1: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
2: %% Trim Size: 9.75in x 6.5in
3: %% Text Area: 8in (include Runningheads) x 5in
4: %% ws-mpla.tex : 20-9-2006
5: %% TeX file to use with ws-mpla.cls written in Latex2E.
6: %% The content, structure, format and layout of this style file is the
7: %% property of World Scientific Publishing Co. Pte. Ltd.
8: %% Copyright 1995, 2002 by World Scientific Publishing Co.
9: %% All rights are reserved.
10: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
11:
12:
13: \documentclass{ws-mpla}
14:
15:
16: \voffset=-1.2mm
17:
18: \begin{document}
19:
20: \markboth{A. M. Gavrilik and A. P. Rebesh}
21: {A $q$-oscillator with 'accidental' degeneracy of energy levels}
22:
23: %%%%%%%%%%%%%%%%%%%%% Publisher's Area please ignore %%%%%%%%%%%%%%
24: % \catchline{}{}{}{}{}
25: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
26:
27: \title{\bf A $q$-OSCILLATOR WITH 'ACCIDENTAL' DEGENERACY OF ENERGY LEVELS}
28:
29: \author{\footnotesize A. M. GAVRILIK}
30:
31: \address{Bogolyubov Institute for Theoretical Physics, Kiev 03680, Ukraine\\
32: omgavr@bitp.kiev.ua
33: }
34:
35: \author{A. P. REBESH}
36:
37: \address{Bogolyubov Institute for Theoretical Physics, Kiev 03680, Ukraine\\
38: omgavr@bitp.kiev.ua
39: }
40:
41: \maketitle
42:
43: \pub{Received (Day Month Year)}{Revised (Day Month Year)}
44:
45: \begin{abstract}
46: We study main features of the exotic case of $q$-deformed oscillators
47: (so-called Tamm-Dancoff cutoff oscillator) and find some special
48: properties:
49: (i) degeneracy of the energy levels $E_{n_1} = E_{n_1+1}$,
50: $n_1\ge 1$, at the {\em real value} $q=\sqrt{\frac{n_1}{n_1+2}}$ of
51: deformation parameter, as well as the occurrence of other degeneracies
52: $E_{n_1} = E_{n_1+k},\ $ for $k \geq 2,\ $ at the corresponding values
53: of $q$ which depend on both $n_{1}$ and $k$;
54: (ii) the position and momentum operators $X$ and $P$ {\em commute
55: on the state} $|n_1\rangle$ if $q$ is fixed as $q={\frac{n_1}{n_1+1}}$,
56: that implies unusual uncertainty relation;
57: (iii) two commuting copies of the creation, annihilation,
58: and number operators of this $q$-oscillator generate %$N_1, N_2$
59: the corresponding $q$-deformation of the {\em non-simple} Lie
60: algebra $su(2)\oplus u(1)$ whose nontrivial $q$-deformed
61: commutation relation is: $[J_+, J_-]=2 J_0~q^{2J_3-1}$ where
62: $J_0\equiv \frac12 (N_1-N_2)$ and $J_3\equiv\frac12 (N_1+N_2)$.
63: \keywords{$q$-deformed oscillator; degeneracy of energy levels;
64: $q$-deformed algebra.}
65: \end{abstract}
66:
67: \ccode{PACS Nos.: 02.20.Uw, 03.65.G, 03.65.Fd, 05.30.Pr}
68:
69: \section{Introduction}
70:
71: Since the famous papers\cite{Bied,Macf},
72: the diverse aspects and applications of
73: Biedenharn-Macfarlane (BM) $q$-deformed
74: oscillators have become a very popular subject.
75: This same can be said about the somewhat earlier introduced
76: and slightly more simple Arik-Coon (AC) version\cite{AC}
77: of $q$-deformed oscillator.
78: The principal difference between these two versions is that
79: unlike the latter, the former one admits not only real,
80: but also the phase-like complex values of the deformation
81: parameter $q$, and this has important %very unusual
82: consequences.
83: For instance, a possibility of 'accidental' degeneracies
84: does appear in case of $q$ being a %which is equal to some
85: root of unity, popular values %that is known to be admissible
86: for the BM type $q$-oscillator.
87: Namely, for $q=\exp(\frac{{\rm i}\pi} {2n+2})$ the following
88: two neighboring energy levels coincide: $E_{n+1}=E_{n}$.
89: This and other coincidences lead to a kind of periodicity and
90: naturally make the corresponding
91: phase space both discrete and finite\cite{bonats}.
92:
93: The question arises: {\it is it possible that analogous property of
94: 'accidental' degeneracy may occur at real value(s)} of the
95: $q$-parameter for, maybe, some other type of $q$-deformed
96: oscillator? Say, is it possible that fixation of the deformation
97: parameter by some real value(s), e.g. $q=\sqrt{0.98}$ or
98: $q=1.02^{-1/2}$, does provide degeneracy of respective energy
99: levels?
100: The goal of this note is %consists %presentation just
101: to present an analysis of such an exotic $q$-oscillator
102: and to show it possesses the already mentioned
103: degeneracy properties, and also a couple of other
104: unusual properties.
105: This type of $q$-oscillator has first appeared in Ref.~\refcite{Odaka,Jagan}
106: where it was called the 'Tamm-Dancoff cutoff'
107: deformed oscillator. Representation-theoretic,
108: unified view, and generalized ($f$-oscillator)
109: aspects of the $q$-oscillator algebra
110: were studied in Ref.~\refcite{Ques1,Melj,Mizrahi}.
111: Our interest to this $q$-oscillator has emerged
112: in the course of our recent study of possible application\cite{SIGMA}
113: of the (set of) two-parameter $q,\!p$-deformed
114: oscillators (those include the already mentioned
115: BM and AC $q$-oscillators as the two particular
116: $p=q^{-1}$ and $p=1$ limiting cases). Namely,
117: the $q,\!p$-deformed oscillators have been utilized
118: in the framework of the model of two-parameter deformed $q,\!p$-Bose
119: gas model, for which the explicit formulas for
120: (intercepts of) general $n$-particle momentum correlation
121: functions have been obtained\cite{AdGa}.
122: Note that these results generalize previously known
123: formulas for two-particle correlations in the AC and BM
124: versions of $q$-Bose gas model.
125:
126: \section{Setup on the $q$-oscillator}
127:
128: The {$q$-deformed oscillator} of our main interest here
129: is defined by the relations:
130: %
131: \begin{equation}
132: a a^\dagger - q a^\dagger a = q^{N} \ ,
133: \end{equation}
134: \begin{equation}
135: [{ N},a]=- a \ , \qquad
136: [{ N},a^\dagger]= a^\dagger \ . %\qquad \qquad \quad
137: \end{equation}
138: %It is unclear, at least from first sight, how
139: To state the crucially important relation between
140: the number operator $N$, on one side, and the operators
141: $a^\dagger a$ or $a a^\dagger$, on the other, we appeal
142: to the well-known generalized two-parameter family\cite{Chakra}
143: of $p,q$-deformed oscillators defined by the relations
144: %
145: \begin{equation}
146: A A^\dagger - q\ A^\dagger A = p^{\cal N} \ , \qquad\quad \quad
147: A A^\dagger - p\ A^\dagger A = q^{\cal N} \ ,
148: \end{equation}
149: supplemented with two relations involving ${\cal N}$
150: (completely analogous to (2)).
151:
152:
153: From the pair of relations (3) invariant under $q\!\leftrightarrow\!p$,
154: the formulas follow:
155: \begin{equation}
156: A^\dagger A = [\![{\cal N}]\!]_{qp} \ ,
157: \quad
158: %{\rm with}
159: \quad
160: A A^\dagger = [\![{\cal N}+1]\!]_{qp} \ ,
161: \quad
162: %{\rm with}
163: \quad
164: [\![X]\!]_{q,p} \equiv \frac{ q^{X}-p^{X} }{ q-p } \ .
165: \end{equation}
166: At $p\!=\!1$ this two-parameter system reduces
167: to the AC-type\cite{AC}
168: of $q$-bosons, while putting $p=q^{-1}$ yields
169: the other already mentioned case\cite{Bied,Macf}
170: of $q$-oscillators.
171: %Moreover,
172:
173: At $p=q$, each of the relations in (3) turns into
174: the single relation (1) what precludes connecting
175: of $A^\dagger A$ or $A A^\dagger $ immediately
176: with ${\cal N}$.
177: Besides, the $q,\!p$-bracket in (4) becomes undefined if $p=q$,
178: for $X$ being either an operator or a number.
179: Only for non-negative integer $k$ it makes sense, since
180: \begin{equation}
181: [\![k]\!]_{q,p} = \frac{ q^{k}-p^{k} }{ q-p } =
182: %q^{k-1}+q^{k-2}p+...+q p^{k-2}+p^{k-1}
183: \sum_{r=0}^{k-1} q^{k-1-r}p^r =
184: q^{k-1}\sum_{r=0}^{k-1}q^{-r} p^r \ \ \ \
185: \stackrel{p\to q}{\Longrightarrow} \ \ \ \ k q^{k-1}\ .
186: \end{equation}
187: Then, let us take the analogue of the last part of relation (5)
188: as the (at first, formal) definition of the new $q$-bracket
189: in all cases, both for operators and for numbers:
190: \begin{equation}
191: \{X\}_q \equiv X q^{X-1} \ .
192: \end{equation}
193: %
194: To find more strict justification for the definition (6), let us
195: consider the one-parameter subfamily of $q,p$-oscillators (3)-(4).
196: Namely, let $p=q^m$ with real $m$:
197: \begin{equation}
198: A A^\dagger - q\ A^\dagger A = q^{m \cal N} \qquad\quad \quad
199: A A^\dagger - q^m \ A^\dagger A = q^{\cal N} \ ,
200: \end{equation}
201: \[
202: A^\dagger A = [\![{\cal N}]\!]_{q,q^m} \ ,
203: \quad
204: \quad
205: \quad
206: A A^\dagger = [\![{\cal N}+1]\!]_{q,q^m} \ .
207: \]
208: In the limit of $m\to 1$ this reduces to the q-oscillator (1),
209: so we consider %the same limit of the expression
210: \begin{equation}
211: {\lim}_{{m}\to 1} [\![X]\!]_{q,q^m} \equiv
212: {\lim}_{{m}\to 1} \frac{ q^{X}-q^{m X} }{ q-q^m } \ .
213: \end{equation}
214: It is useful to merely put $\widetilde{q}\equiv q^{m-1}$,
215: so that $\widetilde{q}\ \stackrel{m\to 1}{\Longrightarrow} \ 1$. Then,
216: %we have
217: \begin{equation}
218: [\![X]\!]_{q,q^m} = %\equiv \frac{ q^{X}-q^{m X} }{ q-q^m }
219: q^{X-1}\ \frac{ \widetilde{q}^{ X} - 1}{\widetilde{q} - 1} \ .
220: \end{equation}
221: So we define
222: \begin{equation}
223: \{X\}_q \equiv
224: q^{X-1} \cdot {\lim}_{\widetilde{q}\to 1}
225: \left( \frac{ \widetilde{q}^{ X} - 1}{\widetilde{q} - 1}
226: \right) = q^{X-1} \ X \
227: \end{equation}
228: getting the same as in (6) (we have used the L'Hospital rule
229: in order to resolve the uncertainty).
230:
231: Taking into account the formulas
232: \begin{equation}
233: a^\dagger a = \{N\}_q \equiv N q^{N-1} \ ,
234: \qquad \quad
235: a a^\dagger = \{N+1\}_q \equiv (N+1) q^{N} \ ,
236: \end{equation}
237: we verify that the defining relation (1) is satisfied.
238:
239: Note that the $q$-oscillator algebra generated by
240: $a$, $a^\dagger$ and $N$ obeying the relations (1)-(2)
241: may be given a Hopf algebra structure, with
242: the comultiplication, antipode and counit
243: written out as in Ref.~\refcite{Jagan,Chung}.
244:
245: \section{Energy spectrum of the $q$-oscillator}
246:
247: Let us study in some detail the properties of the $q$-oscillator (1).
248: To this end we take, in analogy with usual quantum harmonic oscillator,
249: the Hamiltonian of the $q$-oscillator (1) in the form
250: \begin{equation}
251: H = \frac{\hslash\omega}{2} (a a^\dagger + a^\dagger a) \
252: \end{equation}
253: and from now on put ${\hslash\omega}=1$ for simplicity.
254:
255: We adopt the $q$-analogue of the Fock space with the
256: vacuum state $|0 \rangle$. Then,
257: \begin{equation}
258: a |0 \rangle = 0 \ , \qquad \quad
259: |n \rangle =
260: \frac{(a^\dagger )^n}{\sqrt{\{n \}_q!} } |0 \rangle \ , \qquad \quad
261: N |n \rangle = n~|n \rangle \ ,
262: \end{equation}
263: where $\{n \}_q!=\{n \}_q\{n-1 \}_q ... \{2 \}_q\{1 \}_q$,
264: curly brackets are defined in (11), and
265: the creation/annihilation operators act by the formulas %according to
266: \begin{equation}
267: a \ |n \rangle = \sqrt{n q^{n-1}} \ |n-1 \rangle\ , \qquad \quad
268: a^\dagger \ |n \rangle = \sqrt{(n+1) q^{n}} \ |n+1 \rangle \ .
269: \end{equation}
270: For any real nonnegative $q$, the operators $a$ and $a^\dagger$ are
271: indeed conjugates of each other.
272:
273: From (12)-(14), the spectrum $H|n\rangle = E_n |n\rangle$ of
274: the Hamiltonian reads:
275: \begin{equation}
276: E_n = \frac12 \Bigl( (n+1) q^{n} + n q^{n-1} \Bigr)
277: = \frac12 \ q^{n} \Bigl( 1 + n (1+ q^{-1}) \Bigr) \ .
278: \end{equation}
279: At $q\to 1$ we recover the familiar result $E_n = n + \frac12$
280: as it should.
281: Moreover, at $n=0$ we have $E_0 = \frac12$ whatever is the value of $q$.
282:
283: For any $q\ne 1$, the {\em spectrum is not uniformly spaced} (not equidistant).
284: It is easy to see that if $q> 1$ the spacing $E_{n+1}-E_{n}$ gradually
285: increases with growing $n$, so that $E_{n} \to \infty$ as $n \to \infty$.
286: Different, and more interesting situation emerging at $0<q<1$ is
287: the subject of our study in the next sections.
288:
289:
290: \vspace{1mm}
291: \section{ Various 'accidental' degeneracies $E_{n_1}\!=E_{n_2}$ of energy levels }
292:
293: %\vspace{1mm}
294: From now on we deal with the values of $q$ such that
295: \begin{equation}
296: 0 < q < 1 \ .
297: \end{equation}
298: We have to stress that the special feature of the $q$-oscillator
299: defined by the relations (1)-(2), with any fixed value of the
300: $q$-deformation parameter from (16), consists in the asymptotical
301: tending, as $n\to\infty$, of the $n$-th level energy to zero :
302: $E_n\to 0^+$. For that reason the $q$-oscillator (1)-(2) has been
303: named the "Tamm-Dankoff cutoff" oscillator\cite{Odaka,Jagan}. In
304: this letter we use this same term (or "TD" in short).
305:
306: {\bf Remark~1.} If one imposes the requirement\cite{Chung}
307: that for all the energy values the
308: inequality $\Delta E (n)\equiv E_{n+1}-E_n > 0$
309: {\em must hold}, then the spectrum is truncated and
310: becomes finite: the set of energy levels consists
311: of $E_n$ with $n=0, 1, ..., \lfloor\frac{1+q^2}{1-q^2}\rfloor$
312: (here $\lfloor x \rfloor$ denotes integral part of $x$).
313: We however do not require this in our paper, and consider
314: the whole (infinite) spectrum.
315:
316: The energy spectrum given by the expression (15) manifests
317: some sorts of degeneracies, with strong dependence
318: on the particular fixed value of $q$.
319: Let us consider a number of different cases.
320:
321: \vspace{1mm}
322: \subsection{Degeneracy of nearest-neighbor levels: $E_m = E_{m+1}$}
323: %{(I) \em Degeneracy of the type $E_n = E_{n+1}$}
324: \vspace{1mm}
325:
326: Let the $q$ be fixed as $q=\sqrt{\frac13}$.
327: \noindent
328: Then it turns out that %besides the level $E_0=\frac12$,
329: \[
330: E_0=\frac12 < E_1 = E_2 = \frac12 + \frac{1}{\sqrt{3}}\,,
331: %E_2 > E_3=\frac12 +\frac{2}{3\sqrt{3}} \ > \frac12 \ , \qquad \quad
332: \]
333: \[
334: E_2 > E_3=\frac12 +\frac{2}{3\sqrt{3}} \ >
335: E_4=\frac{5}{18} +\frac{2}{3\sqrt{3}} > \frac12\ , \quad
336: %E_5=\frac{1}{2} \frac{\sqrt{3}-2}{9} < \frac12,
337: \]
338: and for the rest (infinite set) of levels we find
339: \[
340: \frac12 > E_5=\frac{1}{2}+\frac{\sqrt{3}-2}{9} > E_6 > \ldots >
341: E_{m} > E_{m+1} > \ldots >0 \ , \quad \qquad m=7, 8, \ldots \ .
342: \]
343: This degeneracy $E_1 = E_2$ is a particular case of the next more general property.
344: %Namely,
345:
346: %%%%%%%%%%%%%%%%%%%%%
347: {\bf Proposition 1.} Let the parameter $q$ be fixed as $q=\sqrt{\frac{m}{m+2}}$
348: where $ \ m\ge 1\ $.
349: Then the degeneracy occurs for the following pair of the energy levels:
350: \begin{equation}
351: E_m =E_{m+1} \,.
352: \end{equation}
353: This is proved by direct check.
354: Figs.~1,2,3 illustrate such a degeneracy for three particular cases,
355: $m=1, 2, 9$.
356:
357:
358: Note that $n=0$ is excluded from the above series (17) since
359: the degeneracy $E_0 = E_1$ is not possible, as this would require $q=0$.
360: However, yet another degeneracies with the $E_0$ involved does occur, see below.
361: %
362: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% figs. 1 %%%%%%%%%%%%%%%%%%%%%%%%%%%%
363: \begin{figure} [hbtp]
364: \centering {
365: %\hspace{2mm}
366: \includegraphics[angle=-90, width=0.65\textwidth] {E1-E2.eps}
367: }
368: %\vspace{-6mm}
369: \caption{ Spectrum of the $q$-oscillator (1) at fixed $q=\sqrt{1/3}$.
370: Observe that $E_0=\frac12$ and %the degeneracy %of the two energy levels:
371: $E_1=E_2$.}
372: \end{figure}
373: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% figs. 1 %%%%%%%%%%%%%%%%%%%%%%%%%%%%
374: \vspace{-1mm}
375: %
376: \vspace{-1mm}
377: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% figs. 2 %%%%%%%%%%%%%%%%%%%%%%%%%%%%
378: \begin{figure} [hbtp]
379: \centering {
380: %\hspace{2mm}
381: \includegraphics[angle=-90, width=0.65\textwidth] {E2-E3.eps}
382: }
383: %\vspace{-6mm}
384: \caption{ Spectrum of the $q$-oscillator (1) at fixed $q=\sqrt{2/4}$.
385: Observe the degeneracy %of the two energy levels:
386: $E_2=E_3$ .}
387: \end{figure}
388: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% figs. 2 %%%%%%%%%%%%%%%%%%%%%%%%%%%%
389: %\vspace{1mm}
390: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% figs. 3 %%%%%%%%%%%%%%%%%%%%%%%%%%%%
391: \begin{figure} [hbtp]
392: \centering {
393: %\hspace{2mm}
394: \includegraphics[angle=-90, width=0.65\textwidth] {E9-E10.eps}
395: }
396: %\vspace{-6mm}
397: \caption{ Spectrum of the $q$-oscillator (1) at fixed $q=\sqrt{9/11}$.
398: Observe the degeneracy %of the two energy levels:
399: $E_{9}=E_{10}$ .}
400: \end{figure}
401: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% figs. 3 %%%%%%%%%%%%%%%%%%%%%%%%%%%%
402: \vspace{1mm}
403: %
404:
405: \subsection{Degeneracy of next to nearest neighbors: $E_m = E_{m+2}$}
406:
407: Let the $q$-parameter be fixed as $q=\frac13$.
408: \noindent
409: %In this case
410: Then, %we have
411: \[ E_0=\frac12\ , \hspace{8mm} E_1=\frac56 > \frac12\ ,
412: %\]
413: %and
414: %\[
415: \hspace{10mm}
416: E_2=E_0=\frac12\ , \hspace{8mm}
417: E_3=\frac{13}{54} < \frac12\ ,
418: \]
419: while for all other levels we find
420: %\[
421: $\ \
422: \frac12 > E_{m} > E_{m+1} , \ \ %\qquad \qquad
423: m\ge 2 \ .
424: $
425: %\]
426: This instance is nothing but a particular case of the next
427: more general situation.
428:
429: %%%%%%%%%%%%%%%%%%%%%
430: {\bf Proposition 2.} Let the parameter $q$ be fixed as
431:
432: \begin{equation}
433: q=\frac{1 + \sqrt{4 m^2 + 12 m + 1}}{2 ( m + 3)} \qquad \qquad
434: {\rm with} \quad \quad \ \ m\ge 0\ .
435: \end{equation}
436: Then among the energy levels $E_n$ the following degeneracy occurs:
437: \begin{equation}
438: E_m =E_{m+2} \ . % \qquad \qquad m\ge 5 \ .
439: \end{equation}
440:
441: This statement is proved by direct verification.
442: %
443: The two particular cases $m=0$ and $m=5$ of the series of
444: degeneracies (19) are shown in Fig.~4 and Fig.~5.
445: \vspace{-1mm}
446: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% figs. 4 %%%%%%%%%%%%%%%%%%%%%%%%%%%%
447: \begin{figure} [hbtp]
448: \centering {
449: %\hspace{2mm}
450: \includegraphics[angle=0, width=0.60\textwidth] {E0-E2.eps}
451: }
452: %\vspace{-6mm}
453: \caption{ Energy levels $E_0$, $E_1$,..., $E_{10}$ of the
454: $q$-oscillator (1) at $q=\frac13$, see (18).
455: Observe $E_0=E_2$.}
456: \end{figure}
457: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% figs. 4 %%%%%%%%%%%%%%%%%%%%%%%%%%%%
458: \vspace{-1mm}
459: %
460: \vspace{-1mm}
461: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% figs. 5 %%%%%%%%%%%%%%%%%%%%%%%%%%%%
462: \begin{figure} [hbtp]
463: \centering {
464: %\hspace{2mm}
465: \includegraphics[angle=0, width=0.60\textwidth] {E5-E7.eps}
466: }
467: %\vspace{-6mm}
468: \caption{ Energy levels $E_0$, ..., $E_{30}$ of the $q$-oscillator (1)
469: at $q=\frac{1+\sqrt{161}}{16}$, see (18).
470: Observe $E_5=E_7$.}
471: \end{figure}
472: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% figs. 5 %%%%%%%%%%%%%%%%%%%%%%%%%%%%
473: %\vspace{1mm}
474:
475: \subsection{Degeneracy of the type $E_0 = E_n$}
476:
477: Now let us study the occurrence of yet another type of
478: degeneracy, $E_0 = E_n$.
479:
480: %%%%%%%%%%%%%%%%%%%%%
481: {\bf Proposition 3.} For any integer $ n = 2, 3, 4, ..., $
482: there exists the appropriate value $q_n = q(n)$ that guarantees
483: validity of the equality
484: \begin{equation}
485: E_0 = E_n \ .
486: \end{equation}
487: The proposition can be proved by graphical treatment.
488: Suppose (20) is valid. Then
489: \begin{equation}
490: (n+1) q^n + n q^{n-1} = 1\ \qquad \qquad (0 < q < 1) \
491: \end{equation}
492: or, equivalently, using for convenience $x$ instead of $n$,
493: \begin{equation}
494: (1 + q^{-1}) x = q^{-x} - 1\ .
495: \end{equation}
496: One can easily check the fact of intersection
497: (at $x$ {\em being the integers} 2, 3, 4, \ldots)
498: of the curves corresponding to the linear resp. exponential
499: functions in the l.h.s. resp. in the r.h.s. of the latter equality.
500: Fig.~6 and Fig.~7 serve to illustrate this fact.
501: %\vspace{-1mm}
502: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% figs. 6 %%%%%%%%%%%%%%%%%%%%%%%%%%%%
503: \begin{figure} [h]
504: \centering{
505: %\hspace{2mm}
506: \includegraphics[angle=0, width=0.45\textwidth] {E0-E3.eps}
507: }
508: %\vspace{-6mm}
509: \caption{ Degeneracy $E_0=E_{3}$ at $q=q_{3}\simeq 0.45541$ from
510: graphical solution of Eq.~(22).}
511: \end{figure}
512: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% figs. 6 %%%%%%%%%%%%%%%%%%%%%%%%%%%%
513: \vspace{-2mm}
514: %%%%%
515: \vspace{-2mm}
516: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% figs. 7 %%%%%%%%%%%%%%%%%%%%%%%%%%%%
517: \begin{figure} [h]
518: \centering {
519: %\hspace{2mm}
520: \includegraphics[angle=0, width=0.47\textwidth] {E0-E10.eps}
521: }
522: %\vspace{-6mm}
523: \caption{ Degeneracy $E_0=E_{10}$ at $q=q_{10}\simeq 0.725405$
524: from graphical solution of Eq.~(22).
525: }
526: \end{figure}
527: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% figs. 7 %%%%%%%%%%%%%%%%%%%%%%%%%%%%
528:
529: \vspace{4mm}
530: It is instructive to have some particular values of the $q$-parameter
531: which provide the degeneracies of the type (20). To this end,
532: we rewrite Eq. (21) as the equation for $z\equiv q^{-1}$:
533: \[
534: z^m - m z - (m+1) = 0\ , \hspace{14mm} m = 2, 3, 4, ... .
535: \]
536: The first three members of this series are solved exactly (in radicals),
537: while for $m \ge 5$ the values $q_m$ are found approximately (recall
538: that only those solutions $q_m$ are admitted that obey $0 < q_m < 1$).
539: Table 1 presents a list of sample values.
540: \begin{table}[h]
541: \tbl{Value of $q_{m}$ that gives $E_0=E_{m}$.}
542: {\begin{tabular}{@{}cc@{}} \toprule %|c|c|} %|l|}
543: {Value of ${m}$} & Value of $q_{m}$ \\ %such that $E_0=E_{m}$ \qquad \\
544: \colrule
545: $m=2$ & \hspace{-7mm} ${q}_2= \frac13\simeq 0.333333$ \\
546: $m=3$ & \hspace{-5mm} ${q}_3 = \left( \sqrt[3]{2+\sqrt{3}}
547: +\sqrt[3]{2-\sqrt{3}}\ \right)^{-1}\simeq 0.45541$ \\
548: $m=4$ & \hspace{18mm} ${q}_4 = 3 \Bigl(1 + \sqrt[3]{64+6 \sqrt{114}}
549: + \sqrt[3]{64-6 \sqrt{114}}\ \Bigr)^{-1}\simeq 0.53156446$ \\
550: $m=5$ & \hspace{-3mm}
551: $q_5\simeq 0.585442$ \\
552: $m=6$ & $q_6\simeq 0.6262253$ \\
553: $m=10$ & $q_{10}\simeq 0.725405$ \\
554: $m=25$ & $q_{25}\simeq 0.851675$ \\
555: $m=50$ & $q_{50}\simeq 0.910968$ \\
556: $m=100$ & $q_{100}\simeq 0.948094$ \\
557: $m=200$ & $q_{200}\simeq 0.9704016$ \\
558: $m=400$ & $q_{400}\simeq 0.98340363$ \\ \botrule
559: \end{tabular}}
560: \end{table}
561:
562: \subsection{General two-fold 'accidental' degeneracy $E_m = E_{m+k}$}
563:
564: Here we briefly discuss the most general form of two-fold
565: degeneracy $E_{m}=E_{m\!+\!k}$.
566: The equation for the $q$-parameter value $q=q(m,k)$
567: responsible for this fact reads
568: %\[
569: $\
570: (m+k+1)q^{m+k} + (m+k)q^{m+k-1}\!-\!(m+1)q^{m}- m q^{m-1} = 0\ ,
571: $
572: %\]
573: or
574: \begin{equation}
575: (m+k+1)~q^{k+1} + (m+k)~q^{k} - (m+1)~q - m = 0\ .
576: \end{equation}
577: It can be proved that for each pair ($m$, $m + k$) there exists
578: such real solution $q\!=\!q(m,k)$ of (23) that $0\!<\!q\!<\!1$.
579: Here we only %Instead, let us
580: comment on few cases of low $k$ values.
581: It is obvious that $k=1$ resp. $k=2$
582: correspond to the particular series of degeneracies already
583: considered above, in subsection 4.1 resp. subsection 4.2.
584: For the next two cases the equations to be solved are:
585: \begin{equation}
586: \hspace{-20mm}
587: \underline{k=3} \hspace{20mm}
588: q^{4} + \frac{m+3}{m+4}~q^{3} - \frac{m+1}{m+4}~q - \frac{m}{m+4} = 0\ ,
589: \end{equation}
590: \begin{equation}
591: \hspace{-6mm}
592: \underline{k=4} \hspace{16mm}
593: q^{4} - \frac{1}{m+5}~q^{3} + \frac{1}{m+5}~q^2 - \frac{1}{m+5}~q - \frac{m}{m+5} = 0\
594: \end{equation}
595: where for $k=4$ it was taken into account that the starting 5-th degree
596: equation divides by $q+1$. Note that the root $q=-1$ also exists in all
597: the cases of higher even $k$ in (23). The equations (24),(25) can be solved in
598: radicals, what yields huge expressions. And, analogously to
599: the above Table~1, a set of values $q=q(m,k)$
600: can be found numerically and as well tabulated.
601: Finally, let us remind that the case $E_0=E_m$, see subsec.~4.3,
602: is obviously covered by the most general situation (23).
603:
604: \section{Fibonacci property of the TD $q$-oscillator}
605:
606: %According to \cite{Arik-Fib},
607: Fibonacci oscillators, see Ref.~\refcite{Arik-fibo},
608: are characterized by the property
609: %by the relation which involves their three
610: %successive energy eigenvalues:
611: that each their energy eigenvalue $E_{n+1}$ is uniquely
612: determined by a linear combination of the two preceding
613: eigenvalues:
614: \begin{equation}
615: E_{n+1} = \alpha\,E_{n} + \beta\,E_{n-1}, \ \qquad \alpha,\beta\in {\bf R}.
616: \end{equation}
617: Taking into account the explicit formula (15) %for the $E_{m}$,
618: one verifies that the TD-oscillator {\em does possess
619: the Fibonacci property} (26) if the coefficients are chosen as
620: \begin{equation}
621: \alpha = 2\,q \ , \hspace{12mm} \beta = - q^2 \ .
622: \end{equation}
623: Now it is of interest to consider the interplay
624: "Fibonacci + degeneracy", that is, to examine
625: how a specified version of the Fibonacci property manifests
626: itself 'locally', i.e. for those three consecutive levels
627: which include some degenerate pair $E_{n_1} = E_{n_2}$.
628:
629: Let us first put
630: $q=\sqrt{\frac{m-1}{m+1}}$, $\ m>1$, which results in $E_{m-1}=E_m$.\
631: Then, %we have:
632: %\[
633: \begin{equation}
634: E_{m+1} = q (2-q) E_m \hspace{6mm} {\rm or} \hspace{6mm}
635: E_{m+1} = q (2-q) E_{m-1}\,, \hspace{6mm}
636: q=\sqrt{\frac{m-1}{m+1}}\,. \hspace{6mm}
637: \end{equation}
638: Likewise, if we
639: let $q=\sqrt{\frac{m}{m+2}}$ with $\ m>0$ then $E_{m+1}=E_m$.
640: In this case we have:
641: \begin{equation}
642: E_{m-1}=\frac{2q-1}{q^2}~E_{m} \hspace{6mm} {\rm or} \hspace{6mm}
643: E_{m-1} = \frac{2q-1}{q^2}\ E_{m+1}\ , \hspace{6mm} q=\sqrt{\frac{m}{m+2}}\ .
644: \hspace{6mm}
645: \end{equation}
646: Yet another situation corresponds to account of the degeneracy
647: of next to nearest neighbors, $E_{m+2}=E_m$, for
648: the $q$-parameter fixed exactly as in (18).
649: In this case:
650: \begin{equation}
651: E_{m+1} = \frac{1 + q^2}{2 q}\ E_{m}\ \hspace{6mm} {\rm or}
652: \hspace{6mm} E_{m+2} = \frac{2 q}{1 + q^2}\ E_{m+1}
653: \hspace{6mm}
654: \end{equation}
655: where
656: $ q=\frac{1 + \sqrt{4 m^2 + 12 m + 1}}{2 ( m + 3)}\,. $
657: In a similar way, other two-fold degeneracies analyzed
658: in Section~4 can be dealt with in conjunction with
659: the Fibonacci property (26).
660:
661:
662: \section{ 'Locally classical' appearance of the commutator of $X$ and $P$}
663:
664: We use standard form for the position/momentum operators:
665: $
666: X\equiv \frac1{\sqrt2}(a + a^\dag), \ % \hspace{18mm}
667: P\equiv \frac{\rm i}{\sqrt2} (a^\dag - a).
668: $
669: With the account of (11), the commutator of these operators is
670: \[
671: -\frac{i}{\sqrt2}\,[ X, P ] = [ a , a^\dag ] = (N+1)q^N - Nq^{N-1} = q^N + (q-1) N q^{N-1}\
672: \]
673: or, on the basis states,
674: \begin{equation}
675: -\frac{i}{\sqrt2}\,[ X, P ]\,|n \rangle = \bigl( q^n + (q-1) n q^{n-1}\bigr)|n \rangle\
676: = q^n\bigl( 1 + n(1-q^{-1})\bigr)|n \rangle\,.
677: \end{equation}
678: If we fix the deformation parameter as the rational number
679: $q = \frac{m}{m+1}$ ,
680: then it turns out that 'locally', i.e. on the individual basis
681: element $|n=m\rangle $, the two operators $X, P$ behave just as
682: the "classical" ones:
683: \begin{equation}
684: [ X, P ]\,|m \rangle = 0
685: \hspace{10mm} {\rm if} \hspace{10mm} q = \frac{m}{m+1}\ .
686: \end{equation}
687: Like in the cases of AC and BM type $q$-oscillators\cite{AC,Bied}
688: whose $X\!-\!P$ uncertainty relations depend on both the $q$-parameter and
689: the state-labeling number $n$
690: (i.e., {\em state-dependent uncertainty relations}\cite{Bied,M-Delgado}),
691: the $q$-deformed oscillator treated in the present letter also yields
692: the commutator $X,P$, and its associated uncertainty relation, with the
693: explicit $|n>$ dependence, see (31). However, unlike the AC $q$-oscillator
694: with {\em real} $q$, the TD type $q$-oscillator admits for each state $|m>$
695: the respective special value (32) of deformation parameter which makes $X$ and $P$
696: 'classical' (commuting) on this, and only this, state.
697: This fact has evident consequence for (the minimum of) associated
698: uncertainty relation.
699:
700:
701: \section{ Quantum algebra realized with TD $q$-oscillators}
702:
703: Using two identical and independent (commuting) copies of
704: the TD $q$-oscillator with the generators
705: $a_1,a_1^{\dag}\ $, $a_2,a_2^{\dag}\ $, $N_1, N_2$ obeying (1), (2), (11)
706: we form the generators
707: \[
708: J_+ := a_1^\dag a_2\ , \hspace{5mm} J_- := a_2^\dag a_1\ , \hspace{5mm}
709: J_0 : =\frac12 \bigr( N_1 - N_2 \bigr)\ ,
710: \hspace{3mm} J_3 : =\frac12 \bigr( N_1 + N_2 \bigr)\ ,
711: \]
712: and find their closing into the following $q$-deformed algebra
713: \begin{equation}
714: [J_0,J_{\pm}]=\pm J_{\pm} \ , \hspace{4mm} [J_+,J_-]= 2 J_0\ q^{2J_3-1} \ ,
715: \hspace{4mm} [J_0,J_3]= 0 \ , \hspace{3mm} [J_\pm,J_3]= 0\ ,
716: \end{equation}
717: which is a particular $q$-deformed version $U_q(su(2)\oplus u(1))$ of the
718: universal enveloping algebra of the {\em non-simple} algebra
719: $su(2)\oplus u(1)$.
720: This is in contrast with the well-known fact that the
721: $q$-analog $U_q(su(2))$ of the (simple) algebra $su(2)$
722: is realizable in terms of two independent modes
723: of the BM $q$-oscillator\cite{Bied,Macf}.
724: The Hopf-algebra structure of the $q$-deformed
725: algebra (33) can be written out along the lines of Ref.~\refcite{Chung},
726: so we do not reproduce it here.
727:
728: Using the $q$-Fock space framework, see (13)-(14) above,
729: one is lead to the following rather simple yet $q$-dependent
730: representation formulas ($j=0,\frac12,1,\frac32,...$; $m=-j, -j+1,...,j-1,j$)
731: for the $q$-algebra $U_q(su(2)\oplus u(1))$:
732: \[
733: J_+ |j,m \rangle = \sqrt{(j-m)(j+m+1) q^{j-m-1} q^{j+m}}\ |j,m+1 \rangle
734: \]
735: \begin{equation} %\hspace{14mm}
736: =q^{j-\frac12}\sqrt{(j-m)(j+m+1)}\ |j,m+1 \rangle \ ,
737: \end{equation}
738: \[
739: J_- |j,m \rangle = \sqrt{(j+m)(j-m+1) q^{j+m-1} q^{j-m}}\ |j,m-1 \rangle
740: \]
741: \begin{equation} %\hspace{14mm}
742: =q^{j-\frac12}\sqrt{(j+m)(j-m+1)}\ |j,m-1 \rangle \ , \\
743: %\hspace{6mm} [J_0,J_3]= 0 \ .
744: \end{equation}
745: \begin{equation}
746: %\[
747: J_0\ |j,m \rangle = m \ |j,m \rangle \ , \hspace{14mm}
748: %\]
749: J_3 \ |j,m \rangle = j \ |j,m \rangle \ .
750: \end{equation}
751:
752: {\bf Remark~2.} As an interesting peculiarity of this quantum algebra,
753: let us emphasize the following fact: %peculiarity:
754: at $q<1$, due to the multiplier $q^{j-\frac12}$ in formulas (34)-(35),
755: the behavior of the matrix elements of the raising/lowering operators
756: in the 'large spin' limit, i.e. for $j\to\infty$,
757: principally differs from that of $J_0$ and $J_3$. Namely, at $j\to\infty$
758: we have $\langle j,m |J_3|j,m \rangle \to \infty$, unlike the limit:
759: $\lim_{j\to\infty} \langle j,m\pm 1 |J_{\pm}|j,m \rangle = 0$.
760: The latter behavior is completely analogous to the (exponential) 'cutoff'
761: property of the TD type $q$-oscillator whose energy values satisfy:
762: $E_n\to 0$ when $n\to\infty$.
763: On the other hand such asymptotics contrasts with large $j$ behavior
764: of representation matrix elements of $J_+, J_-$ of the usual $su(2)$.
765:
766:
767: \section{Concluding remarks}
768:
769: Our goal was to explore the rather unusual and seemingly
770: overlooked properties of the TD type $q$-oscillator.
771: In particular, we have analyzed various types of occurring
772: degeneracies: nearest-neighbor levels, zero level $E_0$ \&
773: some other level $E_m$, $m\ge 2$, etc.
774: The considered degeneracies are different
775: from the two-fold degeneracies of
776: oscillator-like system studied in Ref.~\refcite{Ques2}
777: where due to extra cyclic symmetry, an
778: infinite series of pairs of degenerate energy
779: levels appears. In the case of TD $q$-oscillator,
780: at a properly fixed value of the $q$-parameter
781: the degeneracy is 'accidental' (without underlying
782: symmetry) and involves single pair of levels.
783:
784: For this, and other properties studied above,
785: we expect for the TD $q$-oscillator a number
786: of interesting physical applications.
787: Let us mention the already proposed usage\cite{SIGMA}
788: of TD $q$-oscillators as the base for
789: %the framework of
790: the related "TD $q$-Bose gas" model, the
791: $p\!=\!q$ one-parameter limit of $q,\!p$-Bose gas model. As argued\cite{SIGMA},
792: %in Ref.~\refcite{ }
793: this TD-type $q$-Bose gas model gives as well efficient
794: description of the observed non-Bose properties of the
795: two- and multi-pion (-kaon) correlations in the
796: experiments on relativistic heavy-ion collisions
797: as the other, e.g. the BM, types of $q$-Bose
798: gas model do.\cite{AGI-1,AGI-2}
799:
800: \vspace{1mm}
801:
802: {\bf Acknowledgements}
803:
804: One of the authors (A.M.G) thanks Prof. A.U.Klimyk for valuable
805: discussions and remarks.
806: This research was partially supported by the Grant 10.01/015 of the
807: State Foundation of Fundamental Research of Ukraine.
808: \vspace{1mm}
809:
810: \begin{thebibliography}{0}
811:
812: \bibitem{Bied}
813: L. C. Biedenharn,
814: %The quantum group $SU_q(2)$ and a $q$-analogue of the boson operators,
815: {\it J. Phys. A: Math. Gen.} {\bf 22}, L873, (1989).
816:
817: \bibitem{Macf}
818: A. J. Macfarlane,
819: %On $q$-analogues of the quantum harmonic oscillator and the
820: %quantum group $SU_q(2)$,
821: {\it J. Phys. A: Math. Gen.} {\bf 22}, 4581, (1989).
822:
823: \bibitem{AC}
824: M. Arik and D. D. Coon,
825: %Hilbert spaces of analytic functions and generalized coherent states,
826: {\it J. Math. Phys.} {\bf 17}, 524 (1976).
827: %Meljanac S., Milekovic M., {\it Int. J. Mod. Phys.} {\bf A 11}, 1391 (1996).
828:
829: \bibitem{bonats}
830: D. Bonatsos et al.,
831: %C. Daskaloyannis, D. Ellinas and A. Faessler,
832: {\it Phys. Lett. B} {\bf 331}, 150 (1994).
833:
834: \bibitem{Odaka}
835: K. Odaka, T. Kishi and S. Kamefuchi,
836: %On quantization of simple harmonic oscillators,
837: {\it J. Phys. A: Math. Gen.} {\bf 24}, L591 (1991).
838:
839: \bibitem{Jagan}
840: S. Chaturvedi, V. Srinivasan and R. Jagannathan,
841: %Tamm-Dancoff deformation of bosonic oscillator algebras,
842: {\it Mod. Phys. Lett. A} {\bf 8}, 3727 (1993).
843:
844: \bibitem{Ques1} C. Quesne and N. Vansteenkiste,
845: {\it Helv. Phys. Acta} {\bf 69}, 141 (1996).
846:
847: \bibitem{Melj} S. Meljanac, M. Milekovic and S. Pallua,
848: {\it Phys. Lett. B} {\bf 328}, 55 (1994).
849:
850: \bibitem{Mizrahi}
851: S. S. Mizrahi, J. P. Camargo Lima and V. V. Dodonov,
852: {\it J. Phys. A: Math. Gen.} {\bf 37}, 3707 (2004).
853:
854: \bibitem{SIGMA}
855: A. M. Gavrilik,
856: %"Combined analysis of two- and three-particle correlations %in $qp$-Bose gas model",
857: {\it SIGMA} ({\it Symmetry, Integrability and Geometry: Methods and
858: Applications} ), {\bf 2}, paper 074, 1-12 (2006), {\tt
859: hep-ph/0512357}.
860:
861: \bibitem{AdGa}
862: L. V. Adamska and A. M. Gavrilik,
863: %Multi-particle correlations in $qp$-Bose gas model,
864: {\it J. Phys. A: Math. Gen.} {\bf 37}, 4787 %--4795, N~17,
865: (2004), {\tt hep-ph/0312390}.
866: \bibitem{Chakra}
867: A. Chakrabarti and R. Jagannathan,
868: %A ($p,q$)oscillator realization of two-parameter quantum algebras,
869: {\it J. Phys. A: Math. Gen.} {\bf 24}, L711 (1991).
870:
871: \bibitem{Chung}
872: W.-S. Chung et al.,
873: {\it Phys. Lett. A} {\bf 183}, 363 (1993).
874:
875: \bibitem{M-Delgado}
876: M. A. Martin-Delgado,
877: {\it J. Phys. A: Math. Gen.} {\bf 24}, L1285 (1991).
878:
879: \bibitem{Arik-fibo}
880: M. Arik et al.,
881: {\it Z. Phys. C} {\bf 55}, 89 (1992).
882:
883: \bibitem{Ques2} C. Quesne and N. Vansteenkiste,
884: {\it Helv. Phys. Acta} {\bf 72}, 71 (1999), {\tt math-ph/9901016}.
885:
886: \bibitem{AGI-1} D. V. Anchishkin, A. M. Gavrilik and N. Z. Iorgov,
887: {\it Eur. Phys. J. A} {\bf 7}, 229 (2000), {\tt nucl-th/9906034}.
888:
889: \bibitem{AGI-2}
890: D. V. Anchishkin, A. M. Gavrilik and N. Z. Iorgov,
891: {\it Mod. Phys. Lett. A} {\bf 15}, 1637 (2000), {\tt hep-ph/0010019}.
892:
893:
894: \end{thebibliography}
895: \end{document}
896: