1: \documentclass[aps,preprint]{revtex4}
2: \usepackage [dvips]{graphicx}
3: \newcommand{\ds}{\displaystyle}
4: \newcommand{\be}{\begin{equation}}
5: \newcommand{\ee}{\end{equation}}
6: \newcommand{\ba}{\begin{eqnarray}}
7: \newcommand{\ea}{\end{eqnarray}}
8: \newcommand{\bo}{$^{10}$B }
9: \newcommand{\nit}{$^{14}$N }
10: \newcommand{\ca}{$^{14}$C }
11: \newcommand{\caa}{$^{12}$C }
12: \newcommand{\bi}{\bibitem}
13: \newcommand{\pr}[1]{Phys. Rev. Lett. {\bf#1}}
14: \newcommand{\pc}[1]{Phys. Rev. {\bf#1}}
15: \newcommand{\np}[1]{Nucl. Phys. {\bf#1}}
16: \begin{document}
17: \title{T=0 effective interaction in \nit and \bo.}
18: \author{N. Vinh Mau}
19: \affiliation{Institut de Physique Nucl\'eaire, IN2P3-CNRS,
20: Universite Paris-Sud, F-91406 Orsay Cedex, France}
21:
22: \begin{abstract}
23: We have calculated the 1$^+$ and 3$^+$, T=0 states in \nit and \bo. In a
24: neutron-proton RPA model these two nuclei are described by the same set
25: of equations. We first show that a bare Minnesota
26: interaction leads to too weak binding in both nuclei. Furthermore it does
27: not produce a
28: 3$^+$ ground state in \bo as it should. Including medium
29: effects as an exchange of phonons between
30: the neutron-proton pair cures the desagreement in \nit but still gives a
31: 1$^+$ ground state in \bo with the 3$^+$ as an excited state. The same study
32: with a Gogny effective interaction reproduces nicely the properties of both
33: nuclei: same agreement in \nit as previously when medium effeccts were
34: introduced but now the 3$^+$ in \bo becomes the ground state. This
35: success suggests that through its density dependent term the Gogny
36: interaction takes account of the presence of a three-body force which, in a
37: shell model calculation, has
38: been shown to be essential to give a 3$^+$ ground state in \bo.
39: \end{abstract}
40:
41: \maketitle
42: \section{Introduction.}
43:
44: The T=0 neutron-proton pairing correlations in N=Z nuclei have
45: motivated a number of theoretical and experimental works. In odd-odd N=Z
46: nuclei,
47: T=0 and T=1 states are both present so that they offer a good place to
48: make a comparative study of the two types of correlations. We have focused
49: our attention on the two odd-odd nuclei of \bo (N=Z=5) and \nit (N=Z=7)
50: for several reasons. First \nit and
51: \bo have a 1$^+$, T=0 and a 3$^+$ T=0 ground states
52: respectively
53: while the 0$^+$, T=1 states are excited states and therefore
54: suggesting stronger correlations in the T=0 channel.
55: Furthermore because of charge invariance of the nucleon-nucleon interaction,
56: the neutron-proton and neutron-neutron effective interactions should
57: be the same in the T=1 channel.
58: Therefore an effective interaction and a nuclear model able to explain the
59: 0$^+$, T=1 states in a core plus two neutrons system (like \ca) should also
60: describe well the same T=1 states in the core plus a neutron-proton
61: pair (like \nit). We
62: have then an opportunity to make a further test of our previous work
63: in collaboration with
64: J.C. Pacheco on N=8 nuclei \cite{pv}. With the Gogny effective
65: interaction \cite{goa,gob}or its zero range density dependent
66: substitute \cite{sca},
67: and in a two-neutron RPA model we were able to
68: reproduce the 0$^+$ states of \ca-$^{10}$C, $^{12}$Be-$^{8}$Be
69: and $^{11}$Li-$^7$Li. The nucleus \nit ($^{10}$B)
70: described as a core of $^{12}$C plus (minus) a neutron-proton pair is the
71: analog of \ca ($^{10}$C) described as a core of $^{12}$C plus (minus)
72: a neutron-neutron pair
73: and our
74: first aim is to check if the same interactions and the same RPA model where
75: the two-neutron pair is replaced by a neutron-proton pair, will lead
76: also to a good description of the 0$^{+}$ states in \nit and \bo.
77:
78: In the two-nucleon RPA
79: model, successful in the description of the T=1 states, we calculate the T=0,
80: 1$^+$ and 3$^+$ states in \nit and \bo. First we perform the calculation
81: with a simple zero range force fitted to reproduce the deuteron binding energy.
82: The results obtained
83: with this simple potential show a too weak binding.
84: Because a zero range force
85: is a crude simplification of a realistic one, we have made the same
86: calculation with the Minnesota bare interaction \cite{min} which has one
87: short range
88: component, two long range components and all exchange terms. The results
89: are very close to the previous ones and show again a lack of binding.
90: It is not surprising that bare
91: interactions could not describe well bound many-body systems
92: for which we know that medium effects are important.
93: Medium effects on the bare nucleon- nucleon interaction may arise, at least
94: partly, from the exchange of phonons between the two-nucleon pair.
95: Indeed strong
96: two-body correlations in nuclei manifest themselves as very collective
97: vibrationnal states at low excitation energies. These vibrations, or phonons,
98: may be exchanged between the two nucleons and induce a modification of the
99: nuclear interaction. The presence of these phonons, besides their
100: effect on the nuclear two-body force, has an other manifestation by
101: modifying the average interaction of a single nucleon with the core due to the
102: coupling between the nucleon and the phonons of the core.
103: Such couplings
104: modify the Hartree-Fock average potential and are so strong in
105: nuclei like $^{11}$Be and $^{10}$Li that they are, at least partly,
106: responsible for the
107: inversion of the 1/2$^+$ and 1/2$^-$ states \cite{sa,vm,br}.
108: Both effects, on the
109: two-nucleon force and on the nucleon-core interaction will be included in our
110: two-particle RPA model for the description of \nit and \bo.
111:
112: The spectra obtained in $^{14}$N and $^{10}$B will be compared to those
113: derived with the phenomenological T=0 Gogny effective interaction
114: \cite{goa,gob}. From this comparison and comparison with shell model
115: calculations \cite{arn} using two-body and three-body interactions, we will
116: try to understand better what is implicitly contained in this empirical
117: force.
118:
119: In section II we present briefly the two-particle RPA model applied to a
120: neutron-proton pair and precise
121: our choice of single neutron and proton basis and our choice of two-body
122: interactions. In section III we first present the results for the $0^+$,
123: T=1 states. Then we show the results obtained for the T=0 states with the
124: two bare nucleon-nucleon interactions and discuss the contribution of
125: phonon exchanges between the neutron-proton pair.
126: Then we make the same study with the
127: effective Gogny interaction. Comparing our results with those
128: of shell model calculations we can give a qualitative interpretation of the
129: medium effects contained in this very successful effective interaction.
130: Section IV is devoted to our general conclusions.
131:
132: \section{The neutron-proton RPA model}
133:
134: We describe \nit and \bo as a core of $^{12}$C in its ground state plus or
135: minus a neutron-proton pair respectively and define two-body amplitudes as:
136:
137: \ba
138: X^+_a&=&\langle ^{14}N|A^+_a|^{12}C\rangle \\
139: Y^+_\alpha&=&\langle^{14}N|A^+_\alpha|^{12}C\rangle\\
140: X^-_\alpha&=& \langle^{10}B|A_\alpha|^{12}C\rangle\\
141: Y^-_a&=&\langle^{10}B |A_a|^{12}C\rangle
142: \ea
143: where a,b\ldots and $\alpha$,$\beta$,\ldots represent configurations where
144: the neutron and the proton are respectively in states unoccupied and
145: occupied in the Hartree Fock ground state of the \caa core. A$^+_{a(\alpha)}$
146: (A$_{a(\alpha)}$) are
147: operators which create (annihilate) a neutron-proton pair coupled to given
148: spin and isospin (for simplicity we omit them in our equations). $|^{12}C>$
149: represents the correlated ground state of the $^{12}C$-core.
150: Note that the Y-amplitudes,named the small RPA amplitudes, would be zero for
151: an uncorrelated core.
152:
153: The two-nucleon RPA model has been described and used in a number of papers
154: and we remind briefly that:
155:
156: -the same system of equations determines the amplitudes and energies of \nit
157: and \bo which then are not independent. For a given spin and isospin these
158: equations write as:
159: \ba
160: (\Omega -\epsilon_a)x_a-\sum_b<a|{V}|b>x_b-\sum_{\beta}
161: <a|{V}|\beta>x_\beta=0\\
162: (\Omega-\epsilon_\alpha)x_\alpha+\sum_b<\alpha|{V}|b>x_b+\sum _\beta
163: <\alpha|{V}|\beta>x_\beta=0
164: \ea
165: where the eigenvalues,$x_a$ and $x_\alpha$, are related to the amplitudes of
166: eqs.(1-4) as explained below. The $\epsilon_a$ and $\epsilon_\alpha$ are
167: the unperturbed energies of the neutron-proton pair in states a and $\alpha$
168: respectively.The matrix elements of the neutron-proton
169: interaction V have to be antisymmetrised.
170:
171: -if the model subspace contains N configurations a,b,.. and M
172: configurations $\alpha$, $\beta$ \ldots , the RPA equations
173: have N+M eigenstates with eigenvalues $\Omega$ and eigenvectors x$_a$ and
174: x$_{\alpha}$. N of them correspond to \nit with:
175:
176: \ba
177: &~~&E_n(^{14}N)-E_0(^{12}C)=\Omega_n \nonumber \\
178: &~~&X^{+(n)}_a=x_a^{(n)}\\
179: &~~&Y^{+(n)}_\alpha=x_\alpha^{(n)}\nonumber
180: \ea
181: and M to \bo with:
182: \ba
183: &~~&E_m(^{10}B)-E_0(^{12}C)=-\Omega_m \nonumber \\
184: &~~&X^{-(m)}_\alpha=x_\alpha^{(m)}\\
185: &~~&Y^{-(m)}_a=x_a^{(m)}\nonumber
186: \ea
187: E$_0$(\caa) is the ground state energy of the $^{12}$C-core.
188: The separation into two sets of solutions is unambigous and based on
189: energy considerations and relative values of $x_a $ and $x_{\alpha}$
190: amplitudes.
191:
192: - the amplitudes x$_a$ and x$_\alpha$ are normalised according to:
193:
194: \ba
195: \sum_a |x_a|^2-\sum_\alpha|x_\alpha|^2&=&1 \;\;\; for \;\;^{14}N\\
196: &=&-1 \;\;\; for \;\;^{10}B
197: \ea
198:
199: -the only inputs of the calculation are the individual neutron and proton
200: energies and the effective two-body interaction.
201:
202: \subsection{Choice of the single particle basis}
203:
204: We make a semi-phenomenogical approach to the RPA model. We replace the
205: Hartree Fock average potential by a Saxon-Woods potential plus a spin-orbit
206: force plus a phenomenological surface potential fitted to
207: reproduce
208: the experimental single neutron energies in the field of \caa
209: and write the one
210: neutron hamiltonian as:
211: \be
212: h_n=t_n+V_0\left(f(r)-0.44 r_0^2({\bf l.s}) \frac{1}{r} \frac{df(r)}{dr}\right
213: )+\delta V_n
214: \ee
215: where:
216: \be
217: f(r)=[1+exp(\frac{r-R_0}{a})]^{-1}
218: \ee
219: with V$_0$=-50.5 MeV, a=0.75 fm, R$_0$=r$_0$(12)$^{1/3}$ with r$_0$=1.27 fm.
220: The last term
221: $\delta V_n$ is added to simulate medium effects due to the coupling of the
222: neutron with the phonons of the core.. The
223: shape of $\delta V_n$ is suggested by a semi-microscopic calculation
224: of neutron-
225: phonon couplings \cite{vm,vp}. This contribution to the average one body potential
226: depends on the neutron state and is written as:
227: \be
228: \delta V_n= \alpha_n\left(\frac{df(r)}{dr}\right)^2
229: \ee
230:
231: \begin{table}
232: \begin{center}
233: \begin{tabular}{|c|cc|cc|cc|cc|} \hline
234: &\multicolumn{2}{c|}{1p$_{3/2}$}&
235: \multicolumn{2}{c|}{1p$_{1/2}$}&
236: \multicolumn{2}{c|}{2s}&
237: \multicolumn{2}{c|}{1d$_{5/2}$}\\
238: &exp.&cal.&exp. &cal. &exp. &cal. & exp. & cal.\\ \hline
239: neutrons& -18.72&-18.71&-4.95 & -4.95 & -1.85 & -1.86 & -1.1 &-1.1\\
240: protons&-15.96&-15.54& -1.94 & -1.98 & 0.42 & 0.25 & 1.61 & 1.35 \\ \hline
241: \end{tabular}
242: \caption{Experimental and calculated individual energies (in MeV) for
243: neutrons and protons.}
244: \end{center}
245: \end{table}
246:
247: The coefficients $\alpha_n$ are fitted on the experimental neutron energies
248: for 1p$_{3/2}$, 1p$_{1/2}$, 2s and 1d$_{5/2}$ states and put to zero for
249: higher states. The proton energies are calculated by adding the
250: Coulomb potential to the neutron hamiltonian of eq.(11). In Table I are
251: given the experimental energies and
252: the corresponding calculated energies for the lowest, known,
253: neutrons and protons states. The results presented below have been obtained
254: with the 1p$_{3/2}$ proton energy replaced by the experimental value. However
255: it has no effect on the \nit spectrum and in \bo gives both 1$^+$ and 3$^+$
256: energies are lowered by about 0.5 MeV what does not change qualitatively our
257: discussion of results.
258: The effect of
259: the Coulomb potential on the wave functions is neglected and
260: neutron and proton wave functions are assumed to be the same.
261:
262: \subsection{Choice of two-body interactions.}
263: We have first used a zero range density dependent interaction.
264: The general form of such a force writes as:
265: \be
266: V({\bf{r_1,r_2}})=-V_0\left\{1-\eta \left
267: (\frac{\rho(\bf{(r_1+r_2)}/2)}{\rho_0}\right)^{\alpha}\right \}
268: \delta(\bf{r_1-r_2})
269: \ee
270:
271: With the parameters of Garrido et al. \cite{scb}, $V_0$=500 MeV.fm$^3$,
272: $\alpha$=0.47
273: ,$\eta$=-0.1 we get much too weak binding. These parameters were fitted so
274: that to reproduce the gap in nuclear matter calculated using the Paris
275: potential, then they do not take account of the medium effects which are
276: expected to be important. Therefore we have proceeded in a
277: different way. We first use a density independent zero range neutron-proton
278: interaction
279: where the strength $V_0$ is fitted to give the binding energy of deuteron.
280: The calculation leads to a relation between the T=0 and T=1 strengths given
281: by \cite{be}:
282:
283: \be
284: V_0(T=0)=V_0(T=1)\left \{ 1-\left (\frac{-\epsilon_b}{\epsilon_c}\right )^{1/2}
285: Arctg\left (\frac{\epsilon_c}{-\epsilon_b}\right )^{1/2}\right \}^{-1}
286: \ee
287: where $\epsilon_b$ is the deuteron binding energy and $\epsilon_c$
288: the cut-off on
289: the nucleon energies which in a nucleus should be counted relative to
290: the bottom of the average one-nucleon potential. In our calculations we take
291: for a nucleon in the field of our $^{12}$C-core
292: a cut-off of 10 MeV which is equivalent to $\epsilon_c$=60 Mev and
293: corresponds to:
294: \ba
295: V_0(T=0)&=&1.36V_0(T=1) \\
296: &=&682 MeV.fm^3
297: \ea
298: A similar ratio between the strengths of the T=0 and T=1 pairing
299: interactions has been found by Satula and Wyss \cite{we} in their study of
300: even-even nuclei. However a
301: zero range force can be considered only as a simple substitute to a more
302: realistic one, and before introducing medium effects in our calculations
303: we have tested this substitution and made the calculation with the Minnesota
304: free nucleon-nucleon force \cite{min}. This force has one short range repulsive
305: components and two long range attractive ones with all exchange terms and
306: has been fitted on nucleon-nucleon scattering lengths and deuteron binding
307: energy.
308: The results with the two forces show very similar results with the same
309: underestimation of two-body correlations suggesting the necessity to take
310: account of medium effects on the neutron-proton interaction. Indeed when two
311: nucleons are
312: added to a core their mutual interaction will be modified by the
313: presence of the other nucleons. Strong two-body correlations which
314: manifest themselves as very
315: collective low energy vibrationnal states can induce a
316: modification of the interaction through an exchange of these collective
317: phonons between the pair of nucleons. Such phonon exchange contribution
318: to the two-neutron pairing force have been studied by
319: Barranco et al. \cite{bb}
320: and found to be responsible
321: for about half of the gap in the isotopes $^A$Ca, $^A$Ti and $^A$Sn.
322: In our case the $^{12}$C core has a
323: low 2$^+$, T=0 state at 4.4 MeV with a very strong collective transition
324: amplitude
325: $\beta_2$=0.6 and a less collective 3$^-$, T=0 state at 9.6 MeV with
326: $\beta_3$=0.4 \cite{sag}. We can expect that these two states will give
327: most of the effect and we include both of them in our calculation.
328:
329: \begin{figure}[h]
330: \hspace*{-1.cm}
331: \includegraphics*[scale=0.9]{pict3.ps}
332: \caption{Diagrams corresponding to a one phonon exchange between
333: neutron-proton pairs appearing in the RPA equations.}
334: \end{figure}
335:
336: The diagrams corresponding to the exchange of phonons
337: are represented in Fig.1 for the three types of
338: matrix elements entering in the RPA equations. The diagrams a) -b)
339: concern the matrix elements $<a|V|b>$ while the diagrams c) -d) and e) -f)
340: concern respectively the matrix elements $<\alpha|V|a>$ and
341: $<\alpha|V|\beta>$ of eqs.(5-6)
342: . The calculation of their
343: contribution to the RPA matrix elements is given in the Appendix when
344: each vertex of the diagrams is
345: replaced by a phenomenological anzats. These matrix elements which have
346: to be added to the bare matrix elements depend on the eigenvalues of the RPA
347: equations which therefore are now nonlinear equations and will be solved
348: by iteration.
349:
350: Once one has seen the effect of phonon exchanges on the \nit and \bo
351: spectra, the last step is to look if a phenomenological
352: effective interaction such as the Gogny interaction may be interpreted as a
353: bare interaction corrected by the diagrams of Fig.1. In the T=0 channel the
354: Gogny interactions, D1 \cite{goa} and D1S \cite{gob}, have a density
355: dependent component of zero
356: range. As any zero range interaction, it diverges and its fitted strength
357: depends on the cut-off of the single nucleon energies. Since our cut-off
358: for neutrons and protons is quite low (10 MeV) we have used a
359: slightly increased strength, as will be discussed below.
360:
361:
362:
363: \section{Results}
364:
365: Our configuration subspace is restricted to neutron states up to 10 MeV
366: and to the corresponding proton states.
367:
368: \subsection{0$^+$, T=1 states}
369:
370: Previous calculations \cite{pv} have shown that the 0$^+$, T=1 states
371: in the Li, Be
372: and C isotopes are very well described in the two-neutron RPA model with the
373: Gogny forces or their zero range density dependent substitute \cite{sca}.
374: \nit and \bo are the analogs of $^{14}$C-$^{10}$C where the two-neutron pair
375: is replaced by a neutron-proton pair added or substracted from a $^{12}$C-
376: core. Because of charge invariance of strong interactions the same T=1
377: effective interactions should be able to describe the two kinds of nuclear
378: systems. The calculation has been performed for \nit and \bo and yields to
379: the results reported in Table II for the zero range density dependent force
380: and the Gogny interactions D1 and D1S. We see that the three forces give
381: close results as already found in ref[1] for $^{14}$C and a good agreement
382: with the measured energies \cite{exp,ex,exp1} even though the excited
383: 0$^+$ state is slightly too
384: high as it was in $^{14}$C. We may conclude that we have a good test of the
385: validity of our two-body RPA model as well as a further confirmation of the
386: efficiency of the Gogny effective interactions to describe light nuclei for
387: which it was not designed.
388:
389: \begin{table}
390: \begin{center}
391: \begin{tabular}{|c|cccc|}\hline
392: & a)& b)&c)&exp.)\\ \hline
393: $^{14}$N &-10.12&-10.18&-10.03&-10.14\\
394: & -3.28&-3.10&-3.06&-3.8\\ \hline
395: $^{10}$B&29.7&29.0&29.24&29.15\\ \hline
396: \end{tabular}
397: \caption{Energies(in MeV) of the 0$^+$,T=1 states with respect to the ground
398: state of $^{12}$C in $^{14}$N and $^{10}$B obtained with: a) the zero range
399: density dependent force, b) and c) the D1 and D1S Gogny interaction
400: respectively. In the last column are given the experimental energies. }
401: \end{center}
402: \end{table}
403:
404:
405: \subsection{1$^+$ and 3$^+$, T=0 states}
406:
407: All results of this section are presented in Tables III-V and Figures 2 and
408: 3 for \nit and \bo.
409:
410: With the zero range interaction of eq.(14) and the parameters of ref.[11],
411: V$_0$=500 MeV.fm$^3$, $\eta$ =-0.1 and $\alpha$=0.2, our
412: results desagree with experimental spectra. In particular the lowest
413: 1$^+$ state in \nit is above the lowest 0$^+$ T=1
414: which is then the ground state in desagreement with experiments \cite{exp}.
415:
416:
417: \begin{table}
418: \begin{center}
419: \begin{tabular}{|cc|cccccc|} \hline
420: & J$^\varpi $&a)&b)& c)&d)&e)&exp. \\ \hline
421: $^{14}$N &1$^+$ &-11.7&-11.9&-12.4&-12.8&-12.5\\
422: &1$^+$ & -5.8 &-4.61&-5.75 &-6.9&-6.9&-6.3\\
423: & 3$^+$ &-3.26&-4.27&-5.55&-7.25&-7.2&-6.05\\ \hline
424: $^{10}$B&1$^+$ &27.6&28.7&28.1&29.1&28.4&28.1\\
425: & 3$^+$ & 30.3 & 29. & 28.6&28.3&27.7 & 27.4 \\ \hline
426: \end{tabular}
427: \caption{Energies of the lowest 1$^+$ and 3$^+$,T=0 states in $^{14}$N and
428: $^{10}$B with respect
429: to the ground state energy of $^{12}$C obtained with different interactions:
430: a) the bare zero-range interaction -b) the bare Minnesota interaction -c)the
431: Minnesota force plus the exchange of phonons -d) and e) the Gogny force
432: with respectively the experimental and calculated 1p$_{3/2}$ proton energy .
433: In the last column are given the experimental energies.}
434: \end{center}
435: \end{table}
436:
437:
438: Instead we have first made the calculations using the bare density
439: independent zero range neutron-proton interaction with the strength
440: $V_0$=682 MeV.fm$^3$ of eq.(17)
441: and with the Minesota force \cite{min}. The energies of
442: the lowest states referred to the theoretical ground state energy of
443: $^{12}$C as defined in eqs.(7-8) are given in the table III.
444: We see that the energies are very similar
445: for the two bare interactions but that there is a significant desagreement
446: with experimental energies for both nuclei. The desagreement is still more
447: pronounced for \bo where the ground state is found as a 1$^+$ state while
448: experimentally it is a 3$^+$ state. As expected we see clearly
449: that a free neutron-proton interaction is not sufficient to give the
450: binding of two nucleons inside a nucleus. Then
451: we have introduced medium effects due to the exchange of phonons as
452: explained in section II.B and in the Appendix and have calculated the induced
453: matrix elements
454: according to eqs.(21-29). The results are again very similar for the two
455: interactions and we show
456: in the tables and figures those obtained with the
457: Minnesota interaction. The excitation spectra are shown in Fig.2 for
458: $^{14}$N and Fig.3 for \bo for the bare and bare plus induced
459: interaction. At the bottom of the figures are given the
460: separation energies of a neutron-proton pair in \nit and $^{12}$C which are
461: directly related to the lowest RPA energies in \nit and \bo respectively.
462: \begin{figure}[h]
463: \hspace{-2.5cm}
464: \includegraphics*[angle=-90.,scale=0.5]{fig07_1.ps}
465: \caption{Excited spectrum of \nit calculated with: a) the bare Minnesota
466: interaction-b) the same with the exchange of phonons-c) the Gogny
467: interaction ( D1S with t$_3$=1600 MeV.fm$^3$. The experimental spectrum is
468: given in the last column. Below are given the ccalculated and experimental
469: neutron-proton separation energies. }
470: \end{figure}
471:
472: \begin{figure}[h]
473: \hspace{-2.5cm}
474: \includegraphics*[angle=-90.,scale=0.5]{fig07_2.ps}
475: \caption{Same legend as Fig.2 for \bo. The numbers at the bottom of the
476: figure are the neutron-proton separation energies, calculated and
477: experimental, in $^{12}$C.}
478: \end{figure}
479:
480: We see an improvement for the \nit excited spectrum as well as for the
481: neutron-proton separation energies.
482: The excitation spectrum of \nit (see Figure 2)
483: shows now a quite good agreement with the experimental spectrum.
484: Since we take in the definition of our neutron-proton subspace
485: nucleon states up to 10 MeV we have in the higher part of the \nit spectrum
486: a large number of
487: states but we have given only those called pairing vibrational
488: states which have amplitudes on several neutron-proton configurations.
489: The other
490: levels have a smaller probability to be seen experimentally \cite{vb}.
491: For these high energy levels
492: we have not calculated the contribution of phonon exchange. Our code is
493: inefficient when we have several very close eigenstates but we
494: expect the induced matrix elements to be weak because of energy denominators.
495: The contribution of phonon exchange is very important for the low
496: energy states and improves significantly the energy
497: spectrum where now the ground state energy, or equivalently the
498: neutron-proton separation energy, is very good. At about 6
499: MeV we can reproduce the experimental 1$^+$- 3$^+$ doublet with
500: the right order of levels
501: and energies very close to the experimental ones. The main RPA amplitudes
502: \begin{table}
503: \begin{center}
504: \begin{tabular}{|c|c|cccc|c|}\hline
505: J$^\varpi$ &E$^*$(MeV)&(1p$_{1/2}$)$^2$ & (1d$_{5/2}$)$^2$ & (2s)$^2$ &
506: (2s,1d$_{5/2}$) & (1p${3/2}$)$^2$\\ \hline
507: 1$^+$ & 0 & 0.88 & 0.24 & 0.1 & & -0.24 \\
508: 1$^+$ & 6.6 & -0.22 & 0.50 & 0.74 & & -0.11\\
509: 3$^+$ & 6.8 & & 0.50 & & 0.78 & -0.13 \\ \hline
510: \end{tabular}
511: \caption{RPA amplitudes in \nit }
512: \end{center}
513: \end{table}
514: are given in Table IV for the lowest states . We see that the RPA
515: Y-amplitudes with two
516: nucleons in the 1p$_{3/2}$ shell are not negligible what indicates the
517: presence in the wave function of the $^{12}$C ground state of
518: configurations with at least two holes in this shell, in agreement with the
519: shell model calculation of Cohen and Kurath for example \cite{ck} . In the
520: higher part of the spectrum, above 10 MeV, we get a group of two
521: 1$^+$ and two 3$^+$ states which may be identified with the same
522: experimental group. However we miss the experimental 1$^+$ level at 3.95 MeV
523: which is very likely formed mainly of
524: a $^{12}$C core
525: excited to its 2$^+$ state at 4.4 MeV coupled to the n-p pair in
526: its 1$^+$ ground state. This is also suggested by the analysis of
527: $^{12}$C ($^6$Li,$\alpha$)$^{14}$N \cite{mk} and $^{16}$O($\gamma$,np)\nit
528: reactions \cite{is,gaa}.
529: RPA is not able to describe such state since it relies on the assumption of
530: an inert core in its ground state.
531: In the same way we miss the 1$^+$ and 3$^+$ states at about 9 MeV
532: which again are very likely due to the coupling of the same 2$^+$ in
533: $^{12}$C with the neutron-proton pair in the 1$^+$ or the 3$^+$ excited
534: states at about 6 MeV. This assumption is supported by the presence
535: in the experimental spectrum \cite{exp}of two T=0 levels, a 2$^+$ at
536: 8.98 MeV and
537: a 5$^+$ at 8.96 MeV which very likely belong to the same multiplet.
538:
539:
540: \begin{table}
541: \begin{center}
542: \begin{tabular}{|c|ccc|}\hline
543: & (1p$_{1/2}$)$^2_{1^+,0}$&(1p$_{3/2}$)$^2_{1^+,0}$&
544: (1p$_{3/2}$)$^2_{3^+,0}$ \\ \hline
545: Min& -3.42&-3.74&-5.19\\
546: Min'&-3.42&-4.03&-5.49\\
547: PWBT & -3.45&-4.16&-6.04\\ \hline
548: \end{tabular}
549: \caption{Diagonal matrix elements (in MeV) for the two nucleons in the
550: 1p$_{1/2}$ or
551: 1p$_{3/2}$ shells coupled to J$^{\varpi}$=1$^+$ or 3$^+$ calculated with the
552: bare Minnesota interaction (Min), when the phonon exchange contribution is
553: added (Min') and those fitted in ref.[24] by Warbuton and Brown(PWBT). }
554: \end{center}
555: \end{table}
556:
557:
558: In \bo the results are improved by phonon exchanges but not enough to get
559: the 3$^+$ state as the ground state. The energy of the 1$^+$ when referred
560: to the core ground state energy, is close to the
561: experimental value but the 3$^+$ is too high and appears again as an
562: excited state. This 3$^+$ state
563: is a nearly pure (p$_{3/2}$)$^{-2}$ state
564: weakly affected by
565: RPA correlations because the lowest possible configuration with unoccupied
566: neutron and proton unperturbed states (anomalous RPA configuration)
567: is the (2s,1d$_{5/2}$) configuration, much higher in energy.
568: The correlated energy is therefore mostly due to the diagonal matrix element
569: $\langle (p_{3/2})^2|V|(p_{3/2})^2 \rangle$ which appears to be too weak as
570: is discussed now. Indeed we compare our matrix elements with those of
571: Warburton and Brown \cite{wa}
572: which are determined by least-squares fits to 216 levels in A=10-22
573: nuclei. They use a (1p,2s,1d) shell model space, compatible with our
574: subspace, at least for the lowest states. In Table V are reported
575: their fitted
576: matrix elements and ours calculated with the Minnesota plus phonon exchange
577: when the n-p pair is in the 1p-shell. We show the diagonal matrix
578: elements for the
579: n-p pair in the 1p$_{3/2}$ and 1p$_{1/2}$ shells coupled to spin 1$^+$ and
580: 3$^+$ which are for a large
581: part responsible for the binding energy of the lowest 1$^+$ states in
582: \nit and \bo and the 3$^+$ state in \bo. The same comparison has been made
583: with the matrix elements fitted by Cohen and Kurath \cite{ck}
584: in a smaller model space but the conclusion is qualitatively the same. We
585: see a good agreement when the two nucleons are coupled to J=1$^+$
586: but a too weak binding for J=3$^+$.
587: It shows that something is missing in our force and a possible
588: reason why it is too weak could be found in more recent shell model
589: calculations. Indeed in a no-core shell model and with the
590: Argonne V8' two-nucleon force , Aroua et al \cite{ar} get a good description
591: of the 1$^+$ ground state of \nit while Caurier et al.\cite{cau} and Navratil
592: and Ormand \cite{arn} show that with the same force and the same model
593: it is not possible to get the
594: ground state spin of \bo. The authors of ref.[9] show that it is correctly
595: obtained only when they include
596: the Tucson-Melbourne three-body interaction. The necessity to use a
597: three-body force in the description of \bo is also clear from quantum
598: Monte Carlo shell model calculations of \bo \cite{pip,wir}
599: A genuine three-body force is out of the scoop of the RPA since we work with
600: two nucleons only. However this three-body term could induce an effective
601: two-body contribution which is not included when we use the Minnesota force
602: plus phonon exchange. Intuitively we expect this term to be more important
603: for two 1p$_{3/2}$-nucleons inside the $^{12}$C, therefore for the
604: description of \bo, where they will interact with
605: a third nucleon off the six other nucleons in the same shell than for the
606: two nucleons added in higher shells to describe \nit which will have less
607: interaction with a third nucleon inside $^{12}$C. This might explain why our
608: calculations failed in \bo but are very satisfying in \nit.
609:
610:
611:
612:
613: At last we have made the same RPA calculations
614: with the D1 and D1S Gogny forces \cite{goa,gob}. The two forces give the
615: same results and we discuss only those obtained with D1S.
616: In the T=0 channel this force has a zero range density
617: dependent component with a strength t$_3$=1390.6 MeV.fm$^3$. Any zero range
618: interaction is divergent and the fitted strength depends on the cut-off on
619: single-particle energies. Since our cut-off is somewhat low we have increased
620: t$_3$ in order to get close to the measured energy of the 1$^+$
621: ground state in \nit.
622: With t$_3$=1600 MeV.fm$^3$, a value slightly stronger than the genuine
623: value, we get a binding energy for the n-p pair of -12.7 MeV instead of
624: -12.5 MeV, the experimental value. In the third columns of Figs.2
625: and 3 we show the excited spectra
626: obtained for \nit and \bo respectively. We see
627: that for \nit the levels are in very close agreement with experiment and
628: very close to those obtained with the
629: Minnesota interaction plus the phonon exchange contribution with however
630: an inversion of
631: the 1$^+$- 3$^+$ states at about 6 MeV. This result suggests that the Gogny
632: interaction, often thought as a G-matrix, includes implicitly the
633: phonon exchange contribution. In \bo the 3$^+$ state is obtained as the ground
634: state as it should, contrarly to what we got previously with an effective
635: two-body interaction.
636: The
637: 1$^+$ state is now the
638: first excited state with an excitation energy in agreement with the
639: experimental value and with shell model calculations of ref[9]. Therefore
640: according to our previous discussion on shell model results we may
641: conclude that,
642: through its density dependent term, the Gogny interaction includes
643: implicitly an effective two-body contribution
644: coming from three-body forces.However
645: the neutron-proton separation energy in $^{12}$C given as the
646: difference between the ground state energies in \bo and $^{12}$C is too high
647: by 0.8 MeV but a still larger overestimation is observed in the shell
648: model calculations (see their tables V and VII). Note that if we use the
649: calculated energy for a 1p$_{3/2}$ proton (see Table 1) instead of the
650: experimental energy this discrepancy is strongly reduced.
651:
652: \section{Conclusions}
653:
654: In the framework of a two-particle RPA model applied to the description of
655: \bo and \nit formed of a correlated core of $^{12}$C in its ground state
656: minus or plus a neutron-proton pair, we have performed a detailed analysis of
657: the T=0 effective nucleon-nucleon interaction. First we have shown that for
658: the 0$^+$ T=1 states the results are as good as they were for $^{14}$C-$^{10}$C
659: (where the neutron-proton pair is replaced by a two-neutron pair) when the
660: Gogny interactions, D1 and D1S, or their zero range equivalent are used.
661: The same
662: two-nucleon RPA model has been applied to 1$^+$ and 3$^+$, T=0 states and
663: an attempt to analyse the contains of an effective interaction has been
664: made. We have shown that an effective
665: interaction constructed as a bare interaction, from Minnesota or of zero
666: range, fitted to the deuteron binding energy, complemented by medium effects
667: due to phonon exchanges between the neutron-proton pair gives a very good
668: representation of the \nit levels but fails to reproduce the 3$^+$ ground
669: state of \bo. By comparing with shell model calculations we are able to
670: suggest that this is due to the presence of a three-body component in the
671: interaction which
672: are not included in our derivation. This additionnal component
673: will not spoil our good results
674: in \nit but will improve those in \bo. At last our
675: study suggests that the T=0 Gogny interaction which yields good agreement
676: with measurements in both nuclei, includes empirically both the effect of
677: phonons exchange in the effective
678: interaction and the effect of a
679: two-body component coming from the presence of
680: a three-body interaction and included in the density dependent term.
681:
682: I am very grateful to E. Viggezzi and F. Barranco for pointing to me a crucial
683: problem in the early stage of this work. Many thanks are also due to N.
684: Auerbach and B. Barrett for their interest in this work and for a careful
685: reading of the manuscript.
686:
687:
688:
689: \appendix
690: \section{Matrix elements of the interaction induced by phonon exchange.}
691:
692: The matrix elements of the induced interaction of Figure I are calculated
693: between two neutron-proton pairs coupled to spin J,M and isospin T=0.
694:
695: Each vertex in the diagrams is replaced by a phenomenological expression.
696: Let's call L,M$_L$ the angular momentum of the phonon, the transition density
697: from a zero-phonon state to the one-phonon state is written as:
698: \be
699: <1ph|V|0ph>=\frac{1}{\hat{L}}\beta_L R_0 \frac{dU(r)}{dr} Y_L^{M_L*}(\omega)
700: \ee
701: where U(r) is the average one-body potential assumed to be a
702: Saxon-potential such as:
703: \be
704: \frac{dU(r)}{dr}=-U_0 \frac{df(r)}{dr}=\frac{U_0}{a} g(r)
705: \ee
706:
707: where the function f(r) and the values of the parameters are given in
708: section II.
709: The collective amplitudes, $\beta_L$, can be fitted on the
710: experimental values of the B(EL) or on proton inelastic scattering cross
711: sections.
712:
713: The two-nucleon wave function is constructed by coupling the two-nucleon
714: states to a total spin (JM) as:
715:
716: \be
717: |j_1 j_2,JM>=\sum_{m_1m_2}<j_1j_2m_1m_2|JM> |l_1j_1m_1>|l_2j_2m_2>
718: \ee
719: where the nucleons 1 and 2 are either in occupied states, $\alpha_1$
720: $\alpha_2$, or unoccupied states, $a_1$ $a_2$ with energies
721: $\epsilon_{\alpha_1}$, $\epsilon_{\alpha_2}$ or $\epsilon_{a_1}$
722: $\epsilon_{a_2}$ respectively. The total wave function has to
723: be antisymmetric so that for T=0 states the spin-space wave function has
724: to be symmetric. Therefore the general expression for the
725: antisymmetrised matrix element of the induced interaction, $V_{ind}$, is
726: obtained as:
727: \be
728: <12|V_{ind}|34>=\sum_L\{ V_{dL}D_{dL}+V_{eL} D_{eL}\}
729: \ee
730:
731: with:
732:
733: \ba
734: V_{dL}&=&\frac{1}{2}[1+(-1)^{l_2+l_4+L}]
735: (-1)^{j_1+j_3+J}\frac{\beta_L^2R_0^2U_0^2}{4\pi a^2}
736: \frac{1}{\sqrt{(1+\delta_{12})(1+\delta_{34})}}\hat{\jmath_1}\hat{\jmath_2}
737: \hat{\jmath_3}
738: \hat{\jmath_4} \nonumber \\
739: & & \left(\begin{array}{ccc}
740: j_1&j_3&L\\
741: 0.5&-0.5&0
742: \end{array}
743: \right)
744: \left(\begin{array}{ccc}
745: j_2&j_4&L \\
746: 0.5&-0.5&0
747: \end{array}
748: \right)
749: \left\{\begin{array}{ccc}
750: j_1&j_2&J\\
751: j_4&j_3&L
752: \end{array}
753: \right\}
754: <1|g|3><2|g|4> \\
755: V_{eL}&=&\frac{1}{2}[1+(-1)^{l_2+l_3+L}]
756: (-1)^{j_1-j_3}\frac{\beta_L^2R_0^2U_0^2}{4\pi a^2}
757: \frac{1}{\sqrt{(1+\delta_{12})(1+\delta_{34})}}\hat{\jmath_1}\hat{\jmath_2}
758: \hat{\jmath_3}
759: \hat{\jmath_4} \nonumber\\
760: & & \left(\begin{array}{ccc}
761: j_2&j_3&L\\
762: 0.5&-0.5&0
763: \end{array}
764: \right)
765: \left(\begin{array}{ccc}
766: j_1&j_4&L \\
767: 0.5&-0.5&0
768: \end{array}
769: \right)
770: \left\{\begin{array}{ccc}
771: j_2&j_1&J\\
772: j_4&j_3&L
773: \end{array}
774: \right\}
775: <2|g|3><1|g|4> \\
776: <m|g|n>&=&\int_0^\infty g(r) \phi_{l_mj_m}^*(r)\phi_{l_nj_n}(r)r^2dr
777: \ea
778: where the $\phi$'s are the one-nucleon radial wave functions.
779:
780: While the quantities $V_{dL}$ and $V_{eL}$ have the same expressions
781: whatever are the initial and final two-nucleon states, occupied or
782: unoccupied, the expressions of $D_{dL}$ and $D_{eL}$ which come from energy
783: denominators have to be derived in three different cases corresponding to
784: the diagrams of Fig.I:
785:
786: If (12)=(a$_1$ a$_2$), (34)=(a$_3$ a$_4$) (diagrams a) and b))
787: \ba
788: D_{dL}&=&[\Omega-(\epsilon_{a_2}+\epsilon_{a_3}+\omega_L)]^{-1}+
789: [\Omega-(\epsilon_{a_1}+\epsilon_{a_4}+\omega_L)]^{-1}\\
790: D_{eL}&=&[\Omega-(\epsilon_{a_1}+\epsilon_{a_3}+\omega_L)]^{-1}+
791: [\Omega-(\epsilon_{a_2}+\epsilon_{a_4}+\omega_L)]^{-1}
792: \ea
793:
794: If (12)=(a$_1$ a$_2$), (34)=($\alpha_1$ $\alpha_2$) or the inverse (diagrams
795: c) and d)):
796: \ba
797: D_{dL}&=&-[\epsilon_{a_2}-\epsilon_{\alpha_2}+\omega_L]^{-1}
798: -[\epsilon_{a_1}-\epsilon_{\alpha_1}+\omega_L]^{-1} \nonumber \\
799: D_{eL}&=&-[\epsilon_{a_1}-\epsilon_{\alpha_2}+\omega_L]^{-1}
800: -[\epsilon_{a_2}-\epsilon_{\alpha_1}+\omega_L]^{-1}
801: \ea
802:
803: It is straightforward to show that:
804: \be
805: <a_1 a_2|V_{ind}|\alpha_1 \alpha_2>=<\alpha_1 \alpha_2|V_{ind}|a_1 a_2>
806: \ee
807:
808: If (12)=($\alpha_1$ $\alpha_2$), (34)=($\alpha_3$ $\alpha_4$) (diagrams e)
809: and f)):
810: \ba
811: D_{dL}&=&-[\Omega-\epsilon_{\alpha_2}-\epsilon_{\alpha_3}+\omega_L]^{-1}
812: -[\Omega-\epsilon_{\alpha_1}-\epsilon_{\alpha_4}+\omega_L]^{-1} \nonumber \\
813: D_{eL}&=&-[\Omega-\epsilon_{\alpha_1}-\epsilon_{\alpha_3}+\omega_L]^{-1}
814: -[\Omega-\epsilon_{\alpha_2}-\epsilon_{\alpha_4}+\omega_L]^{-1}
815: \ea
816:
817: In these equations $\Omega$ is the eigenvalue of the RPA eqs.(5-6) and
818: $\omega_L$ is the energy of the phonon L. When we add these phonon exchange
819: contributions to the bare matrix elements, the RPA equations become
820: non-linear and will be solved by iteration.
821:
822: Note that the approximation of eq.(18) implies that one can defined an
823: equivalent two-body term which has to be added to the bare neutron-proton
824: interaction which is the sum of separable terms of the following form:
825:
826: \be
827: \delta V=\frac{d\rho(r)}{dr}\frac{d\rho(r')}{dr'}Y_L^M(\omega)Y_L^M(\omega')
828: \ee
829:
830: \begin{thebibliography}{99}
831: \bi{pv} J.C.Pacheco and N.Vinh Mau, \pc {C65} 044004 (2002)
832: \bi{goa} J. Decharge and D. Gogny \pc {C21} 1568 (1980)
833: \bi{gob} J.F. Berger, M. Girod and D. Gogny Computer Physics Communiccations
834: {\bf 63} 365 (1991)
835: \bi{sca} E.Garrido et al., \pc {C60} 064312(1999)
836: \bi{min} D.R.Thomson, M.Lemere and Y.C. Tang \np {A286} 53 (1977)
837: \bi{sa} H. Sagawa et al. Phys. Lett. {\bf {B309}}, 1, (1993)
838: \bi{vm} N. Vinh Mau \np {A592} 33(1995)
839: \bi{br} G. Gori et al., \pc {C69} 041302 (2004)
840: \bi{arn} P. Navratil and W. E. Ormand \pc {C68} 034305 (2003)
841: \bi{vp} N. Vinh Mau and J.C. Pacheco, \np {A607}, 163 (1996)
842: \bi{scb} E. Garrido et al., \pc {C63}, 037304 (2001)
843: \bi{be} G. F. Bertsch and H. Esbensen Ann. of Phys. (N.Y.), 209 (1991)
844: \bi{we} W. Satula and R. Wyss \pr {86} 4488 (2001)
845: \bi{bb} F. Barranco et al. \pr {83}, 2147 (1999)
846: \bi{sag} S.A. Satchler, \np {A128}, 401 (1969)
847: \bi{exp} F. Ajzenberg-Selove \np {A523}, 1 (1991)
848: \bi{ex}A.H. Wapstra and G. Audi, \np {A565} 1 (1993)
849: \bi{exp1} F. Ajzenberg-Selove \np {A490}, 1 (1988)
850: \bi{vb} A. Bouyssy and N. Vinh Mau, Phys. Lett.{\bf 35B},269 (1971)
851: \bi{ck} S. Cohen and D. Kurath, \np {73}, 1 (1965)
852: \bi{mk} A.J. Mendez et al., \pc {C51}, 651 (1995)
853: \bi{is} L. Isaksson et al., \pr {83}, 3146 (1999)
854: \bi{gaa} K.R. Garrow et al., \pc {C64}, 064602 (2001)
855: \bi{wa} E.K. Warburton and B.A. Brown, \pc {C46}, 923 (1992)
856: \bi{cau} E. Caurier,P. Navratil, W.E. Ormand and J.P. Vary, \pc {C66},
857: 024314 (2002)
858: \bi{ar} S. Aroua et al., \np {A720}, 71 (2003)
859: \bi{pip} S.C. Pieper, K. Varga and R.B. Wiringa, \pc {C66}, 044310 (2002)
860: \bi{wir} R.B. Wiringa and S.C. Pieper, Phys. Rev. Lett. 89,1825501 (2002)
861: \end{thebibliography}
862: \end{document}
863: