1: \documentclass[prc,fleqn,twoside,floatfix]{revtex4}
2: %\documentclass[12pt,aps,floats,prc,epsfig,preprint]{revtex}
3: %\documentclass[aps,floats,prc,epsfig]{revtex}
4: \usepackage{latexsym,exscale}
5: \usepackage{epsfig}
6: \usepackage{amsmath,amssymb,amsfonts}
7: \parindent0ex
8: \def\Journal#1#2#3#4{{\em #1} {\bf #2}, #3 (#4) }
9: \def\NPA{{ Nucl. Phys.} A}
10: \def\NP{{ Nucl. Phys.}}
11: \def\PRL{Phys. Rev. Lett.}
12: \def\PREV{ Phys. Rev.}
13: \def\PREP{ Phys. Rep.}
14: \def\PRC{{Phys. Rev.} C}
15: \def\PL {Phys. Lett.}
16: \def\PLB {Phys. Lett. B}
17: \def\EJA {Eur. Phys. J. A}
18: \newcommand{\matrixel}[3]{\mbox{$\left<#1|#2|#3\right>$}}
19:
20: \begin{document}
21: \title{Sum rules and short-range correlations in
22: nuclear matter at finite temperature}
23: \author{T. Frick and H. M\"uther}
24: \affiliation{Institut f\"ur
25: Theoretische Physik, \\ Universit\"at T\"ubingen, D-72076 T\"ubingen, Germany}
26: \author{ A. Polls}
27: \affiliation{Departament d'Estructura i Constituents de la Mat\`eria,\\
28: Universitat de Barcelona, E-08028 Barcelona, Spain}
29:
30:
31: \begin{abstract}
32: The nucleon spectral function in nuclear matter fulfills an energy
33: weighted sum rule. Comparing two different realistic potential, these sum
34: rules are studied for Green's functions that are derived
35: self-consistently within the $T$~matrix approximation at finite
36: temperature.
37: \end{abstract}
38: %\pacs{21.65.+f, 21.30.Fe}
39: \maketitle
40:
41: %\section{Introduction\label{Introduction}}
42:
43: The microscopic study of the single-particle properties in nuclear matter
44: requires a rigorous treatment of the nucleon-nucleon (NN) correlations
45: \cite{bal1,her1}.
46: In fact, the strong short-range and tensor components, which are needed
47: in realistic NN interactions to fit the NN scattering data, lead to
48: important modifications of the nuclear wave function. A clear indication of
49: the importance of correlations is provided by the observation that a simple
50: Hartree-Fock calculation for nuclear matter at the empirical
51: saturation
52: density using
53: such realistic NN interactions typically results in positive energies
54: rather
55: than the
56: empirical value of $-16\,\mbox{MeV}$ per nucleon \cite{her1}.
57:
58: Correlations do not only manifest themselves
59: in the bulk properties but also modify the single-particle properties in a
60: substantial way. Several recent calculations have shown without
61: ambiguity
62: how the
63: NN correlations produce a partial occupation of the single particle
64: states which would
65: be fully occupied in a mean field description and a wide distribution
66: in energy of
67: the single particle strength. These two features have also been
68: empirically
69: founded
70: in the analysis of the (e,e'p) nucleon knock-out reactions \cite{exp1}.
71: The theoretical studies have been conducted both in finite
72: nuclei \cite{her2} and also in
73: nuclear matter \cite{fan1,ram1,dick1}.
74:
75: An optimal tool to study the single particle properties is provided
76: by the self-consistent
77: Green's function technique (SCGF) \cite{dick2}.
78: This method gives direct access to the single particle
79: spectral function, which should be self-consistently determined at the same time
80: than the effective interactions between the nucleons in the medium.
81: Enormous progress
82: in the SCGF applications to nuclear matter have been reported in the
83: last years, both at zero
84: \cite{dick1} and finite temperature \cite{boz1,boz2,fri03}.
85:
86: The efforts at $T=0$ have mainly been addressed to provide the
87: appropriate theoretical
88: background for the interpretation of the (e,e'p) experiments while the
89: investigation
90: at finite $T$ are mainly oriented to describe the nuclear medium in astrophysical
91: environments or to the interpretation of the dynamics of heavy ion collisions.
92:
93: In any case, the key quantity is the single-particle spectral
94: function,
95: i.e. the distribution
96: of strength in energy when one adds or removes a particle of the
97: system
98: with a given
99: momentum. A possible way to analyze the single-particle spectral
100: function is by means
101: of the energy weighted sum rules. They are well established in the
102: literature and have been numerically analyzed in the case of zero
103: temperature \cite{pol94}.
104:
105: The analysis of the energy weighted sum-rules can give useful
106: insights not only on the
107: numerical accuracy of the many-body approach used to calculate them
108: but also can help to
109: understand the properties and structure of the NN potential.
110:
111: This paper is devoted to study the physical implications of the
112: fulfillment of these
113: sum rules for single particle spectral functions in nuclear matter at
114: finite $T$. This investigation is based on the
115: framework of SCGF employing a fully self-consistent ladder approximation in
116: which the complete
117: spectral function has been used to describe the intermediate states in
118: the Galistkii-Feynman
119: equation.
120:
121: After a brief summary of the definitions of the single-particle
122: spectral function,
123: we give a simple derivation of the sum rules. Then we analyze the
124: results for two types
125: of realistic potentials, the CDBONN and the Argonne V18, and discuss
126: the different behaviors
127: based on the different strength of the short-range and tensor
128: components of both potentials.
129:
130: %\section{Sum rules}
131:
132: For a given Hamiltonian $H$,
133: the Green's function for a system at finite temperature can
134: be defined in a grand-canonical formulation:
135: \begin{equation}
136: \label{green_def}
137: {\mathrm{i}}g({\mathbf{k}}t;{\mathbf{k}}^{\prime}t^{\prime})=
138: {\mathrm{Tr}}\{\rho\,
139: %{\mathbb{T}}
140: {\mathbf{T}}
141: [a^{\phantom{\dagger}}_{\mathbf{k}}(t)
142: a^{\dagger}_{{\mathbf{k}}^{\prime}}(t^{\prime})]\}.
143: \end{equation}
144: $\mathbf{T}$ is the time ordering operator that acts on a product
145: of Heisenberg field operators
146: $a_{\mathbf{k}}(t)=e^{{\mathrm{i}}tH}a_{\mathbf{k}}e^{-{\mathrm{i}}tH}$
147: in such a way that the field operator with the largest time argument $t$
148: is put to the left.
149: The trace is to be taken over all energy eigenstates and all particle number
150: eigenstates of the many body system, weighted by the statistical operator,
151: \begin{equation}
152: \label{eq_statop}
153: \rho=\frac{1}{Z}\,{e^{-\beta(H-\mu N)}}.
154: \end{equation}
155: $\beta$ and $\mu$ denote the inverse temperature and the chemical
156: potential, respectively.
157: N is the operator that counts the total number of
158: particles in the system,
159: \begin{equation}
160: N=
161: \sum_{\mathbf{k}}
162: a_{\mathbf{k}}^{\dagger}(t)a_{\mathbf{k}}^{\phantom{\dagger}}(t)
163: %\int\!{\mathrm{d}}^3\!x\,
164: %\psi^{\dagger}({\mathbf{x}t})
165: %\psi({\mathbf{x}t}).
166: \end{equation}
167: $N$ is independent of time, since it commutes with $H$.
168: The normalization factor in Eq.~(\ref{eq_statop}) is given by the
169: grand partition function of statistical mechanics,
170: \begin{equation}
171: Z={{\mathrm{Tr}}\,e^{-\beta(H-\mu N)}}.
172: \end{equation}
173: For a homogeneous system, the Green's function is diagonal in
174: momentum space and depends only on the absolute value of
175: ${\mathbf{k}}$ and on the difference $\tau=t^{\prime}-t$.
176: Starting from the definition of the Green's function,
177: we first focus on the case $\tau>0$.
178: In order to recover the
179: expression for the ensemble average of the occupation number
180: $n(k)$ for $\tau=0^+$,
181: the following definition of the correlation function, $g^<$,
182: includes an additional factor of $-{\mathrm{i}}$ with respect to the
183: definition of the Green's function $g$,
184: \begin{equation}
185: g^<(k,\tau)=
186: \label{eq_gs1}
187: {\mathrm{Tr}}\{\rho\,
188: e^{{\mathrm{i}}\tau H}a^{\dagger}_{{\mathbf{k}}}
189: e^{-{\mathrm{i}}\tau H}
190: a^{\phantom{\dagger}}_{\mathbf{k}}\}.
191: \end{equation}
192: $g^<(k,\tau)$ can be expressed as a Fourier integral over all frequencies,
193: \begin{equation}
194: \label{eq_momdist}
195: g^<(k,\tau)=\int_{-\infty}^{+\infty}
196: \frac{{\mathrm{d}}\omega}{2\pi}\,e^{-{\mathrm{i}}\omega \tau}A^<(k,\omega)
197: \end{equation}
198: if $A^<(k,\omega)$ is defined by~\cite{kraeft}:
199: \begin{equation}
200: \label{eq_gsmaller}
201: A^<(k,\omega)=2\pi
202: %\frac{1}{Z}
203: \sum_{nm}
204: \frac{e^{-\beta(E_m-\mu N_m)}}{Z}
205: \left|\left<\Psi_n|a_k|\Psi_m\right>\right|^2
206: \delta(\omega-(E_m-E_n)).
207: \end{equation}
208: This can be easily checked by inserting the eigenstates
209: $\left|\Psi_n\right>$ into the expression of the trace in
210: Eq.~(\ref{eq_gs1}). It is important to
211: note that $\left|\Psi_n\right>$ are simultaneous
212: eigenstates of both the number operator and the Hamiltonian.\\
213: A similar analysis can be conducted for $\tau<0$, yielding a
214: function
215: \begin{equation}
216: A^>(k,\omega)=e^{\beta(\omega-\mu)}A^<(k,\omega).\label{eq:connec}
217: \end{equation}
218: The spectral function at finite temperature is defined as the sum
219: of the two positive functions, $A^<$ and $A^>$,
220: \begin{equation}
221: A(k,\omega)=A^<(k,\omega)+A^>(k,\omega).
222: \end{equation}
223: Expression~(\ref{eq_gsmaller}), for $A^<$, can be compared
224: to the result for the hole spectral function
225: at zero temperature, that was reported in Ref~\cite{pol94},
226: \begin{equation}
227: A_h(k,\omega)= 2\pi
228: \sum_{n}
229: \left|\left<\Psi_n^{A-1}|a_k|\Psi_0^{A}\right>\right|^2
230: \delta(\omega-(E_0^{A}-E_n^{A-1})),
231: \end{equation}
232: where
233: $\left|\Psi_0^{A}\right>$ is the ground state of an $A$ particle system and
234: $\left|\Psi_n^{A-1}\right>$ labels the excited energy eigenstates of a system
235: that contains one particle less.
236: The physical interpretation of the hole spectral function in a system
237: at zero temperature is the following: $A_h(k,\omega)$ is the probability
238: to remove a particle from the ground state of the $A$-body system, such,
239: that the residual system
240: is left with an excitation energy $E^{A-1}_n=E^A_0-\omega$.
241: $E^A_0$ is the ground state energy of the $A$ particle system.
242: It is clear that the lowest possible energy of the final state is
243: the ground state energy of the $A-1$ particle system, so that there
244: is an upper limit for the hole spectral function at
245: $\omega=E_0^A-E_0^{A-1}=\mu$. In a similar fashion, the particle spectral
246: function $A_p$ can be defined as the probability to attach a further nucleon
247: to the system in such a way, that the excitation energy of the compound
248: system with respect to the ground state energy of the initial system is
249: $\omega=E^{A+1}-E_0^{A}$. In this case, one can argue that,
250: to add a further particle, one has to pay
251: at least the chemical potential, so that $\mu$ is a lower bound for
252: $\omega$. At zero temperature,
253: this behavior causes a complete separation of the particle and the
254: hole spectral function.
255:
256: The situation is quite different in a grand-canonical formulation at
257: finite temperature
258: To illustrate these changes, the full spectral function A,
259: %The spectral function,
260: as well as $A^>$ and $A^<$
261: %at finite temperature is
262: are shown in Fig.~\ref{fig_sf} for three momenta around the Fermi momentum
263: of a zero temperature system at the same density, $\rho=0.2\,\mbox{fm}^{-3}$.
264: Numerical values for the integrated strength of $A^<$ are listed in Table
265: \ref{tab_dist}.
266: Since thermally excited states $\left|\Psi_m\right>$
267: are always included in the grand-canonical ensemble
268: average according to their weight factor $e^{-\beta(E_m-\mu N_m)}$,
269: one can take out a particle from a thermally excited
270: state and end up in a weakly excited state close to
271: the ground state of the residual system. This leads to a contribution to $A^<$
272: for an energy $\omega$ larger than $\mu$. Also one has to keep in mind that we
273: are considering a grand-canonical average. This implies that with the
274: appropriate weight one also considers systems with a density larger than the
275: mean value. For those systems, a removal of
276: particles from states with single-particle energies
277: above $\mu$ will also be possible.
278:
279: Similarly, a particle can be added to a thermally excited state,
280: leaving the compound system in a state close to its ground state,
281: so that $A^>(k,\omega)$
282: extends to the region below $\mu$.
283: In any case, there is no longer a separation between $A^>$ and $A^<$,
284: and the maxima of both functions can even coincide. This is also quite obvious
285: from the relation (\ref{eq:connec}).
286:
287: For the $T$~matrix approximation to the self energy
288: reported in~\cite{fri03}, one can determine the
289: single-particle Green's function as the solution of
290: Dyson's equation for any complex value of the frequency variable $z$,
291: \begin{equation}
292: \label{dyson_g}
293: g(k,z)=
294: \frac{1}{z-\frac{k^2}{2m}-\Sigma(k,z)}.
295: \end{equation}
296: Using the analytical properties of the finite temperature Green's
297: function along the imaginary time axis, an important relation
298: between the spectral function and the Green's function can be
299: derived and analytically continued to slightly complex values~\cite{kad62}:
300: %Using the periodicity of the finite temperature Green's function along
301: %the real-time axis, the following relation between $g$ and $A$ can be
302: %derived~\cite{kad62},
303: \begin{equation}
304: \label{spec_g}
305: %g(k,z_{\nu})=\int_{-\infty}^{+\infty}
306: %\frac{{\mathrm{d}}\omega}{2\pi}\, \frac{A(k,\omega)}{z_{\nu}-\omega}.
307: g(k,\omega+{\mathrm{i}}\eta)=\int_{-\infty}^{+\infty}
308: \frac{{\mathrm{d}}\omega^{\prime}}{2\pi}\, \frac{A(k,\omega)}
309: {\omega-\omega^{\prime}+{\mathrm{i}}\eta}.
310: \end{equation}
311: One can extract sum rules from the asymptotic
312: behavior at large $\omega$
313: by expanding the real part of both expressions for the Green's function,
314: Eqs.~(\ref{dyson_g}) and (\ref{spec_g}), in powers of
315: $\frac{1}{\omega}$. This yields
316: \begin{equation}
317: {\mathrm{Re}}\,g(k,\omega)=\frac{1}{\omega}
318: %{\mathrm{Re}}\,g(k,\omega+{\mathrm{i}}\eta)=\frac{1}{\omega}
319: \left\{1+\frac{1}{\omega}\left[
320: %\frac{k^2}{2m}+\Sigma^{\infty}(k)
321: \frac{k^2}{2m}+\lim_{\omega\rightarrow\infty}{\mathrm{Re}}\,
322: \Sigma(k,\omega)
323: %\Sigma(k,\omega+{\mathrm{i}}\eta)
324: \right]+\cdots\right\}
325: \end{equation}
326: and
327: \begin{equation}
328: {\mathrm{Re}}\,g(k,\omega)=
329: %{\mathrm{Re}}\,g(k,\omega+{\mathrm{i}}\eta)=
330: %\frac{1}{\omega}
331: \frac{1}{\omega}
332: \left\{
333: \int_{-\infty}^{+\infty}
334: {\mathrm{d}}\omega^{\prime}\,
335: A(k,\omega^{\prime})+
336: \frac{1}{\omega}
337: \int_{-\infty}^{+\infty}
338: {\mathrm{d}}\omega^{\prime}\,
339: \omega^{\prime}A(k,\omega^{\prime})+\cdots\right\}.
340: \end{equation}
341: By comparing the first two expansion coefficients, one finds the $m_0$
342: and the $m_1$ sum rules,
343: \begin{equation}
344: \int_{-\infty}^{+\infty} \frac{{\mathrm{d}}\omega}{2\pi}
345: A(k,\omega)=1,
346: \end{equation}
347: and
348: \begin{equation}
349: \int_{-\infty}^{+\infty} \frac{{\mathrm{d}}\omega}{2\pi}
350: A(k,\omega)\omega=\frac{k^2}{2m}+\lim_{\omega\rightarrow\infty}{\mathrm{Re}}\,
351: \Sigma(k,\omega)
352: \end{equation}
353: Similar sum rules can be obtained from the higher order terms,
354: as it was done in Ref.~\cite{pol94} for $m_2$.
355: Thinking of an arbitrary
356: approximation scheme for $\Sigma(k,\omega)$, it might be
357: interesting to ask whether or to what extend such a scheme fulfills
358: the sum rules. This is, however, not the point we want to address in
359: this paper. In the $T$~matrix approximation, the real part of
360: the self energy can be computed from the imaginary part, using a
361: dispersion relation,
362: \begin{equation}
363: \label{eq_real_sigma}
364: {\mathrm{Re}}\,\Sigma(k,\omega)=\Sigma^{\infty}(k)-\frac{{\mathcal{P}}}{\pi}
365: \int_{-\infty}^{+\infty}
366: {\mathrm{d}}\lambda\,
367: \frac{{\mathrm{Im}}\Sigma(k,\lambda+{\mathrm{i}}\eta)}{\omega-\lambda}.
368: %+{\mathrm{i}}\,{\mathrm{Im}}\Sigma(k,\omega+{\mathrm{i}}\eta),
369: \end{equation}
370: In the derivation of Eq.~(\ref{eq_real_sigma}), the spectral
371: decomposition of the Green's function was already used, so it is a
372: property of the $T$~matrix approach that it automatically fulfills the sum
373: rules. Nevertheless, besides providing a useful consistency
374: check for the numerics, it is interesting to use the sum rules
375: to compare the importance of
376: short-range correlations for different realistic
377: potentials on a quantitative level.
378: The first term on the right hand side of Eq.~(\ref{eq_real_sigma}) is the energy independent part
379: of the self energy,
380: \begin{equation}
381: \label{eq_HF}
382: \Sigma^{\infty}(k)
383: =
384: \int \frac{{\mathrm{d}}^3k^{\prime}}{(2\pi)^3}
385: \left<{\mathbf{k}}{\mathbf{k}}^{\prime}\right|
386: V
387: \left|{\mathbf{k}}{\mathbf{k}}^{\prime}\right>_A
388: n(k^{\prime}).
389: \end{equation}
390: %\begin{equation}
391: %\label{eq_HF}
392: %\Sigma^{\infty}(k)=
393: %\frac{1}{\beta}
394: %\sum_{\nu}
395: %\int \frac{{\mathrm{d}}^3k^{\prime}}{(2\pi)^3}
396: %\left<{\mathbf{k}}{\mathbf{k}}^{\prime}\right|V
397: %\left|{\mathbf{k}}{\mathbf{k}}^{\prime}\right>_A
398: %%\left<{\mathbf{k}}{\mathbf{k}}^{\prime}\right|V
399: %%\left|{\mathbf{k}}^{\prime}{\mathbf{k}}\right>
400: %%\right]
401: %g(k^{\prime},z_{\nu}),
402: %\end{equation}
403: which can be identified with $\lim_{\omega\rightarrow\infty}{\mathrm{Re}}\,
404: \Sigma(k,\omega)$, since the dispersive part decays like
405: $\frac{1}{\omega}$ for $\omega\rightarrow\pm\infty$.
406: Eq.~(\ref{eq_HF}) looks like a Hartree-Fock potential, however, $n(k)$ is
407: the momentum distribution that is determined from a non-trivial
408: spectral function $A^<$ in Eq.~(\ref{eq_momdist}), assuming $\tau=0$.
409: %\begin{equation}
410: %n(k)=
411: %\int_{-\infty}^{+\infty} \frac{{\mathrm{d}}\omega}{2\pi}
412: %A^<(k,\omega)
413: %\end{equation}
414: In contrast, the Hartree-Fock self energy at finite temperature
415: must be determined from an energy spectrum
416: $\epsilon(k)$ and a momentum distribution
417: $n_{HF}(k)=f(\epsilon(k))$, where $f(\omega)$ is the Fermi function.
418: Unlike $n_{HF}(k)$, the non-trivial $n(k)$ accounts for depletion effects of
419: the bound states due to short-range correlations.
420: In this sense, $\Sigma^{\infty}$ is a generalization of a Hartree-Fock
421: potential. Fig.~\ref{fig_tadpole} illustrates the difference between the
422: two pictures with the corresponding Feynman diagrams.
423:
424: %\section{Results and discussion}
425:
426: All results in this paper have been obtained using the
427: iteration procedure that was described in
428: Ref.\cite{fri03}. Fully self-consistent spectral functions were
429: calculated for
430: %at a density of $\rho=0.2\,\mbox{fm}^{-3}$.
431: two realistic potentials,
432: the stiffer Argonne V18 and the softer CDBONN.
433: %, have been used as an input.
434:
435: The $m_0$ sum rule is fulfilled better than $0.1\%$ in the
436: complete momentum range.
437: Results for the $m_1$ sum rule are given in Fig~\ref{fig_m1} for a temperature
438: of $T=10\,\mbox{MeV}$ and a density of $\rho=0.2\,\mbox{fm}^{-3}$.
439: It is satisfied better than $1\%$.
440: Both right hand side and left hand side are plotted,
441: but the curves lie on top of each other and cannot be distinguished
442: (solid lines). The lower dash-dotted line shows the $m_1$-contribution
443: from $A^<$, which is always negative and goes to zero for high momenta,
444: since there $A^<$ is strongly suppressed. The probability to remove a
445: high-momentum particle from the system is simply very small.
446: The upper dash-dotted line displays the contribution from $A^>$. Due to the
447: short-range correlations,
448: there is a high-energy tail present in the spectral function, and so
449: this contribution is already positive at low momenta, furthermore, it is
450: nearly constant in this range,
451: reflecting the fact that the high energy strength distribution is
452: momentum independent. As soon as the quasiparticle peak of the
453: spectral function is located at energies greater than $\mu$,
454: the $A^>$ contribution increases steadily, following the position of this
455: peak. Both contributions add up to $m_1$.
456: It is interesting to remind the fact that for free particles,
457: the sum rules are automatically fulfilled. In this case,
458: $A^<$ and $A^>$ are delta peaks that are located at the same position
459: and their strength adds up to one. Their relative strength is given by the
460: ratio of the phase space factors $f(\epsilon(k))$ and
461: $[1-f(\epsilon(k))]$, respectively, where $\epsilon(k)=\frac{k^2}{2m}$
462: in the free case.\\
463: The results in Fig.~\ref{fig_m1} shows that the sum rule
464: $m_1$ is rather sensitive to the differences in the NN potentials.
465: The $m_1$ results for the CDBONN interaction is about
466: $65\,\mbox{MeV}$
467: more attractive than the Argonne V18 result.
468: A closer examination shows that this is predominantly due to
469: the $A^>$ contribution, which is almost $50\,\mbox{MeV}$ more repulsive
470: for the Argonne V18. This means that the Argonne potential
471: produces more correlations in the sense that the
472: strength that effects $m_1$ is redistributed to higher energies. \\
473: The dotted lines are the simple Hartree-Fock estimate of $m_1$
474: for the same temperature and density. For both potentials,
475: the Hartree-Fock result makes up quite a good approximation to the sum rule.
476: This result is interesting, since it permits a quantitative estimate of the
477: amount of correlations produced by any given NN potential without a
478: sophisticated many-body calculation. \\
479: Fig.~\ref{fig_sat} reports the exhaustion of the sum
480: rules $m_0$ (left panel) and $m_1$ (right panel)
481: versus the upper integration limit $\omega$ for a
482: momentum of $k=500\,\mbox{MeV}$. At this momentum, the quasiparticle
483: peak is located around $100\,\mbox{MeV}$.
484: For both interactions that were considered,
485: the main contribution to $m_0$, more than $80\%$,
486: come from the quasiparticle peak of the spectral function.
487: In the region far above the peak, the CDBONN saturates considerably faster.
488: In Table.~\ref{tab_sat}, the upper integration limits that have to be
489: chosen to exhaust the sum rule to a given percentage are reported for
490: $m_0$ and $m_1$ and compared for both potentials. In the case of
491: $m_0$ and the stiffer Argonne V18, one must integrate almost twice as far as
492: for the softer CDBONN.
493: The saturation of the $m_1$ sum rule is different,
494: because in this case, a somewhat higher energy region of the spectral function
495: is probed.
496: In the right panel of Fig.~\ref{fig_sat}, one can observe that the
497: quasiparticle peak contributes less than $50\%$ to $m_1$, and the
498: high-energy tail becomes much more important, since it is weighted
499: by a factor of $\omega$. While both potentials behave qualitatively
500: similar up to an integration limit of about
501: $700$ or $800\,\mbox{MeV}$, where $m_1$ is already exhausted by
502: about 75\% for the CDBONN potential
503: (cf.Tab.~\ref{tab_sat}), a large contribution
504: of about $40\%$ is still above this energy in the
505: case of the Argonne V18. One can also note, that,
506: to exhaust $m_1$ completely, one has to integrate up to higher
507: energies in the case of the CDBONN.
508: However, these contributions to the spectral
509: function above $\omega\approx 4000\,\mbox{MeV}$ are weak and
510: yield no further repulsion.
511:
512:
513:
514:
515: %$\mu=-25.5\,\mbox{MeV}$
516: %$\mu=-8.9\,\mbox{MeV}$
517:
518: This work has been supported by the German-Spanish exchange program (DAAD,
519: Acciones Integradas Hispano-Alemanas).
520: We also would like to acknowledge financial support from the {\it Europ\"aische
521: Graduiertenkolleg T\"ubingen - Basel} (DFG - SNF) and the DGICYT (Spain) Project
522: No. BFM2002-01868 and from Generalitat de Catalunya Project No. 2001SGR00064.
523:
524: \begin{thebibliography}{99}
525: \bibitem {bal1} M. Baldo, {\it Nuclear Methods and the Nuclear Equation of State,}
526: Int. Rev. of Nucl. Phys, Vol. 9 (World Scientific, Singapore, 1999).
527: \bibitem {her1} H. M\"uther and A. Polls, Prog. Part. Nucl. Phys.
528: {\bf 45}, 243 (2000).
529: \bibitem{exp1} M. F. van Batenburg,
530: Ph.D. thesis, University of Utrecht (2001).
531: \bibitem{her2} H. M\"uther, A. Polls and W. H. Dickhoff, Phys. Rev. {\bf C51},3040
532: (1995).
533: \bibitem{fan1} O. Benhar, A. Fabrocini and S. Fantoni,
534: Nucl. Phys. {\bf A 505}, 267 (1989).
535: \bibitem{ram1} A. Ramos, A. Polls and W. H. Dickhoff,
536: Nucl. Phys. {\bf A 503}, 1 (1989).
537: \bibitem{dick1} Y. Dewulf, W. H. Dickhoff, D. Van Neck,
538: E. R. Stoddard and M. Waroquier,
539: Phys. Rev. Let. {\bf 90}, 152501 (2003).
540: \bibitem{dick2} W. H. Dickhoff and H. M\"uther, Rep. Prog. Phys. {\bf 55},
541: 1947 (1992).
542: \bibitem{boz1} P. Bo\.zek, Phys. Rev. {\bf C 59}, 2619 (1999).
543: \bibitem{boz2} P. Bo\.zek, Phys. Rev. {\bf C 65}, 054306 (2002).
544: \bibitem{fri03} T. Frick and H. M\"uther, Phys. Rev. {\bf C 68}, 034310 (2003).
545: %\Journal{\PRC}{68}{034310}{2003}
546: \bibitem{pol94} A. Polls, A. Ramos, J. Ventura, S. Amari and
547: W. H. Dickhoff,
548: Phys. Rev. {\bf C 49}, 3050 (1994).
549: \bibitem{kraeft} W. D. Kraeft, D. Kremp, W. Ebeling and G. R\"opke,
550: {\em Quantum Statistics of Charged Particle Systems}
551: (Akademie-Verlag, Berlin, 1986).
552: \bibitem{kad62} L. P. Kadanoff and G. Baym,
553: {\em Quantum Statistical Mechanics} (Benjamin, New York, 1962).
554:
555:
556:
557:
558: \end{thebibliography}
559:
560:
561: \begin{table}
562: \begin{center}
563: \begin{tabular}{cccc}
564: \multicolumn{1}{c}{$k$ [MeV]} &
565: \multicolumn{1}{c}{below $\mu$ [\%]} &
566: \multicolumn{1}{c}{above $\mu$ [\%]} &
567: \multicolumn{1}{c}{$n(k)$}
568: \\ \hline\hline
569: 230 & 98 & 2 & 0.706 \\
570: 275 & 77 & 23 & 0.481 \\
571: 320 & 33 & 67 & 0.191 \\
572: 400 & 71 & 29 & 0.025 \\
573: 500 & 95 & 5 & 0.006
574: \end{tabular}
575: \caption{\label{tab_dist}Strength distribution of $A^<$. The
576: numbers give the fraction of the integrated
577: strength above and below the chemical
578: potential~$\mu$. The last column reports the occupation number of the
579: respective state.
580: The parameters are the same as in Fig.~\ref{fig_sf}.}
581: \end{center}
582: \end{table}
583:
584:
585:
586: \begin{table}
587: \begin{center}
588: \begin{tabular}{ccccc}
589: \multicolumn{1}{c}{\% saturation} &
590: \multicolumn{1}{c}{$m_0$ CDB} &
591: \multicolumn{1}{c}{$m_0$ V18} &
592: \multicolumn{1}{c}{$m_1$ CDB} &
593: \multicolumn{1}{c}{$m_1$ V18}
594: \\ \hline\hline
595: 60 & - & - & 277 & 690 \\
596: 75 & - & - & 790 & 1518 \\
597: 90 & 215 & 311 & 2250 & 2860 \\
598: 95 & 403 & 725 & 3756 & 3740 \\
599: 99 & 1388 & 2277 & 8545 & 5720 \\
600: \end{tabular}
601: \caption{\label{tab_sat}Upper integration limits of the running integrals
602: that must be chosen to exhaust the sum rule $m_0$ and $m_1$ up to the
603: fraction given in the first column. The parameters are the same as in
604: Fig.~\ref{fig_sat}.}
605: \end{center}
606: \end{table}
607:
608:
609:
610:
611: \begin{figure}
612: \begin{center}
613: \epsfig{figure=fig_sf.eps,scale=0.4}
614: \end{center}
615: \caption{\label{fig_sf}Spectral function for a density of
616: $\rho=0.2\,\mbox{fm}^{-3}$ and a temperature of $T=10\,\mbox{MeV}$
617: (solid line). Various momenta are considered as indicated in the three panels.
618: $A^<$ (dashed line) and $A^>$ (dash-dotted line) are also displayed.}
619: \end{figure}
620:
621: \begin{figure}
622: \begin{center}
623: \epsfig{figure=fig_tadpole.eps,width=8cm}
624: \end{center}
625: \caption{\label{fig_tadpole}Diagrammatic representation of the HF
626: approximation (left) and the energy independent part of the
627: self-consistently dressed self energy (right).}
628: \end{figure}
629:
630:
631: \begin{figure}
632: \begin{center}
633: \epsfig{figure=fig_m1.eps,scale=0.4,angle=0}
634: \end{center}
635: \caption{\label{fig_m1}Illustration of the energy weighted sum rule
636: $m_1$ (solid lines) for the CDBONN potential (left panel)
637: and the Argonne V18 potential (right panel).
638: Both right hand side and left hand side are displayed,
639: but the sum rule is so well fulfilled that they are on top of each other.
640: The contribution to $m_1$ that comes from $A^>$
641: and $A^<$ is indicated by the upper and the lower dash-dotted lines,
642: the latter approaching zero rapidly for high momenta. The dotted line
643: is the Hartree-Fock single particle spectrum.
644: Density and temperature are the same as in Fig.~\ref{fig_sf}.
645: }
646: \end{figure}
647:
648:
649:
650: \begin{figure}
651: \begin{center}
652: \epsfig{figure=fig_sat.eps,scale=0.4,angle=0}
653: \end{center}
654: \caption{\label{fig_sat}Saturation of the sum rules $m_0$ (right panel)
655: and $m_1$ (left panel) for the CDBONN potential (solid line) and the
656: Argonne V18 potential (dashed line). The momentum is $k=500\,\mbox{MeV}$.
657: Again, temperature and density are the same as in Fig.~\ref{fig_sf}.}
658: \end{figure}
659:
660:
661:
662:
663:
664: \end{document}
665:
666:
667:
668:
669:
670:
671: