cond-mat0610749/njp.tex
1:                                                
2: \documentclass[12pt,a4paper]{iopart}
3: \usepackage{amssymb,graphicx,cite}
4: \eqnobysec
5: \begin{document}                    
6: 
7:  
8: %LATEX FILE OF MANUSCRIPT
9: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
10:  
11:  
12:  
13: %LATEX file of the manuscript
14:  
15: %\documentclass[preprint,eqsecnum,aps,epsf]{revtex4} % PH. REV.
16: %\documentclass[eqsecnum,aps,twocolumn,epsf]{revtex4} % PH. REV.
17: %\documentclass[aps,twocolumn,epsf,showpacs,floatfix]{revtex4} % PH. REV.
18: %\usepackage{graphicx}
19: %\documentclass[eqsecnum,aps,twocolumn]{revtex4} % PH. REV.
20:  
21: %\renewcommand{\baselinestretch}{.89}
22:  
23:  
24: %\documentstyle[aps,epsf]{revtex}
25: %\documentstyle[eqsecnum,aps,epsf]{revtex}
26: %%% <<< epsf commands in the next two lines >>>
27: %\newcommand{\postscript}[2] {\setlength{\epsfxsize}{#2\hsize}
28: %\centerline{\epsfbox{#1}}}
29:  
30:  
31:  
32: %\documentstyle[eqsecnum,aps,epsf]{revtex} % PH. REV. FINAL FORMAT STYLE
33: %\documentstyle[aps,epsf]{revtex} % PH. REV. FINAL FORMAT STYLE
34: %%% <<< epsf commands in the next two lines >>>
35: %\newcommand{\postscript}[2] {\setlength{\epsfxsize}{#2\hsize}
36: %\centerline{\epsfbox{#1}}}
37:  
38: %\documentstyle[preprint,aps]{revtex}
39: %\documentstyle[eqsecnum,aps]{revtex}
40: %\documentstyle[aps]{revtex}
41: %\renewcommand{\baselinestretch}{.945}
42:  
43: %\begin{document}
44:    
45: %\twocolumn[\hsize\textwidth\columnwidth\hsize\csname@twocolumnfalse\endcsname
46:  
47: %\title[Jet formation in a collapsing Bose-Einstein condensate]{Mean-field
48: %model of jet formation in a collapsing Bose-Einstein condensate}
49: 
50: \title[Collapse in a fermion-fermion mixture]{Dynamical collapse 
51: in  a degenerate binary fermion 
52: mixture using a hydrodynamic model}
53:  
54: \author{Sadhan K. Adhikari\footnote{Electronic
55: address: adhikari@ift.unesp.br; \\
56: URL: http://www.ift.unesp.br/users/adhikari/}}
57: %\affiliation
58: \address
59: {Instituto de F\'{\i}sica Te\'orica, UNESP $-$ S\~ao Paulo State
60: University,
61: 01.405-900 S\~ao Paulo, S\~ao Paulo, Brazil\\}
62:  
63:  
64: \date{\today}
65:  
66: %\maketitle
67:  
68: \begin{abstract}
69: 
70:  
71: We  use a time-dependent dynamical hydrodynamic
72: model to study
73: a collapse  in a degenerate fermion-fermion       
74: mixture (DFFM) of different atoms. Due to a strong Pauli-blocking
75: repulsion among
76: identical spin-polarized fermions at short distances there cannot be
77: a collapse  for  repulsive interspecies
78: fermion-fermion interaction.
79: However, 
80: there can be
81: a collapse  for a sufficiently attractive  interspecies
82: fermion-fermion interaction in a DFFM  of different 
83: atoms. Using a variational analysis and numerical solution of the 
84: hydrodynamic model we study
85: different aspects of   collapse in such a DFFM
86: initiated by a jump in the interspecies
87: fermion-fermion interaction (scattering length)
88: to a large negative (attractive) value  using a
89: Feshbach resonance. Suggestion for experiments of collapse in a DFFM
90:  of distinct atoms 
91: is
92: made.
93: 
94:  
95: \end{abstract}
96: 
97: \pacs{03.75.Ss} 
98: %\pacs{03.75.-b, 03.75.Nt}
99: 
100: \maketitle
101:  
102:  
103: 
104: 
105:  \section{Introduction}
106:  
107: Recent successful observation of degenerate  boson-fermion  mixture 
108: (DBFM) and 
109: fermion-fermion  mixture (DFFM) of
110: trapped alkali-metal atoms by different experimental groups
111: \cite{exp1,exp2,exp21,exp211,exp22,exp3,exp4} has initiated the 
112: intensive 
113: experimental
114: studies of different novel phenomena \cite{exp6,exp5,exp5x}.  
115: It has been possible to achieve a degenerate Fermi gas  (DFG) by 
116: sympathetic cooling in the presence of a second boson or fermion 
117: component, as there cannot be an effective evaporative cooling \cite{exp1} 
118: of a 
119: single-component DFG  due to a strong Pauli-blocking
120: repulsion at low 
121: temperature  among spin-polarized fermions.  
122: Among these
123: experiments on a DFG, apart from the study of a DBFM in
124: $^{6,7}$Li \cite{exp3}, $^{23}$Na-$^6$Li
125: \cite{exp4} and $^{87}$Rb-$^{40}$K 
126: \cite{exp5,exp5x,bongs,mur},  
127: there have been studies of a DFFM 
128: in
129: $^{40}$K-$^{40}$K$^*$ \cite{exp1} and $^6$Li-$^6$Li$^*$ 
130: \cite{exp2,exp21,exp211,exp22} 
131: systems, where $^*$ denotes a distinct  hyperfine state. More 
132: recently the
133: formation
134: of a Bardeen-Cooper-Schreiffer (BCS) condensate of fermionic $^6$Li
135: atoms in  $^{23}$Na-$^6$Li \cite{bcs}
136:  and $^6$Li-$^6$Li$^*$ \cite{exp211}
137: mixtures 
138: has been observed experimentally.
139:  The collapse in a DBFM  of 
140: $^{87}$Rb-$^{40}$K atoms has been observed and studied by Modugno {\it et
141: al.}
142: \cite{exp5} and more recently by Ospelkaus  {\it et
143: al.}
144: \cite{bongs}. 
145: Recently, experiments on controlled collapse on $^{87}$Rb-$^{40}$K
146: have been accomplished \cite{ccol}.
147:  
148: A collapse in a Bose-Einstein condensate
149: (BEC) takes place
150: due to an attractive atomic
151: interaction \cite{hulet,don}.  
152: A study of controlled collapse   and explosion
153: has been performed by Donley {\it et al.}
154: \cite{don} on an attractive  $^{85}$Rb BEC, where they
155: manipulated the inter-atomic interaction by varying
156: a background magnetic
157: field exploiting  a nearby Feshbach resonance 
158: \cite{fs}. 
159: There
160: have been many  theoretical \cite{th1,th2} studies to describe different
161: features of this experiment \cite{don}. 
162: More recently, there have been experimental studies on collapse in a DBFM
163: of $^{87}$Rb-$^{40}$K by two different groups
164: \cite{exp5,bongs,ccol} as 
165: well as related theoretical investigations \cite{ska,zzz}.
166: As the interaction in a pure DFG  at 
167: short
168: distances is
169: repulsive due to Pauli-blocking, there cannot be a
170: collapse in
171: it. The Pauli repulsion is responsible for the stability of a
172: neutron star
173: against a (gravitational) collapse. 
174: A collapse is possible  in a DBFM
175: in the presence of a sufficiently strong
176: boson-fermion attraction which can overcome 
177: the Pauli repulsion among
178: identical fermions \cite{exp5,bongs,ska}.
179: 
180: 
181: In this paper we study the  collapse in a DFFM
182: for a sufficiently attractive interspecies fermion-fermion interaction
183: which can overcome the Pauli repulsion.  However, there is already 
184: experimental
185: evidence and theoretical conjecture that a Fermi gas in two (spin) 
186: hyperfine
187: states of the same atom is much more stable \cite{gehm} than expected on
188: the basis of a scattering length approach \cite{exp22,phase} and there
189: is no collapse in such a system. A similar conclusion follows from an
190: examination of the compressibility of such a system \cite{kok}.  A
191: strongly attractive DFFM in two (spin)
192: hyperfine
193: states
194: exhibits universal behavior and should be mechanically stable as a
195: consequence of the quantum-mechanical requirement of unitarity. This
196: requirement limits the maximum attractive force for such a  DFFM
197: to a value smaller than that of the outward Fermi pressure due to Pauli
198: repulsion \cite{gehm}.  It has been demonstrated that a two-component
199: DFFM in different (spin) hyperfine states is stable against collapse
200: \cite{hei}, whereas a multicomponent degenerate fermion mixture
201: \cite{gehm,hei} or a DFFM of different atoms \cite{hulet1} could undergo
202: collapse. Hence, by taking a DFFM of two different fermionic atoms of
203: different atomic mass one can avoid the
204: problem \cite{gehm,hei} of a possible suppression of collapse. Thus one
205: could study collapse in a DFFM in the same manner as in a DBFM. The
206: second component of fermions will
207: then only aid in inducing an attraction among the fermions of the first
208: component responsible for collapse without suppressing the collapse.
209: Although, the past experiment \cite{gehm} on two-component cold Fermi 
210: gas
211: used two (spin) hyperfine states of the same atom, experiments can be 
212: realized
213: with distinct atoms and one can look for collapse in such a system. One 
214: such system is the $^6$Li-$^{40}$K mixture: both $^6$Li \cite{exp211} 
215: and 
216: $^{40}$K \cite{exp1}
217: have been trapped and studied in laboratory. 
218: 
219: 
220: 
221: Here we use a coupled time-dependent mean-field hydrodynamic model which 
222: is
223: inspired by the success of a similar model used by the present author in
224: the investigation of a fermionic collapse \cite{ska} and bright
225: \cite{fbs2} and dark \cite{fds} solitons in a DBFM as well as of
226: mixing-demixing \cite{md} and black solitons \cite{bs} in a DFFM. The
227: conclusions of the study on bright soliton \cite{fbs2} are in agreement
228: with a microscopic study \cite{bong1} and the noted survival of collapse
229: in the numerical study \cite{ska} has been experimentally substantiated
230: later in a DBFM of $^{40}$K-$^{87}$Rb \cite{bongs}.  A very similar
231: model has been used by Jezek {\it et al.} \cite{Jezek} in a successful
232: description of vortex states in a DBFM. Although, a mean-field model is 
233: simple to use and leads to a proper prediction of probability density of 
234: the fermionic system, many true quantum effects are lost in this simplified 
235: model, e. g., it cannot predict the suppression of collapse of a DFFM in 
236: two different (spin) hyperfine states \cite{gehm,hei} as discussed in 
237: the last 
238: paragraph. We recall  many true 
239: quantum effects are also 
240: lost \cite{11}  in the mean-field Gross-Pitaevskii equation for trapped 
241: bosons.
242: 
243: 
244: In
245: our study on collapse in a DFFM we shall consider a strong attraction
246: among fermions which will naturally lead to molecule formation and not
247: to a BCS state. A BCS state is usually formed for a weak attraction
248: among identical fermions. The possibility of molecule formation by 
249: three-body recombination 
250: is  explicitly included in our model by an absorptive nonlinear term. 
251: Apart from the direct experimental interest in
252: trapped cold atoms, a study of strongly interacting Fermi gases and
253: their possible collapse is also relevant \cite{gehm} in condensed matter
254: physics (superconductivity), nuclear physics (nuclear matter), high
255: energy physics (effective theories of strong interaction), and
256: astrophysics (compact stellar objects), which makes the present study of
257: greater interest.
258: 
259: 
260: 
261: 
262: 
263: 
264: 
265: The collapse in a  DFFM
266: is first studied  by a 
267: variational analysis of the present model, which is later substantiated by 
268: a complete numerical solution using the Crank-Nicholson scheme \cite{sk1}. 
269: During a collapse and an explosion of the  
270: DFFM, the system loses atoms as 
271: in a collapsing and exploding BEC \cite{don}. The loss of atoms is 
272: accounted for by three-body 
273: recombination involving two types of fermions. We study the 
274: sensitivity 
275: of our results on the three-body recombination
276: loss rates. We also study the quasi-periodic 
277: oscillation of the sizes of the  
278: DFFM
279: undergoing a
280: collapse. The collapse 
281: is to be
282: initiated by jumping the interspecies scattering length to a large 
283: negative (attractive) value near a fermion-fermion Feshbach resonance 
284: \cite{fsff}. However, the collapse starts after a time delay upon this 
285: jump and we study the variation of this time delay with the final 
286: scattering length. This variation has a behavior similar to that observed 
287: in the bosonic case \cite{don}. The collapsing  
288: DFFM 
289: is found to execute a 
290: quasi-periodic oscillation with a frequency approximately equal to twice 
291: the frequency 
292: of the harmonic trap as in a BEC \cite{don}. 
293: 
294: 
295: 
296: 
297: 
298: 
299: Previously, in addition to the study of a collapse in a pure 
300: BEC \cite{th1,th2},
301: we also investigated \cite{adhi} the collapse   in a
302: two-species
303: BEC initiated by 
304: an interspecies attraction. 
305: The predicted collapse in a two-species  BEC for intra-species
306: repulsion  and interspecies attraction \cite{adhi}
307: is similar to that in a  DBFM
308: of $^{87}$Rb-$^{40}$K
309: studied before 
310: \cite{exp5,bongs,ska} and 
311: that in a  DFFM
312: studied in
313: this paper. 
314: 
315: In section 2 we present the coupled hydrodynamic model for a  DFFM
316:  which we 
317: apply to predict and study a collapse. In this
318: section we also present a variational analysis based on this model
319: which substantiates the
320: collapse in a  DFFM
321: for a sufficiently attractive
322: interspecies
323: fermion-fermion attraction.  In section 3 
324: we 
325: present results of numerical simulation of  our study on collapse. 
326: Finally, in section 4  we present a brief summary  of our 
327: investigation. 
328: 
329:  \section{Coupled Hydrodynamic Model for a Fermion-Fermion Mixture}
330: 
331: \subsection{Model}
332: 
333: A mean-field-hydrodynamic Lagrangian for
334: a DFG   has been used successfully in the study of a
335: DBFM
336: \cite{fbs2,fds,Jezek}
337: which we shall use in the present investigation. The virtue of the
338: hydrodynamic  model for a DFG  over a  microscopic
339: description is its 
340: simplicity and
341: good predictive power.
342: To  develop a set of practical time-dependent hydrodynamic
343: equations for a  DFFM, we consider   the
344: following Lagrangian density \cite{fbs2,fds}
345: \begin{eqnarray}\label{yy}
346: &{\cal L}& = \frac{i}{2}\hbar \sum_{j=1,2}\left(
347: \Psi_j \frac{\partial \Psi_j \- ^*}{\partial t} - \Psi_j \- ^  
348: *
349: \frac{\partial \Psi_j
350: }{\partial
351: t} 
352: \right) \nonumber \\ 
353: &+& \sum_{j=1}^2 \left(\frac{\hbar^2 |\nabla
354:   \Psi_j |^2 }{6m_j}+
355: V_j n_j+\frac{3}{5}  A_j  |n_j |^{5/3}\right)\nonumber \\
356: &+& g_{12}  n_1 n_2-i\hbar \left(K_{31}n_1 n_2^2  
357: +K_{32} n_1^2 n_2  \right),  
358: \end{eqnarray}
359: where  $m_j$ is the
360: mass of component $j (=1,2)$, 
361: $A_j=\hbar^2(6\pi^2)^{2/3}/(2m_j)$, 
362: $\Psi_j$ is a complex probability amplitude,  $n_j=|\Psi_j|^2$ is a
363: real probability density  and
364: $N_j \equiv  \int d{\bf r} n_j({\bf r}) $  the number.  
365: Here the  interspecies
366: coupling is 
367: $g_{12}=2\pi \hbar^2 a_{12}
368: /m_R$
369: with the
370: reduced mass $m_R=m_1m_2/(m_1+m_2),$ and  $ a_{12}$
371: is the interspecies
372: scattering length.
373: The spherically-symmetric potential is taken as  $
374: V_{j}({\bf
375: r})=\frac{1}{2}(3m_j) \omega ^2 r^2$ where
376:  $\omega$ is  the radial ($r$) frequency.
377: The interaction between intra-species fermions in
378: spin-polarized state is highly suppressed due
379: to Pauli blocking
380: and has been neglected in  (\ref{yy})  and will be
381: neglected throughout.
382: The kinetic energy terms in this equation
383: are derived from a hydrodynamic equation for the
384: fermions
385: \cite{capu}. The
386: kinetic energy terms contribute little to this problem compared to the
387: dominating Pauli blocking term 
388: $3A_j|n_j|^{5/3}/5$  in (\ref{yy}).
389: However, its inclusion leads
390: to a smooth  solution for the probability density everywhere
391: \cite{fbs2}. The Lagrangian density of each fermion component in
392:  (\ref{yy}) is
393: identical to that used in Refs.  \cite{fbs2,fds}. The last two terms in
394:  (\ref{yy}) correspond to three-body recombination  due to the
395: following reactions, respectively
396: $F_1+F_2+F_2  \to  (F_1F_2)+F_2,$ and 
397: $F_1+F_2+F_1  \to  (F_1F_2)+F_1,$
398: where $(F_1F_2)$ is a composite structure (resonance/molecule) of
399: fermions $F_1$ and $F_2$ and $K_{31}$ and $K_{32}$ are the corresponding
400: three-body
401: loss rates.  The contribution to the Lagrangian density of the
402: recombination reactions 
403: is proportional to
404: the density of the participating fermions. Here we neglected
405: two-body loss.
406: 
407: 
408: The  dynamical equations for the system are just the 
409: usual  Euler-Lagrange  (EL) equations with the Lagrangian density 
410: (\ref{yy}) \cite{gold}
411: \begin{equation}
412: \frac{\partial}{\partial t}\frac{\partial {\cal L}}{\partial
413: \frac{\partial
414: \Psi_j\- ^*}{\partial t}}+
415: \sum _{k=1}^3 \frac{d}{dx_k}\frac{\partial {\cal L}}{\partial
416: \frac{\partial \Psi_j\- ^*}{\partial x_k}}= \frac{\partial {\cal
417: L}}{\partial
418:  \Psi_j\- ^*},
419: \end{equation}
420: where $x_k, k=1,2,3$ are the three space components, and 
421: $j=1,2$ refer to the fermion components. 
422: Consequently,   the following  EL   equations of 
423: motion are
424: derived:
425: \begin{eqnarray}\label{e} \biggr[ &-& i\hbar\frac{\partial
426: }{\partial t}
427: -\frac{\hbar^2\nabla_{\bf r}^2}{6m_{{1}}}
428: + V_{{1}}({\bf r})+A_1|n_1|^{2/3} 
429: + g_{{12}} n_2
430: \nonumber \\
431: &-&i\hbar \left( K_{31}n_2^2+2K_{32}n_1  n_2 \right)
432:  \biggr]\Psi_1=0. 
433: \end{eqnarray}
434: \begin{eqnarray}\label{f} \biggr[& -& i\hbar\frac{\partial
435: }{\partial t}
436: -\frac{\hbar^2\nabla_{\bf r}^2}{6m_{{2}}}
437: + V_{{2}}({\bf r}) 
438: + A_2 |n_2|^{2/3} 
439: + g_{{12}} n_1
440: \nonumber \\
441: &-& i\hbar \left( 2K_{31}n_1 n_2 + K_{32} n_1 ^2  \right)
442:  \biggr]\Psi_2 =0. 
443: \end{eqnarray}
444: 
445: 
446: 
447: 
448: 
449:     
450:  
451: In the spherically-symmetric state the fermion density  
452:  has the form $\Psi_j({\bf
453: r};t)=\psi_j(r;t).$ 
454: Now  transforming to
455: dimensionless variables
456: defined by $x =\sqrt 2 r/l$,     $\tau=t \omega, $
457: $l\equiv \sqrt {\hbar/(m\omega)}$, $m=3m_1=3m_2$
458: and
459: \begin{equation}\label{wf}
460: \frac{ \phi_j(x;\tau)}{x} =  
461: \sqrt{\frac{4 \pi l^3}{N_j\sqrt 8}}\psi_j(r;t),
462: \end{equation}
463: we obtain from  (\ref{e}) and (\ref{f})
464: \begin{eqnarray}\label{d1}
465: &\biggr[&-i\frac{\partial
466: }{\partial \tau} -\frac{\partial^2}{\partial
467: x^2} 
468: +\frac{x^2}{4} 
469: %+3\left(\frac{3\pi N_1}{2} \right)^{2/3}
470: + {\cal N}_{11}
471: \left|\frac {\phi_1}{x}\right|^{4/3}
472: \nonumber \\          
473: &+& 6 \sqrt 2    {\cal N}_{12}\left
474: |\frac {\phi_2}{x}\right|^2 
475: - i2\xi_{32}  {\cal N}_{12}  {\cal
476: N}_{21} \left|\frac
477: {\phi_1}{x}\right|^2  \left|\frac
478: {\phi_2}{x}\right|^2
479:  \nonumber \\  &-& 
480: i\xi_{31} {\cal N}_{12} ^2  \left|\frac
481: {\phi_2}{x}\right|^4
482: \biggr]\phi_1({ x};\tau)=0,
483: \end{eqnarray}
484: \begin{eqnarray}\label{d2}
485: &\biggr[&-i\frac{\partial
486: }{\partial \tau} -\frac{\partial^2}{\partial
487: x^2} 
488: +\frac{x^2}{4} 
489: %+3\left(\frac{3\pi N_2}{2} \right)^{2/3}
490: + {\cal N}_{22}
491: \left|\frac {\phi_2}{x}\right|^{4/3}
492: \nonumber \\   
493: &+& 6 \sqrt 2    {\cal N}_{21}\left
494: |\frac {\phi_1}{x}\right|^2
495: - i2\xi_{31}  {\cal N}_{12}  {\cal
496: N}_{21} \left|\frac
497: {\phi_1}{x}\right|^2  \left|\frac
498: {\phi_2}{x}\right|^2
499:  \nonumber \\  &-&
500: i\xi_{32} {\cal N}_{21} ^2  \left|\frac
501: {\phi_1}{x}\right|^4    
502:  \biggr] \phi_2({ x};\tau)=0,
503: \end{eqnarray}
504: where
505: $ {\cal N}_{jj}=3(3\pi N_j/2)^{2/3}$, 
506: $ {\cal N}_{12} =   N_2 a_{12} /l,$ 
507: $ {\cal N}_{21} =   N_1 a_{12} /l,$
508: $\xi_{32}=K_{32}/(2\pi^2 a_{12}^2l^4\omega),$ and
509: $\xi_{31}=K_{31}/(2\pi^2 a_{12}^2l^4\omega).$ In the non-absorptive 
510: case without any loss of atoms due to three-body recombination
511: $\xi_{31}=\xi_{32}=0$, the normalization of the wave-function components
512: is
513: given by $\int_0^\infty dx |\phi_j(x;\tau)|^2 =1, j=1,2.$ In the
514: absorptive case $\xi_{31}\ne 0$ and $\xi_{32} \ne 0$ it is possible to
515: have loss  of atoms due to three-body recombination and  the
516: normalization reduces with time due to loss of atoms. 
517: 
518: We solve the coupled hydrodynamic equations (\ref{d1}) and (\ref{d2})
519: numerically using a time-iteration method based on the Crank-Nicholson
520: discretization scheme elaborated in Refs. \cite{sk1,sk2}.  We discretize
521: the
522: hydrodynamic equations using time step $0.00025$ and space step $0.025$
523:  spanning $x$ from 0 to 25. This domain of space was sufficient to
524: encompass the entire fermion function during a collapse and explosion
525: and obtain convergent solution for the total number of fermions.  First
526: we solve (\ref{d1}) and (\ref{d2}) with $\xi_{31}=\xi_{32}=0$ to find an 
527: initial
528: stationary state of the DFFM.  It is true that the three-body loss, 
529: taken care of by  terms 
530: $\xi_{31}$ and $\xi_{32}$, is always present in the system, its effect 
531: is small leading to at best a small loss rate in atoms except for very 
532: large negative values of $a_{12}$. Hence,  for the consideration of the 
533: initial 
534: state, we could as well neglect  three-body loss. (This is why a 
535: Gross-Pitaevskii mean-field equation without three-body loss has been 
536: successfully used to study many features of a repulsive trapped 
537: BEC \cite{11}.)
538: However, in the final 
539: state, when $a_{12}$ is suddenly turned to a large negative value by a 
540: Feshbach resonance, the three-body loss has a dramatic effect 
541: responsible for 
542: a proper description of a 
543: collapse and explosion with  
544: a very rapid loss of atoms in a short interval of time (see figure 3).
545: 
546: 
547: In our numerical investigation we take
548: $l=1$ $\mu$m and consider the equal-mass fermions with the mass of
549: $^{40}$K corresponding to a radial frequency $\omega \approx 2\pi \times
550: 83$ Hz. The present simplified mean-field model cannot predict the
551: suppression of collapse \cite{gehm,hei} of a DFFM in two hyperfine
552: states which is a true quantum many-body effect. Nevertheless, it leads
553: to a proper description of collapse dynamics of a DFFM of two distinct
554: atoms. The use of equal-mass fermions only keeps the algebra simple 
555: specially in section 2.2, but
556: otherwise has no effect on the general qualitative dynamics studied in 
557: this paper.  
558: In this study, the unit of time is
559: $1/\omega \approx 2$ ms, and unit of length $l/\sqrt 2 \approx 0.7$
560: $\mu$m. 
561: Actually, any two different fermionic atoms can be used in
562: experiment, a proper quantitative treatment of which will require the 
563: use of 
564: different mass factors in the dynamical equations.
565: 
566:  
567: 
568: 
569: \subsection{Variational Analysis}  
570: 
571: To understand how the stationary states of a DFFM
572: are formed, we employ
573: a variational method for the solution of  (\ref{d1}) and (\ref{d2}) in
574: the symmetric case $N_1=N_2\equiv N$, while
575: $\phi_1/x=\phi_2/x\equiv \varphi$ satisfies 
576: \begin{eqnarray}\label{v1}
577: \biggr[-i\frac{\partial
578: }{\partial \tau} -\frac{\partial^2}{\partial
579: x^2}-\frac{2}{x}\frac{\partial}{\partial
580: x}
581: +\frac{x^2}{4}
582: + \mu
583: \left| {\varphi}\right|^{4/3}
584: + g \left
585: | \varphi\right|^2\biggr] \varphi=0,
586:  \end{eqnarray}
587: where $g= 6 \sqrt 2 N a_{12}/l$ and $\mu = 3(3\pi N/2)^{2/3}$
588: \cite{varia}. Here
589: we have
590: set the absorptive terms to zero for stationary states.  
591: We consider  the following trial 
592: Gaussian
593: wave function for the solution of   (\ref{v1})
594: \cite{varia}
595: \begin{equation}\label{twf}
596: \varphi(x,t)=  A(t)\exp\left[-\frac{x^2}{2R^2(t)}
597: +\frac{i}{2}{ \beta(t) }x^2+i\alpha(t)
598: \right],
599: \end{equation}
600: where $A(t)$,  $R(t)$, $\beta(t)$, and $\alpha(t)$ are the
601: normalization, width, chirp, and
602: phase, respectively.  The normalization condition $ \int_0 ^\infty dx
603: x^2
604: \varphi^2(x,t) =1$ sets 
605: $A(t)=[\pi^{1/4}R^{3/2}(t)/2]^{-1}$.  The Lagrangian density for 
606: generating  
607: (\ref{v1})   is \cite{varia}
608: \begin{eqnarray}
609: {\cal L}(\varphi)&=&\frac{i}{2}\left(\dot \varphi  \varphi^*
610: - \dot  \varphi^*  \varphi
611: \right)-
612: \left|\frac{\partial
613:   \varphi}{\partial x} \right|^2  -
614: \frac{x^2}{4}| \varphi|^2\nonumber \\
615: &-& \frac{1}{2} g | \varphi|^4- \frac{3}{5} \mu |\varphi|^{10/3} ,
616: \end{eqnarray}
617: where  the
618: overhead dot represents time derivative.
619: The trial wave function (\ref{twf}) is
620: substituted in the Lagrangian density and the
621: effective Lagrangian $L_{\mbox{eff}}$
622: is calculated via  $L_{\mbox{eff}}= \int {\cal
623: L}(\varphi)
624: d ^3 x:$ 
625: \begin{eqnarray} 
626: L_{\mbox{eff}}=\frac{\pi^{3/2}A^2(t)R^5(t)}{2}\biggr[-\frac{3}{2}\dot
627: \beta(t)
628: -\frac{g}{2\sqrt 2} \frac{ A^2(t)}{R^2(t)}\nonumber
629: \\-\frac{9\sqrt 3}{25 \sqrt 5}\mu \frac{A^{4/3}(t)}{R^2(t)}
630: -\frac{2\dot \alpha(t)}{R^2(t)} -\frac{3}{R^4(t)}-3\beta^2(t) -
631: \frac{3}{4} \biggr].
632: \end{eqnarray}
633: 
634: The generalized Lagrange equations for this effective Lagrangian 
635: given by  \cite{gold}
636:  \begin{equation}
637: \frac{d}{d t}\frac{\partial L_{\mbox{eff}}}{\partial\dot \gamma(t)}=
638: \frac{\partial L_{\mbox{eff}}}{\partial \gamma(t)},
639: \end{equation}
640: with $\gamma(t)$ representing   $\alpha(t)$, $A(t), \beta(t),$ and
641: $R(t) $
642: are  written explicitly as 
643: \begin{equation}\label{e1}
644: \pi^{3/2}A^2R^3= \mbox{constant}=4\pi,
645: \end{equation}
646: \begin{eqnarray}\label{e2}
647: 3\dot \beta+\frac{4\dot
648: \alpha}{R^2}+\frac{6}{R^4}+6\beta^2+\frac{3}{2}=- 
649: \frac{\sqrt 2gA^2}{R^2}-\frac{6\sqrt 3}{5 \sqrt 5}\frac{\mu A^{4/3}}{R^2}
650: ,\nonumber \\ \end{eqnarray}
651: \begin{eqnarray}\label{e3}
652: \dot R= 2 R \beta,
653: \end{eqnarray}
654: \begin{eqnarray}\label{e4}
655: 5\dot \beta+\frac{4\dot
656: \alpha}{R^2}+\frac{2}{R^4}+10\beta^2+\frac{10}{4}=-
657: \frac{gA^2}{\sqrt
658: 2R^2}-\frac{18\sqrt 3}{25\sqrt 5}\frac{\mu A^{4/3}}{R^2},\nonumber \\
659: \end{eqnarray}
660: where the time dependence of different observable is suppressed.
661: Eliminating $\alpha$ between   (\ref{e2}) and (\ref{e4}) one obtains
662: \begin{eqnarray}\label{e5}
663: 2\dot \beta= \frac{4}{R^4}-4\beta^2+ \frac{gA^2}{\sqrt
664: 2R^2}+\frac{12\sqrt 3}{25\sqrt 5}\frac{\mu A^{4/3}}{R^2}
665: -1.
666: \end{eqnarray}
667: 
668: 
669: \begin{figure}%[!ht]
670:  
671: \begin{center}
672: \includegraphics[width=1.\linewidth]{fig1.ps}
673: %\postscript{fig1.eps}{1.}
674: \end{center}
675: 
676: \caption{ The effective potential $U(R)$ of  
677: (\ref{eff}) vs. $R$ for different $a_{12}$ and $N=1000$ and $l=1$ $\mu$m. 
678: }
679: 
680: \end{figure}
681: 
682: 
683:  From   (\ref{e3}) and  (\ref{e5}) we get the following
684: second-order
685: differential equation for the evolution of the width $R$
686: \begin{eqnarray}\label{el}\frac{d^2R}{dt^2}
687: =\frac{4}{R^3}+\frac{4 g}{\sqrt{2
688: \pi}R^4}+\frac{12\mu 4^{2/3}\sqrt 3}{25\pi^{1/3} \sqrt 
689: 5R^3}
690: -R,
691: \end{eqnarray}
692: \begin{eqnarray}
693: = -\frac{d}{dR}\left[\frac{2}{R^2}+ \frac{4g}{3\sqrt{2
694: \pi}}\frac{1}{R^3}+ \frac{6\mu  4^{2/3}\sqrt 3}{25\pi^{1/3} \sqrt 
695: 5R^2} 
696: +\frac{R^2}{2}
697: \right]. \label{el2}
698: \end{eqnarray}
699: The quantity in the square brackets of  (\ref{el2}) is the effective
700: potential  $U(R)$
701: of the  equation of motion:
702: \begin{equation}\label{eff}
703: U(R)=\frac{2}{R^2}+ \frac{4g}{3\sqrt{2
704: \pi}}\frac{1}{R^3}+ \frac{6\mu  4^{2/3}\sqrt 3}{25\pi^{1/3} \sqrt
705: 5R^2}
706: +\frac{R^2}{2}. 
707: \end{equation}
708: Small oscillation of a stationary state 
709: around a
710: stable configuration is  possible when there is a minimum in this
711: effective
712: potential determined by a zero of  (\ref{el}). This condition
713: yields the variational width from which the variational solution for
714: the wave function is obtained via  (\ref{twf}).
715: 
716: 
717:  
718: 
719: 
720: \begin{figure}%[!ht]
721:  
722: \begin{center}
723: \includegraphics[width=1.\linewidth]{fig2.ps}
724: %\postscript{fig1.eps}{1.}
725: \end{center}
726: 
727: \caption{The phase diagram for collapse where we plot
728: log$_{10}(N)$ vs. $a_{12}/l$ for fermion-fermion mixture (full
729: line) and bosons (dotted line). The line separates the regions where
730: collapse is present and  absent.}
731: 
732: \end{figure}
733: 
734: 
735: 
736: In figure  1 we plot the effective potential  $U(R)$ of 
737: (\ref{eff})  for   different $a_{12}$, 
738: $N=1000$ and $l=1$ $\mu$m.  For positive (repulsive) $a_{12}=100 $ nm,  
739: $U(R)$ 
740: leads to a confining well with a minimum at $R=R_0=5.3$, so that one 
741: could 
742: have a stable DFG  of width $R_0=5.3  $. The variational
743: profile for 
744: this 
745: function is 
746: \begin{equation} \label{var}
747: \varphi (x) = 
748: \frac{2}{\pi^{1/4}R_0^{3/2}}\exp\left[-\frac{x^2}{2R_0^2}\right].
749: \end{equation}
750: In  figure 1, as $a_{12}$ turns  gradually negative (attractive), the 
751: infinite wall near $R=0$ of  $U(R)$ is gradually lowered and for a 
752: sufficiently attractive scattering length $a_{12}\approx -100$ nm, 
753: this 
754: wall 
755: disappears
756: completely 
757: and one has the possibility of collapse for 
758:  $a_{12}<  -100$ nm. The minimum in $U(R)$ first becomes a
759: point 
760: of inflection for $a_{12}\approx -100$ nm and then disappears.  The
761: profile of the effective potential in  figure  1 is similar to the same in
762: the
763: bosonic case \cite{varia,11}.
764: 
765: 
766: Next we show  the phase diagram for collapse for $N_1=N_2=N$ using the
767: variational
768: approach in  figure  2, where  we plot log$_{10}(N)$ vs. $a_{12}/l$. 
769: The line
770: separates the plot in two regions. In the upper half of the plot collapse
771: is possible and in the lower half we have stable configurations.  The
772: phase diagram of  figure  2 is quite similar to that obtained in
773: Ref. \cite{phase} in a study of the stability of a DFFM.  In case of bosons in a spherically-symmetric
774: trap the line of stability is given by $Na/l=-0.575$ \cite{11} and is also
775: shown in
776:  figure  2. As expected, for a fixed $|a/l|$ a much larger number of 
777: fermions 
778: can be accommodated in a stable state.
779: 
780: 
781: 
782: 
783: \section{Numerical Results} 
784:  
785: 
786: 
787: 
788: In this section we present results on collapse from a numerical solution
789: of  (\ref{d1}) and (\ref{d2}).
790: After some experimentation 
791: we take in the initial  DFFM
792: $N_1=1000$, $N_2=2000$, and $a_{12}= 100$ nm, so that $a_{12}/l=0.1$. 
793: This
794: corresponds to nonlinearities ${\cal N}_{11}=843$, ${\cal
795: N}_{22}=1338$,
796: ${\cal N}_{12}= 200$ and ${\cal N}_{21}=
797: 100$. The collapse dynamics is sensitive to the loss rates
798: $K_{31}$ and $K_{32}$.    
799: As these loss rates 
800: are not experimentally  known,  
801: in  the present simulation we take them to be equal: $K \equiv 
802: K_{31}=K_{32}$,
803: and 
804: consider  different values of $K$.      
805: 
806:  
807: 
808:  
809: 
810: 
811: 
812:  
813: \begin{figure}%[!ht]
814:  
815: \begin{center}
816: \includegraphics[width=1.\linewidth]{fig3.ps}
817: %\postscript{fig1.eps}{1.}
818: \end{center}
819: 
820: \caption{ The evolution of  fermion numbers 
821: $N_j(t)$ of the two components  vs. time 
822:  during a collapse initiated by a jump in scattering
823: length $a_{12}$ from 100 nm to $-200$ nm  in a  DFFM
824: of $N_1=1000$ and
825: $N_2=2000$
826: for three-body loss rates 
827: $K=  10^{-26}$ cm$^6$s$^{-1}$,  $ 10^{-25}$
828: cm$^6$s$^{-1}$, and 
829: $  10^{-24}$ cm$^6$s$^{-1}.$ The dotted (blue) curves
830: refer to fermion 2
831: and the solid (red)  curves to 
832: fermion 1. The
833: curves are labeled by their respective $K$ values.  }
834: 
835: \end{figure}
836: 
837: 
838: 
839:  
840: 
841:  
842: 
843:  
844: Now we consider the collapse of fermions initiated by a sudden jump in the
845: fermion-fermion scattering length from $a_{12}=100$ nm to 
846: $-200$ nm
847: which
848: can be implemented near a fermion-fermion Feshbach resonance, 
849: observed in fermionic systems
850:  \cite{fsff}.
851: This
852: resonance should enable an experimental control of the interspecies
853: interaction \cite{fs} and hence can be used to increase the 
854: attractive
855: force between  interspecies   fermions by varying a background magnetic
856: field,  
857: which in turn increases the
858: attractive
859: nonlinearities $6\sqrt 2 {\cal N}_{12}$ and $6\sqrt 2 {\cal N}_{21}$ 
860: in
861:  (\ref{d1})  and (\ref{d2}). If these attractive nonlinear terms can
862: overcome 
863: the repulsive nonlinearities in these equations it is possible
864: to have a collapse of fermions. 
865: 
866: \begin{figure}%[!ht]
867:  
868: \begin{center}
869: \includegraphics[width=1.\linewidth]{fig4.ps}
870: %\postscript{fig1.eps}{1.}
871: \end{center}
872: 
873: \caption{ The   fermion density $n_j(r)$ at $t=0,20$
874: ms 
875: before and during the  collapse exhibited in  figure  3 
876: for $K=10 ^{-26}$ cm$^6$
877: s$^{-1}$. The density calculated from the variational profile of the wave
878: function (\ref{var}) for $R_0=5.3$ and $N=1000$ is also shown,}
879: 
880: \end{figure}
881: 
882: 
883: Due to the three-body loss  terms in  (\ref{d1}) and (\ref{d2}) the
884: number of 
885: fermions decay with time. When the net
886: nonlinear attraction
887: in these equations is small there is a smooth and steady decay of number
888: of
889: atoms. However, when the  net
890: nonlinear attraction is jumped to a large value,
891: the steady decay of number of
892: atoms develops into a violent decay called collapse. 
893: When this happens, the DFFM
894:  loses a significant fraction of atoms in a
895: small interval of
896: time (milliseconds) after which a remnant DFG  with a
897: reasonably
898: constant number of atoms is formed. Also, during and immediately
899: after collapse, the fermion density function becomes  unsmooth
900: and spiky in nature as opposed to a reasonably smooth  function in the
901: case of a steady decay. This also happened in the collapse of a BEC 
902: \cite{don}.
903: 
904: 
905: We study the evolution of  fermion numbers in the
906: DFFM 
907:  from time $t=0$ to $t=50$ ms after a sudden
908: jump in the
909: scattering length from $a_{12}=100$ nm to $-200$ nm at $t=0$. In agreement
910: with the variational analysis of last section we find that this jump in
911: the scattering length leads to collapse.  The
912: evolution of fermion numbers after the jump in scattering length $a_{12}$
913: depends on the value of the three-body loss
914: rate $K$. We study the sensitivity of the result on $K$ by
915: performing the calculation for different loss rates. In  figure  3 we plot 
916: the
917: evolution of the fermion numbers for loss rates: $K
918: = 10^{-26}$ cm$^6$s$^{-1}$, $10^{-25}$ cm$^6$s$^{-1}$, and
919: $10^{-24}$ cm$^6$s$^{-1}$. 
920: With the increase of $K$,  the decay
921: rate increases, although the results for different $K$ are qualitatively
922: similar. We see in  figure  3 that, in all cases, the 
923: number of  fermions decays
924: rapidly and attain an approximately  constant (remnant)  value  in less
925: than 50
926: ms as in the case of bosons \cite{don,th1}. We used a space step of 
927: 0.025 in the numerical solution of 
928:  (\ref{d1}) and (\ref{d2}) and found that this step size was
929: sufficient to reach a converged result even during collapse. In our
930: previous study on the collapse of a BEC of $^{85}$Rb \cite{ska,th1} we
931: found that 
932: even a much larger step size of 0.1 led to converged result for the number
933: of atoms in the remnant
934: in
935: quantitative agreement with experiment \cite{don}. 
936: This gives assurance on the reliability of the present calculation. 
937: 
938: 
939: 
940: 
941: \begin{figure}%[!ht]
942:  
943: \begin{center}
944: \includegraphics[width=1.\linewidth]{fig5a.ps}
945: \includegraphics[width=1.\linewidth]{fig5b.ps}
946: %\postscript{fig1.eps}{1.}
947: \end{center}
948: 
949: \caption{Evolution of    central density $n_j(0)$
950: of  fermion $j=$ (a) 1 and (b) 2
951:  during the collapse exhibited in
952:  figure  3 for $K=10^{-26}$ cm$^6$s$^{-1}$.} 
953: 
954: 
955: \end{figure}
956: 
957: 
958: 
959:  In  figure  4
960:  we plot the fermion probability densities
961: at times $t=0$ and $t=20$ ms.
962: A close look at  figure  4 reveals that before collapse at $t=0$ the
963: fermion densities are smooth. We have also plotted here the density
964: corresponding to the variational profile (\ref{var}) for $R_0=5.3$ for
965: fermion number $N=1000$. Although the ranges of the exact and variational
966: densities agree with each other, the functions do not agree 
967: well. This is understandable as the exact profile of the 
968: DFG  is very
969: different from the Gaussian trial function used in variational
970: calculation. Although the fermion densities at $t=0$ are smooth, 
971: the fermion densities  during and after collapse  have an entirely 
972: different profile. 
973: As expected the  densities  are highly peaked in the central ($r=0$) 
974: region and develop spikes. Near $r=0$ the densities could be two to three
975: orders of magnitude larger than those for larger $r$ values (see  figure  
976: 5).
977: However, they extend over a large distance
978: too. The final spiky function indicates the collapse, in contrast to a
979: smooth final  function corresponding to a  steady loss of
980: atoms. The collapse is a quick process lasting at most a few tens of
981: milliseconds when a significant fraction of atoms are lost. For example,
982: in  figure  3 for $K=10^{-24}$ cm$^6$s$^{-1}$, the collapse lasts for the
983: first 
984: 25 ms when most of the atoms are lost. After this 
985: interval the rate of loss of atoms is reduced and 
986: remnant a DFG  
987: with a roughly constant number of atoms are formed. 
988: 
989: 
990: 
991: 
992: 
993: To confirm further the  collapse in  figures 3 and 4 for $K=
994: 10^{-26}$
995: cm$^6$s$^{-1}$, we plot 
996: in  figure  5  the evolution of the central probability  density
997: $n_j(0)$
998: of fermion $j$ during collapse. 
999: We note a very strong fluctuation of a very large  central density
1000: reminiscent of
1001: collapse in both  components. The central density is three orders of
1002: magnitude larger than the equilibrium density in  figure  4. 
1003: Similar fluctuations 
1004: were noted in the  collapse of a pure BEC 
1005: \cite{th2}  as well as a DBFM
1006: \cite{ska}. Such a
1007: strong fluctuation of the central density  could not be due to a weak
1008: evaporation of the  DFFM  due to recombination. 
1009: 
1010: 
1011: \begin{figure}%[!ht]
1012:  
1013: \begin{center}
1014: \includegraphics[width=1.\linewidth]{fig6.ps}
1015: %\postscript{fig1.eps}{1.}
1016: \end{center}
1017: 
1018:  
1019: \caption{ The evolution of time to collapse 
1020: $t_{\mbox{collapse}}$ vs. final scattering length $a_{\mbox{collapse}}$
1021: for $a_{\mbox{initial}} = 0$ and 100 nm for a  DFFM
1022: with $N_1=1000$, 
1023: $N_2=2000$ and 
1024: $K=  10^{-26}$ cm$^6$s$^{-1}$.   The
1025: curves are labeled by their respective $a_{\mbox{initial}}$ values.}
1026: 
1027: 
1028: 
1029: 
1030: 
1031: \end{figure}
1032: 
1033: 
1034: From  figure  3  we find that the number of fermions remains practically 
1035: constant during the first 4 ms or so after jumping the scattering 
1036: length from 100 nm to $-200$ nm
1037: indicating that the collapse starts only 
1038: after this interval of time.  
1039: This is confirmed from the plot of central densities in  figure  5 where
1040: we 
1041: see that very large values of central density also appear after 
1042: an  
1043: interval of time called ``time to collapse". A similar phenomenon was also 
1044: observed in the collapse 
1045: of bosons \cite{don}. Next we study an evolution of this time to 
1046: collapse ($t_{\mbox{collapse}}$) with changing initial 
1047: ($a_{\mbox{initial}}$) and final ($a_{\mbox{collapse}}$) scattering 
1048: lengths. This is shown in  figure  6 for two values of
1049: $a_{\mbox{initial}}$, 
1050: where we plot $t_{\mbox{collapse}}$ vs. $a_{\mbox{collapse}}$ for 
1051: $N_1=1000$,  $N_2=2000$ and $K=10^{-26}$ cm$^6$s$^{-1}$. 
1052: The time to collapse is large for a small jump in the scattering length
1053: and reduces when the jump in the scattering length is increased, as also
1054: observed in the
1055: case of bosons \cite{don}.
1056: 
1057:  
1058: 
1059: \begin{figure}%[!ht]
1060:  
1061: \begin{center}
1062: \includegraphics[width=1.\linewidth]{fig7.ps}
1063: %\postscript{fig1.eps}{1.}
1064: \end{center}
1065: 
1066:  
1067: \caption{ The evolution of fermion numbers 
1068: $N_j(t)$
1069:  during collapse initiated by a jump in  scattering
1070: length $a_{12}$ from 100 nm to $-300$ nm  for
1071: $K=  10^{-27}$ cm$^6$s$^{-1}$,  $ 10^{-26}$
1072: cm$^6$s$^{-1}$, and
1073:   $10^{-25}$ cm$^6$s$^{-1}.$ 
1074:  The dotted (blue) curves
1075: refer to fermion 2
1076: and the solid (red)  curves to
1077: fermion 1. The
1078: curves are labeled by their respective $K$ values.}
1079: 
1080: 
1081: 
1082: 
1083: 
1084: \end{figure}
1085:  
1086: 
1087: 
1088: One interesting aspect of  figure  3 is the appearance of a revival of 
1089: collapse. The fermion number after the primary collapse remains
1090: approximately constant 
1091: for an interval of time and then again reduces abruptly. This revival
1092: of collapse takes place several times. A similar revival of 
1093: collapse was noted in the fermion component in a
1094: numerical simulation in a DBFM \cite{ska}
1095: and was confirmed later in an experiment \cite{bongs}
1096: on the
1097: $^{87}$Rb-$^{40}$K DBFM.
1098: To study the  revival of
1099: collapse further
1100: in a  DFFM we considered  a different
1101: jump in the scattering length. 
1102: For the same initial state of  figure  3 we now
1103: consider a jump in $a_{12}$ from $100$ nm to $-300$ nm and the dynamics
1104: is reported in  figure  7 for different $K$ values. We find that the
1105: revival of collapse has practically disappeared  in this case. If the 
1106: collapse is initiated by a  
1107: small jump  in the scattering length, the initial collapse is less 
1108: violent. However, after this initial milder collapse the  
1109: DFFM cannot 
1110: reach an equilibrium state and it remains large and attractive. 
1111: Consequently, the  DFFM undergoes further
1112: collapse(s). On the other hand, 
1113: if the collapse is initiated by a large jump in the scattering length, 
1114: the initial collapse is very violent  through which  the  
1115: DFFM
1116:  gets 
1117: rid of a very large number of atoms. Consequently, the  DFFM reaches  
1118: reasonably small and cold remnant states  which do not further undergo 
1119: collapse and one has
1120: one primary collapse.  
1121: This is clear from  figures  3 and 7. In  figure  3 after the first 
1122: collapse the DFFM loses a smaller percentage of atoms whereas in
1123:  figure  7 a large 
1124: percentage of atoms are lost after the primary collapse. 
1125: 
1126: 
1127: 
1128: \begin{figure}%[!ht]
1129:  
1130: \begin{center}
1131: \includegraphics[width=1.\linewidth]{fig8.ps}
1132: %\postscript{fig1.eps}{1.}
1133: \end{center}
1134: 
1135:  
1136: \caption{ The rms radii of the two components  vs. 
1137: time before and 
1138: during collapse exhibited in  figure  3  for $N_1=1000$ and $N_2=2000$ 
1139: initiated by a jump in  scattering
1140: length $a_{12}$ from 100 nm to $-200$ nm  for
1141: $K=  10^{-26}$ cm$^6$s$^{-1}$. 
1142:  The dotted (blue) curves
1143: refer to fermion 2
1144: and the solid (red)  curves to
1145: fermion 1. }
1146: 
1147: 
1148: 
1149: 
1150: 
1151: \end{figure}
1152: 
1153: 
1154: It was found in the experiment on collapse on a BEC \cite{don} that
1155: during collapse the root mean square (rms) sizes of the condensate
1156: execute periodic oscillation with approximately 
1157: twice the frequency of the trap. A breathing  oscillation of same
1158: frequency was found in a BEC when a small perturbation was applied
1159: \cite{sk1}.
1160: In
1161: dimensionless unit, the angular frequency of the trap is $\omega = 1$,
1162: corresponding to a (linear) frequency of $1/(2\pi)$.  The 
1163: observed frequency of
1164: oscillation of rms size was $1/\pi$ \cite{don} $-$ twice the trap 
1165: frequency. In actual time unit the frequency of oscillation of rms sizes
1166: during collapse  corresponds to 
1167: $1/(2\pi)$ ms$^{-1}$  $\approx 0.16$ ms$^{-1}$. 
1168: We also investigated if such
1169: oscillation existed in the present case in the  
1170: DFFM during the 
1171: collapse. 
1172: In  figure  8 we plot the rms
1173: radii of the two  components of the collapsing  DFFM
1174: and find that they also execute
1175: quasi-periodic oscillation. The calculated frequency from  figure 8 is
1176: 0.145
1177: ms$^{-1}$ close to that found in the case of bosons, e.g.  0.16
1178: ms$^{-1}$. The difference could be due to the coupled nature of the
1179: hydrodynamic 
1180: equations as well as the very large nonlinearity for fermions.
1181: 
1182: 
1183: 
1184: 
1185: \section{Summary}
1186:  
1187: We suggested a coupled set of time-dependent hydrodynamic
1188: equations for a trapped  DFFM including the effect of three-body 
1189: recombination.
1190: The present time-dependent formulation permits us to study the
1191: non-equilibrium dynamics of a  DFFM. Using a
1192: variational analysis as well
1193: as a numerical solution of our model, 
1194: we study, for an attractive
1195: inter-species fermion-fermion interaction, the collapse 
1196: in a DFFM composed of two types of nonidentical atoms.
1197: The collapse of a  DFFM of two different atoms
1198:  can be realized
1199: experimentally 
1200: by jumping the inter-species
1201: scattering length  to a large negative value by exploiting a
1202: fermion-fermion Feshbach resonance \cite{fsff}.  The collapse dynamics is
1203: strongly
1204: dependent
1205: on the three-body loss rate $K$ and we present results for different loss 
1206: rates. We note the possibility of  a revival of collapse in a  
1207: DFFM
1208: as in a previous simulation \cite{ska} on a 
1209: DBFM,
1210: confirmed later in an  
1211: experiment \cite{bongs} on
1212: $^{87}$Rb-$^{40}$K.  We find that a revival of collapse in a 
1213: DFFM takes 
1214: place for a moderate jump in the interspecies scattering length which
1215: disappears for a larger jump. We also study the quasi-periodic 
1216: oscillation of
1217: the  DFFM
1218: with approximately twice the trap
1219: frequency
1220: during collapse and explosion. 
1221: 
1222: A proper treatment of a  DFFM
1223: should be performed
1224: using a fully
1225: antisymmetrized many-body Slater determinant wave function \cite{yyy1} as
1226: in the case of atomic and molecular scattering involving many electrons
1227: \cite{ps}. However, in view of the success of the hydrodynamic model in a
1228: description of a collapse \cite{ska}, the formation of bright \cite{fbs2}
1229: and dark \cite{fds} solitons, and vortex states \cite{Jezek} in a
1230: DBFM,
1231: we
1232: do not believe that the present study on the collapse in a  DFFM
1233: to be so
1234: peculiar as to have no general validity. The present study on
1235: collapse in a  DFFM of nonidentical atoms
1236: can be verified in future
1237: experiments, which can really
1238: validate the present hydrodynamic model and the related numerical study.
1239: 
1240: 
1241:  
1242:  
1243: %\begin{center}
1244:  
1245: \ack
1246: %\acknowledgments
1247: %\end{center} 
1248: 
1249: The work is supported in part by the CNPq and FAPESP  
1250: of Brazil.
1251:  
1252:   
1253: \section*{References}
1254:  
1255: \begin{thebibliography}{99}
1256: %\begin{references}
1257: 
1258: 
1259:  
1260: \bibitem{exp1}DeMarco B and Jin D S 1999 {\it  Science} {\bf 285} 1703
1261:  
1262:  
1263: \bibitem{exp2} O'Hara K M, Hemmer S L, Gehm M E,  Granade S R
1264: and
1265:  Thomas J E 2002 {\it Science} {\bf 298} 2179
1266:  
1267: 
1268: 
1269:  
1270: \bibitem{exp21}  
1271: Jochim S, Bartenstein M, Hendl G, Denschlag J H, Grimm R, Mosk A and 
1272: Weidemuller M
1273:  2002 \PRL {\bf 
1274: 89} 273202
1275: 
1276: 
1277:  Jochim S, Bartenstein M, Altmeyer A, Hendl G, Chin C, Denschlag J H and 
1278: Grimm R  2003 \PRL {\bf
1279: 91} 240402
1280: 
1281: \bibitem{exp211}	 Chin C, Bartenstein M, Altmeyer A, Riedl S, 
1282: Jochim S, Denschlag J H and Grimm R 
1283: 2004 {\it Science} {\bf 305} 1128 
1284: 
1285:  Partridge G B, Li W H, Kamar R I, Liao Y A and  Hulet R G 2006 
1286: {\it Science} {\bf 311} 503
1287: 
1288: 
1289: \bibitem{exp22}Houbiers M, Ferwerda R, Stoof H T C, McAlexander W I, 
1290: Sackett C A and Hulet R G  1997
1291: \PR A {\bf 56} 4864
1292: 
1293: 
1294: \bibitem{exp3}Schreck F,  Khaykovich L, Corwin K L,  Ferrari G,
1295:  Bourdel T,
1296: Cubizolles J  and  Salomon C  2001 {\it
1297: Phys. Rev. Lett.} {\bf 87} 080403
1298:  
1299:  Truscott A G,  Strecker K E,  McAlexander W I,
1300: Partridge G B and  Hulet R G 2001 {\it  Science} {\bf 291} 2570
1301: 
1302: 
1303:  
1304:  
1305: \bibitem{exp4}  Hadzibabic Z,  Stan C A,
1306:  Dieckmann K,  Gupta S,  Zwierlein M W,  Gorlitz A and  Ketterle W 2002
1307: {\it Phys. Rev. Lett.} {\bf 88} 160401
1308: 
1309: 
1310: 
1311: %\bibitem{exp4} Z. Hadzibabic, C. A. Stan, K. Dieckmann, S. Gupta, M. W.
1312: %Zwierlein, A. Gorlitz, and W. Ketterle, Phys. Rev. Lett. {\bf 88}, 160401
1313: %(2002); Z. Hadzibabic, S. Gupta, C. A. Stan, C. H. Schunck, M. W.
1314: %Zwierlein, K. Dieckmann, and W. Ketterle, {\it ibid.} {\bf 91}, 160401
1315: %(2003).
1316: 
1317: \bibitem{exp6}Strecker K E,  Partridge G B and  Hulet R G 2003
1318: {\it Phys. Rev. Lett.} {\bf 91} 080406
1319:  
1320:  Hadzibabic Z,
1321:  Gupta S,  Stan C A,   Schunck C H,   Zwierlein M W,
1322:   Dieckmann K  and  Ketterle W 2003 {\it Phys. Rev. Lett.} {\bf 91}
1323: 160401
1324: 
1325: 
1326:  
1327:  
1328: \bibitem{exp5}Modugno G,  Roati G,  Riboli F,  Ferlaino F,
1329:  Brecha R J
1330: and  Inguscio M 2002 {\it  Science} {\bf 297} 2240
1331: 
1332: 
1333:  
1334:  
1335: \bibitem{exp5x}  Roati G,  Riboli F,  Modugno G and  Inguscio M 2002
1336: {\it Phys. Rev. Lett.} {\bf 89} 150403
1337: 
1338:  
1339:  
1340:  
1341: 
1342: 
1343:  
1344:  
1345: \bibitem{bongs} Ospelkaus C,  Ospelkaus S,  Sengstock K and
1346:  Bongs K  2006 \PRL {\bf 96} 020401
1347: 
1348: \bibitem{mur} 
1349: Gunter K, Stoferle T, Moritz H, Kohl M and  Esslinger T
1350: 2006 \PRL {\bf 96} 180402
1351: 
1352: 
1353:  Goldwin J, Papp S  B, DeMarco B and  Jin D S
1354: 2002 \PR A {\bf 65}  021402
1355: 
1356: 
1357: 
1358: \bibitem{bcs}Zwierlein M W,  Abo-Shaeer J R, Schirotzek A,  
1359: Schunck C H
1360: and  Ketterle  W 2005 {\it  Nature} {\bf 435} 1047 
1361: 
1362: 
1363: Zwierlein M W, Schirotzek A, Schunck C H and Ketterle W
1364:   2006 {\it 
1365: Science}  {\bf 311} 492
1366: 
1367: \bibitem{ccol}Zaccanti M, D'Errico C,  Ferlaino F,  Roati G,  
1368: Inguscio  M and   Modugno G 2006  ``Control of the interaction in a 
1369: Fermi-Bose 
1370: misture"  cond-mat/0606757 
1371: 
1372:   Ospelkaus S,  Ospelkaus C, 
1373: Humbert L,  Sengstock K and  Bongs K  2006 
1374: \PRL {\bf 97} 120403
1375: 
1376:  
1377: \bibitem{hulet}  Gerton J M,   Strekalov D,   Prodan I   and
1378:   Hulet R G 2001
1379: {\it Nature} {\bf 408} 692
1380: 
1381: 
1382: 
1383: 
1384:  
1385:  
1386: \bibitem{don}  Donley E A, Claussen N R,   Cornish S L,
1387: Roberts J L,  Cornell E A and  Wieman C E 2001 {\it  Nature} {\bf 412}
1388: 295
1389: 
1390: 
1391:  
1392: \bibitem{fs}  Inouye S,   Andrews M R,   Stenger J,   Miesner H J,
1393:  
1394:  Stamper-Kurn D M and   Ketterle W 1998    {\it  Nature } {\bf 392} 151
1395: 
1396:  
1397:  
1398:  
1399: \bibitem{th1}  
1400: Adhikari S K 2002 \PR A       {\bf 66} 043601
1401:  
1402: Adhikari S K 2005 \PR A      {\bf 71} 053603
1403: 
1404: 
1405:  
1406: \bibitem{th2}  Santos L  and Shlyapnikov G V 2002 \PR   A
1407:   {\bf 66} 011602(R)
1408:  
1409:   Saito H  and   Ueda M 2002  \PR A       {\bf 65} 033624
1410:  
1411:  Savage C M,  Robins N P  and   Hope J J 2003  \PR A
1412:   {\bf 67} 014304
1413:  
1414:  
1415: Duine R A  and  Stoof H T C 2003  \PR A
1416:   {\bf 68} 013602
1417: 
1418:  
1419:  Calzetta E A  and  Hu B L  2003 \PR A
1420:  {\bf 68} 043625
1421:  
1422:  
1423:  Bao W,  Jaksch D   and  Markowich P A  2004  \jpb  {\bf 37} 329
1424:  
1425:  
1426:  Adhikari S K   2004 \jpb   {\bf 37} 1185
1427:  
1428:  
1429: \bibitem{ska} Adhikari S K 2004 {\it Phys. Rev. } A {\bf 70} 043617
1430: 
1431:  
1432: \bibitem{zzz} Modugno M,  Ferlaino F,  Riboli F,  Roati G,
1433:  Modugno G
1434: and
1435:  Inguscio M 2003 \PR A
1436:  {\bf 68} 043626
1437: 
1438: \bibitem{gehm} Gehm M E,  Hemmer S L,  Granade S R,  O'Hara K M
1439: and
1440:  Thomas J E 2003 {\it Phys. Rev.} A {\bf 68} 011401(R) 
1441: 
1442: 
1443:  
1444: 
1445: 
1446: 
1447: 
1448: 
1449:  
1450:  
1451: \bibitem{phase} Roth R and
1452:  Feldmeier H 2001 \jpb
1453:  {\bf 34}  4629
1454: 
1455:  
1456: 
1457: \bibitem{kok} Kokkelmans S J J M F, Milstein J N, Chiofalo M L, Walser 
1458: R and  
1459: Holland M J
1460:   2002
1461: \PR A {\bf 65} 053617
1462: 
1463: \bibitem{hei}Heiselberg H 2001
1464: \PR A {\bf 63}  043606
1465: 
1466: \bibitem{hulet1} Hulet R G 2006 private communication.  I thank Dr. 
1467: Hulet 
1468: for a discussion, which clarified this point,  suggesting the strong 
1469: possibility of collapse in a
1470: DFFM of different atoms.   
1471: 
1472: 
1473: \bibitem{fbs2} Adhikari S K 2005 {\it Phys. Rev. } A {\bf 72} 053608
1474: \bibitem{fds} Adhikari S K 2005 \jpb  {\bf 38} 3607
1475: 
1476: 
1477: \bibitem{md}Adhikari S K 2006 {\it Phys. Rev.} A {\bf 73} 043619 
1478: \bibitem{bs}
1479: Adhikari S K 2006  Laser Phys. Lett.  in press
1480: 
1481: Adhikari S K 2006  Laser Phys. Lett.  in press
1482: 
1483:  
1484: \bibitem{bong1}Karpiuk T, Brewczyk M, Ospelkaus-Schwarzer S, Bongs K,
1485: Gajda M
1486: and Rz\c a\.zewski K 2004 \PRL {\bf 93} 100401
1487: 
1488:  
1489: \bibitem{Jezek}
1490:    Jezek D M,  Barranco M,  Guilleumas M,  Mayol R and
1491:  Pi M 2004 \PR A
1492:  {\bf 70}
1493: 043630
1494:  
1495: %D. M. Jezek, M. Barranco, M. Guilleumas, R. Mayol, and M. Pi,
1496: %Phys. Rev. A {\bf 70}, 043630 (2004);
1497:  
1498:  Guilleumas M,  Centelles M,  Barranco M,  Mayol R and  Pi M 2005
1499: \PR A
1500: {\bf 72} 053602
1501: 
1502: \bibitem{11}  Dalfovo F,  Giorgini S,  Pitaevskii L P and
1503:   Stringari S 1999
1504: {\it Rev. Mod. Phys.}  {\bf 71} 463
1505: 
1506: 
1507: 
1508:  Yukalov V I 2004 {\it Laser
1509: Phys. Lett.}  {\bf 1} 435 
1510:  
1511:  
1512: \bibitem{sk1}  Adhikari S K  and  Muruganandam P 2002 \jpb
1513: {\bf 35} 2831
1514: 
1515: \bibitem{fsff}  O'Hara K M,  Hemmer S L,   Granade S R,  Gehm M E,
1516:  Thomas J E,  Venturi V,  Tiesinga E and  Williams C J 2002 {\it
1517: Phys. Rev.} A
1518: {\bf 66} 041401(R)
1519:  
1520:  Dieckmann K,
1521:  Stan C A,   Gupta S,  Hadzibabic Z,  Schunck C H  and  Ketterle W
1522: 2002
1523: {\it Phys. Rev. Lett.} {\bf 89} 203201
1524:  
1525:  Loftus T,  Regal C A,
1526: Ticknor C,  Bohn J L and  Jin D S 2002 \PRL
1527:  {\bf 88} 173201
1528:  
1529:   Regal C A,  Greiner M and
1530:  Jin D S 2004   \PRL         {\bf 92}  083201
1531: 
1532:  
1533:  
1534: \bibitem{adhi}Adhikari S K 2001 {\it  Phys. Rev.} A {\bf 63} 043611
1535:  
1536:  
1537: Adhikari S K 2001 \jpb
1538: {\bf 34} 4231
1539:  
1540: 
1541: 
1542:   
1543:  \bibitem{capu} 
1544:  Capuzzi P,  Minguzzi A and  Tosi M P  2003
1545: {\it  Phys. Rev. } A
1546: {\bf 67} 053605
1547:  
1548:  Capuzzi P,  Minguzzi A and  Tosi M P  2003
1549: {\it  Phys. Rev. } A
1550: {\bf 68} 033605
1551: 
1552: 
1553:  
1554:  \bibitem{gold} Goldstein H 1980  {\it Classical Mechanics},
1555: (Reading, Addison
1556: Wesley)
1557: 
1558: 
1559:  
1560: \bibitem{sk2}
1561:  Muruganandam P  and   Adhikari S K 2003 \jpb           {\bf
1562: 36} 2501
1563: 
1564:  
1565: \bibitem{varia}  P\'erez-Garc\'ia V M,  Michinel H,
1566:  Cirac J I,  Lewenstein M  and Zoller P 1996
1567: {\it Phys. Rev. Lett.}  {\bf 77} 5320
1568:  
1569: Stoof H T C 1997 {\it  J. Stat. Phys.}  {\bf 87} 1353
1570:  
1571:  
1572:  Adhikari S K 2005 \jpb  {\bf 38} 579
1573:  
1574:   Anderson D 1983
1575: {\it Phys. Rev.} A {\bf 27}  3135
1576: 
1577: 
1578:  
1579:  
1580:  
1581: 
1582:  
1583:  
1584: \bibitem{yyy1}  Molmer K 1998 \PRL
1585: {\bf 80} 1804
1586:  
1587:  
1588:  
1589:  
1590:  Liu X-J  and  Hu H 2003 \PR A
1591:    {\bf 67} 023613
1592:  
1593:  Roth R and
1594:  Feldmeier H 2002   \PR A {\bf 65} 021603R
1595:  
1596:   Miyakawa T,
1597:  Suzuki T and  Yabu H 2001 \PR A   {\bf 64} 033611
1598: 
1599: 
1600: 
1601:  
1602: \bibitem{ps}
1603: Biswas P K and   Adhikari S K 2000  \jpb  {\bf 33} 1575
1604:  
1605: %Biswas P K and   Adhikari S K 1998  \jpb  {\bf 31} L737
1606:  
1607: Biswas P K and   Adhikari S K 1998  \jpb  {\bf 31} L315
1608:  
1609: %Biswas P K and   Adhikari S K 1998  \jpb  {\bf 31} 3147
1610:  
1611:   Adhikari S K 1979 \PR C {\bf 19}  1729
1612: 
1613:   Adhikari S K and Ghosh A 1997 \JPA {\bf 30} 6553
1614: 
1615: Tomio L and  Adhikari S K 1980 \PR C {\bf 22}  28
1616: 
1617:  
1618: \end{thebibliography} 
1619:  
1620: \end{document}
1621: 
1622: 
1623: