Self-consistent description of finite nuclei based on a relativistic quark model
Extracted source text ยท arXiv nucl-th/9606020
\documentstyle[12pt,epsfig]{article}
%\documentstyle[12pt,eclepsf]{article}
%\documentstyle[12pt]{article}
%
\topmargin = 0cm
\oddsidemargin = 0cm \evensidemargin = 0cm
\textheight = 23cm \textwidth = 16cm
\def\baselinestretch{1.6}
%
\def\bge{\begin{equation}}
\def\ene{\end{equation}}
\def\bg{\begin{eqnarray}}
\def\en{\end{eqnarray}}
\def\nn{\nonumber}
\def\ra{\rightarrow}
\def\la{\leftarrow}
\def\ua{\uparrow}
\def\da{\downarrow}
\def\del{\partial}
\def\aleq{\stackrel{<}{\sim}}
\def\ageq{\stackrel{>}{\sim}}
\def\g{\gamma^*}
\def\vn{V N N}
\def\gtuv{\widetilde{g}^{\mu\nu}}
\def\gtab{\widetilde{g}^{\alpha\beta}}
\def\ktu{\widetilde{k}^{\mu}}
\def\ktv{\widetilde{k}^{\nu}}
\def\Ord{O\left( \frac{1}{P_L^2} \right)}
\def\What{\widehat{W}}
\def\WTzero{\widehat{W}^0_T}
\def\WTone{\widehat{W}^1_T}
\def\WTtwo{\widehat{W}^2_T}
\def\pq{p \cdot q}
\def\Tr{{\rm Tr}}
\def\qsl{\not\!q}
\def\pslash{\not\!p}
\def\PPsl{\not\!\!P}
\def\bi{\bibitem}
%
\begin{document}
\renewcommand{\thefootnote}{\fnsymbol{footnote}} %use symbolic footnote
%
\begin{flushright}
ADP-96-17/T220
\end{flushright}
\vspace{0.5cm}
%
\begin{LARGE}
\begin{center}
Self-consistent description of finite nuclei based on a
relativistic quark model
\end{center}
\end{LARGE}
%
\vspace{2cm}
%
\begin{center}
\begin{large}
Koichi~Saito\footnote{ksaito@nucl.phys.tohoku.ac.jp} \\
{\sl Physics Division, Tohoku College of Pharmacy \\
Sendai 981, Japan} \\
%
\vspace{0.5cm}
Kazuo~Tsushima\footnote{ktsushim@physics.adelaide.edu.au} and
Anthony~W.~Thomas~\footnote{athomas@physics.adelaide.edu.au} \\
{\sl Department of Physics and Mathematical Physics, \\
and Institute for Theoretical Physics, \\
University of Adelaide, South Australia 5005, Australia}
\end{large}
\end{center}
%
\vspace{4.0cm}
%
PACS numbers: 12.39.Ba, 21.60.-n, 21.10.-k, 24.85.+p \\
Keywords: Relativistic mean-field theory, finite nuclei,
quark degrees of freedom, MIT bag model, nuclear charge distribution,
energy spectra
%
\newpage
\begin{abstract}
Relativistic Hartree equations for spherical nuclei have been derived
from a
relativistic quark model of the structure of bound nucleons which
interact through the (self-consistent) exchange of scalar ($\sigma$) and
vector ($\omega$ and $\rho$) mesons.
The coupling constants and the mass of the $\sigma$-meson are determined
from the properties of
symmetric nuclear matter and the rms charge radius in $^{40}$Ca.
Calculated properties of static, closed-shell nuclei from $^{16}$O to
$^{208}$Pb are compared with experimental data and with results of Quantum
Hadrodynamics (QHD). The dependence of the results
on the nucleon size and the quark mass is
investigated. Several possible
extensions of the model are also discussed.
\end{abstract}
%
%
\newpage
\section{Introduction}
\label{Intro}
%
{\em Do quarks play an important role in finite nuclei ?}
This is one of the central questions
in nuclear physics. We now know that explicit quark
degrees of freedom are certainly needed to understand deep-inelastic
scattering at momentum transfers of several GeV, e.g., the nuclear EMC
effect\cite{emc}. However, it is perhaps not surprising that such
physics seems irrelevant at a scale of a few tens of MeV, and the idea
that the quark structure of the nucleon should be of any importance in
nuclear physics is often dismissed out of hand.
Within the traditional approach to nuclear physics
these tens of MeV are the result of a fairly fine-tuned cancellation
between short-range repulsion, fairly strong intermediate range
attraction and
large kinetic energy. Nearest
neighbour nucleons in nuclear matter tend to be an average 1.8 fm apart,
which corresponds to the intermediate range attraction associated with
two-pion-exchange. This attraction is mainly generated by the coupling to
the $\Delta$ (through the process NN $\ra$ N$\Delta \ra$ NN),
which is just the first excited state of the internal quark structure of
the nucleon, namely spin-flip. Of course, one can treat the
NN $\ra$ N$\Delta$ system in a coupled channels
formalism without ever mentioning quark structure. One can even replace the
N$\Delta$ coupling by simple scalar meson ($\sigma$) exchange without
significantly altering the binding energy of nuclear systems. If this were
the full story quarks would be irrelevant.
However, it is well known that two-body interactions alone cannot reproduce
either the saturation properties of nuclear matter or the binding energies
and charge radii of finite nuclei. One needs some form of three-body force,
and this leads to a nightmare of badly known form-factors and coupling
constants, of inconsistent treatments of the $\pi$N, NN and NNN
systems, and so on\cite{formf}.
Conventional descriptions of finite nuclei are usually based on the
Schr{\"o}dinger equation which involves nucleons interacting through
static potentials.
These effective potentials are sometimes determined partly by the
Brueckner-Hartree-Fock equations, and often involve some phenomenological
dependence on the nuclear density to include the effect of higher-order
contributions to the N-N force in nuclei\cite{nonrel}. Those
density-dependent
potentials are however {\em purely phenomenological}. Although those
conventional approaches give rather successful results for the ground
states of nuclei, discrepancies between experimental data and theoretical
calculations still remain.
Furthermore, there are several problems at a
low energy scale, which might not be explained by the traditional approach.
One example is the Okamoto-Nolen-Schiffer (ONS) anomaly\cite{nsa},
which is a well known, long-standing problem in nuclear physics.
Conventional nuclear contributions to the anomaly are thought to be at
the few percent level and cannot explain the experimental
findings\cite{nsa2}. The effects of charge symmetry breaking in
the nuclear force, especially $\rho$-$\omega$ mixing, seem to reduce the
discrepancy\cite{nsa,nsa2,hs}. However, recent investigations of
the off-shell variation of the $\rho$-$\omega$ mixing amplitude have put
this explanation into question\cite{ght}.
Another example concerns the unitarity of the Cabibbo-Kobayashi-Maskawa
(CKM) matrix elements.
There is also a well known discrepancy between the value of the
Fermi decay constant extracted from super-allowed Fermi beta-decay of nuclear
isotriplets and that required by unitarity of the
CKM matrix. This discrepancy remains at the level
of a few tenths of a percent (but beyond two standard deviations from the
experimental data) after the most rigorous investigation of
conventional nuclear and radiative corrections\cite{weak}.
These facts seem to indicate that the traditional
approach may have its limitations and suggest a need for
the study of alternate approaches including
sub-nucleonic degrees of freedom.
Exciting a single quark in the nucleon costs about 400 MeV. This is not
significantly different from the energy required to excite a $\Delta$.
It is the same order of magnitude as the scalar and vector potentials
required in Quantum Hadrodynamics (QHD)\cite{qhd}.
Furthermore, the quarks are very light and should be able to
respond rapidly to their environment.
{\em We know of no physical argument why this response should be ignored.}
About eight years ago Guichon\cite{guichon} proposed an entirely different
model for nuclear matter, based on a mean-field description of nucleons
described by the non-overlapping MIT bag model, namely the quark-meson
coupling (QMC) model. Later this model was refined by Saito and
Thomas\cite{st1} who clarified its relationship to QHD.
Related work has been carried out by a number of groups\cite{yazaki}.
The QMC model may be viewed as an extension of QHD in which the
nucleons still interact through the exchange of $\sigma$ and $\omega$
mesons. However, the mesons couple not to point-like nucleons but to
confined quarks. In studies of infinite nuclear matter\cite{guichon,st1}
it was found that
the extra degrees of freedom provided by the internal structure of the
nucleon mean that one gets quite an acceptable value for the
incompressibility once the coupling constants are chosen to
reproduce the correct saturation energy and density for symmetric
nuclear matter. This is a significant improvement on QHD
at the same level of sophistication.
A wonderful feature of this picture is that most specific details
of the model are irrelevant. For example, it does not matter that one
uses the MIT bag model, all that is needed is that the quarks are
relativistic. It does not matter that the $\omega$ generates the
repulsion, all that is needed is that it has a vector character.
There have been several interesting applications to the properties of
finite nuclei using the local-density approximation.
Surprisingly, the model can provide a semi-quantitative explanation
of the ONS anomaly when quark mass differences are
included\cite{st2,juelich}. An application of the model, including quark mass
differences, has also suggested a previously unknown correction to the
extraction of the CKM matrix element, $V_{ud}$, from super-allowed Fermi
beta-decay, which would bring the discrepancy in the unitarity problem of the
CKM matrix back to only one standard deviation\cite{st3,osaka}.
Furthermore, in the light of current experimental work in relativistic
heavy ion collisions\cite{brwn,expts},
which produce nuclear matter at densities several times
normal, there has been some initial work on the variation of baryon and
meson properties with density\cite{hadrons}.
Finally the model has been applied to the case where quark degrees of freedom
are undisputedly involved -- namely the nuclear EMC effect\cite{st4}.
However, the inherent problems of the local-density approximation mean
that these applications can at best be semi-quantitative and it was
clearly very important that the extension to finite nuclei be developed.
Recently we have succeeded in developing a formulation
of the quark-meson coupling
model for finite nuclei\cite{finite}, based on the Born-Oppenheimer
approximation. The effective equation of motion for the nucleon, as
well as self-consistent equations for the meson mean fields have been
derived. We have also shown some initial results for $^{16}$O and
$^{40}$Ca, and discussed, in particular, the spin-orbit force in the
model and its relation to the corresponding force in conventional models
involving meson exchange between point-like nucleons.
Related work has been carried out by several groups\cite{jin,blun}.
Our aim in this paper is to
show details of our
calculations for the properties of spherically symmetric, closed-shell
nuclei from $^{16}$O to $^{208}$Pb.
In Sect.\ref{qmc} we briefly summarize
the QMC model for both infinite nuclear
matter and finite nuclei. Starting from the Lagrangian density proposed in
Ref.\cite{finite} we derive the Dirac equation for the nucleon and the
equations for the meson fields in mean-field approximation and discuss
the effective $\sigma$-N coupling constant. The properties of infinite
nuclear matter are then reviewed briefly in Sect.\ref{matter}.
We introduce the
couplings of the $\rho$ meson and the photon to the quarks,
in Sect.\ref{fnt}, in order to describe
realistic finite nuclei, and again derive explicit forms for the equations
of motion of the nucleon and the mesons. In Sect.\ref{results}, the
parameters in our calculations are determined so as to reproduce the rms
charge radius of $^{40}$Ca, and then our calculated results are
compared with those of QHD and the experimental data.
Our results agree with
data favorably. We give a summary and discuss some remaining problems
and some possible extensions in Sect.\ref{discuss}.
%
%
\section{Relativistic formulation of the QMC model}
\label{qmc}
%
The solution of the general problem of a composite, quantum particle
moving in background scalar and vector fields that vary with position is
extremely difficult. One has a chance to solve the particular problem of
interest to us, namely light quarks confined in a nucleon which is
itself bound in a finite nucleus, only because the nucleon motion is
relatively slow and the quarks highly relativistic. Thus the
Born-Oppenheimer approximation, in which the nucleon internal
structure has time to adjust to the local fields, is naturally suited to
the problem. It is relatively easy to establish that the method should
be reliable at the level of a few percent\cite{finite}.
Even within the Born-Oppenheimer approximation, the nuclear surface
gives rise to external fields that may vary appreciably
across the finite size of the nucleon. In Ref.\cite{finite}, our
approach was to start with a classical nucleon
and to allow its internal structure (quark wavefunctions and bag radius)
to adjust to minimise the energy of three quarks in the ground-state of
a system consisting of the bag plus constant scalar and vector fields,
with values equal to those at the centre of the nucleon.
Of course, the major problem with
the MIT bag (as with many other relativistic models of nucleon structure)
is that it is difficult to boost. We therefore solve the bag
equations in the instantaneous rest frame (IRF) of the nucleon -- using a
standard Lorentz transformation to find the energy and momentum of the
classical nucleon bag in the nuclear rest frame.
Having solved the problem using the meson fields at the centre of
the $\lq\lq$nucleon'' (which is a quasi-particle with nucleon quantum
numbers),
one can use perturbation theory to correct for the variation of the
scalar and vector fields across the nucleon bag. In first order perturbation
theory only the spatial components of the vector potential
give a non-vanishing contribution. (Note that, although in the nuclear
rest frame only the time component of the vector field is non-zero,
in the IRF of the nucleon there are also non-vanishing spatial components.)
This extra term is a correction to the spin-orbit force, which is
however very
small\cite{finite}. Blunden and Miller\cite{blun} have also
investigated
the QMC model for finite nuclei recently.
They too reported
that the correction due to the variation of meson fields inside the
nucleon is not large.
As shown in Ref.\cite{finite}, the basic result in the QMC model is that,
in the scalar ($\sigma$) and vector ($\omega$)
meson fields, the nucleon behaves essentially as a point-like
particle with an effective mass
$M_N^{\star}$, which depends on the position through only the $\sigma$
field, moving in a vector potential generated by the $\omega$ meson.
Because of the vector character, the vector interactions have {\em no effect
on the nucleon structure} except for an overall phase in the wave function,
which gives a shift in the nucleon energy.
If we restrict ourselves to consider static, spherically
symmetric nuclei, the effective Lagrangian density involving the quark
degrees of freedom and the meson fields in mean-field approximation would
be given by\cite{finite}
%
\bg
{\cal L}_{QMC}^0 &=& \overline{\psi} [i \gamma \cdot \partial
- M_N^{\star}(\sigma({\vec r})) - g_\omega \omega({\vec r}) \gamma_0]
\psi \nn \\
&-& \frac{1}{2}[ (\nabla \sigma({\vec r}))^2 +
m_{\sigma}^2 \sigma({\vec r})^2 ]
+ \frac{1}{2}[ (\nabla \omega({\vec r}))^2 + m_{\omega}^2
\omega({\vec r})^2 ] ,
\label{relat}
\en
%
where $\psi({\vec r})$, $\sigma({\vec r})$ and $\omega({\vec r})$ are
respectively the
nucleon, $\sigma$ and the time component of $\omega$ fields. (In the mean
field approximation, the meson fields are given by their time independent
expectation values in the ground state of the nucleus. For a symmetric,
parity-eigenstate, the spatial component of the $\omega$ field vanishes.)
$m_\sigma$ and $m_\omega$ are respectively the masses of the
$\sigma$ and $\omega$
mesons. $g_\omega$ is the $\omega$-N coupling constant (which is related to
the corresponding quark-$\omega$ coupling constant,
$g_\omega^q$, as $g_\omega =
3 g_\omega^q$\cite{guichon,st1,finite}).
The effective nucleon mass $M_N^{\star}$ is given by a model describing the
nucleon structure and it depends on the position through only the $\sigma$
field (as mentioned above). We use the MIT bag model in this paper.
(A relativistic oscillator model has also been used in Ref.\cite{blun} as an
alternative model.) If we define the field-dependent $\sigma$-N coupling
constant, $g_\sigma(\sigma)$, by
%
\bge
M_N^{\star}(\sigma({\vec r})) \equiv M_N - g_\sigma(\sigma({\vec r}))
\sigma({\vec r}) , \label{coup}
\ene
%
where $M_N$ is the free nucleon mass, it is easy to compare with
QHD\cite{qhd}. The Lagrangian density then becomes
%
\bg
{\cal L}_{QMC}^0 &=& \overline{\psi} [i \gamma \cdot \partial
- M_N + g_\sigma (\sigma({\vec r})) \sigma({\vec r})
- g_\omega \omega({\vec r}) \gamma_0] \psi \nn \\
&-& \frac{1}{2}[ (\nabla \sigma({\vec r}))^2 +
m_{\sigma}^2 \sigma({\vec r})^2 ]
+ \frac{1}{2}[ (\nabla \omega({\vec r}))^2 + m_{\omega}^2
\omega({\vec r})^2 ] .
\label{relat2}
\en
%
The difference from QHD clearly lies in the fact that the internal
structure of the nucleon has forced a {\em known} dependence of the
$\sigma$-N coupling constant on the scalar field itself. (Note that this
dependence is not the same as that adopted in the density-dependent
hadron field theory\cite{fuchs} where the meson-nucleon vertices are
assumed to depend directly on the baryonic densities.)
Equation (\ref{relat}) yields the Dirac equation for the nucleon
%
\bge
[i\gamma \cdot \partial -M_N^{\star}(\sigma)-
g_\omega \gamma_0 \omega ] \psi = 0 ,
\label{dirac}
\ene
as well as the equations for the meson mean-fields
%
\bg
(-\nabla^2_r+m^2_\sigma)\sigma(\vec{r})&=& - \left(
\frac{\partial}{\partial \sigma}M_N^{\star}(\sigma) \right)
\rho_s({\vec r}), \label{sig} \\
(-\nabla^2_r+m^2_\omega) \omega(\vec{r}) &=&
g_\omega \rho_B({\vec r}), \label{omg}
\en
%
where $\rho_s$ and $\rho_B$ are respectively the scalar and baryon densities
of the nucleon in the nucleus, $A$, which are defined by the expectation
values in the ground state of $A$:
%
\bge
\rho_s({\vec r}) = \langle A|\overline{\psi}\psi(\vec{r})|A \rangle \ \
\mbox{and} \ \ \rho_B({\vec r}) = \langle A|\psi^\dagger\psi(\vec{r})|A
\rangle . \label{density}
\ene
%
One can easily see that the derivative of $M_N^{\star}$ with respect to
$\sigma$ in Eq.(\ref{sig}) is the response of the nucleon to the
external scalar field, and that it is given by the scalar density of a
quark in the nucleon\cite{st1}:
%
\bge
\left( \frac{\partial M_N^{\star}}{\partial \sigma} \right)
= -3g_{\sigma}^q \int_{bag} d{\vec r} \ {\overline \psi}_q \psi_q
\equiv -3g_{\sigma}^q S(\sigma) , \label{deriv}
\ene
%
where $g_\sigma^q$ is the quark-$\sigma$ coupling constant and
$\psi_q$ is the quark wave function in the nucleon bag with
radius $R_B^{\star}$ in matter (we denote the bag radius of the free nucleon
by $R_B$). We define $S(\sigma)$ by Eq.(\ref{deriv}),
which is a function of scalar field itself.
Using the MIT bag model\cite{finite}, $S$ is explicitly given by
%
\bge
S(\sigma(\vec{r})) = \frac{\Omega/2 + m_q^{\star}R_B^{\star}(\Omega-1)}
{\Omega(\Omega-1) + m_q^{\star}R_B^{\star}/2}, \label{sss}
\ene
%
where $\Omega = \sqrt{x^2 + (R_B^{\star}m_q^{\star})^2}$ is the kinetic
energy of
the quark and $m_q^{\star}({\vec r})$ is the effective quark mass defined
by $m_q - g_\sigma^q \sigma({\vec r})$, with the quark mass $m_q$
inside the free nucleon.
Furthermore, we define the scalar density ratio, $S(\sigma)/S(0)$,
by $C(\sigma)$ and the $\sigma$-N coupling constant at $\sigma=0$ by
$g_\sigma$ (i.e., $g_\sigma \equiv g_\sigma(\sigma=0)$)\footnote{
Note the change in notation from the earlier works of Saito and Thomas where
$C$ was used for what we now call $S$. }:
%
\bge
C(\sigma) = S(\sigma)/S(0) \ \ \mbox{and} \ \
g_{\sigma} = 3g_{\sigma}^q S(0) . \label{cn}
\ene
%
Comparing with Eq.(\ref{coup}), we find
%
\bge
\left( \frac{\partial M_N^{\star}}{\partial \sigma} \right)
= -g_{\sigma} C(\sigma) = - \frac{\partial}{\partial \sigma}
\left[ g_\sigma(\sigma) \sigma \right],
\label{deriv2}
\ene
%
and the equation for the $\sigma$ field becomes
%
\bge
(-\nabla^2_r+m^2_\sigma)\sigma(\vec{r}) = g_\sigma C(\sigma)
\rho_s({\vec r}) . \label{sig2}
\ene
%
We will discuss the $\sigma$-N coupling constant further in the next
subsection.
%
\subsection{Nuclear matter limit}
\label{matter}
%
In isospin-symmetric, infinite nuclear matter the sources of the fields
are constant and can be related to the nucleon Fermi momentum
$k_F$\cite{qhd}:
%
\bg
\rho_B &=& \frac{4}{(2\pi)^3}\int d\vec{k} \theta (k_F - k) =
\frac{2 k_F^3}{3\pi^2} ,
\label{rhoB}\\
\rho_s &=& \frac{4}{(2\pi)^3}\int d\vec{k} \theta (k_F - k)
\frac{M_N^{\star}}{\sqrt{M_N^{\star 2}+\vec{k}^2}},
\label{rhos}
\en
%
where $M_N^{\star}$ means the constant value of the effective nucleon
mass given by the MIT bag model (detailed derivations are given in
Ref.\cite{finite}).
As in Ref.\cite{finite}, the bag constant $B$ and the
parameter $z_0$ (which accounts for the sum of the c.m. and
gluon fluctuation corrections) in the familiar form of the MIT bag
model Lagrangian are
fixed to reproduce the free nucleon mass ($M_N$ = 939 MeV)
under the stationary
condition in terms of the free bag radius, $R_B$. In the following we
treat the free bag radius as a variable parameter to test
the sensitivity of our
results to the free nucleon size. The results for $B$ and $z_0$ are
shown in Table~\ref{b,z}. We choose the free quark mass $m_q$ = 0, 5,
10 MeV.
%
\begin{table}[htbp]
\begin{center}
\caption{Bag constant and parameter $z_0$ for several bag radii and quark
masses.}
\label{b,z}
\begin{tabular}[t]{c|ccc|ccc|ccc}
\hline
$m_q$(MeV) & & 0 & & & 5 & & & 10 & \\
\hline
$R_B$(fm) & 0.6 & 0.8 & 1.0 & 0.6 & 0.8 & 1.0 & 0.6 & 0.8 & 1.0 \\
\hline
$B^{1/4}$(MeV) &211.3&170.3&144.1&210.9&170.0&143.8&210.5&169.6&143.5 \\
$z_0$ &3.987&3.273&2.559&4.003&3.295&2.587&4.020&3.317&2.614 \\
\hline
\end{tabular}
\end{center}
\end{table}
%
Let ($\overline{\sigma}, \overline{\omega}$) be the {\em constant}
mean-values of the meson fields. From Eqs.(\ref{omg}) and (\ref{sig2})
we find
%
\bg
\overline{\omega}&=&\frac{g_\omega \rho_B}{m_\omega^2} , \label{omgf}\\
\overline{\sigma}&=&\frac{g_\sigma }{m_\sigma^2}C(\overline{\sigma})
\frac{4}{(2\pi)^3}\int d\vec{k} \theta (k_F - k)
\frac{M_N^{\star}}{\sqrt{M_N^{\star 2}+\vec{k}^2}} , \label{sigf}
\en
%
where $C(\overline{\sigma})$ is now the constant value of $C$ in the
scalar field.
This self-consistency equation for $\overline{\sigma}$
is the same as that in QHD, {\it except} that in the latter model one has
$C(\overline{\sigma})=1$ \cite{st1,blun}. (This corresponds to the
heavy-quark mass limit of the QMC model.)
%
%\begin{figure}[htb]
%\centering{\
%%\epsfile{file=cmc1.ps,height=7cm}
%\epsfig{file=cmc1.ps,height=7cm}
%\caption{Mean-field values of the $\sigma$ meson for various bag radii
%as a function of $\rho_B$. The solid, dotted and dashed curves show
%$g_{\sigma} {\overline \sigma}$ for $R_B$ = 0.6, 0.8 and
%1.0 fm, respectively. The quark mass is chosen to be 5 MeV.}
%\label{fig:sigma}}
%\end{figure}
%
\begin{figure}[htb]
\centering{\
%\epsfile{file=cmc2.ps,height=7cm}
\epsfig{file=cmc2.ps,height=7cm}
\caption{Effective nucleon mass as a function of density (for $m_q$=5MeV).
The solid, dotted and dashed curves correspond to
$R_B$ = 0.6, 0.8 and 1.0 fm, respectively.}
\label{fig:mass}}
\end{figure}
%
%\begin{figure}[htb]
%\centering{\
%\epsfile{file=cmc3.ps,height=7cm}
%\epsfig{file=cmc3.ps,height=7cm}
%\caption{Scalar density ratio, $C({\overline \sigma})$, as a function of
%$g_{\sigma} {\overline \sigma}$ ($m_q$ = 5 MeV). The curves are
%labelled as in Fig.1. }
%\label{fig:C}}
%\end{figure}
%
Once the self-consistency equation for $\overline{\sigma}$ has been solved,
one can evaluate the total energy per nucleon\cite{guichon,st1,finite}:
%
\bge
E^{total}/A=\frac{4}{(2\pi)^3 \rho_B}\int d\vec{k}
\theta (k_F - k) \sqrt{M_N^{\star 2}+
\vec{k}^2}+\frac{m_\sigma^2\overline{\sigma}^2}{2 \rho_B}+
\frac{g_\omega^2\rho_B}{2m_\omega^2} .
\label{toten}
\ene
%
We determine the coupling constants, $g_{\sigma}$ and $g_{\omega}$, so as
to fit the binding energy ($-15.7$ MeV) per nucleon and the saturation
density, $\rho_0$ = 0.15 fm$^{-3}$ ($k_F^0$ = 1.305 fm$^{-1}$), for
symmetric nuclear matter.
The coupling constants and some calculated properties
of nuclear matter (for $m_{\sigma}=550$ MeV and
$m_{\omega}=783$ MeV) at the saturation density are listed in
Table~\ref{c.c.}.
The last three columns show the relative changes (from their values at
zero density) of the bag radius
($\delta R_B^{\star}/R_B$), the lowest eigenvalue ($\delta x/x_0$)
and the rms radius of the nucleon calculated by the quark wave function
($\delta r_q^{\star}/r_q$) at saturation density.
%
\begin{table}[htbp]
\begin{center}
\caption{Coupling constants and calculated properties for
symmetric nuclear matter at normal nuclear density.
The effective nucleon mass, $M_N^{\star}$, and the nuclear
incompressibility, $K$, are quoted in MeV.
The bottom row is for QHD.}
\label{c.c.}
\begin{tabular}[t]{c|ccccccccc}
\hline
$m_q$(MeV)&$R_B$(fm)&$g_{\sigma}^2/4\pi$&$g_{\omega}^2/4\pi$&
$M_N^{\star}$&$K$&$\delta R_B^{\star}/R_B$&$\delta x/x_0$&$\delta
r_q^{\star}/r_q$ \\
\hline
& 0.6 & 5.84 & 6.29 & 730 & 293 & $-0.02$ & $-0.13$ & 0.01 \\
0 & 0.8 & 5.38 & 5.26 & 756 & 278 & $-0.02$ & $-0.16$ & 0.02 \\
& 1.0 & 5.04 & 4.50 & 774 & 266 & $-0.02$ & $-0.19$ & 0.02 \\
\hline
& 0.6 & 5.86 & 6.34 & 729 & 295 & $-0.02$ & $-0.13$ & 0.01 \\
5 & 0.8 & 5.40 & 5.31 & 754 & 280 & $-0.02$ & $-0.16$ & 0.02 \\
& 1.0 & 5.07 & 4.56 & 773 & 267 & $-0.02$ & $-0.19$ & 0.02 \\
\hline
& 0.6 & 5.87 & 6.37 & 728 & 295 & $-0.02$ & $-0.13$ & 0.02 \\
10 & 0.8 & 5.42 & 5.36 & 753 & 281 & $-0.02$ & $-0.16$ & 0.02 \\
& 1.0 & 5.09 & 4.62 & 772 & 269 & $-0.02$ & $-0.18$ & 0.03 \\
\hline
QHD & & 7.29 & 10.8 & 522 & 540 & & & \\
\hline
\end{tabular}
\end{center}
\end{table}
%
The most notable fact is that the calculated incompressibility, $K$,
is well within the experimental range: $K = 200 \sim
300$ MeV. Also our effective nucleon mass is much larger than in the
case of QHD.
Although the bag radius shrinks a little at finite density, the rms radius
of the quark wavefunction actually {\em increases} by a few percent
at saturation density.
In Fig.\ref{fig:mass}, we show the
effective nucleon mass in medium, as a function of density,
for $m_q$ = 5 MeV.
Its dependence on the quark mass is weak.
The scalar density ratio $C({\overline \sigma})$
decreases linearly (to a very good approximation) with
$g_{\sigma} {\overline \sigma}$\cite{finite}.
It is very useful to have a simple parametrization for $C$ and
the form:
%
\bge
C({\overline \sigma}) = 1 - a \times (g_{\sigma}{\overline \sigma}) ,
\label{paramC}
\ene
%
with $g_{\sigma} {\overline \sigma}$ in MeV (recall $g_{\sigma} =
g_{\sigma}(\sigma=0)$) and
$a$ = (6.6, 8.8, 11) $\times 10^{-4}$ for $R_B$ = (0.6, 0.8, 1.0) fm,
respectively, is quite accurate up to $2 \sim 3\rho_0$.
(These values for the slope parameter, $a$, are valid for
all quark masses listed.)
As a practical matter, it is easy to solve Eq.(\ref{deriv2}) for
$g_\sigma(\sigma)$ in the
case where $C(\sigma)$ is linear in $g_{\sigma} \overline{\sigma}$ as
Eq.(\ref{paramC}). In fact, we find
%
\bge
M^{\star}_N = M_N - g_\sigma \left[ 1 - \frac{a}{2} (g_\sigma
{\overline \sigma}) \right] {\overline \sigma} .
\label{mstaR}
\ene
%
The effective $\sigma$-N coupling constant, $g_{\sigma}(\sigma)$,
decreases at half the rate of $C(\sigma)$.
Eq.(\ref{mstaR}) is quite accurate up to twice nuclear matter density.
Having explicitly solved the nuclear matter problem by
self-consistently solving for the quark wave functions in the bag in the
mean scalar field, {\em one can solve for the properties of finite nuclei
without explicit reference} to the internal structure of the nucleon.
All one needs is Eqs.(\ref{paramC}) and (\ref{mstaR}) for $C(\sigma)$ and
$M^{\star}_N$ as a function of $g_{\sigma} \overline{\sigma}$.
\clearpage
%
\subsection{Finite nuclei}
\label{fnt}
%
To describe a nucleus with different numbers of protons and neutrons
($Z \neq N$), it is necessary to consider the contributions of the $\rho$
meson. Any realistic treatment of nuclear structure also requires that one
introduces the Coulomb force.
The interaction Lagrangian density for the $\rho$- and
$\gamma$-quark couplings in the nucleon IRF is given by
%
\bge
{\cal L}_I^{\rho+\gamma} = - g^q_{\rho} \bar{q}' \frac{\tau_{\alpha}^q}{2}
\gamma_{\mu} q' \rho^{\mu}_{\alpha,IRF} - e \bar{q}' (\frac{1}{6} +
\frac{\tau_3^q}{2}) \gamma_{\mu} q' A^{\mu}_{IRF} ,
\label{rhogamma}
\ene
%
where $q$ is the quark field (the prime means the IRF) and
$\rho^\mu_{\alpha,IRF}$ and $A^{\mu}_{IRF}$
are respectively the $\rho$ meson field with isospin component $\alpha$
and the photon field.
$\tau_{\alpha}^q/2$ is the isospin operator for the quarks. $g_{\rho}^q$
and $e$ are the quark-$\rho$ coupling constant and the electric charge.
We saw in Ref.\cite{finite} that this leads to both a central and a
spin-orbit potential for the nucleon in the nuclear rest frame. For the
isoscalar, $\omega$ meson these potentials are well represented by a
vector $\omega$-nucleon coupling because the isoscalar magnetic moment
of the nucleon is near unity. However, for the $\rho$, the relativistic
formulation {\em at the nucleon level} requires a strong tensor coupling
($\sim \sigma^{\mu \nu} q_\nu$) if it is to reproduce the interaction
given by the quark model\cite{finite}.
In Hartree approximation a tensor coupling gives only a spin-orbit force
for a nucleon bound in a static, spherical nucleus. However, in
Hartree-Fock it can also give rise to a central force which would be
necessary to reproduce the bulk symmetry energy. For simplicity, in the
present work we follow the usual procedure adopted in the Hartree
treatment of QHD of using only a vector coupling for the $\rho$, with
$g_\rho$ adjusted to give the bulk symmetry energy in MFA -- where
$g_\rho = g_\rho^q$. (The accuracy
of this approximation will be examined in detail in future work.)
In this case, in mean field approximation, only $\alpha =3$ contributes
in Eq.(\ref{rhogamma}).
If we denote by $b(\vec{r})$ and $A({\vec r})$ the mean values of the
time component of $\rho$ field and the Coulomb field in the nuclear
rest frame,
we can transpose our results for the $\omega$ field by trivial isospin
factors:
%
\bge
3g^q_\omega \omega({\vec r}) \to g_\rho\frac{\tau^N_3}{2} b({\vec r})
\ \ \mbox{or} \ \ \frac{e}{2}(1 + \tau^N_3) A({\vec r}) ,
\label{subst}
\ene
%
where $\tau^N_3/2$ is the third component of the nucleon
isospin operator.
In summary, our effective Lagrangian density in mean field
approximation takes the form:
%
\bg
{\cal L}_{QMC}&=& \overline{\psi} [i \gamma \cdot \partial
- M_N + g_\sigma (\sigma({\vec r})) \sigma({\vec r})
- g_\omega \omega({\vec r}) \gamma_0 \nn \\
&-& g_\rho \frac{\tau^N_3}{2} b({\vec r}) \gamma_0
- \frac{e}{2} (1+\tau^N_3) A({\vec r}) \gamma_0 ] \psi \nn \\
&-& \frac{1}{2}[ (\nabla \sigma({\vec r}))^2 +
m_{\sigma}^2 \sigma({\vec r})^2 ]
+ \frac{1}{2}[ (\nabla \omega({\vec r}))^2 + m_{\omega}^2
\omega({\vec r})^2 ] \nn \\
&+& \frac{1}{2}[ (\nabla b({\vec r}))^2 + m_{\rho}^2 b({\vec r})^2 ]
+ \frac{1}{2} (\nabla A({\vec r}))^2.
\label{qmclag}
\en
%
The variation of the Lagrangian results in the
following equations for static, spherically symmetric nuclei (see also
Eqs.(\ref{omg}) and (\ref{sig2})):
%
\bg
%
\frac{d^2}{dr^2} \sigma(r) + \frac{2}{r} \frac{d}{dr} \sigma(r)
- m_\sigma^2 \sigma(r) &=& - g_\sigma C(\sigma(r)) \rho_s(r) \nn \\
&\equiv& - g_\sigma C(\sigma(r)) \sum_\alpha^{occ} d_\alpha(r)
(|G_\alpha(r)|^2 - |F_\alpha(r)|^2), \label{sig3} \\
%
\frac{d^2}{dr^2} \omega(r) + \frac{2}{r} \frac{d}{dr} \omega(r)
- m_\omega^2 \omega(r) &=& - g_\omega \rho_B(r) \nn \\
&\equiv& - g_\omega \sum_\alpha^{occ} d_\alpha(r)
(|G_\alpha(r)|^2 + |F_\alpha(r)|^2), \label{omg3} \\
%
\frac{d^2}{dr^2} b(r) + \frac{2}{r} \frac{d}{dr} b(r)
- m_\rho^2 b(r) &=& - \frac{g_\rho}{2} \rho_3(r) \nn \\
&\equiv& - \frac{g_\rho}{2} \sum_\alpha^{occ}
d_\alpha(r) (-)^{t_\alpha -1/2}
(|G_\alpha(r)|^2 + |F_\alpha(r)|^2), \label{rho3} \\
%
\frac{d^2}{dr^2} A(r) + \frac{2}{r} \frac{d}{dr} A(r)
&=& - e \rho_p(r) \nn \\
&\equiv& - e \sum_\alpha^{occ} d_\alpha(r)
(t_\alpha + \frac{1}{2})
(|G_\alpha(r)|^2 + |F_\alpha(r)|^2), \label{phtn3}
%
\en
%
where $d_\alpha(r)= (2j_\alpha+1)/4\pi r^2$ and
%
\bg
\frac{d}{dr} G_\alpha(r) + \frac{\kappa}{r} G_\alpha(r) -
\left[ \epsilon_\alpha - g_\omega \omega(r) - t_\alpha g_\rho b(r)
\right.
&-& \left. (t_\alpha + \frac{1}{2}) e A(r) + M_N \right. \nn \\
&-& \left. g_\sigma(\sigma(r)) \sigma(r) \right] F_\alpha(r) = 0 ,
\label{qwave1} \\
%
\frac{d}{dr} F_\alpha(r) - \frac{\kappa}{r} F_\alpha(r) +
\left[ \epsilon_\alpha - g_\omega \omega(r) - t_\alpha g_\rho b(r)
\right.
&-& \left. (t_\alpha + \frac{1}{2}) e A(r) - M_N \right. \nn \\
&+& \left. g_\sigma (\sigma(r)) \sigma (r) \right] G_\alpha(r) = 0 .
\label{qwave2}
\en
%
Here $iG_\alpha(r)/r$ and $-F_\alpha(r)/r$ are respectively the
radial part of the upper and lower
components of the solution to the Dirac equation for the nucleon
($\alpha$ labelling the quantum numbers and $\epsilon_\alpha$ being
the energy) under the normalization condition:
%
\bge
\int dr (|G_\alpha(r)|^2 + |F_\alpha(r)|^2) =1 . \label{norm}
\ene
%
As usual, $\kappa$ specifies the angular quantum numbers and $t_\alpha$
the eigenvalue of the isospin operator $\tau^N_3/2$. $C(\sigma)$ and
$g_\sigma (\sigma)$ are given by Eqs.(\ref{sss}) and
(\ref{coup}), respectively -- or practically,
by Eqs.(\ref{paramC}) and (\ref{mstaR}), i.e.,
%
\bge
g_\sigma (\sigma({\vec r})) = g_\sigma \left[ 1 - \frac{a}{2}
g_\sigma \sigma({\vec r}) \right] . \label{gsgm}
\ene
%
The total energy of the system is then given by
%
\bg
E_{tot} &=& \sum_\alpha^{occ} (2j_\alpha + 1) \epsilon_\alpha
- \frac{1}{2} \int d{\vec r} \ [ -g_\sigma C(\sigma (r)) \sigma(r)
\rho_s(r) \nn \\
&+& g_\omega \omega(r) \rho_B(r) + \frac{1}{2} g_\rho b(r) \rho_3(r)
+ eA(r) \rho_p(r) ]. \label{ftoten}
\en
%
%
\section{Numerical results for finite nuclei}
\label{results}
%
Equations (\ref{sig3}) to (\ref{norm}) give a set of coupled non-linear
differential equations, which may be solved by a standard iteration
procedure\cite{horowitz}. In this paper the numerical calculation was
carried out by modifying the technique described by Horowitz
{\it et al.}\cite{horowitz,hor2}. Changing the initial condition for
the meson fields, the $\sigma$-N coupling constant and the scalar density
in their fortran program\cite{hor2}, the calculation can be performed
very easily\footnote{If the numerical convergence is slow
it may be improved by mixing appropriately the meson potentials
given by the $i$-th iteration and those by the ($i-1$)-th iteration -- as is
usually done in non-relativistic calculations.}.
The calculation is achieved in at most
20 iterations when it is performed with a maximum radius of 12 (15) fm on
a mesh of 0.04 fm for medium mass (Pb) nuclei.
%
\subsection{Determination of parameters}
\label{param}
%
There are seven parameters to be determined: $g_\sigma$, $g_\omega$,
$g_\rho$, $e$, $m_\sigma$, $m_\omega$ and $m_\rho$. We take the
experimental values: $m_\omega$ = 783 MeV, $m_\rho$ = 770 MeV and
$e^2/4\pi$ = 1/137.036.
The coupling constants $g_\sigma$ and
$g_\omega$ are fixed to describe the nuclear matter properties
with $m_\sigma$ = 550 MeV in Sec.\ref{matter} (see Table~\ref{c.c.}).
The $\sigma$-meson mass however determines the range of the attractive
interaction and changes in $m_\sigma$ affect the nuclear-surface slope and
its thickness. Therefore, as in the paper of Horowitz and
Serot\cite{horowitz}, we adjust $m_\sigma$ to fit the rms charge radius of
$^{40}$Ca, $r_{ch}$($^{40}$Ca) = 3.48 fm, the experimental
value\cite{sinha}.
We should notice here that {\em variations of $m_\sigma$ at fixed}
($g_\sigma / m_\sigma$)
{\em have no effect on the infinite nuclear matter properties}. Therefore,
keeping the ratio ($g_\sigma / m_\sigma$) constant we vary $m_\sigma$ to fit
the rms charge radius of $^{40}$Ca. We
expect that $m_\sigma$ ranges around 400 $\sim$ 550 MeV\cite{blun,mach}.
The last parameter, $g_\rho$, is adjusted to yield the bulk symmetry
energy per
baryon of 35 MeV\cite{horowitz}. We summarize the parameters in
Table~\ref{c.c.2}.
%
\begin{table}[htbp]
\begin{center}
\caption{Model parameters for finite nuclei.}
\label{c.c.2}
\begin{tabular}[t]{cccccc}
\hline
$m_q$(MeV)&$R_B$(fm)&$g_{\sigma}^2/4\pi$&$g_{\omega}^2/4\pi$&
$g_{\rho}^2/4\pi$&$m_\sigma$(MeV) \\
\hline
& 0.6 & 3.55 & 6.29 & 6.79 & 429 \\
0 & 0.8 & 2.94 & 5.26 & 6.93 & 407 \\
& 1.0 & 2.51 & 4.50 & 7.03 & 388 \\
\hline
& 0.6 & 3.68 & 6.34 & 6.78 & 436 \\
5 & 0.8 & 3.12 & 5.31 & 6.93 & 418 \\
& 1.0 & 2.69 & 4.56 & 7.02 & 401 \\
\hline
& 0.6 & 3.81 & 6.37 & 6.78 & 443 \\
10 & 0.8 & 3.28 & 5.36 & 6.92 & 428 \\
& 1.0 & 2.91 & 4.62 & 7.02 & 416 \\
\hline
\end{tabular}
\end{center}
\end{table}
%
%
\begin{figure}[htb]
\centering{\
%\epsfile{file=pbd58.ps,height=9cm}
\epsfig{file=pbd58.ps,height=9cm}
\caption{Model predictions for the effective nucleon mass, the baryon and
scalar densities in $^{208}$Pb (for $m_q$ = 5 MeV and $R_B$ = 0.8 fm).
The scale on the right vertical axis is for $M_N^{\star}/M_N$. }
\label{pbd58}}
\end{figure}
%
In Fig.\ref{pbd58}, we show the baryon and scalar densities as well as
the effective nucleon mass in lead. We expect that the baryon density
in the interior of lead would be close to the saturation density of infinite
nuclear matter. As seen in the figure, the calculated baryon density
at the center is quite close to 0.15 fm$^{-1}$, which supports our choice of
the parameters.
\clearpage
%
\subsection{Comparison with experimental data}
\label{comp}
%
First we show calculated charge density distributions, $\rho_{ch}$,
in comparison with those of QHD\cite{qhd,horowitz} and the experimental
data in Figs.\ref{ca08}$-$\ref{ch48a1}.
Having solved Eqs.(\ref{sig3}) $\sim$ (\ref{norm}), we obtain the
{\em point-}proton and neutron densities in a nucleus.
In addition to those densities, we should include
the effects of c.m. corrections, nucleon form factors, meson-exchange
currents, etc. However, one knows that the dominant correction comes
from the proton-form factor\cite{horowitz}. We calculate the charge
density as a
convolution of the point-proton density, $\rho_p({\vec r})$, with
the proton charge
distribution, $\rho^p_{ch}({\vec r})$:
%
\bge
\rho_{ch}({\vec r}) = \int d{\vec r}\, ' \ \rho^p_{ch}({\vec r} -
{\vec r}\, ') \rho_p({\vec r}\, ') , \label{charge}
\ene
%
where we have used a gaussian form for $\rho^p_{ch}$
%
\bge
\rho^p_{ch}({\vec r}) = (\beta/\pi)^{3/2} \exp (-\beta r^2) .
\label{gauss}
\ene
%
The parameter $\beta$, which determines the proton size, is chosen so
as to reproduce the experimental rms charge radius of the proton, 0.82 fm
(i.e., $\beta$ = 2.231 fm$^{-2}$).
As we noted before, in the present model the rms radius of the nucleon
in nuclear matter increases a little. However, this
amount is quite small and it can be ignored in the numerical calculations
of nuclear parameters.
The charge density distributions for $^{40}$Ca (for $m_q$ = 0, 5, 10 MeV
and $R_B$ = 0.8 fm) are presented in Figs. \ref{ca08}$-$\ref{ca18}.
%
\begin{figure}[htb]
\centering{\
%\epsfile{file=ca08.ps,height=7cm}
\epsfig{file=ca08.ps,height=7cm}
\caption{Charge density distribution for $^{40}$Ca (for $m_q$ = 0 MeV and
$R_B$ = 0.8 fm) compared with the experimental data and that of QHD. }
\label{ca08}}
\end{figure}
%
\begin{figure}[htb]
\centering{\
%\epsfile{file=ca58.ps,height=7cm}
\epsfig{file=ca58.ps,height=7cm}
\caption{Same as Fig.5 (for $m_q$ = 5 MeV and $R_B$ = 0.8 fm). }
\label{ca58}}
\end{figure}
%
\begin{figure}[htb]
\centering{\
%\epsfile{file=ca18.ps,height=7cm}
\epsfig{file=ca18.ps,height=7cm}
\caption{Same as Fig.5 (for $m_q$ = 10 MeV and $R_B$ = 0.8 fm). }
\label{ca18}}
\end{figure}
%
The experimental data is taken from Ref.\cite{cadata}.
Once the rms charge radius of $^{40}$Ca is fitted by adjusting $m_\sigma$
the QMC model can reproduce $\rho_{ch}$($^{40}$Ca) quite well.
As seen in the figures, the calculated charge densities lie almost within
the experimental (hatched) area. We note that the dependence of
$\rho_{ch}$($^{40}$Ca) on the bag radius is quite weak.
It is interesting to see the quantum oscillations of the interior
density in lead. The dependence of $\rho_{ch}$($^{208}$Pb) on the
quark mass is illustrated in Figs.\ref{pb08}$-$\ref{pb18}.
%
\begin{figure}[htb]
\centering{\
%\epsfile{file=pb08.ps,height=7cm}
\epsfig{file=pb08.ps,height=7cm}
\caption{Charge density distribution for $^{208}$Pb (for $m_q$ = 0 MeV and
$R_B$ = 0.8 fm) compared with the experimental data and that of QHD. }
\label{pb08}}
\end{figure}
%
\begin{figure}[htb]
\centering{\
%\epsfile{file=pb58.ps,height=7cm}
\epsfig{file=pb58.ps,height=7cm}
\caption{Same as Fig.8 (for $m_q$ = 5 MeV and $R_B$ = 0.8 fm). }
\label{pb58}}
\end{figure}
%
\begin{figure}[htb]
\centering{\
%\epsfile{file=pb18.ps,height=7cm}
\epsfig{file=pb18.ps,height=7cm}
\caption{Same as Fig.8 (for $m_q$ = 10 MeV and $R_B$ = 1.0 fm). }
\label{pb18}}
\end{figure}
%
We note that the dependence of $\rho_{ch}$($^{208}$Pb)
on the free nucleon size is also quite
weak. As seen in the figures, our model gives charge densities
very close to those of QHD and still somewhat larger in the central
region than those observed experimentally\cite{frois}.
In Figs.\ref{o58} and \ref{zr58}, we show respectively the charge density
distributions for $^{16}$O and $^{90}$Zr. For zirconium the calculated
$\rho_{ch}$ lies in between that of the non-relativistic density-dependent
Hartree-Fock calculations\cite{nonrel} and that of QHD.
%
\begin{figure}[htb]
\centering{\
%\epsfile{file=o58.ps,height=7cm}
\epsfig{file=o58.ps,height=7cm}
\caption{Charge density distribution for $^{16}$O (for $m_q$ = 5 MeV and
$R_B$ = 0.8 fm) compared with the experimental data and that of QHD. }
\label{o58}}
\end{figure}
%
\begin{figure}[htb]
\centering{\
%\epsfile{file=zr58.ps,height=7cm}
\epsfig{file=zr58.ps,height=7cm}
\caption{Charge density distribution for $^{90}$Zr (for $m_q$ = 5 MeV and
$R_B$ = 0.8 fm) compared with the experimental data and that of QHD. }
\label{zr58}}
\end{figure}
%
The experimental data for oxygen and zirconium are taken
from Refs.\cite{odata} and \cite{zdata}. (For both cases the dependence of
$\rho_{ch}$ on $m_q$ and $R_B$ is weak.)
To see the isotope shift in charge density we have plotted $r^2$ times the
difference between $\rho_{ch}$($^{40}$Ca) and $\rho_{ch}$($^{48}$Ca) in
Figs.\ref{ch48a05} and \ref{ch48a1}. Its dependence on the bag radius is
again weak for a small quark mass while it becomes a little stronger
for $m_q$ = 10 MeV, as shown in Fig.\ref{ch48a1}.
%
\begin{figure}[htb]
\centering{\
%\epsfile{file=ch48a05.ps,height=7cm}
\epsfig{file=ch48a05.ps,height=7cm}
\caption{Isotope shift between $\rho_{ch}$($^{40}$Ca) and
$\rho_{ch}$($^{48}$Ca) compared with the experimental data and that of QHD
(for $m_q$ = 0 and 5 MeV with $R_B$ = 0.8 fm). }
\label{ch48a05}}
\end{figure}
%
\begin{figure}[htb]
\centering{\
%\epsfile{file=ch48a1.ps,height=7cm}
\epsfig{file=ch48a1.ps,height=7cm}
\caption{Same as Fig.13 (for $m_q$ = 10 with $R_B$ = 0.6, 0.8 and 1.0 fm).}
\label{ch48a1}}
\end{figure}
%
The experimental data is taken from Ref.\cite{ch48data}. (Note that in
this case we also checked that including the charge distribution of the
neutron had a small effect.)
In Figs.\ref{can58}$-$\ref{pbn58}, we present the point-neutron density
distributions, $\rho_n$, in calcium and lead.
For $^{40}$Ca, since the dependence of $\rho_n$ on $m_q$ and $R_B$ is again
fairly weak, only the result for $m_q$ = 5 MeV and $R_B$ = 0.8 fm is
shown, together with the empirical fit\cite{ncadata} to proton
scattering data.
%
\begin{figure}[htb]
\centering{\
%\epsfile{file=can58.ps,height=7cm}
\epsfig{file=can58.ps,height=7cm}
\caption{Point-neutron density distribution in $^{40}$Ca
(for $m_q$ = 5 MeV and
$R_B$ = 0.8 fm) compared with that of QHD and the empirical fit. }
\label{can58}}
\end{figure}
%
\begin{figure}[htb]
\centering{\
%\epsfile{file=nd58.ps,height=7cm}
\epsfig{file=nd58.ps,height=7cm}
\caption{Difference between $\rho_{n}$($^{48}$Ca) and $\rho_{n}$($^{40}$Ca)
compared with that of QHD and the empirical fit
(for $m_q$ = 5 and $R_B$ = 0.8 fm). }
\label{nd58}}
\end{figure}
%
\begin{figure}[htb]
\centering{\
%\epsfile{file=pbn58.ps,height=7cm}
\epsfig{file=pbn58.ps,height=7cm}
\caption{Point-neutron density distribution in $^{208}$Pb
(for $m_q$ = 5 MeV and $R_B$ = 0.8 fm)
compared with that of QHD and the empirical fit. }
\label{pbn58}}
\end{figure}
%
We again find reasonable agreement with the data.
For the isotope shift of $\rho_n$($^{48}$Ca)$-\rho_n$($^{40}$Ca),
the calculated difference is closer to those of non-relativistic results
than to those of QHD.
The neutron density distribution in lead is shown in Fig.\ref{pbn58}.
Its behavior is again similar to that of QHD.
Table~\ref{sum} gives a summary of the calculated binding energy per nucleon
($E/A$), rms charge radii and the difference between nuclear rms radii for
neutrons and protons ($r_n - r_p$), for several closed-shell nuclei.
%
\begin{table}[htbp]
\begin{center}
\caption{Binding energy per nucleon $E/A$ (in MeV), rms charge radius
$r_{ch}$ (in fm) and difference between $r_n$ and $r_p$ (in fm).
$m_q$ = 5 MeV and $R_B$ = 0.8 fm. ($^*$ fit) }
\label{sum}
\begin{tabular}[t]{c|ccc|ccc|ccc}
\hline
& & $-E/A$ & & & $r_{ch}$ & & & $r_n-r_p$ & \\
\hline
Model & QMC & QHD & Exp. & QMC & QHD & Exp. & QMC & QHD & Exp. \\
\hline
$^{16}$O &5.84&4.89&7.98&2.79&2.75&2.73&$-0.03$&$-0.03$&0.0 \\
$^{40}$Ca&7.36&6.31&8.45&3.48$^*$&3.48$^*$&3.48&$-0.05$&$-0.06$&0.05$\pm$0.05\\
$^{48}$Ca&7.26&6.72&8.57&3.52&3.47&3.47&0.23&0.21&0.2$\pm$0.05 \\
$^{90}$Zr&7.79&7.02&8.66&4.27&4.26&4.27&0.11&0.10&0.05$\pm$0.1 \\
$^{208}$Pb&7.25&6.57&7.86&5.49&5.46&5.50&0.26&0.27&0.16$\pm$0.05 \\
\hline
\end{tabular}
\end{center}
\end{table}
%
Since the calculated properties do not depend strongly on $m_q$ and $R_B$,
we only list the values for the QMC model with $m_q$ = 5 MeV
and $R_B$ = 0.8 fm. References for the experimental values can be found in
Ref.\cite{horowitz}. While there are still some discrepancies between
the results and data, the present model provides quite reasonable results.
In particular, it gives much larger binding energies per nucleon
than those of
QHD while still reproducing the rms charge radii for medium
and heavy nuclei quite well.
In Figs.\ref{casp2} and \ref{pbsp2}, the calculated spectra of
$^{40}$Ca and $^{208}$Pb are presented.
%
\begin{figure}[htb]
\centering{\
%\epsfile{file=casp2.ps,height=13cm}
\epsfig{file=casp2.ps,height=13cm}
\caption{Model predictions for the energy spectrum for $^{40}$Ca.
QMC(H) denotes the case for $m_q$ = 5 (300) MeV and $R_B$ = 0.8 fm. }
\label{casp2}}
\end{figure}
%
\begin{figure}[htb]
\centering{\
%\epsfile{file=pbsp2.ps,height=13cm}
\epsfig{file=pbsp2.ps,height=13cm}
\caption{Same as Fig.18 (for $^{208}$Pb). }
\label{pbsp2}}
\end{figure}
%
Because of the relatively smaller scalar and vector fields in the present
model than in QHD, the spin-orbit splittings are smaller.
The improvement over QHD in the binding energy per nucleon comes
at the expense of a reduction in the spin-orbit force.
We should note that there is a strong correlation between the effective
nucleon mass and the spin-orbit force. This problem is also discussed
in Refs.\cite{jin,blun}.
As a test of the sensitivity of the spin-orbit splitting to features of
the model,
we consider the case of a larger quark mass (noting that the Born-Oppenheimer
approximation requires that the nucleon motion is
relatively slow and the quarks are highly relativistic). For example,
we have calculated the case $m_q$ = 300 MeV (and $R_B$ = 0.8 fm) which
is a typical constituent quark mass.
The calculated spectra for $m_q$ = 300 MeV are also illustrated
in Figs.\ref{casp2} and \ref{pbsp2} (QMCH).
In this case, the various parameters were
$g_\sigma^2/4\pi$ = 5.58, $g_\omega^2/4\pi$ = 8.51 (to
satisfy the saturation condition), $g_\rho^2/4\pi$ = 6.45,
$m_\sigma$ = 497 MeV (to fit the rms charge radius of $^{40}$Ca),
$K$ = 334 MeV and $M_N^{\star}$ = 674 MeV at saturation density.
We should record that the bag radius in this case
increases by 7\% at saturation density (which is uncomfortably large).
The slope parameter in Eq.(\ref{paramC}) is $a=3.9 \times 10^{-4}$.
One can expect that a heavy quark mass gives a spectrum closer to
those of QHD\cite{st1,finite}. We can see from the figures
that the calculated spectra are somewhat closer to the experimental data.
We note that the charge density distributions for $^{40}$Ca and $^{208}$Pb
are also reproduced well in this case.
\clearpage
%
%
\section{Summary, discussion and further applications}
\label{discuss}
%
Starting with the quark-meson coupling model,
in which quarks
confined in nucleon bags interact through the exchange of scalar and
vector mesons, we have presented a generalisation of QHD with a $\sigma$
field-dependent scalar coupling to describe finite nuclei quantitatively.
The physical origin of this field-dependence,
which provides a new saturation mechanism for nuclear matter,
is the relatively rapid increase of the lower Dirac component of the
wavefunction of the confined, light quark. We have then derived a set of
coupled non-linear differential equations which must be solved
self-consistently but which are not much more difficult to solve than the
relativistic Hartree equations of QHD. Our calculations for static,
closed-shell nuclei from $^{16}$O to $^{208}$Pb reproduce fairly well the
observed charge density distributions, neutron density distributions etc.
It will be very interesting to explore the connection between the $\sigma$
field-dependence of the variation of the effective $\sigma$-N
coupling constant, which arises so naturally here, and the variation
found empirically in earlier work. We note, in particular, that while
our numerical results depend on the particular model chosen here (namely,
the MIT bag model), the qualitative features which we find (such as the
$\sigma$ field-dependent decrease of the scalar coupling etc.) will apply in
{\em any model} in which the nucleon contains light quarks and the
attractive N-N
force is a Lorentz scalar. Of course, it will be important to
investigate the degree of variation in the numerical results for other
models of nucleon structure\cite{blun}.
In the present model there are, however, still some discrepancies in energy
spectra of nuclei, in particular, the spin-orbit splittings.
In the previous section we have briefly discussed one possibility to
partly overcome this defect in the present model.
As an alternative approach,
Jin and Jennings\cite{jin} and Blunden and Miller\cite{blun} have
considered variations of the bag constant $B$ and $z$ parameter in
nuclear matter, which have been suggested by the fact that quarks are
presumably deconfined at high enough densities.
This was taken to suggest that
$B$ might decrease with increasing density. Adopting this idea,
Blunden and Miller\cite{blun} have studied properties of both nuclear
matter and finite nuclei. In their approach, $B$ in matter, $B^{\star}$,
is given by
%
\bge
B^{\star} = B \left[ 1 - \alpha_B \frac{U_s({\vec r})}
{M_N} \right] , \label{B*}
\ene
%
where $U_s$ is an average scalar potential and $\alpha_B$ is an arbitrary
parameter. For finite nuclei the results move towards those of QHD, and
there is an improvement in the spin-orbit splittings. However, the
nuclear incompressibility, $K$, is again larger than
the experimental data, and the bag radius at saturation density increases
by 5\% $\sim$ 13{\%}\cite{blun} or more\cite{jin}, which seems uncomfortably
large\cite{emc}.
We probably ought to construct a formalism
beyond the Hartree approximation for this class of models in the future.
In this work we have not considered any effect of a decrease of the meson
masses in nuclear matter. (In Ref.\cite{hadrons} these
effects were investigated
in infinite nuclear matter, using an earlier version of the
model.) The $\sigma$-meson mass which we found it necessary to use in
this paper ranges from 400 $\sim$ 450 MeV. It seems rather small
compared with
the empirical $\sigma$-mass in the OBEP analyses of the free NN force
\cite{formf,mach,obep}.
On the other hand, noting its origin in two-pion-exchange with N-$\Delta$
intermediate states, we observe that the N-$\Delta$ mass difference
would decrease in medium by a similar percentage.
It will be interesting, in future work, to incorporate a self-consistent
reduction of the vector-meson mass \cite{hadrons} (as well
as the scalar-meson mass\cite{sgmms})
into the calculation of the properties of finite nuclei.
The successful generalisation of the QMC model to
finite nuclei opens a tremendous number of opportunities for further
work. For example, earlier results for the Okamoto-Nolen-Schiffer
anomaly\cite{st2}, the nuclear EMC effect\cite{st4},
the charge-symmetry violating correction to
super-allowed Fermi beta-decay\cite{st3}
and so on, can now be treated in a truly quantitative way.
Since the present model can provide self-consistent quark wave functions in
a deeply bound nucleon, e.g., $1s_{1/2}$ in $^{208}$Pb, it would also be very
interesting to investigate variations in magnetic moments, $g_A$ and so on,
for such nucleons\cite{yamazaki}.
For hypernuclei the
natural extension (see Ref.\cite{hadrons}) is to assume that the
$\sigma$ and $\omega$ mesons couple only to the non-strange
constituents. From our discussion of the spin-orbit force in
Ref.\cite{finite} and
the fact that the spin of the $\Lambda$ is carried entirely by the
strange quark, one can easily see that the $\Lambda$ spin-orbit force
will arise entirely from the Thomas precession term.
This means that the $\Lambda$ spin-orbit force is very naturally suppressed
in this model -- as observed experimentally. It will be important to
follow this observation with quantitative results.
Finally, we list a number of important ways in which this model could be
extended: replacing the
MIT bag by a model respecting PCAC (e.g., the
cloudy bag model\cite{cbm}); replacing $\sigma$-exchange by
two-pion exchange; replacing $\omega$ exchange by nucleon
overlap at short distance, and so on.
In terms of further theoretical development it will also be interesting to
compare the present model with more phenomenological, non-linear
extensions of QHD -- as reviewed recently in Ref.\cite{delf}.
On the practical
side, we stress that the present model can be applied to all the
problems for which QHD has proven so attractive, with very little extra
effort.
\vspace{1cm}
\flushleft{Acknowledgements} \\
\noindent
The authors are pleased to thank Pierre A.M. Guichon for valuable
discussions and comments.
This work was supported by the Australian Research Council.
%
\newpage
\begin{thebibliography}{99}
%
\bibitem{emc} D.F. Geesaman, K. Saito and A.W. Thomas, Annu. Rev. Nucl.
Part. Sci. {\bf 45} (1995) 337; \\
M. Arneodo, Phys. Rep. {\bf 240} (1994) 301.
%
\bibitem{formf} K. Holinde and A.W. Thomas, Phys. Rev. {\bf C42} (1990)
R1195; \\
S. Deister et al., Few Body Systems {\bf 10} (1991) 1; \\
J. Haidenbauer, K. Holinde and A.W. Thomas, Phys. Rev. {\bf C49} (1994)
2331.
%
\bibitem{nonrel} J.W. Negele, Phys. Rev. {\bf C1} (1970) 1260; \\
X. Campi and D.W.L. Sprung, Nucl. Phys. {\bf A194} (1972) 401; \\
J.L. Friar and J.W. Negele, Advan. Nucl. Phys. {\bf 8} (1975) 219.
%
\bibitem{nsa} K. Okamoto, Phys. Lett. {\bf 11} (1964) 150; \\
J.A. Nolen, Jr. and J.P. Schiffer, Annu. Rev. Nucl. Sci. {\bf 19} (1969) 471.
%
\bibitem{nsa2} S. Shlomo, Rep. Prog. Phys. {\bf 41} (1978) 957; \\
G.A. Miller, B.M.K. Nefkens and I. $\check{\mbox{S}}$laus, Phys. Rep.
{\bf 194} (1990) 1.
%
\bibitem{hs} H. Sato, Nucl. Phys. {\bf A296} (1976) 378.
%
\bibitem{ght} T. Goldman, J.A. Henderson and A.W. Thomas, Few Body Systems
{\bf 12} (1992) 123; \\
G. Krein, A.W. Thomas and A.W. Williams, Phys. Lett. {\bf B317} (1993) 293; \\
J. Piekarewicz and A.W. Williams, Phys. Rev. {\bf C47} (1993) R2461; \\
T. Hatsuda, E.M. Henley, Th. Meissner and G. Krein, Phys. Rev. {\bf C49}
(1994) 452.
%
\bibitem{weak} D.H. Wilkinson, Nucl. Phys. {\bf A571} (1994) 87; Z.
Phys. {\bf A348} (1994) 129; \\
I.S. Towner and J.C. Hardy, in {\it The Nucleus as a Laboratory for Studying
Symmetries and Fundamental Interaction}, ed. by E.M. Henley and W.C. Haxton,
(nucl-th/9504015) (World Scientific, 1996); \\
G. Savard {\it et al.}, Phys. Rev. Lett. {\bf 74} (1995) 1521.
%
\bibitem{qhd} J.D. Walecka, Ann. Phys. (N.Y.) {\bf 83} (1974) 491; \\
B.D. Serot and J.D. Walecka, Advan. Nucl. Phys. {\bf 16} (1986) 1.
%
\bibitem{guichon} P.A.M. Guichon, Phys. Lett. {\bf B200} (1988) 235.
%
\bibitem{st1} K. Saito and A.W. Thomas, Phys. Lett. {\bf B327}
(1994) 9; \\
Phys. Rev. {\bf C52} (1995) 2789.
%
\bibitem{yazaki} S. Fleck et al., Nucl. Phys. {\bf A510} (1990) 731; \\
V.K. Mishra et al., Phys. Rev. {\bf C46} (1992) 1143; \\
M.K. Banerjee, Phys. Rev. {\bf C45} (1992) 1359; \\
E. Naar and M.C. Birse, J. Phys. {\bf G19} (1993) 555.
%
\bibitem{st2} K. Saito and A.W. Thomas, Phys. Lett. {\bf B335} (1994) 17.
%
\bibitem{juelich} A.W. Thomas and K. Saito, ``The Role of Nucleon
Structure in Nuclear Physics'', in Proc. Int. Conf. on Physics with GeV
Beams, eds. H. Machner and K. Sistemich, World Scientific, Singapore
(1995) p.175.
%
\bibitem{st3} K. Saito and A.W. Thomas, Phys. Lett. {\bf B363} (1995) 157.
%
\bibitem{osaka} A.W. Thomas and K. Saito, ``Charge Symmetry Violation in
Nuclear Physics", in Proc. of the Int. Conf. on Weak
and Electromagnetic Interactions in Nuclei (WEIN '95), eds. H. Ejiri,
T. Kishimoto and T. Sato, World Scientific, Singapore (1995) p.495.
%
\bibitem{brwn} G.Q. Li, C.M. Ko and G.E. Brown, Phys. Rev. Lett. {\bf 75}
(1995) 4007.
%
\bibitem{expts} G. Agakichiev {\it et al.} (CERES Collaboration),
Phys. Rev. Lett. {\bf 75} (1995) 1272; \\
M. Masera {\it et al.} (HELIOS-3 Collaboration), Nucl. Phys. {\bf A590}
(1995) 93c.
%
\bibitem{hadrons} K. Saito and A.W. Thomas, Phys. Rev. {\bf C51} (1995)
2757.
%
\bibitem{st4} K. Saito, A. Michels and A.W. Thomas, Phys. Rev. {\bf C46}
(1992) R2149; \\
A.W. Thomas, K. Saito and A. Michels, Aust. J. Phys. {\bf 46} (1993) 3; \\
K. Saito and A.W. Thomas, Nucl. Phys. {\bf A574} (1994) 659.
%
\bibitem{finite} P.A.M. Guichon, K. Saito, E. Rodionov and A.W. Thomas,
nucl-th/9509034, to appear in Nucl. Phys. {\bf A} (1996); \\
P.A.M. Guichon, K. Saito and A.W. Thomas, nucl-th/9602022, to appear in
Australian Journal of Physics (1996).
%
\bibitem{jin} Xuemin Jin and B.K. Jennings, Phys. Lett. {\bf B374} (1996)
13; \\
TRIUMF preprint, TRI-PP-96-8 (nucl-th/9604018).
%
\bibitem{blun} P.G. Blunden and G.A. Miller, nucl-th/9602031 (1996), to
appear in Phys. Rev. {\bf C} (1996).
%
\bibitem{fuchs} C. Fuchs, H. Lenske and H.H. Wolter, Phys. Rev. {\bf C52}
(1995) 3043; \\
H. Lenske and C. Fuchs, Phys. Lett. {\bf B345} (1995) 355.
%
\bibitem{horowitz} C.J. Horowitz and B.D. Serot, Nucl. Phys. {\bf A368}
(1981) 503.
%
\bibitem{hor2} C.J. Horowitz, D.P. Murdoch and B.D. Serot, Computational
Nuclear Physics 1, eds. K. Langanke, J.A. Maruhn and S.E. Koonin
(Springer-Verlag, Berlin, 1991), p.129.
%
\bibitem{sinha} B.B.P. Sinha {\it et al.}, Phys. Lett. {\bf B53} (1971) 217.
%
\bibitem{mach} R. Machleidt, Advan. in Nucl. Phys. {\bf 19} (1989) 189.
%
\bibitem{cadata} I. Sick {\it et al.}, Phys. Lett. {\bf B88} (1979) 245.
%
\bibitem{frois} B. Frois {\it et al.}, Phys. Rev. Lett. {\bf 38} (1977) 152.
%
\bibitem{odata} I. Sick and J.S. McCarthy, Nucl. Phys. {\bf A150} (1970) 631.
%
\bibitem{zdata} L.A. Fajardo {\it et al.}, Phys. Lett. {\bf B37} (1971) 363.
%
\bibitem{ch48data} B. Frois, Lecture notes in Physics {\bf 108} (1979) p.52.
%
\bibitem{ncadata} L. Ray, Phys. Rev. {\bf C19} (1979) 1855; \\
L. Ray and P.E. Hodgson, Phys. Rev. {\bf C20} (1979) 2403.
%
\bibitem{obep} K. Erkelenz, Phys. Rep. {\bf C13} (1974) 194.
%
\bibitem{sgmms} K. Saito, T. Maruyama and K. Soutome, Phys. Rev. {\bf C40}
(1989) 407; \\
T. Hatsuda and T. Kunihiro, Phys. Rep. {\bf 247} (1994) 221.
%
%\bibitem{smas} N.A. T{\" o}rnqvist and M. Roos, Phys. Rev. Lett.
%{\bf 76} (1996) 1575.
%
\bibitem{yamazaki} T. Yamazaki, Mesons in Nuclei, vol.II, eds. M. Rho and
D.H. Wilkinson (North Holland, Amsterdam, 1979), p.651.
%
\bibitem{cbm} S.\ Th\'eberge et al., Phys.\ Rev.\ {\bf D22} (1980) 2838;
{\bf D23} (1981) 2106(E); \\
A.\ W.\ Thomas, Adv.\ Nucl.\ Phys.\ {\bf 13} (1984) 1; \\
G.\ A.\ Miller, in Int.\ Rev.\ Nucl.\ Phys.\ {\bf 2} (1984).
%
\bibitem{delf} A. Delfino et al., nucl-th/9602004 (1996).
%
\end{thebibliography}
%
\end{document}