1: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2: %% LATEST VERSION
3: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
4: %\documentstyle[preprint,aps]{revtex}
5: %\documentstyle[twocolumn,prl,aps]{revtex}
6: \documentclass[showpacs,showkeys,amssymb,aps,prl,epsfig]{revtex4}
7: \topmargin 0.1mm
8: \newcommand{\be}{\begin{equation}} \newcommand{\ee}{\end{equation}}
9: \newcommand{\bea}{\begin{eqnarray}}\newcommand{\eea}{\end{eqnarray}}
10: \newcommand{\bm}[1]{\mbox{\boldmath$#1$}}
11: \newcommand{\grad}{\bm \nabla}
12: \usepackage{graphicx}
13: \begin{document}
14: \preprint{SINP-TNP/02-23}
15: \title{Inequivalent Quantizations of the Rational Calogero Model}
16: \author{B. Basu-Mallick,\footnote{Email: biru@theory.saha.ernet.in}
17: Pijush K. Ghosh \footnote{Email: pijush@theory.saha.ernet.in} and
18: Kumar S. Gupta\footnote{Email: gupta@theory.saha.ernet.in ~ (corresponding
19: author)}}
20: %\footnote{Email: biru@theory.saha.ernet.in},
21: %Pijush K. Ghosh\footnote{Email: pijush@theory.saha.ernet.in}
22: %and Kumar S. Gupta\footnote{Email: gupta@theory.saha.ernet.in }}
23: \affiliation{ Theory Division\\
24: Saha Institute of Nuclear Physics\\
25: 1/AF Bidhannagar, Calcutta - 700064, India.\\}
26:
27: \begin{abstract}
28: We show that the self-adjoint extensions of the
29: rational Calogero model with suitable boundary conditions leads to
30: inequivalent quantizations of the system. The corresponding spectrum is
31: non-equispaced, consisting of infinitely many positive energy states and at
32: most a single negative energy state. These new states appear for arbitrary
33: number of particles and for specific range of coupling constant.
34: \end{abstract}
35:
36: \pacs{ 02.30.Ik, 03.65.-w, 03.65.Ge}
37: \keywords{ Calogero model, Self-adjoint extension, Bound states}
38: \maketitle
39:
40: The rational Calogero model is described by $N$ identical particles interacting
41: with each other through a long-range inverse-square and harmonic
42: interaction on the line \cite{calo3}. This is one of the most
43: celebrated examples of exactly solvable many-particle quantum mechanical
44: systems \cite{pr}. This model and its variants \cite{pr} are
45: relevant to the study of many branches of contemporary physics, including
46: generalized exclusion statistics \cite{poly},
47: quantum hall effect \cite{qhe}, Tomonaga-Luttinger liquid \cite{ll}, quantum
48: chaos \cite{rmt}, quantum electric transport in mesoscopic system \cite{qet},
49: spin-chain models \cite{hs}, Seiberg-Witten theory \cite{sw} and
50: black holes \cite{black}.
51:
52: The spectrum of the $N$-particle rational Calogero model was first obtained
53: almost three decades ago, which has since
54: been analyzed using a variety of different techniques \cite{brink}. In
55: his original work \cite{calo3}, Calogero used the boundary condition
56: that the wavefunction and the current vanish when any two or more particles
57: coincide. With this boundary condition the Hamiltonian is self-adjoint,
58: which ensures the reality of eigenvalues as well as the completeness of the
59: states. The central issue that we
60: address in this Letter is whether the spectrum obtained for rational Calogero
61: model is unique or does the system admit inequivalent quantizations leading
62: to different spectra? One way to address this issue is to look for
63: more general boundary conditions for which the Calogero Hamiltonian is
64: self-adjoint. The possible boundary conditions for an operator are encoded in
65: the choice of its domains, which are classified by the self-adjoint extensions
66: \cite{reed} of the operator. We are thus naturally led to the study of the
67: self-adjoint extensions of the Calogero model. It may be noted that
68: self-adjoint extensions are known to play important roles in a variety of
69: physical contexts
70: including Aharonov-Bohm effect \cite{gerbert}, two and three dimensional
71: delta function potentials \cite{jackiw}, anyons \cite{manuel}, anomalies
72: \cite{esteve}, $\zeta$-function renormalization \cite{falo},
73: particle statistics in one dimension \cite{bal} and black
74: holes \cite{trg}. Indeed, the self-adjoint extensions of the rational
75: Calogero model in absence of the confining interaction has recently
76: been studied \cite{we}.
77:
78: In this Letter we shall show that the Calogero model in presence of the
79: confining interaction can indeed be consistently quantized with choices of
80: boundary conditions different than what was considered in Ref. \cite{calo3}.
81: It will be shown that under certain conditions, the corresponding Hamiltonian
82: admits self-adjoint extensions labelled by $e^{iz}$ where
83: $z \in R$ (mod $2 \pi$). The parameter $z$ classifies the possible boundary
84: conditions for which the Hamiltonian is self-adjoint.
85: In any given situation, the physical interpretation of $z$
86: depends on the details of the particular problem \cite{reed}. For example,
87: in a description of black holes in terms of the Calogero model \cite{black},
88: $z$ is related to the mass and entropy of the black hole \cite{trg}.
89: As another example, in the realization of generalized exclusion statistics
90: within the framework of Calogero model \cite{poly}, $z$ is related to the
91: statistical parameter.
92:
93: As a consequence of the self-adjoint extensions, we get a new
94: class of bound states for the Calogero model. Unlike the known spectrum
95: obtained by Calogero \cite{calo3}, we find infinitely many
96: energy states which are not equispaced except for special
97: values of $z$. Moreover, the spectrum in general includes a single
98: negative energy bound state. This is the first time in the
99: literature that the existence of non-equispaced energy levels
100: with a single negative energy state have been found for the
101: the rational Calogero model. The spectrum depends explicitly on the value
102: of $z$, leading to inequivalent quantizations of this system.
103:
104: The Hamiltonian of the rational Calogero model is given by
105: \be
106: H = - \sum^{N}_{i=1} \frac{{\partial}^2}{\partial x_i^2} +
107: \sum_{i \neq j} \left [ \frac{a^2 - \frac{1}{4}}{(x_i - x_j)^2} +
108: \frac{\Omega^2}{16} (x_i - x_j)^2 \right ]
109: \label{e0}
110: \ee
111: where $a$, $\Omega$ are constants,
112: $x_i$ is the coordinate of the $i^{\rm th}$ particle and
113: units have been chosen such that $2 m {\hbar}^{- 2} = 1$.
114: We are interested in finding normalizable solutions of the
115: eigenvalue problem
116: \be
117: H \psi = E \psi.
118: \label{e1}
119: \ee
120: Following \cite{calo3}, we consider the above eigenvalue equation
121: in a sector of configuration
122: space corresponding to a definite ordering of particles given by
123: $x_1 \geq x_2 \geq
124: \cdots \geq x_N$. The translation-invariant
125: eigenfunctions of the Hamiltonian $H$ can be written as
126: \be
127: \psi = \prod_{i <j} \left (x_i - x_j \right )^{a + \frac{1}{2}} \
128: \phi (r) \ P_k (x),
129: \label{e2}
130: \ee
131: where $x \equiv (x_1, x_2, \dots, x_N)$,
132: \be
133: r^2 = \frac{1}{N} \sum_{i < j} (x_i - x_j)^2 \ \
134: \label{e3}
135: \ee
136: and $P_k (x)$ is a translation-invariant as well as homogeneous
137: polynomial of degree $k(\geq 0)$ which satisfies the
138: equation
139: \be
140: \left[ \sum^{N}_{i=1}\frac{{\partial}^2}{\partial x_i^2}
141: + \sum_{i \neq j} \frac{ 2 (a + \frac{1}{2}) }{(x_i -
142: x_j)} \frac{{\partial}}{\partial x_i}
143: %\frac{{\partial}}{\partial x_j} \right)
144: \right] P_k (x) = 0.
145: \label{e4}
146: \ee
147: The existence of
148: complete solutions of (\ref{e4}) has been discussed by Calogero
149: \cite{calo3}.
150: Substituting Eqn. (\ref{e2}) in Eqn. (\ref{e1}) and using Eqns. (\ref{e3}-
151: \ref{e4}) we get
152: \be
153: \tilde{H} \phi
154: = E \phi,
155: \label{e5}
156: \ee
157: where
158: \be
159: \tilde{H} = \left [ - \frac{d^2}{dr^2} - (1 + 2 \nu )
160: \frac{1}{r} \frac{d}{d r} + w^2 r^2 \right ]
161: \label{e9}
162: \ee
163: with $w^2 = \frac{1}{8} \Omega^2 N $ and
164: \be
165: \nu = k + \frac{1}{2}(N - 3) + \frac{1}{2} N (N-1)(a + \frac{1}{2}).
166: \label{e6}
167: \ee
168: $\tilde{H}$ is the effective Hamiltonian in the ``radial'' direction.
169: Following \cite{we}, it can be easily shown that $\phi(r) \in L^2[R^+,d\mu]$
170: where the measure is given by $d\mu = r^{1 + 2 \nu} dr$.
171:
172: The Hamiltonian $\tilde{H}$ is a symmetric (Hermitian) operator on the domain
173: $D(\tilde{H}) \equiv \{\phi (0) = \phi^{\prime} (0) = 0,~
174: \phi,~ \phi^{\prime}~ {\rm absolutely~ continuous} \} $.
175: To determine whether
176: $\tilde{H}$ is self-adjoint \cite{reed} in $D(\tilde{H})$,
177: we have to first look for square integrable solutions of the equations
178: \be
179: \tilde{H^*} \phi_{\pm} = \pm i \phi_{\pm},
180: \label{e10}
181: \ee
182: where $\tilde{H^*}$ is the adjoint of $\tilde{H}$ (note that $\tilde{H^*}$
183: is given by the same differential operator as $\tilde{H}$ although their
184: domains might be different).
185: Let $n_+(n_-)$ be the total number of square-integrable, independent solutions
186: of (\ref{e10})
187: with the upper (lower) sign in the right hand side. Now $\tilde{H}$
188: falls in one of the following categories \cite{reed} :\\
189: 1) $\tilde{H}$ is (essentially) self-adjoint iff
190: $( n_+ , n_- ) = (0,0)$.\\
191: 2) $\tilde{H}$ has self-adjoint extensions iff $n_+ = n_- \neq 0$.\\
192: 3) If $n_+ \neq n_-$, then $\tilde{H}$ has no self-adjoint extensions.\\
193:
194: The solutions of Eqn. (\ref{e10}) are given by
195: \be
196: \phi_{\pm} (r) = {\mathrm e}^{- \frac{w r^2}{2}}
197: U \left ( d_\pm, c, w r^2 \right ),
198: \label{e11}
199: \ee
200: where $d_{\pm} = \frac{1+ \nu }{2} \mp \frac{i}{4 w}$, $c = 1+ \nu $ and $U$
201: denotes the confluent hypergeometric function of the
202: second kind \cite{abr}. The asymptotic behaviour of $U$ \cite{abr} together
203: with the exponential factor in Eqn. (\ref{e11}) ensures that
204: $\phi_{\pm} (r)$ vanish at infinity. The solution in Eqn. (10) have
205: different short distance behaviour for $\nu \neq 0$ and $\nu = 0$.
206: From now onwards, we shall restrict our discussion to the case for
207: $\nu \neq 0$, the analysis for $\nu = 0$ being similar. When $\nu \neq 0$,
208: $U(d_{\pm},c,w r^2)$ can be written as
209: \be
210: U \left ( d_{\pm},c,w r^2 \right )=
211: C
212: \bigg [ \frac{M \left ( d_{\pm}, c, w r^2 \right )}
213: {\Gamma (b_{\pm}) \Gamma (c)}
214: - \left ( w r^2 \right )^{1 -c}
215: \frac{M \left ( b_{\pm}, 2-c, w r^2 \right )}
216: {\Gamma (d_{\pm}) \Gamma (2-c)} \bigg ],
217: \label{e12}
218: \ee
219: where $b_\pm = \frac{1-\nu}{2} \mp \frac{i}{4w}$, $C = \frac{\pi}{{\mathrm
220: sin} (\pi + \nu \pi)}$ and
221: $M$ denotes the confluent hypergeometric function of the first kind
222: \cite{abr}.
223: In the limit $r \rightarrow 0$, $M(d_{\pm},c,wr^2) \rightarrow 1$.
224: This together with Eqns. (\ref{e11}) and (\ref{e12}) implies that as
225: $r \rightarrow 0$,
226: \be
227: |\phi_{\pm} (r)|^2 d \mu \rightarrow
228: \left [ A_1 r^{(1 + 2 \nu )} + A_2 r + A_3 r^{(1 - 2 \nu )} \right ] dr,
229: \label{e14}
230: \ee
231: where $A_1, A_2$ and $A_3$ are constants independent of $r$.
232: From Eqn. (\ref{e14}) it is
233: now clear that in the limit $r \rightarrow 0$, the functions
234: $\phi_{\pm} (r)$ are not square-integrable if ${\mid \nu \mid} \geq 1 $.
235: In that case, $n_+ = n_- = 0$ and $\tilde{H}$ is essentially self-adjoint
236: in the domain $D(\tilde{H})$. However, if either $ 0 < \nu < 1$ or
237: $ -1 < \nu < 0$, the functions $\phi_{\pm} (r)$ are indeed square-integrable.
238: Thus if $\nu$ lies in these ranges, we have $n_+ = n_- = 1$ and
239: Hamiltonian $\tilde{H}$ is not self-adjoint in $D(\tilde{H})$ but
240: admits self-adjoint extensions.
241: The domain $D_z(\tilde{H})$ in which $\tilde{H}$ is self-adjoint contains all
242: the elements of $D(\tilde{H})$ together with elements of the form
243: $\phi_+ + {\mathrm e}^{iz} \phi_-$, where $ z \in R$ (mod $2 \pi$) \cite{reed}.
244: We can similarly show that $n_+ = n_- = 1$ for $\nu = 0$ as well. Thus the
245: self-adjoint extensions of this model exist when $-1 < \nu < 1$. It may be
246: noted that the values of $n_+$ and $n_-$ as well as the allowed range of
247: $\nu$ obtained above is the same as that found in Ref. \cite{we}, which
248: discussed the Calogero model without the confining term. In both these
249: cases,
250: the existence of the self-adjoint extension is essentially determined by the
251: nature of the singularity at $r=0$. However, the domain
252: $D_z(\tilde{H})$ obtained above is very different from the corresponding
253: domain found in Ref. \cite{we}. This is due to the fact that the
254: presence of the confining
255: potential affects the expressions of $\phi_{\pm}(r)$, which in turn determine
256: the allowed domain of $\tilde{H}$. As discussed below,
257: this difference in the structure of the domains leads to a completely
258: different spectrum in presence of the confining potential.
259:
260: The range of $\nu$ required for the existence of the self-adjoint extension
261: together with Eqn. (8) implies that for given
262: values of $N$ and $k$, $a + \frac{1}{2}$ must lie on the range
263: \be
264: - \frac{ N - 1 + 2 k}{N ( N-1)} < a + \frac{1}{2} < -
265: \frac{ N - 5 + 2 k}{N ( N-1)}.
266: \label{pp}
267: \ee
268: For $N \geq 3$, we have the following
269: classifications of the boundary conditions depending on the value of the
270: parameter $a+ \frac{1}{2}$.\\
271: (i) $a + \frac{1}{2}\geq \frac{1}{2}$ : This corresponds to the boundary
272: condition considered by Calogero for
273: which both the wave-function and the current vanish as $x_i \rightarrow x_j$.
274: In this case, $\nu > 1$ for all values of $k \geq 0$.
275: The corresponding Hamiltonian is essentially self-adjoint in the domain
276: $D(\tilde{H})$, leading to a unique quantum theory.\\
277: (ii) $ 0 < a + \frac{1}{2} < \frac{1}{2}$ :
278: The wave-function vanishes in the limit $x_i \rightarrow x_j$, though
279: the current may show a divergent behaviour in the same limit. Such a boundary
280: condition on the wave-function is quite similar to what one encounters for
281: strongly repulsive $\delta$-function Bose gas. In this case
282: $\nu$ is positive and $k$ must be equal to zero so that $\nu$ may belong to
283: the range
284: $0 < \nu < 1$. The corresponding constraint on $a + \frac{1}{2}$ is given by
285: $0 < a + \frac{1}{2} < \frac{5 - N }{N(N-1)}$, which can only be
286: satisfied for $N = 3$ and $4$.\\
287: (iii) $ -\frac{1}{2} < a + \frac{1}{2} < 0$ :
288: The lower bound on $a+ \frac{1}{2}$ is obtained from the condition that the
289: wavefunction be square-integrable. The parameter $a + \frac{1}{2} $ in this
290: range leads to a singularity in the wavefunction resulting from the
291: coincidence of any two or more particles.
292: Using permutation symmetry, such an eigenfunction can be extended to the
293: whole of configuration space, although not in a smooth fashion.
294: The new quantum states in this case exist for arbitrary $N$ and even for
295: non-zero values of $k$. In fact,
296: imposing the condition that the upper bound on $a + \frac{1}{2}$ should be
297: greater than $-\frac{1}{2}$, we find from Eqn. (13) that $k$ is restricted as
298: $k < \frac{1}{4} \left ( N^2 - 3 N + 10 \right )$. It can also be
299: shown that there are only two allowed values of $k$ when both $N$ and
300: $a + \frac{1}{2}$ are kept fixed.
301:
302: In order to determine the spectrum
303: we note that the solution to Eqn. (\ref{e5}) which is
304: bounded at infinity is given by
305: \be
306: \phi (r) = B {\mathrm e}^{- \frac{w r^2}{2}}
307: U(d,c,w r^2),
308: \label{e15}
309: \ee
310: where $d = \frac{ 1+ \nu }{2} - \frac{E}{4 w}$ and $B$ is a constant.
311: In the limit $r \rightarrow 0$,
312: \be
313: \phi (r) \rightarrow
314: B C \left [ \frac{1}{\Gamma (b) \Gamma (c)}
315: - \frac{w ^{-\nu} r^{-2 \nu }}{\Gamma (d) \Gamma (2-c)} \right ],
316: \label{e16}
317: \ee
318: where $b = \frac{1 - \nu}{2} - \frac{E}{4w}$.
319: On the other hand, as $r \rightarrow 0$,
320: \be
321: \phi_+ + {\mathrm e}^{iz} \phi_- \rightarrow C \bigg [
322: \frac{1}{\Gamma (c)} \left (
323: \frac{1}{\Gamma (b_+) }
324: +\frac{{\mathrm e}^{iz}}{\Gamma (b_- ) } \right )
325: - \frac{ w ^{-\nu} r^{-2 \nu }}{\Gamma (2-c)}
326: \left ( \frac{1}{\Gamma (d_+) }
327: +\frac{{\mathrm e}^{iz}}{\Gamma (d_-) } \right ) \bigg ].
328: \label{e17}
329: \ee
330: If $\phi (r) \in D_z(\tilde{H})$, then the coefficients of different powers of
331: $r$ in Eqns. (\ref{e16}) and (\ref{e17}) must match. Comparing the coefficients
332: of the constant term and $r^{-2 \nu }$ in
333: Eqns. (\ref{e16}) and (\ref{e17}) we get
334: \be
335: f(E) \equiv \frac{\Gamma \left ( \frac{ 1 - \nu }{2} - \frac{E}{4 w} \right )}
336: {\Gamma \left (\frac{ 1 + \nu }{2} - \frac{E}{4 w} \right ) } =
337: \frac{\xi_2 {\mathrm cos}(\frac{z}{2} - \eta_1)}
338: {\xi_1 {\mathrm cos}(\frac{z}{2} - \eta_2)},
339: \label{e18}
340: \ee
341: where $\Gamma \left ( \frac{ 1 + \nu }{2} + \frac{i}{4 w} \right )
342: \equiv \xi_1 {\mathrm e}^{i \eta_1}$
343: and
344: $\Gamma \left ( \frac{ 1 - \nu }{2} + \frac{i}{4 w} \right )
345: \equiv \xi_2 {\mathrm e}^{i \eta_2}$.
346: For given values of the parameters $\nu$ and $w$,
347: the bound state energy $E$ is obtained from Eqn. (\ref{e18}) as a function of
348: $z$. The corresponding eigenfunctions are
349: obtained by substituting $\phi(r)$ from Eqn. (\ref{e15}) into Eqn. (3).
350: Different choices of $z$ thus leads to
351: inequivalent quantizations of the many-body Calogero model.
352: Moreover from Eqn. (\ref{e18}) we see that for fixed value of $z$,
353: the Calogero model with
354: parameters $(w, \nu)$ and $(w,-\nu)$ produces identical energy
355: spectrum although the corresponding wavefunctions are different.
356:
357: \begin{figure}
358: \begin{center}
359: \includegraphics[width=7cm]{graph1.eps}
360: \end{center}
361: \caption { \label{fig1} A plot of Eqn. (\ref{e18}) using Mathematica
362: with $w = 0.25$, $\nu = 0.25 $ and
363: $z = -1.5$. The horizontal straight line corresponds the value of the r.h.s of
364: Eqn. (\ref{e18}).}
365: \end{figure}
366:
367: The following features about the spectrum may be noted: \\
368: 1) We have obtained the spectrum analytically when the r.h.s.
369: of Eqn. (\ref{e18})
370: is either 0 or $\infty$. When the
371: r.h.s. of Eqn. (\ref{e18}) is $0$, we must have the situation where
372: $\Gamma \left (\frac{ 1 + \nu }{2} - \frac{E}{4 w} \right )$ blows up, i.e.
373: $E_n = 2 w ( 2 n + \nu + 1)$ where $n$ is a positive integer.
374: This happens for the special choice of $z = z_1 = \pi + 2 \eta_1$.
375: These
376: eigenvalues and the corresponding eigenfunctions are analogous to those
377: found by Calogero although for a different parameter range.
378: Similarly, when the r.h.s. of Eqn. (\ref{e18}) is $\infty$,
379: an analysis similar
380: to the one above shows that $E_n = 2 w ( 2 n - \nu + 1)$.
381: This happens for the special value of $z$ given by
382: $z = z_2= \pi + 2 \eta_2$. \\
383: 2) For choices of $z$ other than $z_1$ or $z_2$, the nature of the spectrum
384: can be understood from Figure 1, which is a plot of Eqn. (\ref{e18}) for
385: specific values of $\nu, z$ and $w$.
386: In that plot, the horizontal straight line corresponds to the
387: r.h.s of Eqn. (\ref{e18}). The energy eigenvalues are obtained from the
388: intersection of $f(E)$ with the horizontal straight line.
389: Note that the spectrum
390: generically consists of infinite number of positive energy solutions and at
391: most one negative energy solution. The existence of the negative energy
392: states can be understood in the following way. For large negative values of
393: $E$, the asymptotic value of $f(E)$ is given by $(\frac{E}{4w})^{- \nu}$
394: \cite{abr}, which monotonically tends to 0 or $+ \infty$
395: for $\nu > 0$ or $\nu < 0$ respectively. When $\nu > 0$, the negative energy
396: state will exist provided r.h.s. of Eqn. (\ref{e18}) lies between 0 and
397: $\frac{\Gamma ( \frac{ 1 - \nu }{2})}
398: {\Gamma (\frac{ 1 + \nu }{2} )}$. Similarly, when $\nu < 0$, the
399: negative energy state will exist when the r.h.s. of Eqn. (\ref{e18})
400: lies between
401: $\frac{\Gamma ( \frac{ 1 - \nu }{2})}
402: {\Gamma (\frac{ 1 + \nu }{2} ) }$ and $+ \infty$.
403: For any given values of $\nu$ and $w$, the
404: position of the horizontal straight line in Fig. 1 can always be adjusted
405: to lie anywhere between $-\infty$ and $+\infty$ by
406: suitable choices of $z$. Thus the spectrum would always contain a negative
407: energy state for some choice of the parameter $z$. \\
408: 3) Contrary to the spectrum of the rational Calogero model,
409: the energy spectrum obtained from Eqn. (\ref{e18}) is not equispaced for
410: finite
411: values of $E$ and for generic values of $z$. For example, it is seen from
412: Eqn. (\ref{e18}) that the ratio
413: \be
414: \frac{f(E + 4w)}{f(E)} = \frac{\frac{E}{4w} + \frac{1 - \nu}{2}}
415: {\frac{E}{4w} + \frac{1 + \nu}{2}}
416: \ee
417: in general is not unity except when $E \rightarrow \infty$.
418: This may seem surprising with the presence of $SU(1,1)$ as
419: the spectrum generating algebra in this system \cite{fub}, which demands
420: that the eigenvalues be evenly spaced. In order to address this issue, we
421: consider the action of the dilatation generator $D = \frac{1}{2} \left ( r
422: \frac{d}{dr} + \frac{d}{dr}r \right )$ on an element
423: $\phi(r) = \phi_+(r) + {\mathrm e}^{iz} \phi_-(r)$.
424: In the limit $r \rightarrow 0$, we have
425: \be
426: D \phi = \frac{C}{2} \bigg [
427: \frac{1}
428: {\Gamma (c)} \left (
429: \frac{1}{\Gamma (b_+) }
430: +\frac{{\mathrm e}^{iz}}{\Gamma (b_-) } \right )
431: - \frac{r^{-2 \nu } (1 - 4 \nu) }
432: {\Gamma (2-c)}
433: \left ( \frac{1}{\Gamma (d_+) }
434: +\frac{{\mathrm e}^{iz}}{\Gamma (d_-) } \right ) \bigg ].
435: \label{e19}
436: \ee
437: We therefore see that $D \phi (r) \in D_z (\tilde{H})$ only for $z = z_1$ or
438: $z= z_2$.
439: Thus the generator of dilatations does not in
440: general leave the domain of the Hamiltonian invariant
441: \cite{dh,esteve,jackiw,we}. Consequently, $SU(1,1)$ cannot be implemented as
442: the spectrum generating algebra except for $z = z_1$, $z_2$.\\
443: 4) For $N \geq 3$, the range of $a + \frac{1}{2}$ for which the
444: new quantum states have been found is different from what was used in
445: Ref. \cite{calo3}.
446: The $N=2$ Calogero model however admits new quantum states even in
447: the range of $a + \frac{1}{2}$ considered in Ref. \cite{calo3}.
448: When $N=2$, $k$ must be equal to
449: zero and Eqn. (8) gives $\nu = a$. In this case,
450: the system therefore admits self-adjoint
451: extensions and new quantum states when $-1 < a < 1$. It may be noted that the
452: eigenvalue problem for $N=2$ was solved in Ref.\cite{calo3} with the
453: condition that $a > 0$. Thus when $0 < a < 1$,
454: our analysis predicts a larger family of solutions labelled by the
455: parameter $z$. This set of solutions reduces to that
456: found in Ref. \cite{calo3} for $z = z_1$.\\
457: 5) It may be interesting to compare the spectrum obtained above with that
458: found in Ref. \cite{we}, where the
459: self-adjoint extension of the Calogero model
460: without the confining term was discussed. In the latter case, the spectrum
461: consists of at most
462: one negative energy bound state and infinite number of scattering states
463: with momentum dependent phase shifts. In the presence of the confining
464: potential, as discussed above, we get at most one negative energy bound
465: state and an infinite number of positive energy bound states which are in
466: general not equispaced. It may also be noted that the spectrum found in Ref.
467: \cite{we} cannot be obtained as the $w \rightarrow 0$ limit of that
468: obtained in this paper. This is due to the fact that Eqn. (17), which
469: determines the spectrum in the present case, becomes singular in the
470: $w \rightarrow 0$ limit.
471:
472: In conclusion, we have presented a new quantization scheme for the rational
473: Calogero model. The non-equispaced nature of the energy levels and the
474: existence of a negative energy bound state are some of the salient features
475: that emerge from our analysis. It is expected that the generaized exclusion
476: statistics parameter\cite{poly} of this model would be a function of both
477: $z$ and $\nu$, since the energy spectrum depends on these parameters.
478: We can ascertain this for the special case of $z=z_1$ and $z_2$,
479: when Eq. (\ref{e18}) can be solved exactly.
480: The generalized exclusion statistics is believed to
481: play an important role in one dimensional non-fermi liquids as well as in
482: the edge excitations in the fractional quantum Hall effect.
483: Thus, it would be interesting to
484: investigate the generalized exclusion statistics and the thermodynamic
485: properties of Calogero models for arbitrary $z$ with the spectrum as
486: described here.
487:
488:
489: \noindent
490: \acknowledgments{ The work of PKG is supported(DO No. SR/FTP/PS-06/2001) by
491: the SERC, DST, Govt. of India, under the Fast Track Scheme for Young
492: Scientists:2001-2002. }
493:
494: \begin{thebibliography}{99}
495: \bibitem{calo3} F. Calogero, Jour. Math. Phys. {\bf 10}, 2191 (1969); Jour.
496: Math. Phys. {\bf 10}, 2197 (1969); Jour. Math. Phys. {\bf 12}, 419 (1971).
497:
498: \bibitem{pr} M. A. Olshanetsky and A. M. Perelomov, Phys. Rep. {\bf 71}, 314
499: (1981); {\it ibid} {\bf 94}, 6 (1983).
500:
501: \bibitem{poly} M. V. N. Murthy and R. Shankar, Phys. Rev. Lett. {\bf 73},
502: 3331 (1994); Z. N. C. Ha, {\it Quantum Many-Body Systems in One
503: Dimension}, Series on Advances in Statistical Mechanics, Vol. 12, (World
504: Scientific, 1996); A. P. Polychronakos, hep-th/9902157;
505: B. Basu-Mallick and A. Kundu, Phys. Rev. {\bf B62}, 9927
506: (2000).
507:
508: \bibitem{qhe} H. Azuma and S. Iso, Phys. Lett. {\bf B331}, 107(1994).
509:
510: \bibitem{ll} N. Kawakami and S.-K. Yang, Phys. Rev. Lett. {\bf 67}, 2493
511: (1991).
512:
513: \bibitem{rmt} B. D. Simons, P. A. Lee, and B. L. Altshuler,
514: Phys. Rev. Lett. {\bf 72}, 64(1994); S. Jain, Mod. Phys. Lett. {\bf A11},
515: 1201(1996).
516:
517: \bibitem{qet} C. W. J. Beenakker and B. Rejaei, Phys. Rev. {\bf B49}, 7499
518: (1994); M. Caselle, Phys. Rev. Lett. {\bf 74}, 2776 (1995).
519:
520: \bibitem{hs} F. D. M. Haldane, Phys. Rev. Lett. {\bf 60}, 635 (1988);
521: B. S. Shastry, Phys. Rev. Lett. {\bf 60}, 639 (1988); A. P. Polychronakos,
522: Phys. Rev. Lett. {\bf 70}, 2329 (1993).
523:
524: \bibitem{sw} E. D'Hoker and D. H. Phong, hep-th/9912271; A. Gorsky and
525: A. Mironov, hep-th/0011197; A. J. Bordner, E. Corrigan and R. Sasaki, Prog.
526: Theor. Phys. {\bf 100}, 1107 (1998).
527:
528: \bibitem{black}
529: G. W. Gibbons and P. K. Townsend, Phys. Lett. {\bf B454},
530: 187 (1999).
531:
532: \bibitem{brink} A. P. Polychronakos, Phys. Rev. Lett. {\bf 69}, 703 (1992);
533: L. Brink, T. H. Hansson and M. A. Vasiliev, Phys. Lett.
534: {\bf B286}, 109 (1992); N. Gurappa and P. K. Panigrahi, Phys. Rev. {\bf B
535: 59}, R2490 (1999).
536:
537: \bibitem{reed} M. Reed and B. Simon, {\it Methods of Modern Mathematical
538: Physics}, volume 2, (Academic Press, New York, 1972).
539:
540: \bibitem{gerbert} P. Gerbert, Phys. Rev. {\bf D 40}, 1346 (1989).
541: \bibitem{jackiw} R. Jackiw in M.A.B. Beg Memorial Volume, A. ALi and P.
542: Hoodbhoy, eds. (World Scientific, Singapore, 1991).
543: \bibitem{manuel} C. Manuel and R. Tarrach, Phys. Lett. {\bf B 268}, 222
544: (1991); M. Bourdeau and R. D. Sorkin, Phys. Rev. {\bf D 45}, 687 (1992).
545: \bibitem{esteve} J. G. Esteve, Phys. Rev. {\bf D 34}, 674 (1986); J. G.
546: Esteve, Phys. Rev. {\bf D 66}, 125013 (2002).
547: \bibitem{falo} H. Falomir, P. A. G. Pisani and A. Wipf,
548: Jour. Phys. {\bf A 35}, 5427 (2002).
549: \bibitem{bal} C. Aneziris, A. P. Balachandran and Diptiman Sen, Int. Jour.
550: Mod. Phys. {\bf 6}, 4721 (1991).
551: \bibitem{trg} T. R. Govindarajan, V. Suneeta and S. Vaidya,
552: Nucl. Phys. {\bf B583}, 291 (2000); D. Birmingham, Kumar S. Gupta and
553: Siddhartha Sen, Phys. Lett. {\bf B505}, 191 (2001);
554: Kumar S. Gupta and Siddhartha Sen, Phys. Lett. {\bf B 526}, 121 (2002);
555: Kumar S. Gupta, hep-th/0204137.
556: \bibitem{we} B. Basu-Mallick and K. S. Gupta, Phys. Lett. {\bf A292},
557: 36(2001);
558: B. Basu-Mallick, P. K. Ghosh and K. S. Gupta, hep-th/0207040 (to appear in
559: Nucl. Phys. {\bf B}.)
560: \bibitem{abr} {\it Handbook of Mathematical Functions},
561: M. Abromowitz and I. A. Stegun (Dover Publications, New York, 1974).
562: \bibitem{fub} V. de Alfaro, S. Fubini and G. Furlan, Nuovo Cim. {\bf 34A},
563: 569 (1976).
564: \bibitem{dh} E. D'Hoker and L. Vinet, Comm. Math. Phys. {\bf 97}, 391
565: (1985).
566: \end{thebibliography}
567: \end{document}
568:
569: