similar to the ones 
281: obtained in \cite{ours}. The current $I_{eS}$ is zero when
282: $\delta < 2 \Delta$ since pair breaking is required for 
283: single particle excitations. 
284: Next we concentrate on the Josephson current -- this is non-zero 
285: also for detunings $\delta$ that are smaller than twice the gap energy. 
286: 
287: The Josephson current becomes 
288: \begin{widetext}
289: \begin{equation} 
290: I_{eJ} = -2 {\rm Im}
291: \left[ e^{-i(\tilde{\delta} + \tilde{\delta}')t} \sum_{n,m}\int 
292: d\vec{r}d\vec{r}'\Omega^{*}(\vec{r}) \Omega'^{*} (\vec{r}') 
293: u_{n}^{g*}(\vec{r})u^e_{m}(\vec{r}')v_{m}^{e*} 
294: (\vec{r})v^g_{n}(\vec{r}')
295: \left(\frac{1}{\tilde{\delta}' + 
296: \epsilon^g_{n} +
297: \epsilon^e_{m} + i\eta} -\frac{1}{\tilde{\delta}' - \epsilon^g_{n} -
298: \epsilon^e_{m} + i\eta} \right)  \right].\label{doibre} 
299: \end{equation}
300: \end{widetext}
301: The current $I_{e'J}$ is the same only that $\tilde{\delta}$ and $\tilde{\delta'}$ are
302: interchanged. Note that the oscillating term is proportional to both of
303: the detunings whereas the rest of the expression is proportional 
304: only to $\tilde{\delta'}$. For
305: the choice of a homogeneous system (large trap, local density
306: approximation) and a constant laser profile the expression simplifies
307: into \begin{eqnarray} I_{eJ} &=& I_0(\tilde{\delta}') \sin 
308: [(\tilde{\delta} + \tilde{\delta}')t]
309: \\ I_{e'J} &=& I_0(\tilde{\delta}) \sin [(\tilde{\delta} + 
310: \tilde{\delta}')t] . \end{eqnarray}
311: Both partners of the pair thus oscillate in phase, with the same
312: frequency $\tilde{\delta} + \tilde{\delta}'$. But the amplitudes are 
313: different whenever
314: the detunings $\tilde{\delta}$ and $\tilde{\delta}'$ differ. This means 
315: that more atoms
316: are transferred, say, in the $|g\rangle - |e\rangle$ oscillation than in
317: the $|g'\rangle - |e'\rangle$ one. 
318: 
319: A simple expression for $I_0(\delta)$
320: can be derived when we assume identical superfluids, that is $\Delta =
321: \Delta'$ and $\mu_{g} = \mu_{e} \equiv \mu$: 
322: \begin{eqnarray} 
323: I_0(\delta)
324: = \frac{\sqrt{2m^3}V}{\pi^{2}}\Delta^2 \Omega^2 \int_{-\mu}^\infty  
325: \frac{d\xi\sqrt{\mu +
326: \xi}}{\sqrt{\xi^2 + \Delta^2} (4\xi^2 + 4 \Delta^2 - 
327: \delta^2)}, \nonumber
328: \end{eqnarray} 
329: where $V$ is the volume of the sample and the variable $\xi$ is the 
330: continuous version of $\xi_{k}=\frac{k^{2}}{2m}-\mu$. Since $\Delta \ll 
331: \mu$, the result can also be written as
332: \begin{eqnarray}
333: I_0(\delta)
334: = \frac{\sqrt{2m^3\mu}V}{\pi^{2}}\Delta^2 \Omega^2 
335: \int_{-\infty}^\infty
336: \frac{d\xi}{\sqrt{\xi^2 + \Delta^2} (4\xi^2 + 4 \Delta^2 -   
337: \delta^2)}.\nonumber
338: \end{eqnarray}
339: A plot of the intensity $I_{0}$ as a function of the detuning $\delta$ 
340: is shown in Fig. {\ref{fj_fig2}}.
341: This result shows a divergence at $\delta = 2\Delta$, which 
342: reflects the divergence of the density of states for the two superconductors 
343: at the gap. For quite a large range of detunings, 
344: the amplitude $I_{0}$ of the Josephson current is approximately 
345: constant --- thus no asymmetry effect will be visible. The asymmetry is
346: most pronounced when the timescale of the oscillation, that is, $1/(\delta
347: + \delta')$ is close to $1/(2\Delta)$. Note that $1/(2\Delta)$ 
348: can also be understood as the Cooper pair correlation time based
349: on the uncertainty principle.
350: 
351: \begin{figure}
352: \includegraphics[scale=0.37]{fj_fig2.eps}
353: \caption{The current $I_{0}$ (arbitrary units) as a function of detuning
354: $\delta$, given in units of $\Delta$.}
355: \label{fj_fig2}
356: \end{figure}
357: 
358: According to the conventional intuitive picture of the Josephson effect,
359: the particles forming a Cooper-pair tunnel ``together'' through the junction.
360: Therefore our result seems counterintuitive at first glance.
361: The physics becomes, however, more transparent by a closer look at the equation
362: (\ref{doibre}). For simplicity, we consider here the transfer process to one direction only,
363: from the superfluid $(gg')$ to $(ee')$, which corresponds to the first denominator
364: in (\ref{doibre}). In the initial state $|g\rangle$ is paired with $|g'\rangle$, in the final
365: state $|e\rangle$ with $|e'\rangle$. The process has, however, an intermediate state as indicated
366: by the second-order form of the observable, and the intermediate states corresponding to
367: the observables $I_{eJ}$ and $I_{e'J}$ are different: For $I_{eJ}$,  
368: $|g'\rangle$ has been transferred into $|e'\rangle$. Therefore its pairing partner
369: $|g\rangle$ is left as an excitation in the superfluid $(gg')$ with the energy 
370: $\epsilon_{n}^{g}$ and $|e'\rangle$ becomes an excitation in the superfluid $(ee')$ with the energy
371: $\epsilon_{m}^{e'}=\epsilon_{m}^{e}$. 
372: In contrast, for $I_{e'J}$, the atom in $|g'\rangle$ remains as a quasiparticle of the energy $\epsilon_{n}^{g'}
373: = \epsilon_{n}^{g}$ in the superfluid $(gg')$ and
374: $|e\rangle$ becomes an excitation in the superfluid $(ee')$. 
375: For $I_{eJ}$, the initial 
376: energy of the Cooper pair was $\left(\mu_{g} - \frac{\delta}{2}\right) + 
377: \left(\mu_{g}-\frac{\delta '}{2}\right)$ and the energy of the 
378: intermediate state is 
379: $\left[\left(\mu_{g}-\frac{\delta }{2}\right) + 
380: \epsilon_{n}^{g}\right] + 
381: \left[\left(\mu_{e} + \frac{\delta '}{2}\right) 
382: +\epsilon_{m}^{e}\right]$ (for explanation, see Fig.{\ref{fj_fig1}}). 
383: The relative energy of the intermadiate state with respect to the initial state is 
384: $\epsilon_{n}^{g} +
385: \epsilon_{m}^{e} +\tilde{\delta}'$, which is precisely the first denominator in 
386: (\ref{doibre}). For $I_{e'J}$, the initial 
387: energy of the pair is again $\left(\mu_{g} - \frac{\delta}{2}\right) +
388: \left(\mu_{g}-\frac{\delta '}{2}\right)$, but the intermediate state
389: has an energy $\left[\left(\mu_{g}-\frac{\delta '}{2}\right) +
390: \epsilon_{n}^{g}\right] +
391: \left[\left(\mu_{e} + \frac{\delta }{2}\right)
392: +\epsilon_{m}^{e}\right]$, 
393: or a relative energy
394: $\epsilon_{n}^{g} +
395: \epsilon_{m}^{e} +\tilde{\delta}$. In summary, the intermediate states
396: of the transfer processes for ``spin up'' and ``spin down'' atoms 
397: have different energies and this results in different amplitudes for $I_{eJ}$ and $I_{e'J}$.  
398: 
399: The asymmetry in the currents implies the
400: existence of excitations in the superfluids. 
401: We have analyzed the many-body wavefunction of the system in the
402: Schr\"odinger picture and indeed it contains excitations corresponding
403: to the asymmetry. Specifically, the analysis confirms that 
404: the so called Fermi surface polarization
405: $(\langle N_e \rangle - \langle N_{e'} \rangle)/(\langle N_e \rangle + \langle N_{e'} \rangle)
406: \simeq (\langle N_e \rangle - \langle N_{e'} \rangle)/
407: (\langle N_e \rangle_0 + \langle N_{e'} \rangle_0)$ is non-zero
408: and oscillates as $f(\delta, \delta') \cos((\delta+\delta')t)$ where
409: $f(\delta, \delta) = 0$. We also found that {\it time-independent} perturbation
410: theory is not sufficient to reveal the asymmetry in the amplitudes: the
411: simple treatment of \cite{timeind} applied to our system results in symmetric
412: currents because the ansatz used does not allow any excitations.
413: This and the fact that the oscillation is most pronounced for timescales
414: of the order of the Cooper-pair correlation time indicates that the effect
415: is related to the dynamics of the superfluid state. 
416: 
417: To observe the Josephson effect one should be able to measure the number
418: of particles in two of the states, e.g.\ $|e \rangle$ and $|e' \rangle$,
419: at different stages of the oscillations, either destructively or
420: non-destructively. The scale of the gap energy is for typical systems
421: 1-100kHz, which means that the highest time resolution needed should be
422: just somewhat above $10 \mu s$. Measuring the number of particles
423: accurately is the more challenging part of the observation.  
424: In \cite{ours} we considered laser probing of the superfluid
425: Fermi-gas, where the laser was creating excitations in the BCS state.
426: The number of particles transferred was directly reflected in the
427: absorption of the light. Here one can use similar techniques to detect
428: the Josephson oscillations in a simple way.
429: %: the total light absorbtion 
430: %is directly proportional to the total current $I=I_e + I_{e'}$. 
431: %The currents $I_e$ and $I_{e'}$ can be obtained separately by 
432: %repeated measurements. 
433: 
434: 
435: 
436: In summary, we propose a method to realize Josephson oscillations in
437: superfluid atomic Fermi-gases. The coupling between two superfluids is
438: provided by laser light, and the laser detuning plays the same role as
439: voltage over metallic superconductor junctions. Detunings that affect
440: the two atomic internal states involved in pairing can be chosen to be
441: different -- this would correspond to different voltage for spin-up and
442: spin-down electrons. This leads to asymmetry in the oscillation
443: amplitudes of the two states. The asymmetry is pronounced when the
444: time-scale of the oscillation is the same order of magnitude as the
445: Cooper-pair correlation time. 
446: This is an effect unique to atomic Fermi-gases in the superfluid state.
447: 
448: 
449: \begin{acknowledgments} 
450: We thank the Academy of Finland for support
451: (projects 42588, 48845, 47140 and 44897). Gh.-S. P. also acknowledges
452: the grant NSF DMR 99-86199. We wish to thank Prof. G. Baym and Prof. A.
453: J. Leggett for useful discussions.
454: \end{acknowledgments}
455: 
456: \begin{thebibliography}{99}
457: 
458: \bibitem{Debbie}
459: B.\ DeMarco and D.S.\ Jin, Science \textbf{285}, 1703 (1999); 
460: M.J.\ Holland, B.\ DeMarco, and D.S.\ Jin, 
461: Phys.\ Rev.\ A \textbf{61}, 053610 (2000); B.\ DeMarco, S.B.\ Papp,
462: and D.S.\ Jin, Phys.\ Rev.\ Lett.\ {\bf 86}, 5409 (2001).
463: 
464: \bibitem{Hulet}
465: A.G. Truscott, K. E. Strecker, W. I. McAlexander,
466: G. P. Partridge, and R. G. Hulet, Science \textbf{291}, 2570 (2001).
467: 
468: \bibitem{Salomon}
469: M.O.\ Mewes, G.\ Ferrari, F.\ Schreck, A.\ Sinatra, and C.\ Salomon, Phys.\ 
470: Rev.\ A \textbf{61}, 011403 (R) (2000); F.\ Schreck, L.\ Khaykovich, K.L.\ 
471: Corwin, G.\ Ferrari, T.\ Bourdel, J.\ Cubizolles, and C.\ Salomon,
472: Phys.\ Rev.\ Lett.\ {\bf 87}, 080403 (2001).
473: 
474: \bibitem{Thomas}
475: K.M.\ O'Hara, M.E.\ Gehm, S.R.\ Granade, S.\ Bali, and J.E.\ Thomas, 
476: Phys.\ Rev.\ Lett.\ \textbf{85}, 2092 (2000);
477: S.R.\ Granade, M.E.\ Gehm, K.M.\ O'Hara, and J.E.\ Thomas,
478: Phys.\ Rev.\ Lett.\ {\bf 88},120405 (2002). 
479: 
480: \bibitem{KetterleF}
481: Z.\ Hadzibabic, C.A.\ Stan, K.\ Dieckmann, S.\ Gupta, M.W.\ Zwierlein,
482: A.\ G\"orlitz, and W.\ Ketterle, Phys.\ Rev.\ Lett.\ \textbf{88}, 160401 (2002).
483: 
484: \bibitem{Stoof}
485: H.T.C.\ Stoof,  M.\ Houbiers, C.A.\ Sackett, and R.G.\ Hulet, 
486: Phys.\ Rev.\ Lett.\ \textbf{76}, 10 (1996); 
487: M.\ Houbiers, R.\ Ferwerda, H.T.C.\ Stoof, W.I.\  McAlexander, C.A.\ 
488: Sackett, and R.G.\ Hulet, Phys.\ Rev.\ A \textbf{56}, 4864, (1997); 
489: R.\ Combescot, Phys.\ Rev.\
490: Lett.\ {\bf 83}, 3766 (1999). 
491: 
492: \bibitem{Holland}
493: M.\ Holland, S.J.J.M.F.\ Kokkelmans, M.L.\ Chiofalo, 
494: and R.\ Walser, Phys.\ Rev.\ Lett.\ {\bf 87}, 120406 (2001);
495: J.N.\ Milstein, S.J.J.M.F.\ Kokkelmans, and M.J.\ Holland, cond-mat/0204334
496: (2002). 
497: 
498: \bibitem{Zoller}
499: W.\ Hofstetter, J.I.\ Cirac, P.\ Zoller, E.\ Demler, and M.D.\ Lukin,
500: cond-mat/0204237 (2002).
501: 
502: \bibitem{Pethick}
503: H.\ Heiselberg, C.J.\ Pethick, H.\ Smith, and L.\ Viverit, Phys.\ Rev.\
504: Lett.\ \textbf{85}, 2418 (2000). 
505: 
506: \bibitem{alli}
507: W.\ Zhang, C.A.\ Sackett, and R.G.\ Hulet, Phys.\ Rev.\ A \textbf{60}, 504 (1999); 
508: J.\ Ruostekoski, Phys.\ Rev.\ A \textbf{60}, R1775 (1999); 
509: F.\ Weig and W.\ Zwerger, Europhys.\ Lett.\ \textbf{49}, 282 (2000); M.A.\ Baranov
510:  and D.S.\ Petrov, 
511: Phys.\ Rev.\ A \textbf{62}, 041601(R) (2000); M.\ Farine, P.\ Schuck,
512: and X.\ Vi\~nas,
513:  Phys.\ Rev.\ A \textbf{62}, 013608 (2000); G.M.\ Bruun and C.W.\
514:  Clark, J.\ Phys.\ B \textbf{33}, 3953 (2000).
515: 
516: \bibitem{ours}
517: P.\ T\"orm\"a and P.\ Zoller, Phys.\ Rev.\ 
518: Lett.\ \textbf{85}, 487 (2000); G.M.\ Bruun, P.\ T\"orm\"a,
519: M.\ Rodriguez, and P.\ Zoller,,
520: Phys.\ Rev.\ A \textbf{64}, 033609 (2001); Gh.-S.\ Paraoanu, M.\ Rodriguez, 
521: and P. T\"orm\"a, J.\ Phys.\ B: At.\ Mol.\
522: Opt.\ Phys.\ \textbf{34}, 4763 (2001). 
523: 
524: \bibitem{bcsbec}
525: See M.\ Randeria and references therein in {\it Bose-Einstein Condensation},
526: eds.\ A.\ Griffin, D.W.\ Snoke, and S.\ Stringari (Cambridge Un.\ Press,
527: Cambridge, 1995).
528: 
529: \bibitem{feshbach} E.\ Tiesinga, B.\ J.\ Verhaar, and H.\ T.\ C.\ Stoof, 
530: Phys. Rev. 
531: A \textbf{47}, 4114 (1993); S.\ Inouye, M.R.\ Andrews, J. Stenger, H.-J.\ Miesner,
532: D.M.\ Stamper-Kurn, and W.\ Ketterle, Nature {\bf 392}, 
533: 151 (1998); J.L.\ Roberts, N.R.\ Claussen, S.L.\ Cornish, E.A.\ Donley, E.A.\ Cornell,
534: and C.E.\ Wieman, Phys.\ Rev.\ Lett.\ \textbf{86}, 4211; 
535: E. Timmermans, K.\ Furuya, P.W.\ Milonni, and A.K.\ Kerman, Phys. Lett. A {\bf 285}, 228 
536: (2001).  
537: 
538: \bibitem{vortex}
539: M.\ Rodriguez, G.-S.\ Paraoanu, and P.\ T\"orm\"a, Phys.\ Rev.\ Lett.\ {\bf 
540: 87}, 100402 (2001).
541: 
542: \bibitem{internal} see
543: A.\ J.\ Leggett, Rev. Mod. Phys. {\bf 73}, 307 (2001) and references 
544: therein.
545: 
546: \bibitem{leggett} A.\ J.\ Leggett, Rev. Mod. Phys.{\bf 47}, 331 (1975);
547: J.\ C.\ Weatley, Rev. Mod. Phys. {\bf 47}, 415 (1975). 
548: 
549: \bibitem{ferro}
550: T.\ Kontos, M.\ Aprili, J.\ Lesueur, F.\ Genet, B.\ Stephanidis, and R.\ Boursier,
551: cond-mat/0201104 (2002); F.S.\ Bergeret, A.F.\ Volkov, and K.B.\ Efetov, 
552: Phys.\ Rev.\ Lett.\ \textbf{86}, 3140 (2001); E.A.\ Demler, G.B.\ Arnold, and M.R.\ Beasley,
553: Phys.\ Rev.\ B \textbf{55}, 15174 (1997).
554: 
555: \bibitem{boogie}
556: P.\ de Gennes, {\it Superconductivity of metals and alloys} (Addison- Wesley,
557: New York, 1966).
558: 
559: \bibitem{timeind}
560: R.A.\ Ferrell and R.E.\ Prange, Phys.\ Rev.\ Lett.\ \textbf{10}, 479 (1963).
561: 
562: \end{thebibliography}
563: 
564: \end{document}
565: 
566: 
567: 
568: 
569: 
570: 
571: