1: %documentclass[aps,prl,preprint,psfig,groupedaddress]{revtex4}
2: %\documentclass[aps,prl,preprint,psfig,superscriptaddress]{revtex4}
3: \documentclass[aps,prl,twocolumn,psfig,groupedaddress]{revtex4}
4: \def\bc{\begin{center}}
5: \def\ec{\end{center}}
6: \def\be{\begin{equation}}
7: \def\ee{\end{equation}}
8: \usepackage{longtable}
9: \usepackage{psfig}
10:
11: % You should use BibTeX and apsrev.bst for references
12: % Choosing a journal automatically selects the correct APS
13: % BibTeX style file (bst file), so only uncomment the line
14: % below if necessary.
15: %\bibliographystyle{apsrev}
16:
17: \begin{document}
18:
19: % Use the \preprint command to place your local institutional report
20: % number in the upper righthand corner of the title page in preprint mode.
21: % Multiple \preprint commands are allowed.
22: % Use the 'preprintnumbers' class option to override journal defaults
23: % to display numbers if necessary
24: %\preprint{}
25: %Title of paper
26: \title{Fractional statistics in the fractional quantum Hall effect}
27: \author{Gun Sang Jeon, Kenneth L. Graham, and Jainendra K. Jain}
28: % repeat the \author .. \affiliation etc. as needed
29: % \email, \thanks, \homepage, \altaffiliation all apply to the current
30: % author. Explanatory text should go in the []'s, actual e-mail
31: % address or url should go in the {}'s for \email and \homepage.
32: % Please use the appropriate macro foreach each type of information
33: % \affiliation command applies to all authors since the last
34: % \affiliation command. The \affiliation command should follow the
35: % other information
36: % \affiliation can be followed by \email, \homepage, \thanks as well.
37: %\author{J.K. Jain}
38: %\email[]{Your e-mail address}
39: %\homepage[]{Your web page}
40: %\thanks{}
41: %\altaffiliation{}
42: \affiliation{Physics Department, 104 Davey Laboratory, The Pennsylvania State University,
43: University Park, Pennsylvania 16802}
44:
45: %Collaboration name if desired (requires use of superscriptaddress
46: %option in \documentclass). \noaffiliation is required (may also be
47: %used with the \author command).
48: %\collaboration can be followed by \email, \homepage, \thanks as well.
49: %\collaboration{}
50: %\noaffiliation
51:
52: \date{\today}
53:
54: \begin{abstract}
55: A microscopic confirmation of the fractional statistics
56: of the {\em quasiparticles} in the fractional quantum Hall effect has so
57: far been lacking. We calculate the statistics of the composite-fermion
58: quasiparticles at $\nu=1/3$ and $\nu=2/5$ by evaluating the Berry phase for a
59: closed loop encircling another composite-fermion quasiparticle.
60: A careful consideration of subtle perturbations in the trajectory
61: due to the presence of an additional quasiparticle is crucial for
62: obtaining the correct value of the statistics. The conditions for the
63: applicability of the fractional statistics concept are discussed.
64: \end{abstract}
65: % insert suggested PACS numbers in braces on next line
66: \pacs{71.10.Pm,73.43.-f}
67: % insert suggested keywords - APS authors don't need to do this
68: %\keywords{}
69:
70: %\maketitle must follow title, authors, abstract, \pacs, and \keywords
71:
72: \maketitle
73:
74: % body of paper here - Use proper section commands
75: % References should be done using the \cite, \ref, and \label commands
76: % Put \label in argument of \section for cross-referencing
77: %\section{\label{}}
78:
79:
80: The fractional statistics concept of Leinaas and Myrheim\cite{Leinaas}
81: relies on the property that when particles with infinitely strong short range
82: repulsion are confined in two dimensions, paths with different winding numbers are
83: topologically distinct and cannot be deformed into one another.
84: The particles are said to have statistics $\theta$ if a path
85: independent phase $2\pi\theta$ results
86: when one particle goes around another in a complete loop.
87: A half loop is equivalent to an exchange of particles, assuming
88: translational invariance, which produces a phase factor $e^{i\pi \theta}=(-1)^{\theta}$.
89: Non-integral values of $\theta$ imply fractional statistics.
90: There are no fundamental particles in nature that obey fractional statistics.
91: Any fractional statistics objects will have to be emergent collective
92: particles of a non-trivial condensed matter state. Furthermore, they will
93: be necessarily confined to two dimensions: in higher dimensions the
94: notion of a particle going around another is topologically
95: ill defined, because any loop can be shrunk to zero without
96: ever crossing another particle.
97:
98:
99: Even though the explanation of the fractional quantum Hall effect\cite{Tsui} (FQHE)
100: and numerous other remarkable phenomena
101: follows from the composite fermion theory with no mention of fractional
102: statistics \cite{Heinonen}, fractional statistics is believed to be one of the
103: consequences of incompressibility at a fractional filling \cite{Arovas,Halperin,Su},
104: and may possibly be observable in an experiment specifically designed for this
105: purpose. For Laughlin's quasiholes\cite{Laughlin} at
106: $\nu=1/m$, $m$ odd, the statistics was derived explicitly by Arovas, Schrieffer, and
107: Wilczek\cite{Arovas} in a Berry phase calculation, but a similar demonstration of
108: fractional statistics has been lacking at other fractions, or even for
109: the quasiparticles at $\nu=1/m$.
110: The need for a microscopic confirmation was underscored by
111: Kj{\o}nsberg and Myrheim\cite{Kjonsberg1}
112: who showed that, with Laughlin's wave function, the quasiparticles at $\nu=1/m$
113: do {\em not} possess well-defined statistics.
114: The reason for the discrepancy remains unclear, but it illustrates that
115: the fractional statistics is rather fragile and cannot be taken for granted.
116:
117:
118: The objective of this article is to revisit the issue armed with
119: the microscopic composite-fermion (CF) theory of the FQHE \cite{Jain}.
120: A step in that direction has been taken by Kj{\o}nsberg and
121: Leinaas\cite{Kjonsberg2}, whose calculation of the statistics of
122: the ``unprojected" CF quasiparticle of $\nu=1/m$, the wave function for which is
123: different from that of Laughlin's, produced a definite value,
124: the sign of which, however, was inconsistent with general considerations.
125: We confirm below that the statistics is robust to projection into the lowest
126: Landau level (LL), and provide a non-trivial resolution to the sign
127: enigma, which has its origin in very small perturbations in the
128: trajectory due to the insertion of an additional CF quasiparticle.
129: The calculation is extended to $\nu=2/5$ for further verification of
130: the generality of the concept.
131:
132:
133: Because the CF theory provides an accurate account of the low energy
134: physics, including incompressibility at certain fractional fillings,
135: it must also contain the physics of fractional statistics, which indeed
136: is the case. The fractional statistics
137: can be derived heuristically in the CF theory as follows \cite{Goldhaber}.
138: Composite fermions are bound states of electrons and an even number ($2p$)
139: of vortices. When a composite fermion goes around a closed path
140: encircling an area $A$, the total phase associated with this path is given by
141: \begin{equation}
142: \Phi^*=-2\pi (BA/\phi_0- 2p N_{enc}) \;,
143: \label{Phi*}
144: \end{equation}
145: where $N_{enc}$ is the number of composite fermions inside the
146: loop and $\phi_0=hc/e$ is called the flux quantum.
147: The first term on the right hand side is the usual Aharonov-Bohm
148: phase for a particle of charge $-e$ going around in a counterclockwise loop.
149: The second term is the contribution from the vortices bound to composite fermions,
150: indicating that each enclosed
151: composite fermion effectively reduces the flux by $2p$ flux quanta.
152: (A note on convention: We will take the magnetic field in the $+z$ direction,
153: the electron charge to be $-e$, and consider the counterclockwise direction for
154: the traversal of trajectories.)
155:
156:
157: Eq.~(\ref{Phi*}) summarizes the origin of the FQHE.
158: The phase in Eq.~(\ref{Phi*}) is interpreted as the
159: Aharonov-Bohm phase from an effective magnetic field:
160: $\Phi^* \equiv -2\pi B^*A/\phi_0$.
161: Replacing $N_{enc}$ by its expectation value $\langle N_{enc}\rangle=\rho A$,
162: where $\rho$ is the two-dimensional density of electrons, we get
163: \be
164: B^*=B-2p\phi_0\rho\;.
165: \label{B*}
166: \ee
167: The integral quantum Hall effect\cite{Klitzing} (IQHE) of composite fermions
168: at CF filling $\nu^*=n$ produces the FQHE of electrons at
169: $\nu=n/(2pn+1)$. At these special filling factors, the effective magnetic
170: field is $B^*=B/(2pn+1)$.
171:
172:
173: The fractional statistics is also an immediate corollary of Eq.~(\ref{Phi*}).
174: Let us consider the state with CF filling $n<\nu^*<n+1$ and denote by
175: $\eta_\alpha=x_\alpha-iy_\alpha$ the positions where the composite
176: fermions in the topmost
177: partially filled CF level are localized in suitable wave packets.
178: One may imagine a density lump centered at each $\eta_\alpha$.
179: An ``effective" description in terms of $\eta_\alpha$, which will
180: be called CF quasiparticles (CFQP's), can in principle be obtained by
181: integrating out $z_j=x_j-iy_j$. We can conjecture the winding properties
182: of the CFQP's from the underlying CF theory as follows. Consider
183: two CFQP's, sufficiently far from one another that the overlap
184: between them is negligible. According to Eq.~(\ref{Phi*}) the phase a CFQP acquires
185: for a closed loop depends on whether the loop encloses the other CFQP or not.
186: When it does not, the phase is $\Phi^*=-2\pi eB^*A/hc$.
187: The change in the phase due to the presence of the enclosed CFQP is
188: \be
189: \Delta\Phi^*=2\pi 2p \Delta \langle N_{enc}\rangle=2\pi \frac{2p}{2pn+1}
190: \ee
191: because a CFQP has an excess of
192: $1/(2pn+1)$ electrons associated with it relative to the uniform state
193: [producing a local charge of $q^*=-e/(2pn+1)$].
194: With $\Delta\Phi^*=2\pi\theta^*$ we get the CFQP statistics parameter
195: \be
196: \theta^*=\frac{2p}{2pn+1}\;.
197: \label{theta*}
198: \ee
199: This value is consistent, $mod$ 1, with those quoted previously\cite{Halperin,Su}.
200:
201:
202: Our goal is to confirm Eq.~(\ref{theta*}) in a microscopic calculation
203: of the Berry phases. The statistics is given by
204: \be
205: \theta^* =
206: \oint_{\cal C} \frac{d\theta}{2\pi} \frac{\left<\Psi^{\eta,\eta'}| i\frac{d}{d\theta}
207: \Psi^{\eta,\eta'}\right>} {\left<\Psi^{\eta,\eta'}|\Psi^{\eta,\eta'}\right>}-
208: \oint_{\cal C} \frac{d\theta}{2\pi}
209: \frac{\left<\Psi^\eta|
210: i\frac{d}{d\theta} \Psi^{\eta}\right>}{
211: \left<\Psi^\eta|\Psi^{\eta}\right>}\;,
212: \label{Berry}
213: \ee
214: where $\Psi^\eta$ is the wave function containing a single CFQP at $\eta$, and
215: $\Psi^{\eta,\eta'}$ has two CFQP's at $\eta$ and $\eta'$.
216: Here we take $\eta=R e^{-i\theta}$, and ${\cal C}$ refers to the path
217: with $R$ fixed and $\theta$ varying from $0$ to $2\pi$ in the counterclockwise direction.
218: For convenience, we will take $\eta'=0$.
219:
220:
221: The calculation of $\theta^*$ requires microscopic wave functions
222: which are constructed as follows.
223: The composite fermion theory maps the problem of interacting electrons at
224: $\nu$ into that of weakly interacting composite fermions at
225: $\nu^*$. In order to put these composite fermions at $\eta_\alpha$, we first
226: construct the electronic wave function at $\nu^*$ with the
227: electrons in the partially filled level at $\eta_\alpha$; these are
228: placed in the coherent state wave packets
229: \be
230: \bar \phi^{(n)}_{\eta}(\vec{r})= \phi^{(n)}_{\eta}(\vec{r})
231: \exp[-|z|^2/4l^{*2}]
232: \ee
233: \be
234: \phi^{(n)}_{\eta}(\vec{r})= (\bar{z}-\bar{\eta})^n
235: \exp[\bar{\eta}z/2l^{*2} -|\eta|^2/4l^{*2}]
236: \ee
237: where $l=\sqrt{\hbar c/eB}$ and $l^*=(2pn+1)^{1/2}l$ are the magnetic lengths at $B$ and
238: $B^*$. We then make a mapping into composite fermions {\em in a manner that preserves
239: distances} (to zeroth order) by multiplying by $\Phi_1^{2p}=
240: \prod_{j<k=1}^N(z_j-z_k)^{2p}\exp[-2p\sum_i|z_i|^2/4l_1^2]$
241: with $l_1^2=\hbar c/eB_1=\hbar c/e \rho\phi_0$, followed by projection into
242: the lowest LL.
243:
244:
245: To give an explicit example, consider two CFQP's at $\nu=1/(2p+1)$.
246: The electron wave function at $\nu^*=1$ with fully occupied lowest LL
247: and two additional electrons in the second LL at $\eta$ and $\eta'$ is
248: \be
249: \Phi_1^{\eta,\eta'}=\left|\begin{array}{ccccc}
250: \phi^{(1)}_{\eta}(\vec{r}_1) & \phi^{(1)}_{\eta}(\vec{r}_2) & . &.&.\\
251: \phi^{(1)}_{\eta'}(\vec{r}_1) & \phi^{(1)}_{\eta'}(\vec{r}_2) & . &.&.\\
252: 1 & 1 & . &.&. \\
253: z_{1}&z_{2}&. &.&.\\
254: .&.&.&.&.\\
255: .&.&.&.&. \\
256: z_1^{N-3} & z_2^{N-3} & . &.&.
257: \end{array}
258: \right|e^{-\sum_j|z_j|^2/4l^{*2}} \;.
259: \ee
260: This leads to the (unnormalized) wave function for two CFQP's at $\nu=1/(2p+1)$:
261: \begin{eqnarray}
262: \Psi^{\eta,\eta'}_{1/(2p+1)} &=& {\cal P} \prod_{i<k=1}^N(z_i-z_k)^{2p}
263: \left|\begin{array}{ccccc}
264: \phi^{(1)}_{\eta}(\vec{r}_1) & \phi^{(1)}_{\eta}(\vec{r}_2) & . &.&.\\
265: \phi^{(1)}_{\eta'}(\vec{r}_1) & \phi^{(1)}_{\eta'}(\vec{r}_2) & . &.&.\\
266: 1 & 1 & . &.&. \\
267: z_{1}&z_{2}&. &.&.\\
268: .&.&.&.&.\\
269: .&.&.&.&. \\
270: z_1^{N-3} & z_2^{N-3} & . &.&.
271: \end{array}
272: \right| \nonumber \\
273: & & \;\;\;\;\times e^{-\sum_j|z_j|^2/4l^2}\;.
274: \label{2CFQP}
275: \end{eqnarray}
276: Here, ${\cal P}$ is the lowest Landau level projection operator,
277: and we have used $l^{*-2}+2pl_1^{-2}=l^{-2}$ which is equivalent to Eq.~(\ref{B*}).
278: Wave functions for one or many CFQP's at arbitrary filling
279: factors can be written similarly. The lowest LL projection can be performed
280: in either one of two ways described in the literature\cite{JK}.
281: Our wave functions are similar to those considered
282: in Ref.~\onlinecite{Kjonsberg2}, but not identical.
283:
284:
285: The integrands in Eq.~(\ref{Berry}) involve $2N$ dimensional integrals over
286: the CF coordinates, which we evaluate by Monte Carlo method. To determine the
287: O(1) difference between two O($N$) quantities on the right hand side
288: with sufficient accuracy,
289: we use the same importance sampling for both the quantities on the
290: right hand side, which reduces statistical fluctuations in the difference.
291: The two-CFQP wave function $\Psi^{\eta,\eta'}$ is used as the weight function
292: for both terms in Eq.~(\ref{Berry}). Approximately $4 \times 10^{8}$
293: iterations are performed for each point. For $\nu=1/3$ we have studied systems with
294: $N=50$, 100, and 200 particles, and the projected wave function is used.
295: In this case, a study of fairly large systems is possible because
296: no explicit evaluation of the determinant is required at each step.
297: For $\nu=2/5$, it is much more costly to work with the projected wave
298: function, and we have studied only the unprojected wave function
299: for $N=50$ and 100. The calculation at $\nu=1/3$ explicitly demonstrates
300: that $\theta^*$ is independent of whether the projected or the
301: unprojected wave function is used, or which projection method is used;
302: we assume the same is true at $\nu=2/5$.
303:
304: \begin{figure}
305: \centerline{\psfig{figure=stat1_3.ps,width=3.0in,angle=0}}
306: \vspace{-1.0cm}
307: \centerline{\psfig{figure=stat2_5.ps,width=3.0in,angle=0}}
308: %\hspace{5cm}
309: \caption{The statistical angle $\tilde\theta^*$ for the CF quasiparticles at
310: $\nu=1/3$ (upper panel) and $\nu=2/5$ (lower panel) as a function of
311: $d \equiv |\eta-\eta'|$. $N$ is the total number of composite fermions, and $l$ is the
312: magnetic length. (The symbol $\tilde\theta^*$ is used rather than $\theta^*$
313: for the statistical angle to remind that the correct
314: interpretation of the results gives $\theta^*=-\tilde\theta^*$.) The error bar from
315: Monte Carlo sampling is not shown explicitly when it is smaller than the symbol size.
316: The deviation at the largest $d/l$ for each $N$ is due to proximity to the edge.}
317: \label{fig1}
318: \end{figure}
319:
320: The statistics parameter $\theta^*$ is shown in Fig.~\ref{fig1} for
321: $\nu=1/3$ and $\nu=2/5$. $\theta^*$ takes a well-defined value for large
322: separations.
323: At $\nu=1/3$ it approaches the asymptotic value of $\theta^*=-2/3$,
324: which is consistent with that obtained in Ref.~\onlinecite{Kjonsberg2}
325: without lowest LL projection. At $\nu=2/5$ the system size is
326: smaller and the statistical uncertainty bigger, but the asymptotic value
327: is clearly seen to be $\theta^*=-2/5$. At short separations there are
328: substantial deviations in $\theta^*$; it reaches the asymptotic
329: value only after the the two CFQP's are separated by more than
330: $\sim$ 10 magnetic lengths.
331:
332:
333: The microscopic value of $\theta^*$ obtained above has the same magnitude as
334: $\theta^*$ in Eq.~(\ref{theta*}) {\em but the opposite sign}.
335: The sign discrepancy, if real, is profoundly disturbing because it cannot be
336: reconciled with Eq.~(\ref{Phi*}) and would cast doubt on the
337: fundamental interpretation of the CF physics in terms of an effective
338: magnetic field.
339:
340: \begin{figure}
341: \centerline{\psfig{figure=shift.ps,width=3.0in,angle=0}}
342: \vspace{-5mm}
343: \caption{Density profiles for $\Psi^{\eta}$ (dashed line) and
344: $\Psi^{\eta,\eta'}$ (solid line) along the $x$ axis at
345: $\nu=1/3$, with $\eta=13 l$ and $\eta'=0$.
346: (The uniform state has density $\rho=\nu/2\pi$.)
347: The noise on the curves is a measure of the statistical uncertainty in the
348: Monte Carlo simulation. The CFQP in the second level
349: has a smoke ring shape, with a minimum at its center. The CFQP is located at
350: $x=13 l$ in $\Psi^{\eta}$, but is shifted outward in $\Psi^{\eta,\eta'}$.
351: The inset shows the density profiles for $\chi^{\eta}$ (dashed line) and
352: $\chi^{\eta,\eta'}$ (solid line), describing CFQP's in
353: the lowest LL (see the text for definition).
354: \label{fig2}}
355: \end{figure}
356:
357: To gain insight into the issue,
358: consider two composite fermions in the otherwise empty lowest LL,
359: for which various quantities can be obtained analytically.
360: When there is only one composite fermion at $\eta=Re^{-i\theta}$,
361: it is the same as an electron, with the wave function given by
362: \be
363: \chi^{\eta}=\exp[\bar{\eta}z/2 -R^2/4-|z|^2/4]\;.
364: \ee
365: For a closed loop,
366: \be
367: \oint_{\cal C} \frac{d\theta}{2\pi}
368: \frac{\left<\chi^\eta|
369: i\frac{d}{d\theta} \chi^{\eta}\right>}{
370: \left<\chi^\eta|\chi^{\eta}\right>}=-\frac{R^2}{2l^2}=-\frac{\pi R^2B}{\phi_0}\;.
371: \ee
372: Two composite fermions, one at $\eta$ and the other at
373: $\eta'=0$, are described by the wave function
374: \be
375: \chi^{\eta,0}=(z_1-z_2)^{2p}(e^{\bar\eta z_1/2}-e^{\bar\eta z_2/2})
376: e^{-(R^2+|z_1|^2+|z_2|^2)/4}
377: \ee
378: Here, we expect $\theta^*=2p$.
379: However, an explicit evaluation of the Berry phase shows, neglecting O($R^{-2}$) terms
380: \be
381: \oint_{\cal C} \frac{d\theta}{2\pi}
382: \frac{\left<\chi^{\eta,0}|
383: i\frac{d}{d\theta} \chi^{\eta,0}\right>}{
384: \left<\chi^{\eta,0}|\chi^{\eta,0}\right>}=-\frac{R^2}{2l^2}-2p\;,
385: \label{wrong}
386: \ee
387: which gives $\theta^*=-2p$ for large $R$. Again, it apparently has
388: the ``wrong" sign.
389:
390:
391: A calculation of the density for $\chi^{\eta,0}$ shows that
392: the actual position of the outer composite fermion is not $R=|\eta|$ but
393: $R'$, given by
394: \be
395: R'^2/l^2=R^2/l^2+4\cdot 2p
396: \label{R2}
397: \ee
398: for large $R$. This can also be seen in the inset of Fig.~(\ref{fig2}).
399: The correct interpretation of Eq.~(\ref{wrong}) therefore is
400: \be
401: \oint_{\cal C} \frac{d\theta}{2\pi}
402: \frac{\left<\chi^{\eta,0}|
403: i\frac{d}{d\theta} \chi^{\eta,0}\right>}{
404: \left<\chi^{\eta,0}|\chi^{\eta,0}\right>}=-\frac{R'^2}{2l^2}+2p
405: \ee
406: which produces $\theta^*=2p$. The O(1) correction to the area enclosed
407: thus makes a non-vanishing correction to the statistics.
408: (It is noted that the CF quasiparticle at $\eta=0$ is also a little off center,
409: and executes a tiny circular loop which provides another correction to the
410: phase, but this contribution vanishes in the limit of large $R$.)
411:
412:
413: This exercise tells us that an implicit assumption made in the earlier analysis,
414: namely that the position of the outer CFQP labeled by $\eta$ remains
415: unperturbed by the insertion of another CFQP, leads to an incorrect value
416: for $\theta^*$. In reality, inserting another CFQP inside the loop pushes
417: the CFQP at $\eta$ very slightly outward.
418:
419:
420: To determine the correction at $\nu=n/(2pn+1)$, we note that the
421: mapping into composite fermions preserves distances to zeroth order,
422: so Eq.~(\ref{R2}) ought to be valid also at $\nu=n/(2pn+1)$.
423: This is consistent with the shift seen in Fig.~\ref{fig2} for the position of the
424: CFQP. Our earlier result
425: \be
426: \oint_{\cal C} \frac{d\theta}{2\pi}
427: \frac{\left<\Psi^{\eta,0}|
428: i\frac{d}{d\theta} \Psi^{\eta,0}\right>}{
429: \left<\Psi^{\eta,0}|\Psi^{\eta,0}\right>}=-\frac{R^2}{2l^{*2}}-\frac{2p}{2pn+1}
430: \ee
431: ought to be rewritten, using $l^{*2}/l^2=B/B^*=2pn+1$, as
432: \be
433: \oint_{\cal C} \frac{d\theta}{2\pi}
434: \frac{\left<\Psi^{\eta,0}|
435: i\frac{d}{d\theta} \Psi^{\eta,0}\right>}{
436: \left<\Psi^{\eta,0}|\Psi^{\eta,0}\right>}=-\frac{R'^2}{2l^{*2}}+\frac{2p}{2pn+1}
437: \ee
438: When the contribution from the closed path without the other CFQP,
439: $-R'^2/2l^{*2}$, is subtracted out, $\theta^*$ of Eq.~(\ref{theta*}) is obtained.
440: The neglect of the correction in the radius of the loop introduces an error
441: which just happens to be twice the negative of the correct answer.
442:
443:
444: The fractional statistics of the CFQP should not be confused with
445: the fermionic statistics of composite fermions. The wave
446: functions of composite fermions
447: are single-valued and antisymmetric under particle exchange;
448: the fermionic statistics of composite fermions has been
449: firmly established through a variety of facts, including the observation
450: of the Fermi sea of composite fermions, the observation of FQHE at fillings
451: that correspond to the IQHE of composite fermions, and also by the fact that
452: the low energy spectra in exact calculations on finite systems
453: have a one-to-one correspondence with those of weakly interacting
454: fermions\cite{Heinonen}.
455: There is no contradiction, however. After all, any fractional statistics in nature
456: {\em must} arise in a theory of particles that are either fermions or bosons
457: when an {\em effective} description is sought in terms of certain collective
458: degrees of freedom. The fractional statistics appears in the CF theory when
459: the original particles $\{z\}$ are treated in an average, mean field
460: sense to formulate an effective description in terms of the
461: CFQP's at $\{\eta\}$.
462:
463:
464: The fractional statistics is equivalent to the existence of an
465: effective locally pure gauge vector potential, with
466: no magnetic field associated with it except
467: at the particle positions\cite{Arovas}. In the present case,
468: the substantial deviation of $\theta^*$ from its asymptotic
469: value at separations of up to $10$ magnetic lengths indicates
470: a core region where the induced vector potential is not pure gauge,
471: thereby imposing a limitation on
472: a model in which the the CFQP's are approximated by ideal,
473: point-like particles with well-defined fractional statistics (anyons).
474: Such an idealization is valid only to the extent that the relevant CFQP trajectories
475: do not involve a significant overlap of CFQP's.
476: Given that there does not exist a strong repulsion between the CFQP's --
477: the inter-CF interaction is very weak and often {\em attractive}\cite{Lee} --
478: such trajectories are not precluded energetically, and the
479: anyon model is therefore not a justifiable approximation, except, possibly,
480: for very dilute systems of CFQP's in a narrow filling
481: factor range around $\nu=n/(2pn+1)$.
482: Any experimental attempt to measure the fractional statistics of the CFQP's
483: must ensure that they remain sufficiently far apart during the
484: measurement process.
485:
486:
487:
488: Partial support of this research by the National Science Foundation under grants
489: no. DGE-9987589 (IGERT) and DMR-0240458 is gratefully acknowledged.
490: We thank Profs. A.S. Goldhaber and J.M. Leinaas for comments.
491:
492:
493: \begin{references}
494:
495: \bibitem{Leinaas} J.M. Leinaas and J. Myrheim, Nuovo Cimento {\bf 37 B}, 1 (1977).
496:
497: \bibitem{Tsui} D.C. Tsui, H.L. Stormer, and A.C. Gossard, Phys. Rev.
498: Lett. {\bf 48}, 1559 (1982)
499:
500: \bibitem{Heinonen} O. Heinonen (Ed.) {\em Composite Fermions}
501: (World Scientific, New York, 1998);
502: S. Das Sarma and A. Pinczuk (Eds.), {\em Perspectives in Quantum Hall Effects} (Wiley,
503: New York, 1997); J.K. Jain, Physics Today {\bf 53(4)}, 39 (2000).
504:
505: \bibitem{Arovas} D. Arovas, J.R. Schrieffer, and F. Wilczek, Phys. Rev.
506: Lett. {\bf 53}, 722 (1984).
507:
508: \bibitem{Halperin} B.I. Halperin, Phys. Rev. Lett. {\bf 52}, 1583
509: (1984).
510:
511: \bibitem{Su} W.P. Su, Phys. Rev. B {\bf 34}, 1031 (1986).
512:
513: \bibitem{Laughlin} R.B. Laughlin, Phys. Rev. Lett. {\bf 50}, 1395 (1983).
514:
515: \bibitem{Kjonsberg1} H. Kj{\o}nsberg and J. Myrheim,
516: Int. J. Mod. Phys. A {\bf 14} 537 (1999).
517:
518: \bibitem{Jain} J.K. Jain, Phys. Rev. Lett. {\bf 63}, 199 (1989);
519: Phys. Rev. B {\bf 40}, 8079 (1989).
520:
521: \bibitem{Kjonsberg2} H. Kj{\o}nsberg and J.M. Leinaas, Nucl. Phys. B {\bf 559},
522: 705 (1999).
523:
524: \bibitem{Goldhaber} A.S. Goldhaber and J.K. Jain, Phys. Lett. A {\bf
525: 199}, 267 (1995).
526:
527: \bibitem{Klitzing} K.v. Klitzing, G. Dorda, and M. Pepper, Phys. Rev. Lett.
528: {\bf 45}, 494 (1980).
529:
530: \bibitem{JK} J.K. Jain and R.K. Kamilla, Int. J. Mod. Phys. B {\bf 11}, 2621
531: (1997).
532:
533: \bibitem{Lee} S.-Y. Lee, V.W. Scarola, and J.K. Jain, Phys. Rev. Lett. {\bf
534: 87}, 256803 (2001); A. W\'{o}js and J. J. Quinn, Phys. Rev. B {\bf 61}, 2846 (2000).
535:
536: \end{references}
537:
538:
539:
540: \end{document}
541:
542:
543: