nucl-th0111041/cam.tex
1: \documentstyle[12pt,epsfig]{article}
2: 
3: \setcounter{topnumber}{2}
4: \def\topfraction{.85}
5: \setcounter{bottomnumber}{1}
6: \def\bottomfraction{.85}
7: \setcounter{totalnumber}{3}
8: \def\textfraction{.14}
9: \def\floatpagefraction{.85}
10: \setcounter{dbltopnumber}{2}
11: \def\dbltopfraction{.7}
12: \def\dblfloatpagefraction{.5}
13: 
14: \renewcommand{\baselinestretch}{1.0}
15: \newcommand{\bea}{\begin{eqnarray}}
16: \newcommand{\eea}{\end{eqnarray}}
17: \setlength{\textheight}{21truecm}
18: \setlength{\textwidth}{16.0truecm}
19: \setlength{\oddsidemargin}{0.5truecm}
20: \setlength{\evensidemargin}{0.5truecm}
21: \setlength{\topmargin}{0.0truecm}
22: 
23: \begin{document}
24: 
25: \title{Ground state angular momenta of random nuclei} 
26: 
27: \author{R. Bijker$^{1,2}$ and A. Frank$^{2,3}$\\
28: \mbox{}\\
29: $^1$ Dipartimento di Fisica, Universit\`a degli Studi di Genova, \\
30: Via Dodecaneso 33, I-16146 Genova, Italy 
31: \footnote{Sabattical leave} \\ 
32: $^2$ Instituto de Ciencias Nucleares, UNAM, AP 70-543, \\ 
33: 04510 M\'exico, DF, M\'exico \\
34: $^3$Centro de Ciencias F{\'{\i}}sicas, UNAM, AP 139-B, \\ 
35: 62251 Cuernavaca, Morelos, M\'exico}
36: 
37: \date{}
38: 
39: \maketitle
40: 
41: \begin{abstract} 
42: We investigate the phenomenom of emerging regular spectral 
43: features from random interactions. In particular, we address the dominance 
44: of $L=0$ ground states in the context of the vibron model and the 
45: interacting boson model. A mean-field analysis links different regions of 
46: the parameter space with definite geometric shapes. The results that 
47: are, partly, obtained in closed analytic form, provide a clear and 
48: transparent interpretation of the distribution of ground state angular 
49: momenta as observed before in numerical studies.   
50: \end{abstract}
51: 
52: \section{Introduction}
53: 
54: Recent shell model calculations of even-even nuclei in the $sd$ shell 
55: and the $pf$ shell with random interactions showed a remarkable 
56: statistical preference for ground states with $L=0$, despite the 
57: random nature of the two-body matrix elements, both in sign and in relative 
58: magnitude \cite{JBD}. 
59: A similar preponderance of $L=0$ ground states was found in an analysis 
60: of the Interacting Boson Model (IBM) with random interactions \cite{BF1}. 
61: In addition, in the IBM evidence was found for both vibrational and 
62: rotational band structures \cite{BF1,BF2}. 
63: These are surprising results in the sense that, according to the conventional 
64: ideas in the field, the occurrence of $L=0$ ground states and the existence 
65: of vibrational and rotational bands are due to very specific forms of the 
66: interactions. The studies with random interactions show that the class of 
67: Hamiltonians that lead to these regular features is much larger than usually 
68: thought. 
69: These unexpected results have sparked a large number of investigations to 
70: explain and further explore the properties of random nuclei, both for the 
71: nuclear shell model \cite{BFP1,JBDT,BFP2,MVZ,DD,DW,ZA,Zuker} and the IBM 
72: \cite{BF1,BF2,BFP2,KZC,DK,BF3}.   
73: 
74: In this contribution, we investigate the origin of the dominance of 
75: $L=0$ ground states that emerges from random interactions in two different 
76: models: the vibron model and the IBM. The vibron model is mathematically 
77: simpler than the IBM, but exhibits many of the same qualitative 
78: features. In this case, most of the results can be obtained in closed 
79: analytic form. For both models, we suggest to use a mean-field analysis 
80: to address the problem of the probability distribution of the ground state 
81: angular momenta \cite{BF3}. 
82: 
83: \section{The vibron model} 
84: 
85: The vibron model was introduced to describe the rotational and 
86: vibrational excitations of diatomic molecules \cite{vibron}. Its building 
87: blocks are a dipole boson $p^{\dagger}$ with $L^P=1^-$ and a 
88: scalar boson $s^{\dagger}$ with $L^P=0^+$. The total number of bosons $N$ 
89: is conserved by the vibron Hamiltonian. We only consider 
90: one- and two-body interactions 
91: \bea
92: H &=& \frac{1}{N} \left[ H_1 + \frac{1}{N-1} H_2 \right] ~, 
93: \label{h12}
94: \eea
95: with 
96: \bea
97: H_1 &=& \epsilon_s \, s^{\dagger} \cdot \tilde{s}  
98: - \epsilon_p \, p^{\dagger} \cdot \tilde{p} ~, 
99: \nonumber\\
100: H_2 &=& u_0 \, \frac{1}{2} \, 
101: (s^{\dagger} \times s^{\dagger})^{(0)} \cdot 
102: (\tilde{s} \times \tilde{s})^{(0)} 
103: + u_1 \, (s^{\dagger} \times p^{\dagger})^{(1)} \cdot 
104: (\tilde{p} \times \tilde{s})^{(1)} 
105: \nonumber\\ 
106: &+& \sum_{\lambda=0,2} c_{\lambda} \, \frac{1}{2} \, 
107: (p^{\dagger} \times p^{\dagger})^{(\lambda)} \cdot  
108: (\tilde{p} \times \tilde{p})^{(\lambda)} 
109: \nonumber\\
110: &+& v_0 \, \frac{1}{2\sqrt{2}} \, \left[ 
111:   (p^{\dagger} \times p^{\dagger})^{(0)} \cdot 
112:   (\tilde{s} \times \tilde{s})^{(0)} 
113: + H.c. \right] ~, 
114: \label{hvib}
115: \eea
116: Here $\tilde{s}=s$ and $\tilde{p}_m=(-1)^{1-m}p_{-m}$~. We have scaled 
117: $H_1$ by $N$ and $H_2$ by $N(N-1)$ in order to remove the $N$ dependence 
118: of the matrix elements. The seven parameters of the Hamiltonian,  
119: $\epsilon_s$, $\epsilon_p$, $u_0$, $u_1$, $c_0$, $c_2$, $v_0$,  
120: altogether denoted by $\vec{x}$, are taken as independent random numbers on 
121: a Gaussian distribution with zero mean. In this way, the interaction terms 
122: are arbitrary and equally likely to be attractive or repulsive. 
123: 
124: The connection between the vibron Hamiltonian, potential energy surfaces, 
125: equilibrium configurations and geometric shapes can be 
126: investigated by means of mean-field methods \cite{onno}. For the vibron 
127: model, the coherent state can be written as a condensate of deformed bosons, 
128: which are superpositions of scalar and dipole bosons 
129: \bea
130: \left| \, N,\alpha \, \right> \;=\; \frac{1}{\sqrt{N!}} \, 
131: \left( \cos \alpha \, s^{\dagger} + \sin \alpha \, p_0^{\dagger} 
132: \right)^N \, \left| \, 0 \, \right> ~, 
133: \label{trial}
134: \eea
135: with $0 \leq \alpha \leq \pi/2$. The potential energy surface is then given 
136: by the expectation value of the vibron Hamiltonian of Eqs.~(\ref{h12}) 
137: and (\ref{hvib}) in the coherent state 
138: \bea
139: E(\alpha) &=&  a_4 \, \sin^4 \alpha + a_2 \, \sin^2 \alpha + a_0 ~. 
140: \label{surface}
141: \eea
142: The coefficients $a_i$ are linear combinations of the parameters 
143: of the Hamiltonian 
144: \bea
145: a_4 &=& \vec{r} \cdot \vec{x} \;=\; \frac{1}{2} u_0 + u_1 
146: + \frac{1}{6} c_0 + \frac{1}{3} c_2 + \frac{1}{\sqrt{6}} v_0 ~, 
147: \nonumber\\
148: a_2 &=& \vec{s} \cdot \vec{x} \;=\; -\epsilon_s + \epsilon_p 
149: - u_0 - u_1 - \frac{1}{\sqrt{6}} v_0 ~, 
150: \nonumber\\
151: a_0 &=& \vec{t} \cdot \vec{x} \;=\; \epsilon_s + \frac{1}{2} u_0 ~. 
152: \label{coef}
153: \eea
154: For random interaction strengths, we expect the trial wave function of 
155: Eq.~(\ref{trial}) and the energy surface of Eq.~(\ref{surface}) to 
156: provide information on the distribution of shapes that the model can 
157: acquire. The value of $\alpha_0$ that characterizes the equilibrium 
158: configuration of the potential energy surface only depends on the 
159: coefficients $a_4$ and $a_2$. The $a_2 a_4$ plane 
160: can be divided into different areas according to the three possible 
161: equilibrium configurations: $S_1$ for the $s$-boson or spherical 
162: condensate ($\alpha_0=0$), $S_2$ for the deformed condensate 
163: ($0 < \alpha_0 < \pi/2$), and $S_3$ for the $p$-boson condensate 
164: ($\alpha_0=\pi/2$). The distribution of shapes for an ensemble of  
165: Hamiltonians depends on the joint probability distribution of the 
166: coefficients $a_4$ and $a_2$ which is given by a bivariate normal 
167: distribution 
168: \bea
169: P(a_4,a_2) &=& \frac{1}{2\pi \sqrt{|\det M|}} \, 
170: \mbox{exp} \left[-\frac{1}{2} 
171: \left( \begin{array}{cc} a_4 & a_2 \end{array} \right) M^{-1} 
172: \left( \begin{array}{c} a_4 \\ a_2 \end{array} \right) \right] ~, 
173: \label{bivariate}
174: \eea
175: with 
176: \bea
177: M &=& \left( \begin{array}{ccc} 
178: \vec{r} \cdot \vec{r} &\;\;\;& \vec{r} \cdot \vec{s} \\
179: \vec{r} \cdot \vec{s} && \vec{s} \cdot \vec{s} \end{array} 
180: \right) \;=\; \frac{1}{18} \left( \begin{array}{rrr} 
181: 28 &\;\;\;& -30 \\ -30 && 75 \end{array} \right) ~. 
182: \eea
183: The vectors $\vec{r}$ and $\vec{s}$ are defined in Eq.~(\ref{coef}). 
184: The probability that the equilibrium shape of an ensemble of Hamiltonians 
185: is spherical can be obtained by integrating $P(a_4,a_2)$ over 
186: the appropriate range $S_1$ ($a_2>0$ and $a_4>-a_2$) 
187: \bea
188: P_1 &=& \frac{1}{4\pi} \left[ \pi+2 \arctan \sqrt{\frac{27}{16}} \right]  
189: \;=\; 0.396 ~. 
190: \label{P1}
191: \eea
192: Similarly, the probability for the occurrence of a deformed shape is 
193: obtained by integrating $P(a_4,a_2)$ over $S_2$ ($-2a_4<a_2<0$)  
194: \bea
195: P_2 &=& \frac{1}{2\pi} \arctan \sqrt{\frac{64}{3}} \;=\; 0.216 ~. 
196: \label{P2}
197: \eea
198: Finally, the probability for finding the third solution, a $p$-boson 
199: condensate, is 
200: \bea
201: P_3 \;=\; 1-P_1-P_2 \;=\; 0.388 ~. 
202: \label{P3}
203: \eea
204: The rotational spectrum can be obtained from the angular momentum content 
205: of the condensate in combination with the Thouless-Valatin formula for 
206: the moment of inertia \cite{duke}. The ordering of the rotational energy 
207: levels is determined by the sign of the moment of inertia 
208: \bea
209: E_{\rm rot} &=& \frac{1}{2{\cal I}} L(L+1) ~.
210: \eea
211: 
212: In Table~\ref{percvib} we show the probability distribution of the ground 
213: state angular momenta as obtained in the mean-field analysis. There is a 
214: statistical preference for $L=0$ ground states. This is largely due to the 
215: occurrence of a spherical shape (whose angular momentum content is just 
216: $L=0$) for almost 40 $\%$ of the cases (see Eq.~(\ref{P1})). The deformed 
217: shape corresponds to a rotational band with $L=0,1,\ldots,N$, whose ground 
218: state has $L=0$ for positive values of the moment of inertia ${\cal I}>0$, 
219: and $L=N$ for ${\cal I}<0$. The third solution, the $p$-boson condensate, 
220: has angular momenta $L=N,N-2,\ldots,1$ or $0$. For ${\cal I}>0$, the ground 
221: state has $L=0$ or $L=1$, depending whether the total number of vibrons 
222: $N$ is even or odd, whereas for ${\cal I}<0$ it has the maximum value 
223: $L=N$. The sum of the $L=0$ and $L=1$ percentages is constant. 
224: 
225: \begin{table}
226: \centering
227: \caption[]{\small 
228: Percentages of ground states with $L=0$, $1$ and $N$, obtained 
229: in a mean-field analysis of the vibron model.}
230: \label{percvib}
231: \vspace{15pt}
232: \begin{tabular}{crrrrl}
233: \hline
234: & & & & & \\
235: \multicolumn{2}{c}{Shape} & $L=0$ & $L=1$ & $L=N$ & \\
236: & & & & & \\
237: \hline
238: & & & & & \\
239: $  \alpha_0=0$       & 39.6 $\,\%$ 
240: & 39.6 $\,\%$ &  0.0 $\,\%$ &  0.0 $\,\%$ & \\
241: & & & & & \\
242: $0< \alpha_0 <\pi/2$ & 21.6 $\,\%$ 
243: & 13.3 $\,\%$ &  0.0 $\,\%$ &  8.3 $\,\%$ & \\
244: & & & & & \\
245: $  \alpha_0=\pi/2$   & 38.8 $\,\%$ 
246:  & 17.9 $\,\%$ &  0.0 $\,\%$ & 20.9 $\,\%$ & $N=2k$ \\
247: &&  0.0 $\,\%$ & 17.9 $\,\%$ & 20.9 $\,\%$ & $N=2k+1$ \\
248: & & & & & \\
249: \hline
250: & & & & & \\
251: Total & 100.0 $\,\%$ & 70.8 $\,\%$ &  0.0 $\,\%$ & 29.2 $\,\%$ & $N=2k$ \\
252:       &              & 52.9 $\,\%$ & 17.9 $\,\%$ & 29.2 $\,\%$ & $N=2k+1$ \\
253: & & & & & \\
254: \hline
255: \end{tabular}
256: \end{table}
257: 
258: Fig.~\ref{vibgs} shows that the mean-field results for the 
259: percentages of $L=0$ and $L=1$ ground states 
260: are in excellent agreement with the exact ones. 
261: As is clear from the results presented in Table~\ref{percvib}, the 
262: fluctuations in the percentages of $L=0$ and $L=1$ ground states with $N$ 
263: are due to the contribution from the $p$-boson condensate solution. 
264: 
265: \begin{figure}
266: \centerline{\hbox{\epsfig{figure=cam_fig1.eps} }}
267: \caption[]{\small 
268: Percentages of ground states with $L=0$ and $L=1$ in the vibron 
269: model with random one- and two-body interactions calculated exactly for 
270: 100000 runs (solid lines) and in mean-field approximation (dashed lines).}
271: \label{vibgs}
272: \end{figure}
273: 
274: \section{The interacting boson model}
275: 
276: The interacting boson model (IBM) describes collective excitations in nuclei 
277: in terms of a system of $N$ interacting bosons \cite{IBM}. Its building 
278: blocks are a quadrupole boson $d^{\dagger}$ with $L^P=2^+$ and a 
279: scalar boson $s^{\dagger}$ with $L^P=0^+$. The total number of bosons 
280: $N$ is conserved by the IBM Hamiltonian. We consider the 
281: one- and two-body Hamiltonian of Eq.~(\ref{h12}) with 
282: \bea
283: H_1 &=& \epsilon_s \, s^{\dagger} \cdot \tilde{s}  
284: + \epsilon_d \, d^{\dagger} \cdot \tilde{d} ~, 
285: \nonumber\\
286: H_2 &=& \frac{1}{2} \, u_0 \, 
287: (s^{\dagger} \times s^{\dagger})^{(0)} \cdot 
288: (\tilde{s} \times \tilde{s})^{(0)} 
289: + u_2 \, (s^{\dagger} \times d^{\dagger})^{(2)} \cdot 
290: (\tilde{d} \times \tilde{s})^{(2)} 
291: \nonumber\\ 
292: &&+ \sum_{\lambda=0,2,4} \frac{1}{2} \, c_{\lambda} \, 
293: (d^{\dagger} \times d^{\dagger})^{(\lambda)} \cdot  
294: (\tilde{d} \times \tilde{d})^{(\lambda)} 
295: \nonumber\\
296: &&+ \frac{1}{2\sqrt{2}} \, v_0 \, \left[ 
297:   (d^{\dagger} \times d^{\dagger})^{(0)} \cdot 
298:   (\tilde{s} \times \tilde{s})^{(0)} 
299: + H.c. \right]
300: \nonumber\\
301: &&+ \frac{1}{2} \, v_2 \, \left[ 
302:   (d^{\dagger} \times d^{\dagger})^{(2)} \cdot 
303:   (\tilde{d} \times \tilde{s})^{(2)} 
304: + H.c. \right] ~.  
305: \label{hibm}
306: \eea
307: The nine coefficients $\epsilon_s$, $\epsilon_d$, $u_0$, $u_2$, 
308: $c_0$, $c_2$, $c_4$, $v_0$, $v_2$, are chosen independently from a Gaussian 
309: distribution of random numbers with zero mean \cite{BF1,BF2}. Just as for 
310: the vibron model, the connection between the IBM, potential energy surfaces, 
311: equilibrium configurations and geometric shapes, can be studied with 
312: mean-field Hartree-Bose techniques by means of coherent states \cite{duke,cs}. 
313: The coherent state can be written as an axially symmetric condensate 
314: \bea
315: \left| \, N,\alpha \, \right> \;=\; \frac{1}{\sqrt{N!}} 
316: \left( \cos \alpha \, s^{\dagger} + \sin \alpha \, d_0^{\dagger} 
317: \right)^N \, \left| \, 0 \, \right> ~, 
318: \eea
319: with $-\pi/2 < \alpha \leq \pi/2$. The angle $\alpha$ is related to 
320: the deformation parameters in the intrinsic frame, $\beta$ and $\gamma$ 
321: \cite{IBM,cs}. The potential energy surface is then 
322: given by the expectation value of the Hamiltonian in the coherent state 
323: \bea
324: E_N(\alpha) \;=\; a_4 \, \sin^4 \alpha + 
325: a_3 \, \sin^3 \alpha \cos \alpha + a_2 \, \sin^2 \alpha + a_0 ~.  
326: \eea
327: For the IBM, the structure of the energy surface 
328: is a bit more complicated than for the vibron model due to the presence 
329: of the $a_3$ term. This precludes an analytic treatment as presented 
330: for the vibron model, but qualitatively the results are very similar. 
331: In practice, for each Hamiltonian the minimum of the energy surface 
332: $E(\alpha)$ is determined numerically. Again, the equilibrium configurations 
333: can be divided into three different classes: an $s$-boson or spherical 
334: condensate ($\alpha_0=0$), a deformed condensate with prolate 
335: ($0 < \alpha_0 < \pi/2$) or or oblate symmetry ($-\pi/2 < \alpha_0 < 0$), 
336: and a $d$-boson condensate ($\alpha_0=\pi/2$). 
337: Each equilibrium configuration has its own characteristic angular momentum 
338: content. Even though we do not explicitly project the angular momentum 
339: states from the coherent state, the angular momentum of the ground state 
340: can, to a good approximation, be obtained from the rotational structure of 
341: the condensate in 
342: combination with the Thouless-Valatin formula for the corresponding moments 
343: of inertia \cite{duke}. The results are summarized in Table~\ref{percibm}. 
344: 
345: \begin{table}
346: \centering
347: \caption[]{\small 
348: Percentages of ground states with $L=0$, $2$ and $2N$, obtained 
349: in a mean-field analysis of the interacting boson model.}
350: \label{percibm}
351: \vspace{15pt}
352: \begin{tabular}{crrrrl}
353: \hline
354: & & & & & \\
355: \multicolumn{2}{c}{Shape} & $L=0$ & $L=2$ & $L=2N$ & \\
356: & & & & & \\
357: \hline
358: & & & & & \\
359: $  \alpha_0=0$       & 39.4 $\,\%$ 
360: & 39.4 $\,\%$ &  0.0 $\,\%$ &  0.0 $\,\%$ & \\
361: & & & & & \\
362: $0<|\alpha_0|<\pi/2$ & 36.8 $\,\%$ 
363: & 23.7 $\,\%$ &  0.0 $\,\%$ & 13.1 $\,\%$ & \\
364: & & & & & \\
365: $  \alpha_0=\pi/2$   & 23.8 $\,\%$ 
366:  & 13.5 $\,\%$ &  0.0 $\,\%$ & 10.3 $\,\%$ & $N=6k$ \\
367: &&  0.2 $\,\%$ & 13.2 $\,\%$ & 10.4 $\,\%$ & $N=6k+1,6k+5$ \\
368: &&  4.4 $\,\%$ &  9.0 $\,\%$ & 10.4 $\,\%$ & $N=6k+2,6k+4$ \\
369: &&  9.3 $\,\%$ &  4.0 $\,\%$ & 10.5 $\,\%$ & $N=6k+3$ \\
370: & & & & & \\
371: \hline
372: & & & & & \\
373: Total & 100.0 $\,\%$ 
374:  & 76.6 $\,\%$ &  0.0 $\,\%$ & 23.4 $\,\%$ & $N=6k$ \\
375: && 63.3 $\,\%$ & 13.2 $\,\%$ & 23.5 $\,\%$ & $N=6k+1,6k+5$ \\
376: && 67.5 $\,\%$ &  9.0 $\,\%$ & 23.5 $\,\%$ & $N=6k+2,6k+4$ \\
377: && 72.4 $\,\%$ &  4.0 $\,\%$ & 23.6 $\,\%$ & $N=6k+3$ \\
378: & & & & & \\
379: \hline
380: \end{tabular}
381: \end{table}
382: 
383: (i) The $s$-boson condensate corresponds to a spherical shape. 
384: Whenever such a condensate occurs (in 39.4 $\%$ of the cases), 
385: the ground state has $L=0$.
386: 
387: (ii) The deformed condensate corresponds to an axially symmetric deformed 
388: rotor. The ordering of the rotational energy levels $L=0,2,\ldots,2N$ is 
389: determined by the sign of the moment of inertia 
390: \bea
391: E_{\rm rot} \;=\;  \frac{1}{2{\cal I}_3} L(L+1) ~. 
392: \eea
393: The deformed condensate occurs in 36.8 $\%$ of the cases. For ${\cal I}_3>0$ 
394: the ground state has $L=0$ (23.7 $\%$), while for ${\cal I}_3<0$ the ground 
395: state has the maximum value of the angular momentum $L=2N$ (13.1 $\%$). 
396: 
397: (iii) The $d$-boson condensate corresponds to a quadrupole oscillator with 
398: $N$ quanta. Its rotational structure has a more complicated structure 
399: than the previous two cases. It is characterized by the labels $\tau$, 
400: $n_{\Delta}$ and $L$. The boson seniority $\tau$ is given by 
401: $\tau=3n_{\Delta}+\lambda=N,N-2,\ldots,1$ or 
402: $0$ for $N$ odd or even, and the values of the angular momenta are 
403: $L=\lambda,\lambda+1,\ldots,2\lambda-2,2\lambda$ \cite{IBM}. In this case, 
404: the rotational excitation energies depend on two moments of inertia 
405: \bea
406: E_{\rm rot} \;=\; \frac{1}{2{\cal I}_5} \tau(\tau+3) 
407: + \frac{1}{2{\cal I}_3} L(L+1) ~. 
408: \eea
409: The $d$-boson condensate occurs in 23.8 $\%$ of the cases. 
410: For ${\cal I}_5>0$ the ground state has $\tau=0$ 
411: for $N$ even or $\tau=1$ for $N$ odd ($\sim$ 4 $\%$), while for 
412: ${\cal I}_5<0$ the ground state has the maximum value of the boson 
413: seniority $\tau=N$ ($\sim$ 19 $\%$). For $\tau=0$ and $\tau=1$ 
414: there is a single angular momentum state with $L=0$ and $L=2$, 
415: respectively. For  the $\tau=N$ multiplet, the angular momentum of the 
416: ground state depends on the sign of the moment of inertia ${\cal I}_3$. 
417: For ${\cal I}_3>0$ the ground state has $L=0$ for $N=3k$ or 
418: $L=2$ for $N \neq 3k$ ($\sim$ 9 $\%$), while for ${\cal I}_3<0$ the ground 
419: state has the maximum value of the angular momentum $L=2N$ ($\sim$ 10 $\%$). 
420: 
421: \begin{figure}
422: \centerline{\hbox{\epsfig{figure=cam_fig2.eps} }}
423: \caption[]{\small 
424: Percentages of ground states with $L=0$ and $L=2$ in the IBM 
425: with random one- and two-body interactions calculated exactly for 
426: 10000 runs (solid lines) and in mean-field approximation (dashed lines).}
427: \label{ibmgs}
428: \end{figure}
429: 
430: In Fig.~\ref{ibmgs} we show the percentages of ground states with 
431: $L=0$ and $L=2$ as a function of the total number of bosons $N$. 
432: A comparison of the results of the mean-field analysis (dashed lines) 
433: and the exact ones (solid lines) shows a good agreement. There is a 
434: dominance of ground states with $L=0$. The oscillations of the $L=0$ and 
435: $L=2$ percentages with $N$ are due to the contribution of the $d$-boson 
436: condensate (see Table~\ref{percibm}). For $N=3k$ we see an enhancement 
437: for $L=0$ and a corresponding decrease for $L=2$. In the mean-field 
438: analysis, the sum of the two hardly depends on the number of bosons, 
439: in agreement with the exact results. 
440: 
441: \section{Summary and conclusions}
442: 
443: In this contribution, we have studied the origin of the regular features 
444: that were obtained before in numerical studies of the IBM with random 
445: interactions. In particular, we addressed the dominance of $L=0$ ground 
446: states in the context of the vibron model and the IBM. In a mean-field 
447: analysis it was found that different regions of the parameter space can 
448: be associated with definite geometric shapes: a spherical shape, a deformed 
449: shape and a condensate of dipole (quadrupole) bosons for the vibron model 
450: (IBM). For both models, we obtained a good description of the probability 
451: distribution of ground state angular momenta. 
452: 
453: The studies with random interactions indicate 
454: that there is a significantly larger class of Hamiltonians that leads to 
455: regular, ordered behavior at the low excitation energies than was commonly 
456: assumed. The fact that these properties are shared by different models, 
457: seems to exclude an explanation solely in terms of the angular momentum 
458: algebra, the connectivity of the model space, or the many-body dynamics of 
459: the model, as has been suggested before. The present analysis points, at 
460: least for systems of interacting bosons, to a more general phenomenon that 
461: does not depend so much on the details of the angular momentum coupling, 
462: but rather on the occurrence of definite, robust geometric phases such as 
463: spherical and deformed shapes. For the nuclear shell model the situation 
464: is less clear. Although a large number of investigations to explain and 
465: further explore the properties of random nuclei have shed light on various 
466: aspects of the original problem, i.e. the dominance of $L=0$ ground states, 
467: in our opinion, no definite answer is yet available, and the full 
468: implications for nuclear structure physics are still to be clarified.  
469: 
470: In conclusion, the results presented in this article for the vibron model 
471: and the IBM may help to understand the origin of ordered spectra arising 
472: from random interactions, as has been observed in numerical studies of 
473: nuclear structure models.  
474: 
475: \section*{Acknowledgements}
476: 
477: This work was supported in part by CONACyT under projects 
478: 32416-E and 32397-E, and by DPAGA under project IN106400. 
479: 
480: \begin{thebibliography}{aa}
481: 
482: \bibitem{JBD}
483: C.W. Johnson, G.F. Bertsch and D.J. Dean,
484: Phys. Rev. Lett. {\bf 80}, 2749 (1998).
485: 
486: \bibitem{BF1}
487: R. Bijker and  A. Frank,
488: Phys. Rev. Lett. {\bf 84}, (2000), 420.
489: 
490: \bibitem{BF2} 
491: R. Bijker and A. Frank, 
492: Phys. Rev. C {\bf 62}, 014303 (2000).
493:  
494: \bibitem{BFP1}
495: R. Bijker, A. Frank and S. Pittel,
496: Phys. Rev. C {\bf 60}, 021302 (1999).
497: 
498: \bibitem{JBDT}
499: C.W. Johnson, G.F. Bertsch, D.J. Dean and I. Talmi,
500: Phys. Rev. C {\bf 61}, 014311 (2000).
501: 
502: \bibitem{BFP2}
503: R. Bijker, A. Frank and S. Pittel, 
504: Rev. Mex. F{\'{\i}}s. {\bf 46 S1}, 47 (2000). 
505: 
506: \bibitem{MVZ} 
507: D. Mulhall, A. Volya and V. Zelevinsky, 
508: Phys. Rev. Lett. {\bf 85}, 4016 (2000). 
509: 
510: \bibitem{DD}
511: D. Dean, 
512: Nucl. Phys. A {\bf 682}, 194c (2001). 
513: 
514: \bibitem{DW}
515: S. Dro\v{z}d\v{z} and M. W\'ojcik, 
516: preprint nucl-th/0007045.
517: 
518: \bibitem{ZA}
519: Y.M. Zhao and A. Arima, 
520: Phys. Rev. C {\bf 64}, 041301 (2001). 
521: 
522: \bibitem{Zuker}
523: V. Vel\'azquez and A.P. Zuker, 
524: preprint nucl-th/0106020. 
525: 
526: \bibitem{KZC}
527: D. Kusnezov, N.V. Zamfir and R.F. Casten, 
528: Phys. Rev. Lett. {\bf 85}, 1396 (2000). 
529: 
530: \bibitem{DK}
531: D. Kusnezov, 
532: Phys. Rev. Lett. {\bf 85}, 3773 (2000); \\
533: R. Bijker and A. Frank, 
534: Phys. Rev. Lett {\bf 87}, 029201 (2001); \\
535: D. Kusnezov, 
536: Phys. Rev. Lett {\bf 87}, 029202 (2001). 
537: 
538: \bibitem{BF3} 
539: R. Bijker and A. Frank, 
540: Phys. Rev. C {\bf 64}, 061303 (2001). 
541: 
542: \bibitem{vibron}
543: F. Iachello and R.D. Levine, 
544: {\it Algebraic theory of molecules} 
545: (Oxford University Press, Oxford, 1995). 
546: 
547: \bibitem{onno}
548: O.S. van Roosmalen and A.E.L. Dieperink, 
549: Ann. Phys. (N.Y.) {\bf 139}, 198 (1982); \\
550: O.S. van Roosmalen, Ph.D. thesis, Univ. of Groningen (1982). 
551: 
552: \bibitem{duke}
553: J. Dukelsky, G.G. Dussel, R.P.J. Perazzo, S.L. Reich and H.M. Sofia, 
554: Nucl. Phys. A {\bf 425}, 93 (1984). 
555: 
556: \bibitem{IBM}
557: F. Iachello and A. Arima,
558: {\it The interacting boson model}
559: (Cambridge University Press, Cambridge, 1987).
560: 
561: \bibitem{cs}
562: J.N. Ginocchio and M. Kirson, 
563: Phys. Rev. Lett. {\bf 44}, 1744 (1980); 
564: A.E.L. Dieperink, O. Scholten and F. Iachello, 
565: Phys. Rev. Lett. {\bf 44}, 1747 (1980). 
566: 
567: \end{thebibliography}
568: 
569: \end{document}
570: