Phi Meson Production in Heavy-Ion Collisions at SIS Energies
Extracted source text ยท arXiv nucl-th/9704002
\documentstyle[preprint,aps,epsf]{revtex}
\newcommand{\del}{\mbox{$\Delta$}}
\newcommand{\arr}{\mbox{$\longrightarrow$}}
\newcommand{\pinphin}{\mbox{$\pi N \arr \phi N$}}
\newcommand{\pidelnphi}{\mbox{$\pi \del \arr \phi N$}}
\newcommand{\nnnnphi}{\mbox{$N N \arr N N \phi$}}
\newcommand{\ndelnnphi}{\mbox{$N \del \arr N N \phi$}}
\newcommand{\deldelnnphi}{\mbox{$\del \del \arr N N \phi$}}
\newcommand{\kkphi}{\mbox{$K \bar K \arr \phi$}}
\newcommand{\pirhophi}{\mbox{$\pi \rho \phi$}}
\newcommand{\phinpin}{\mbox{$\phi N \arr \pi N$}}
\newcommand{\phinpiDel}{\mbox{$\phi N \arr \pi \Delta$}}
\newcommand{\phinnrho}{\mbox{$\phi N \arr \rho N$}}
\newcommand{\rhonnphi}{\mbox{$\rho N \arr \phi N$}}
\newcommand{\phinlambdak}{\mbox{$\phi N \arr K\Lambda $}}
\newcommand{\lambdakphin}{\mbox{$K \Lambda \arr \phi N$}}
\newcommand{\ppppphi}{\mbox{$p p \arr p p \phi$}}
\begin{document}
\title{Phi Meson Production in Heavy-Ion Collisions at SIS Energies}
\author{W. S. Chung$^{1}$, G. Q. Li$^{1,2}$, and C. M. Ko$^{1}$}
\address{$^{1}$Department of Physics and Cyclotron Institute,
Texas A\&M University, \\
College Station, Texas 77843, U.S.A.\\
$^{2}$Department of Physics, State University of New York at Stony Brook,\\
Stony Brook, New York 11794, U.S.A.}
\maketitle
\begin{abstract}
Phi meson production in heavy-ion collisions at SIS/GSI energies
($\sim 2$ GeV/nucleon) is studied in the relativistic transport model.
We include contributions from baryon-baryon, pion-baryon, and
kaon-antikaon collisions. The cross sections for the first two processes
are obtained in an one-boson-exchange model, while that for the last process
is taken to be of Breit-Wigner form through the phi meson resonance.
The dominant contribution to phi meson production in heavy ion collisions
at these energies is found to come from secondary pion-nucleon collisions.
Effects due to medium modifications of kaon masses are also studied and
are found to reduce the phi meson yield by about a factor of two, mainly
because of increased phi decay width as a result of dropping kaon-antikaon
masses. In this case, the $\phi/K^-$ ratio is about 4\%, which is a
factor of 2-3 below preliminary experimental data from the FOPI
collaboration at GSI. Including also the reduction of phi meson mass
in medium increases this ratio to about 8\%, which is then in reasonable
agreement with the data.
\end{abstract}
\newpage
\section{Introduction}
\label{intro}
There are increasing interests in phi meson production from heavy ion
collisions since the suggestion by Shor \cite{shor} that an enhanced
production may signal the formation of a quark-gluon plasma
in these collisions as its production in hadronic matter is usually
suppressed by the OZI rule \cite{ozi}. However, recent
theoretical studies have suggested that because of chiral symmetry
restoration hadron properties are expected to change in hot dense hadronic
matter \cite{brown,hat,asa} (For recent reviews see
Refs. \cite{brownrho,mfrev,kkl}). For phi meson in dense matter, QCD
sum rule calculations have shown that its mass decreases by about 2-3\%
at normal nuclear matter density \cite{hat}. Reduced in-medium phi meson
mass has also been predicted in vector dominance model with dropping kaon
masses \cite{ko92}, in hadronic models that include the particle-antiparticle
vacuum polarization \cite{hat95}, and in hidden gauge theory
\cite{song}. A reduced phi meson mass in medium is expected to enhance
its yield from heavy ion collisions even in the absence of the
quark-gluon plasma \cite{kosa}. In this case, one can also study the
phi meson in-medium properties via the dilepton spectrum from these
collisions as pointed out in Refs. \cite{hat,karsch,kkonly}.
At high temperature, the reduction of phi meson mass may
become more appreciable because of the decrease of
strangeness condensate $\langle{\bar s}s\rangle$ as a result of the
presence of strange hadrons \cite{asa}. It has thus been suggested that
a low mass phi peak besides the normal one may appear in the dilepton
spectrum from ultrarelativistic heavy ion collisions if a first-order
phase transition or a slow cross-over between the quark-gluon plasma
and the hadronic matter occurs in the collisions \cite{ak94}.
Experimentally, phi meson production has been
studied at SPS/CERN by the NA38 collaboration \cite{na38} and the HELIOS-3
collaboration \cite{cern} using a proton or sulfur projectile at
200 AGeV and a tungsten or uranium target. A factor of 2 to 3 enhancement
in the double ratio $( \phi / (\omega+\rho^{0} ))_{SU(W)} /( \phi /
(\omega+\rho^{0}))_{pW}$ has been observed. Various theoretical attempts have
been made to understand this enhancement \cite{kosa,greiner,Koch}. In
particular, an enhancement of phi meson yield may be a signature
of the formation of a quark gluon plasma in the collisions \cite{shor}.
However, the enhancement can also be explained
in hadronic models if one takes into account the reduced phi meson mass
in medium \cite{kosa}. In Ref. \cite{greiner}, the color rope fragmentation
and hadronic rescattering have been introduced to understand
the observed phi meson enhancement.
Phi mesons have also been measured at AGS/BNL by the E-802 collaboration
in central collisions between a 14.6 AGeV/c Si beam and a Au target
\cite{AGS}. They are identified from the invariant mass spectrum
of $K^{+} K^{-}$ pairs, and the measured phi meson mass and width
are consistent with that in free space. The ratio of the
phi meson yield to the $K^{-}$ yield has been found to be about 10\%, which
can be understood if thermal and chemical equilibrium are assumed at
freeze out with a temperature of about 110 MeV \cite{bs95,fireball}.
On the other hand, calculations by Baltz and Dover \cite{Dover}
based on the coalescence model, in which the phi meson is formed from
the kaon and antikaon at freeze out, underestimate the data by a large factor.
Phi mesons from heavy-ion collisions at SIS/GSI energies are being
studied by the FOPI collaboration \cite{FOPI} also through
the $K^{+} K^{-}$ invariant mass distribution. About 30 phi
mesons have been identified in central collisions of Ni+Ni at
1.93 AGeV. Based on these preliminary results it has been concluded
that the phi meson yield is about 10\% of the $K^{-}$ yield,
which is very similar to that observed at the AGS energies.
This is somewhat surprising since the SIS energies are below the phi meson
production threshold in the nucleon-nucleon collision in free space, while
the AGS energies are well above the threshold.
In this paper, we shall carry out a detailed transport model study on
phi meson production in heavy-ion collisions at SIS energies. There are
mainly two reasons to work at these energies. First,
particle production at subthreshold energies is sensitive to
medium modifications of hadron properties \cite{mfrev,kkl,cass90,mos91}.
Secondly, the reaction dynamics of heavy-ion collisions at SIS energies
is relatively simple compared to that at higher energies as it involves
mainly the nucleon, delta, and pion. In a previous study
of phi meson production at SIS energies \cite{kkonly}, only the contribution
from the reaction $K \bar K \to\phi$ has been included, and the
$\phi / K^{-}$ ratio is predicted to be less than 1\%, which is an order
of magnitude smaller than the experimental result from the FOPI
collaboration. To explain this large discrepancy we include here
phi meson production also from baryon-baryon and pion-baryon collisions. The
cross sections for these processes are calculated using an
one-boson-exchange model with parameters fitted to available experimental
data. Some results from this study have been reported
earlier \cite{phibrief}. We find that in heavy ion collisions the dominant
contribution to phi meson production is from the $\pi N$ collision,
and its inclusion in the transport mode brings the predicted $\phi/K^{-}$
ratio closer to the measured one by the FOPI collaboration. Furthermore,
we study the effect of medium modifications of kaon properties on phi meson
production. Kaplan and Nelson have pointed out
that the kaon mass changes in nuclear matter \cite{N&K}. In chiral
perturbation theory, the kaon mass is predicted to increase slightly while
the antikaon mass decreases significantly as the density increases \cite{lee}.
Overall, the total mass of a kaon-antikaon pair is lowered in dense matter.
This affects appreciably the phi decay width $\Gamma_{\phi\rightarrow
K{\bar K}}$, which is only about 3.7 MeV in free space,
since the phi meson mass is very close to twice the
kaon mass. A reduced kaon in-medium mass thus increases
$\Gamma_{\phi\rightarrow K{\bar K}}$, so phi mesons are more likely to
decay before freeze out, leading to a lower final phi meson yield.
This reduction is, however, compensated by effects from
the dropping phi meson in-medium mass.
The paper is organized as follows: In section \ref{xsect} we calculate the
elementary cross sections for phi meson production and absorption using
an one-boson-exchange model. Section \ref{npik} describes the nucleon,
pion and kaon dynamics in heavy ion collisions using the relativistic
transport model. Results on the phi meson yield and distributions
from these collisions are given in section \ref{phiresult}. Section
\ref{summary} is the summary and outlook.
\section{Elementary Cross Sections}
\label{xsect}
For heavy ion collisions at SIS energies, the most abundant particles
are the nucleon, $\Delta$ resonance, and pion.
Phi meson can thus be produced from
reactions such as \nnnnphi, $\ndelnnphi$, $\deldelnnphi$, $\pinphin$,
and $\pidelnphi$. The production of phi meson from $\pi \pi$ scattering
is expected to be unimportant due to its higher production threshold
than that for the other reactions,
and is thus ignored. Although both kaon and antikaon are only scarcely
produced, phi meson production from the reaction $\kkphi$ turns out
nonnegligible in the case of using in-medium kaon
masses and will be included.
Unfortunately, the cross sections for these reactions are not
well-known empirically. There exist some data for the $\pi N$ channel near
threshold \cite{pindata}, but the $N N$ channel has been studied
only at high energies \cite{nndata}. Because of the scarcity of experimental
data, an one-boson exchange model has been introduced in Ref. \cite{phibrief}
to extrapolate from the data at high energies to make predictions at
energies near the production threshold, and also to calculate
the cross sections for reactions involving the delta
resonance as they cannot be measured experimentally. In that study,
only the lowest order Born amplitude for phi meson production are
considered, and this should be
a reasonable approximation to the full amplitude because phi meson production
has a very small probability in these reactions.
In the following, we discuss this model and its predictions
for both phi production and absorption cross sections.
The interaction Lagrangians needed in the one-boson exchange model
are well-known, and they are given by
\begin{eqnarray}
{\cal L} _{\pi NN}& =& - {f_{\pi NN}\over m_\pi} {\bar \psi} \gamma ^5 \gamma
^\mu\vec\tau \psi\cdot \partial _\mu \vec\pi,\\
{\cal L} _{\rho NN}& =& -g_{\rho NN} {\bar \psi} \gamma ^\mu\vec\tau\psi
\cdot\vec\rho_\mu - {f_{\rho NN}\over 4m_N} {\bar \psi} \sigma ^{\mu\nu}
\vec\tau\psi\cdot (\partial _\mu\vec\rho_\nu -\partial _\nu\vec\rho_\mu),\\
{\cal L} _{\rho N \Delta}& =& i{f_{\rho N\Delta}\over m_\rho} {\bar \psi}
\gamma ^5 \gamma ^\mu\vec T \psi ^\nu\cdot(\partial_\mu\vec\rho_\nu
-\partial_\nu\vec\rho_\mu) + {\rm h.c.},\\
{\cal L_{\pirhophi}}& =& {f_{\pi \rho \phi} \over m_{\phi} }
\epsilon_{\mu \nu \alpha \beta}
\partial^{\mu} \vec\phi^{\nu} \partial^{\alpha} \rho^{\beta}\cdot\vec\pi.
\end{eqnarray}
In the above, $\psi$ is the nucleon field with mass $m_N$; $\psi _\mu$ is
the Rarita-Schwinger field for the spin-3/2 delta resonance with mass
$m_\Delta$; and $\pi$, $\rho_\mu$, $\phi_\mu$ are meson fields with masses
$m_\pi$, $m_\rho$, $m_\phi$, respectively.
The isospin matrices, Dirac $\gamma$-matrices, and Levi-Civita tensor
are denoted by $\vec\tau$, $\gamma^\mu$, and $\epsilon^{\mu\nu\alpha\beta}$,
respectively.
The coupling constants and cutoff parameters in ${\cal L}_{\pi NN}$,
${\cal L}_{\rho NN}$, and ${\cal L}_{\rho N\Delta}$
are taken from the Model II of Table B.1 in Ref. \cite{Mach}.
From the measured width $\Gamma_{\phi\to\pi\rho}\approx 0.6$ MeV,
the coupling constant $f_{\pirhophi}\approx 1.04$ is determined.
\subsection{Phi meson production from pion-baryon collisions}
The Feynman diagrams for the reactions $\pinphin \/$ and $\pidelnphi$ are
shown in Fig. \ref{FeynmanpiB}. A monopole
form factor is introduced at the $\pi\rho\phi$ vertex, i.e.,
\begin{equation}
F(q_{\pi},q_{\rho}) = { {(\Lambda^\pi _{\pirhophi})^2 - m_{\pi}^{2}} \over
{(\Lambda^\pi _{\pirhophi})^2 + q_{\pi}^{2}} }.
\end{equation}
The cut-off parameter $\Lambda_{\pi\rho\phi}^\rho$ is adjusted to give a
reasonable fit to the experimental data from the reaction
$\pi^{-}p \to n \phi$ \cite{pindata}, and a value of 1.2 GeV is obtained.
The comparison between calculated results and the experimental data
as well as the parameterization by Sibirtsev \cite{sib}
can be found in Ref. \cite{phibrief}.
Since the isospin degree of freedom is usually not explicitly treated
in transport model, we have thus calculated the isospin-averaged cross
sections for $\pinphin \/$ and $\pidelnphi$.
The results are shown in Fig. \ref{isopb}.
Near threshold the cross section for $\pinphin$ is seen to be much larger
than the cross section for $\pidelnphi$ as a result of the strong tensor
coupling at the $\pi NN$ vertex.
To determine the momentum of produced phi meson in the
transport model, we also need the differential cross sections
for these reactions. Since only two particles are in the final state,
the magnitude of the
phi momentum is fixed by kinematics. The differential cross
section $d\sigma /dcos\theta$ for these reactions at an
energy of 0.2 GeV above the threshold is shown in Fig. \ref{diffXpiB}.
\subsection{phi meson production from baryon-baryon collisions}
\label{pbb}
The Feynman diagrams for the reactions
$\nnnnphi$, $\ndelnnphi$, and $\deldelnnphi$ are shown in Fig. \ref{bb}.
Exchange diagrams are also included as the
two final nucleons are identical.
There is some ambiguity in treating the $\pirhophi$ vertex
in Fig. \ref{bb}, which involves two
virtual particles. Such a vertex does not exist in the Bonn potential
model, and there is also no theoretically well-defined criteria for
prescribing its form. The following form has been used in Ref. \cite{phibrief},
\begin{equation}
F(q_{\pi},q_{\rho}) =
\left( { {(\Lambda^\pi _{\pirhophi})^2 - m_{\pi}^{2}} \over
{(\Lambda^\pi _{\pirhophi})^2 + q_{\pi}^{2}} } \right)
\left( {{(\Lambda^\rho _{\pirhophi})^2 - m_{\rho}^{2}} \over
{(\Lambda^\rho _{\pirhophi})^2 + q_{\rho}^{2}}} \right),
\end{equation}
where $q_{\rho}$ and $q_{\pi}$ are four momenta of the virtual
$\rho$ and $\pi$ meson, respectively. The cut-off parameter
$\Lambda^{\rho}_{\pirhophi}$ is the same one introduced earlier in
calculating the cross section for $\pinphin$.
The other cut-off parameter $\Lambda^{\pi}_{\pirhophi}$
is introduced to take into account the virtuality of the pion.
The cross section for $pp \arr p p \phi \/$ has been measured at several
energies \cite{nndata}, and the one at the lowest energy has been used
in Ref. \cite{phibrief}
to fix $\Lambda^{\pi}_{\pirhophi}$, and a value of 0.95 GeV has been obtained.
Comparisons of calculated results with both experimental data
and the results by Sibirtsev based on an one-pion exchange model \cite{sib}
have been given in
Ref. \cite{phibrief}. We note that both $\Lambda^\rho_{\pi\rho\phi}$
and $\Lambda^{\pi}_{\pirhophi}$ are in the order of 1 GeV,
similar to typical values for cut-off parameters in both the Bonn potential
model and other hadronic models.
For the reactions $\ndelnnphi$ and $\deldelnnphi$, the exchanged pion
can be on shell, so the cross sections are singular in certain kinematical
region. This singularity can be regulated by the Peierls method
\cite{peierls}, in which the energy of the delta resonance is taken
to be complex to account for its finite lifetime. As a result,
the four-momentum of the exchanged pion acquires an imaginary part,
which thus renders the pion propagator finite at the pion pole.
The isospin averaged cross sections for phi meson production from
baryon-baryon collisions are shown in Fig. \ref{isobb}.
There are significant differences between the cross sections for
the reactions $\nnnnphi$, $\ndelnnphi$, and $\deldelnnphi$.
Near threshold,
all are much smaller than the cross section for phi meson production from
pion-nucleon collisions.
Since there are three particles in the final state in baryon-baryon
collisions, the momentum of produced phi meson is given by
the triple differential cross sections
$d^3\sigma /dp_xdp_ydp_z$ or double differential cross
section $d^2\sigma /p^2dpd\Omega$ if there is an azimuthal symmetry.
To simplify the numerical
procedure, we introduce the assumption that the phi meson momentum is
isotropic in baryon-baryon center-of-mass frame. We then
need only the momentum spectrum $d\sigma /dp$, which is shown
in Fig. \ref{pspectrum} for these reactions again at an energy of 0.2 GeV
above the threshold.
\subsection{Phi meson production from kaon-antikaon annihilation}
The total cross section for $\kkphi$ is assumed to have
a Breit-Wigner form as in Ref. \cite{kkonly}, i.e.,
\begin{equation}\label{kkp}
\sigma ( K \bar K \arr \phi) = {3 \pi \over k^2}
{ ( m_{\phi} \Gamma_{\phi} )^2 \over
(M^2-m_{\phi}^2)^2+(m_{\phi}\Gamma_{\phi})^2 },
\end{equation}
\noindent where $\Gamma_{\phi}$ is the phi meson decay width to
$K\bar K$ and is given by
\begin{equation}
\Gamma_{\phi} = { g^{2}_{\phi K {\bar K}} \over 4 \pi }
{ (m_\phi^{2}-4m_{K}^{2})^{3/2} \over 6 m_\phi^2},
\label{phiwidth}
\end{equation}
with the coupling constant ${g^{2}_{\phi K {\bar K}}/ 4 \pi } \approx
1.69$ determined from the empirical width of 3.7 MeV. In using Eq.
(\ref{kkp}), we have thus neglected the small effect due to the decay of
phi meson into other channels.
\subsection{Phi meson scattering by nucleons}
Once a phi meson is produced, there are three kinds of reactions it can
take part in: decay, inelastic and elastic scattering by nucleons. The
dominant decay mode of phi meson is $\phi \to K {\bar K}$, with a decay width
given by Eq. (\ref{phiwidth}). Here we only consider phi meson absorption
due to $\phi N$ collisions as that due to $\phi\pi$ and $\phi \Delta$
scattering is expected to be insignificant because of the low
pion and delta densities compared with that of nucleons (see Fig.
\ref{cden}). We have included the reactions $\phinlambdak$, $\phinnrho$,
$\phinpin$, and $\phinpiDel$. The cross section for the first reaction is
related via detailed balance to that of $\Lambda K\to\phi N$, which has
already been calculated in Ref. \cite{kosa} based on a kaon exchange model
with parameters taken from Ref. \cite{holzen}. The reaction
$\phi N\rightarrow K\Sigma$ is neglected since the
$N\Sigma K$ coupling constant ($g_{N\Sigma K}^2/4\pi \sim 0.6$) is much
smaller than that of $N\Lambda K$ ($g_{N\Lambda K}^2/4\pi \sim 16$)
\cite{holzen}.
The cross section for the other three reactions can be obtained
from those for $\rho N\to\phi N$, $\pi N\to\phi N$, and $\pi\Delta\to\phi N$
using detailed balance relations. We have already determined in the
above the cross sections for phi meson production from pion-baryon collisions.
For the reaction $\rho N\to\phi N$, we simply interchange the pion and
rho meson in the first diagram of Fig. \ref{FeynmanpiB}.
The cross sections for these phi meson absorption processes are
summarized in Fig. \ref{phiabs}. We see that $\phinlambdak$
is the dominant one as already pointed out in Refs.
\cite{shor,kosa}, because this is the only one
among the four reactions that is not suppressed by the OZI rule .
The phi-nucleon elastic scattering is also included. Experimentally, the
cross section for this process can be extracted from data on phi meson
photoproduction using the vector meson dominance model \cite{joo}.
In the energy range $3.5$ GeV$<p_{\rm lab}<5.8$ GeV, a value of
\mbox{0.56 mb} has been obtained. Here we make the rough approximation
that the cross section is a constant of 0.56 mb in the energy range
considered. We note that a slightly larger value has been obtained in Ref.
\cite{barger}. The phi-nucleon scattering does not change the phi meson
yield, but affects its momentum distribution. Since this cross section
is rather small, the effect is not appreciable as shown below.
\subsection{Discussions}
So far we have calculated exclusive phi meson production cross sections
with no pions in the final state. In principle, phi meson can also be
produced in association with one or more pions, and one needs for
heavy-ion collisions inclusive cross sections such
as $NN\rightarrow \phi X$. Since the beam energies we consider
in this work are below the phi meson production threshold
in nucleon-nucleon collision, we expect that exclusive
reactions such as $NN\rightarrow NN\phi$, which have lower
thresholds, are more important than the ones with pions
in the final state, which require additional energies
and are thus suppressed.
We would like to point out that in spite of the success
of the one-boson-exchange model at low energies, this model has not been
carefully verified at energies above the phi meson production threshold.
It is often believed that the quark degree of freedom becomes relevant
at a kinetic energy of 1-2 GeV, so it is not known if our procedure of using
the data point at $p_{\rm lab}=10$ GeV to fix the cross section for
$\ppppphi$ near the threshold is reliable. Nevertheless, this model
has been used quite often in calculating the cross sections for
pion production \cite{pion},
eta production \cite{mosel}, kaon production \cite{laget,liko}, and
baryon resonance excitations \cite{huber} from nucleon-nucleon collisions
in this energy region. Measurements of $pp \to pp\phi$ at low
energies are being carried out at SATURNE \cite{saturne}, and these data
will be very useful in testing the validity of our model.
With this caveat in mind, we proceed to our
transport model study using the above elementary cross sections.
\section{The nucleon, pion and kaon dynamics}
\label{npik}
\subsection{The relativistic transport model}
For the description of heavy-ion collisions at SIS energies, we use the
relativistic transport model (RVUU) developed in Ref. \cite{RVUU}
from the non-linear $\sigma-\omega$ model \cite{qhd}
and extensively used in previous studies. At these
energies, the colliding system consists mainly of nucleons, delta
resonances, and pions. While medium effects on pions are neglected
as in most transport models, nucleons and delta resonances propagate
in a common mean-field potential according to Hamilton equations
of motion, i.e.,
\begin{eqnarray}\label{EOMB}
{d{\bf x}\over dt}=~{{\bf p}^*\over E^*} \qquad\qquad
{d{\bf p}\over dt}=~-\nabla_{\bf x} (E^*+(g_\omega /m_\omega )^2\rho _B),
\end{eqnarray}
where $E^*=(m^{*2}+{\bf p}^{*2})^{1/2}$. The effective mass and kinetic
momentum of a baryon are given, respectively, by
$m^*=m-g_\sigma\langle\sigma\rangle$ and
${\bf p}^*={\bf p}-g_\omega\langle{\bbox\omega}\rangle$, where
the expectation values $\langle\sigma\rangle$ and
$\langle{\bbox \omega}\rangle$
are related to the attractive scalar potential
and the vector current of a baryon in nuclear matter. We use in this
work the parameter set that corresponds to the soft equation of
state \cite{liko94} with a compressibility $K=200$ MeV
and an effective nucleon mass $m_N^*=0.83~m_N$ at normal nuclear density.
These particles also undergo stochastic two-body
collisions, including both elastic ($NN\rightarrow NN$,
$N\Delta\rightarrow N\Delta$,
$\Delta\Delta \rightarrow \Delta\Delta$) and inelastic ($NN\leftrightarrow
N\Delta$, $\Delta\leftrightarrow N\pi$) scattering.
The standard Cugnon parameterization
\cite{bert} and proper detailed-balance prescription \cite{dan} are
used for describing these reactions.
Besides pions, other particles are also produced in heavy-ion collisions.
However, their production probabilities are very small at these energies,
and they can thus be treated perturbatively, in the sense that their
production and interactions do not affect the dynamics of nucleons, delta
resonances, and pions. This is carried out using the perturbative test
particle method of Ref. \cite{fang93,Fang}. In the case of kaons, they are
produced from pion-baryon and baryon-baryon collisions whenever the energy
is above the threshold. Then, a probability factor given by the kaon
production cross section to the total pion-baryon or baryon-baryon cross
section is assigned to the produced kaon. The motion and collision of
the kaon are then followed. However, only the kaon momentum is changed
after a kaon-nucleon collision. This method allows us to include easily
the medium effects on the kaon as shown in Ref. \cite{fang93}. We use
in this study the kaon in-medium mass obtained from the mean-field
approximation to chiral Lagrangian
\cite{N&K}. As in Ref. \cite{Kflow}, the kaon mass $m_K^*$
at nuclear matter density $\rho$ is given by
\begin{equation}\label{kmass}
m_K^*=m_K\left[1-\frac{\Sigma_{KN}}{f^2m_K^2}\rho_s+(\frac{3}{8}\frac{\rho}
{f^2})^2\frac{1}{m_K^2}\right]^{1/2}+\frac{3}{8}\frac{\rho}{f^2}.
\end{equation}
In the above, $m_K$ is the kaon mass in free space; $f\approx 93$ MeV
is the pion decay constant; and $\rho_s$ is the nuclear scalar density. The
kaon-nucleon sigma term is denoted by $\Sigma_{KN}$ and is taken to be
about 350 MeV. At normal nuclear density, the kaon mass is then increased
by about 10 MeV. For antikaon, its vector interaction is opposite to that
of kaon, so the last term in Eq. (\ref{kmass}) has instead a negative sign.
The antikaon mass at normal nuclear density is reduced by about 100
MeV. Corrections to the mean-field results may not be negligible
\cite{lee,waas} and are still under debate. Since there are other
uncertainties in the treatment of kaons such as their production cross
sections from baryon-baryon collisions, we will not address the sensitivity
of our results on these corrections.
Our treatment of kaon follows essentially Refs. \cite{Fang,Kflow,lieos},
where the baryon-baryon production channels $B B \to N\Lambda K$ and $B B
\to N\Sigma K$ are included with their cross sections given by
the Randrup and Ko parameterization \cite{ran}. We have also
included the pion-baryon channels $\pi N \to\Lambda K$ and $\pi N
\to\Sigma K$. For these cross sections, the parameterizations due to
Cugnon \cite{Cugnon} are used. Kaon production from $\pi\Delta$ collisions
is neglected as its cross section has been shown to be much smaller
than that for $\pi N$ collisions \cite{fae}.
The production of antikaon in heavy-ion collisions at SIS
energies was first studied in Ref. \cite{akaon} using the
relativistic transport model and the Zwermann-Sch\"urmann \cite{zwer}
parameterization for the antikaon production cross section
from the nucleon-nucleon collision. However, it has been recently
found that this parameterization overestimates the cross sections
near the threshold by about one order of magnitude \cite{sibirtsev,cass},
and contributions from pion-hyperon interactions \cite{ko83},
which were neglected in Ref. \cite{akaon},
are needed to account for the measured $K^-$ yield.
In this work, we have thus modified our earlier treatment
of antikaon production by using instead the parameterizations proposed
in Ref. \cite{sibirtsev} for antikaon production cross sections
in both pion-baryon and baryon-baryon collisions. We have
also included contributions from the pion-hyperon collisions.
\subsection{The pion and nucleon yield and distributions}
In the upper window of Fig. \ref{cden}, we show the time evolution
of central baryon and pion densities in Ni+Ni collisions
at an energy of 1.93 AGeV and zero impact parameter. It is
seen that the maximum baryon density reached in the
central region is around three times normal nuclear
density, while the maximum delta and pion densities are about 1/6
of the maximum nucleon density. The abundance of pions and deltas
is given in the lower window of Fig. \ref{cden}, which shows
that the maximum number of delta is around 18. Since the
number of pions reaches around 30 at freeze out, the
pion/nucleon ratio is thus about 0.25, which is in agreement
with the experimental data \cite{piond} as well
as other transport model results \cite{Wolf}.
Our results for proton and $\pi^-$
rapidity distributions are compared with the experimental
data from the FOPI collaboration \cite{FOPI} in Fig. \ref{npiy}, where
${\bar y}=y/y_{\rm projectile}$ is the normalized center-of-mass rapidity.
The theoretical results are obtained for the impact parameter range $b\le 2$
fm in order to compare appropriately with the data which have a centrality
selection of 100 mb. Also, only protons that have a local density less
than 0.02 fm$^{-3}$ at freeze out are included. Otherwise, they are
considered as bounded in the deuteron and other light fragments. The
agreement between theoretical results and data, especially in
mid-rapidity region, is quite good, so we feel that
our transport model describes properly the dynamics of nucleons, deltas,
and pions. The proton rapidity distribution shows a peak at
${\bar y}=0$, thus indicating that considerable amount of stopping is
achieved in the collisions.
In Fig. \ref{npimt}, the nucleon and pion transverse mass
distributions at mid-rapidity region $(-0.2<y<0.2)$ are shown.
We find that the inverse slope parameter for nucleons
is about 180 MeV, while that of pions is about 120 MeV. The difference is due
to radial flow, which has a larger effect on
nucleons than pions.
\subsection{The kaon yield and distribution}
\label{kaon}
The abundance of kaons and antikaons is shown in Fig. \ref{kabnd}.
Without kaon medium effects the $K^+$ number reaches
about 0.18, while that of $K^-$ reaches about $1.4 \times 10^{-3}$.
Including kaon medium effects, the kaon yield is reduced by about
20\% to about 0.15, while the $K^-$ yield increases by about
a factor of 3 to about $4 \times 10^{-3}$.
In our model, $K^+$ feels a weak repulsive potential in nuclear
matter, which increases its production threshold and thus reduces
the yield slightly. On the other hand, $K^-$ has a
strong attractive potential, so its production threshold is reduced
and a much larger yield is obtained.
The kaon and antikaon rapidity distributions for central Ni+Ni
collisions are shown in Fig. \ref{ky} for both cases with
and without kaon medium effects. As in the case of pions
and nucleons, the rapidity distributions peak at $y=0$.
In Fig. \ref{kmt}, the kaon and antikaon transverse mass
distributions in mid-rapidity region $(-0.2 < y < 0.2)$ are
shown. For kaon, the inverse slope parameter is found to be about
100 MeV without kaon medium effects, and increases to about
120 MeV after kaon medium effects are included. The kaon momentum thus
increases as a result of propagation in the repulsive mean-field potential.
For antikaon, an inverse
slope parameter of about 110 MeV is found in the case without
kaon medium effects. This is somewhat larger than that of the
kaon transverse mass spectra, mainly due to the absorption of
low-momentum antikaons by nucleons because of a larger absorption
cross section \cite{akaon}. Including kaon medium effects,
the inverse slope parameter decreases to about 100 MeV. The propagation
in the mean-field potential thus reduces the antikaon momentum
as a result of the attractive potential. Therefore, from the ratio
of kaon and antikaon transverse mass spectra it is possible to learn
about the kaon medium effects as pointed out in Ref. \cite{koch}.
\section{Phi meson production from heavy ion collisions}
\label{phiresult}
\subsection{Transport model for the phi meson}
Phi mesons are also treated in our model by the perturbative test particle
method \cite{fang93}. Whenever a pion-baryon or a baryon-baryon collision
is above the threshold for phi meson production, a phi meson is produced and
is assigned a probability factor that is given by the phi meson production
cross section to the total pion-baryon or baryon-baryon cross section.
The motion of the phi meson, including its collisions with other nucleons,
is then followed. As in the case of kaon, effects of these collisions
are included for phi mesons but not nucleons. The numerical simulation
of heavy ion collisions is stopped when the number of collisions is
small. The number
of phi mesons at this time is taken to be the yield accessible
to measurement by identifying $K \bar K$ pairs. As shown in
Refs. \cite{Fang,akaon}, $K N$ and ${\bar K} N$ scattering are
significant before freeze out, so the $K \bar K$ pair coming
from the phi meson decay inside the fireball loose their
correlation before they get out and cannot be identified as a phi
meson from the $K \bar K$ invariant mass spectrum.
In $\pi N$ or $\pi \Delta$ channels phi meson is produced in a
two-body final state, so the magnitude of its momentum
is fixed by kinematics. The polar angle of the phi meson momentum
depends on the differential cross section and is thus determined
by a probability function given by the ratio of the differential
cross section to the total cross section. For the azimuthal angle of the
momentum, it has a uniform distribution between zero and $2\pi$.
The baryon-baryon channels are more complex as the phi meson is produced
in a three-body final state. In this case, the final momentum of the
phi meson follows a multivariable probability function given by the
ratio of the differential cross section to the total cross section.
To simplify the numerical procedure,
we assume that the phi meson is produced isotropically in the
baryon-baryon center of mass frame as discussed in Section \ref{pbb}. This
assumption is often used in
transport model study when no data exist for the differential cross section.
The magnitude of phi meson momentum is then determined by
the momentum spectrum such as that shown in Fig. \ref{pspectrum}
When including medium effects, we only change the available phase
space due to the modification of hadron masses.
Modifications on coupling constants, cut-off parameters, and the functional
form of cross sections are neglected in this study. In the case of a
reaction $A + B \to C + D$, the functional form of the cross section as a
function of $E(=\sqrt{s^*}-\sqrt{s_0^*})$ is assumed to be the same as in
free space, where the modified center of mass $\sqrt{s^{*}}$ energy is
given by $\sqrt{s^{*}} = \sqrt{ p_{A}^{*2}+m_{A}^{*2} } +
\sqrt{p_{B}^{*2}+m_{B}^{*2} }$, while the modified production threshold is
$\sqrt{s^{*}_{o}}=m_{C}^{*2}+m_{D}^{*2}$,
with $p^*$ and $m^*$ being the momentum and mass of
a particle in nuclear medium.
In our study medium modifications of pion properties are
neglected, while those on nucleons are always
included. For the treatment of kaon, antikaon and phi meson
production, we will consider three scenarios. In the first
scenario, we neglect medium effects on both kaons and phi mesons.
It is already known from previous studies of antikaon production
\cite{akaon,cass} that the neglect of kaon medium effects
underestimates the antikaon yield. In the second scenario,
we include medium effects on kaons but neglect those on phi mesons. The
kaon medium effects are obtained from the chiral perturbation
calculation as discussed earlier. Inclusion of kaon medium effects improves
the agreement of the antikaon yield with the experimental data
as discussed in Section \ref{kaon}.
Since a phi meson decays mainly into a kaon-antikaon pair,
medium modifications of kaon properties
affect the final phi meson yield determined from the
reconstruction of $K^+K^-$ pairs. In the last scenario,
we include both kaon and phi meson medium effects.
For the in-medium phi meson
mass we use the results from QCD sum rule calculations by Hatsuda and
Lee \cite{hat},
\begin{eqnarray}
{m_\phi ^*\over m_\phi} \approx 1.0-0.0225 {\rho\over \rho_0},
\end{eqnarray}
which is obtained by taking the nucleon strangeness content to be about 0.15.
\subsection{The phi meson yield}
The results for phi meson yield in central Ni+Ni collisions
at 1.93 AGeV are shown in Fig. \ref{phiabnd}.
Phi mesons are mostly produced during the first 10 fm/c, so
the yield increases with time. Afterwards,
phi meson decay and reabsorption become more important, and the phi
meson yield is seen to decrease with time. We find that at
$t\approx 22$ fm/c two-body scattering becomes rather scarce, so
we take this as the freeze-out time and determine the final phi meson yield
by the number of phi mesons which have not decayed.
In the scenario that both kaon and phi meson medium effects
are neglected, the phi meson yield has a value of $3.3
\times 10^{-4}$, while that of
$K^-$ is about $1.4\times 10^{-3}$. The $\phi /K^{-} $ ratio in central
collision is thus about 25\% and is about a factor of 2 too large
compared with preliminary FOPI data. However, as pointed out earlier
\cite{akaon,cass}, the $K^{-}$ yield without medium effects is
inconsistent with experimental results. In order to explain the
observed $K^-$ yield, kaon medium effects are needed, and this will
increase $K^-$ yield by about a factor of 3.
The $\phi /K^-$ ratio would then be about 8\% and is in
reasonable agreement with the FOPI data.
On the other hand, the inclusion of kaon medium effects
inevitably affects the phi decay width. With a lower in-medium
$K{\bar K}$ threshold, the phi meson decay width increases significantly
\cite{kkonly}, so the probability for a phi meson to decay inside the
dense matter is enhanced. Because of strong final-state interactions of kaons
and antikaons, these phi mesons
cannot be identified from $K^+K^-$ reconstruction.
Including this effect we find that the
final phi meson yield is reduced by about a factor of 2 to
about $1.6 \times 10^{-4}$. The $\phi /K^- $ ratio is now about 4\% and
is about a factor of three too small compared to the FOPI data.
Including also the dropping of phi meson mass in dense
matter leads to an increase of phi meson yield, as a result
of lower production threshold.
For example, around 2$\rho_0$ where most phi mesons
are produced, the reduction in the production threshold is
about 45 MeV. The final phi meson yield after including both
kaon and phi meson medium effects is about
$3\times 10^{-4}$, so the $\phi /K^-$ ratio is brought back to about 8\%,
and is again in reasonable agreement with preliminary data from the FOPI
collaboration.
In Figs. \ref{phipp}, \ref{phimp}, and \ref{phimm},
the phi meson yield from different production
channels is shown for the three scenarios mentioned above.
In all three cases the $\pi N$ channel is the dominant
one, contributing about 60\% of the total phi meson yield,
The importance of the $\pi N$ channel is due to its large
cross section near the threshold, although pions
have a lower abundance than that of nucleons and the
production threshold for this process is also the highest.
The baryon-baryon channels add up to
about 15-20\% of the total phi meson yield, with $N \Delta$
contributing the most. If phi meson production
cross sections in $N\Delta$ and $\Delta\Delta$
collisions are assumed to be the same as that in $NN$ collisions,
we find that the final phi meson yield is reduced
by about 10\%, with the baryon-baryon contribution
being only about 5-10\% of the total.
In the scenario that both kaon and phi meson medium
effects are neglected, the $K \bar K$ channel
contributes only about 7\% of the total phi meson yield.
Inclusion of kaon medium effects increases it
slightly to about 12\% of total phi meson yield.
The impact parameter dependence of phi meson yield in Ni+Ni
collisions at 1.93 AGeV is shown in Fig. \ref{phiimpa}.
From this, the minimum-biased phi meson production cross section
can be estimated. It is about 0.15 mb
when both kaon and phi medium effects are neglected.
This reduces to about 0.05 mb if kaon medium effects
are included but those on phi mesons are neglected.
Including both kaon and phi meson medium effects increases
the production cross section to about 0.13 mb.
Both phi meson decay and absorption have significant effects on its
final yield. Without kaon medium effects, about 20\% of produced phi
mesons are absorbed in $\phi N $ scattering, and about another
20\% decay before freeze out. Including kaon medium
effects, absorption by nucleons still removes about 20\% of produced
phi mesons, but phi meson decay is significantly
enhanced, leading to more than 50\% reduction of the phi meson yield.
\subsection{The phi meson rapidity distribution and transverse
mass spectrum}
The rapidity distribution and transverse mass spectrum
of phi mesons at freeze out are shown in Fig. \ref{phiy}
and Fig. \ref{phimt}, respectively. The phi meson rapidity distribution
is seen to peak at $y=0$ as those of nucleons, pions,
and kaons. Their transverse mass distribution has
a typical exponential form with
an inverse slope parameter of about 110 MeV for
all three scenarios considered.
The results shown in the above include $\phi N$ scattering with a cross
section of 0.56 mb. To see more clearly the effects due to scattering
we show in Fig. \ref{phirs} the phi meson transverse mass spectra
with and without $\phi N$ elastic scattering in the scenario of
no medium effects on kaons and phi mesons. To facilitate the comparison,
different results are normalized to have the same value at
$m_t-m_\phi=25$ MeV. We find that the inverse slope parameter for phi mesons
in the mid-rapidity region, obtained by fitting the spectrum up to 0.4 GeV,
changes only slightly when rescattering is included.
As the extraction of $\phi N$ elastic cross section from the experimental
data is model-dependent, its value may not be well determined. To see
the effects
due to a different cross section, we have also repeated the calculation
using a $\phi N$ elastic cross section of 8.3 mb, which
is the total $\phi N$ cross section estimated from the
phi meson photoproduction data \cite{phipho} and
serves as an upper bound on the elastic cross section. The
inverse slope parameter in this calculation is found to be
about 130 MeV. The increase of the slope parameter due to scattering
can be understood as follows. At SIS energies most
produced phi mesons have low momenta, so scattering
with nucleons increases their momenta and helps them in achieving
thermalization with nucleons. Similar effects have been found for
kaons in heavy ion collisions at subthreshold energies \cite{Fang,aich96}.
\section{Summary and Outlook}
\label{summary}
We have carried out a transport model study of phi meson production
from heavy ion collisions
at SIS energies. The production channels included in our study are $\pinphin$,
$\pidelnphi$, $\nnnnphi$, $\ndelnnphi$, $\deldelnnphi$, and $\kkphi$.
The cross sections for the first five reactions are obtained from the
one-boson-exchange model, while that for the last one is taken to be
of Breit-Wigner form through the phi meson resonance. More than half of
the phi mesons are found to be produced from $\pi N$ collisions.
The predicted $\phi /K^{-}$ ratio depends on whether medium effects are
included for kaons and phi mesons. Neglecting both kaon and phi medium
effects, the ratio is about 25\%, which is about a factor of two
larger than that from the FOPI data. However, the $K^-$ yield in this case is
about a factor of 3 below the measured value. To account for the
observed $K^-$ yield requires the inclusion of medium effects on
kaons, such as that predicted by the chiral perturbation theory,
so that the antikaon production threshold is reduced.
The $\phi /K^-$ ratio turns out to be about a factor
of 2-3 smaller than the experimental results.
Including also dropping phi meson mass
in nuclear medium as predicted by the QCD sum rules increases the phi meson
yield. The final $\phi/K^-$ ratio becomes about 8\% and agrees
reasonably with the preliminary FOPI data.
It will be of interest to also measure
phi mesons through their dilepton decay channel, which will
be carried out in the near future by the HADES collaboration
\cite{Hades} at GSI. Since both the $K \bar K$ and
the dilepton channels are subject to same uncertainties
in the production mechanism, the ratio of the phi meson yield
measured from the $K \bar K$ channel to that measured from
the dilepton channel is expected to be less sensitive to such
uncertainties. In addition, the
dilepton channel can also reveal directly the properties
of vector mesons in nuclear medium, especially the change in
their masses \cite{Wolf,likobrown,cass95}. For phi mesons, this
is not possible via the $K \bar K$ channel since the correlation of a
$K \bar K$ pair is lost due to $K N$ scattering
in the nuclear medium. Furthermore, the shape
of the phi meson peak in the dilepton spectrum
can also indirectly give information
on the kaon medium effects \cite{shuryak}. For this purpose, a careful study of
the background is needed to determine whether the phi
meson peak will still be observable. Previous study \cite{Wolf}
shows that the background due to nucleon and
pion bremsstrahlung as well as the eta Dalitz decay
is small near the phi meson peak.
The major background is thus expected to come from decays of
rho and omega mesons. A careful study of the production
of vector mesons in heavy ion collisions is needed in order to learn the phi
meson properties from the dilepton decay channel. Such a study is in progress.
\begin{center}
\noindent {\bf Acknowledgement}
\end{center}
We acknowledge useful discussions with R. Machleidt. We also thanks
J. A. M. Vermaseren for his program FORM (version 1.1)
and P. Lepage for his subroutine VEGAS, which we have used extensively
in evaluating the elementary cross sections for phi meson production.
This work was
supported in part by the National Science Foundation under
Grant No. PHY-9509266. GQL was also supported in part by
the Department of Energy under Contract No. DE-FG02-88Er40388.
\newpage
\begin{thebibliography}{99}
\bibitem{shor} A. Shor, Phys. Rev. Lett., 54 (1985) 1122.
\bibitem{ozi} S. Okubo, Phys. Lett. 5 (1963) 165; J. Iizuka, Prog. Theo.
Phys. Suppl. 37-38 (1966) 21.
\bibitem{brown} G. E. Brown and M. Rho, Phys. Rev. Lett.
{\bf 66} (1991) 2720.
\bibitem{hat} T. Hatsuda and S. H. Lee, Phys. Rev. C 46 (1992) R34;
T. Hatsuda, Nucl. Phys. A 544 (1992) 27c.
\bibitem{asa} M. Asakawa and C.M. Ko, Nucl. Phys. A 572 (1994) 732.
\bibitem{brownrho} G. E. Brown and M. Rho, Phys. Rep. 269 (1996) 333.
\bibitem{mfrev} C. M. Ko and G. Q. Li, J. of Phys. G 22 (1996) 1673.
\bibitem{kkl} C. M. Ko, V. Koch, and G. Q. Li, Ann. Rev. Nucl. Part.
Sci., in press.
\bibitem{ko92} C. M. Ko, P. L\'evai, X. J. Qiu, and C. T. Li,
Phys. Rev. C 45 (1992) 1400.
\bibitem{hat95} H. Kuwaraba and T. Hatsuda, Prog. Theor. Phys.
94 (1995) 1163c.
\bibitem{song} C. S. Song, Phys. Lett. B 388 (1996) 1410.
\bibitem{kosa} C. M. Ko and B. H. Sa, Phys. Lett. B 258 (1991) 6.
\bibitem{karsch} F. Karsch, K. Redlich, and L. Turko, Z. Phys. C 60 (1993)
519.
\bibitem{kkonly} G. Q. Li and C. M. Ko, Nucl. Phys. A 582 (1995) 731.
\bibitem{ak94} M. Asakawa and C. M. Ko, Phys. Lett. B 322 (1994) 33;
Phys. Rev. C 50 (1994) 3064.
\bibitem{na38} R. Ferreira, Nucl. Phys. A 544 (1992) 497c.
\bibitem{cern} M. A. Mazzoni, Nucl. Phys. A 566 (1994) 95c.
\bibitem{greiner} M. Berenguer, H. Sorge, and W. Greiner,
Phys. Lett. B 332 (1994) 15.
\bibitem{Koch} P. Koch, U. Heinz, and J. Pis\'ut,
Phys. Lett. B 243 (1990) 149.
\bibitem{AGS} B. A. Cole, E802 Collaboration, Nucl. Phys. {\bf A590}
(1995) 179c; ~Y. Akiba {\it et al}, Phys. Rev. Lett. 76 (1996) 2021.
\bibitem{bs95} P. Braun-Munziger, J. Stachel, J. P. Wessels, and
N. Xu, Phys. Lett. B 344 (1995) 43.
\bibitem{fireball} J. Cleymans, D. Elliott, H. Satz, and R.L. Thews,
nucl-th/9603004.
\bibitem{Dover} A. J. Baltz and C. Dover, Phys. Rev. C 53 (1996) 362.
\bibitem{FOPI} N. Herrmann for FOPI Collaboration, in: Proc. Quark
Matter '96, Nucl. Phys. A 610 (1996) 49c.
\bibitem{cass90} W. Cassing, V. Metag, U. Mosel, and K. Niita,
Phys. Rep. 188 (1990) 363.
\bibitem{mos91} U. Mosel, Ann. Rev. Nucl. Part. Sci. 41 (1991) 29.
\bibitem{phibrief} W. S. Chung, G. Q. Li, and C. M. Ko,
Phys. Lett. B, in press.
\bibitem{N&K} D. B. Kaplan and A. E. Nelson, Phys. Lett. B 175 (1986) 57;
A. E. Nelson and D. B. Kaplan, Phys. Lett. B 192 (1987) 193.
\bibitem{lee} C. H. Lee, G. E. Brown, D. P. Min, and M. Rho,
Nucl. Phys. A 585 (1995) 401.
\bibitem{pindata} A. Baldini {\it et al.}, Total cross sections
of high energy particles, (Springer-Verlag, Heidelberg, 1988).
\bibitem{nndata} R. Baldi {\it et al.}, Phys. Lett. B68 (1977) 38;
M. W. Arenton {\it et al.}, Phys. Rev. D 25 (1970) 2241;
V. Blobel {\it et. al.}, Phys. Lett. B 59 (1975) 88.
\bibitem{Mach} R. Machleidt, Adv. Nucl. Phys. 19 (1989) 189.
\bibitem{sib} A. A. Sibirtsev, Nucl. Phys. A 604 (1996) 455.
\bibitem{peierls} R. F. Peierls, Phys. Rev. Lett. 6 (1961) 641.
\bibitem{holzen} B. Holzenkamp, K. Holinde, and J. Speth,
Nucl. Phys. A 500 (1989) 485.
\bibitem{joo} H. Joos, Phys. Lett. B24 (1967) 103.
\bibitem{barger} V. Barger and R. J. N. Phillips, Phys. Lett. B58 (1975) 197.
\bibitem{pion} A. Engel, R. Shaym, U. Mosel, and A. K. Dutt-Mazumder, Nucl.
Phys. A 603 (1996) 387.
\bibitem{mosel} W. Peters, U. Mosel, and A. Engel, Z. Phys. A {\bf 353}
(1995) 333.
\bibitem{laget} J. M. Laget, Phys. Lett. B 259 (1991) 24.
\bibitem{liko} G. Q. Li and C. M. Ko, Nucl. Phys. A 594 (1995) 439.
\bibitem{huber} S. Huber and J. Aichelin, Nucl. Phys. A 573 (1994) 587.
\bibitem{saturne} R. Bertini, Nucl. Phys. A 585 (1995) 265c.
\bibitem{RVUU} C. M. Ko, Q. Li, and R. Wang, Phys. Rev. Lett. 59 (1987) 1084;
C. M. Ko and Q. Li, Phys. Rev. C 37 (1988) 2270;
Q. Li, J. Q. Wu, and C.M. Ko. Phys. Rev. C 39 (1989) 849;
C. M. Ko, Nucl. Phys. A 495 (1989) 321c.
\bibitem{qhd} B. D. Serot and J. D. Walecka, Adv. Nucl. Phys.
16 (1986) 1
\bibitem{liko94} G. Q. Li, C. M. Ko, X. S. Fang, and Y. M. Zheng,
Phys. Rev. C 49 (1994) 1139.
\bibitem{bert} G. F. Bertsch and S. Das Gupta, Phys. Rep. 137 (1988) 189.
\bibitem{dan} P. Danielewicz and G. F. Bertsch, Nucl. Phys. A533 (1991)
712.
\bibitem{fang93} X. S. Fang, C. M. Ko, and Y. M. Zheng, Nucl. Phys.
A556 (1993) 499.
\bibitem{Fang} X. S. Fang, C. M. Ko, G. Q. Li, and Y.M. Zheng,
Nucl. Phys. A575 (1994) 766.
\bibitem{Kflow} G. Q. Li and C. M. Ko, Nucl. Phys. A594 (1995) 460.
\bibitem{waas} T. Waas, N. Kaiser, and W. Weise, Phys. Lett. B 365 (1996)
12; 379 (1996) 34.
\bibitem{lieos} G. Q. Li and C. M. Ko, Phys. Lett. B 349 (1995) 405.
\bibitem{ran} J. Randrup and C. M. Ko, Nucl. Phys. A 343 (1980) 579;
~A 411 (1983) 537.
\bibitem{Cugnon} J. Cugnon and R. M. Lombard, Nucl. Phys. A 422 (1984) 635.
\bibitem{fae} K. Tsushima, S.W. Huang, and A. Faessler,
Phys. Lett. B 337 (1994) 245; J. Phys. G 21 (1995) 33.
\bibitem{akaon} G. Q. Li, C. M. Ko, and X. S. Fang,
Phys. Lett. B 329 (1994) 149.
\bibitem{zwer} W. Zwermann and B. Sch\"urmann, Phys. Lett.
B 145 (1984) 315.
\bibitem{sibirtsev} A. Sibirtsev, W. Cassing, and C. M. Ko,
Z. Phys. A, in press.
\bibitem{cass} W. Cassing, E.L. Bratkovskaya,
U. Mosel, S. Teis and A. Sibirtsev, Nucl. Phys. A614 (1997) 415.
\bibitem{ko83} C. M. Ko, Phys. Lett. B 120 (1983) 294.
\bibitem{piond} J. W. Harris {\it et al.,} Phys. Lett. B 153 (1985) 377;
V. Metag, Proc. VI Int. Conf. on Nuclear Reaction Mechanisms,
Varenna, June, 1991, ed. E. Gadioli, p. 683.
\bibitem{Wolf} Gy. Wolf, W. Cassing, and U. Mosel,
Nucl. Phys. A 552 (1993) 549.
\bibitem{koch} X. S. Fang, C. M. Ko, G. E. Brown, and V. Koch,
Phys. Rev. C 47 (1993) 1678.
\bibitem{phipho} H. J. Behrend {\it et al.}, Phys. Lett. B56 (1975) 408.
\bibitem{aich96} C. David, C. Hartnack, M. Kerveno, J.-Ch. Le Pallec, and
J. Aichelin, nucl-th/9611016.
\bibitem{Hades} W. Koenig, in: Proc. Workshop on Dilepton
Production in Relativistic Heavy-Ion Collisions,
ed. H. Bokemeyer(GSI, Darmstadt, 1994).
\bibitem{likobrown} G. Q. Li, C. M. Ko, and G. E. Brown,
Phys. Rev. Lett. 75 (1995) 4007; Nucl. Phys. A606 (1996) 568.
\bibitem{cass95} W. Cassing, W. Ehehalt, and C. M. Ko,
Phys. Lett. B 363 (1995) 35.
\bibitem{shuryak} D. Lissauer and E. Shuryak, Phys. lett. B 253 (1991) 15;
E. Shuryak and V. Thorsson, Nucl. Phys. A 536 (1992) 739.
\end{thebibliography}
\begin{figure}[p]
\begin{center}
\vfill
\mbox{\epsfxsize=14truecm\epsffile{pib.ps}}
\caption{Feynman diagrams for $\pinphin$ and $\pidelnphi$.
\label{FeynmanpiB} }
\vfill
\end{center}
\end{figure}
\begin{figure}[p]
\begin{center}
\vfill
\mbox{\epsfxsize=14truecm\epsffile{isoavepib.ps}}
\caption{Isospin-averaged cross sections for
$\pinphin$ (solid curve) and $\pidelnphi$ (dotted curve).
\label{isopb} }
\vfill
\end{center}
\end{figure}
\begin{figure}[p]
\begin{center}
\vfill
\mbox{\epsfxsize=14truecm\epsffile{pibdx.ps}}
\caption{Differential cross sections for $\pinphin$ (solid curve)
and $\pidelnphi$ (dashed curve) at 0.2 GeV above the threshold.
\label{diffXpiB} }
\vfill
\end{center}
\end{figure}
\pagestyle{empty}
\begin{figure}[p]
\begin{center}
\vfill
\mbox{\epsfxsize=14truecm\epsffile{bbbbphi.ps}}
\caption{Feynman diagrams for $\nnnnphi$, $\ndelnnphi$, and $\deldelnnphi$.
\label{bb} }
\vfill
\end{center}
\end{figure}
\begin{figure}[p]
\begin{center}
\vfill
\mbox{\epsfxsize=14truecm\epsffile{isoavebb.ps}}
\caption{Isospin-averaged cross sections for $\nnnnphi$ (solid curve),
$\ndelnnphi$ (dashed curve), and $\deldelnnphi$ (dotted curve).
\label{isobb} }
\vfill
\end{center}
\end{figure}
\begin{figure}[p]
\begin{center}
\vfill
\mbox{\epsfxsize=14truecm\epsffile{nnps.ps}}
\caption{Phi meson momentum spectrum from the reactions $\nnnnphi$,
$\ndelnnphi$, and $\deldelnnphi$ at 0.2 GeV above the threshold.
\label{pspectrum} }
\vfill
\end{center}
\end{figure}
\begin{figure}[p]
\begin{center}
\vfill
\mbox{\epsfxsize=14truecm\epsffile{phiabs.ps}}
\caption{Cross sections for $\phinlambdak$ (solid curve),
$\phinpin$(dotted curve), $\phinnrho$ (dashed curve), and
$\phinpiDel$ (dash-dotted curve).
\label{phiabs}}
\vfill
\end{center}
\end{figure}
\begin{figure}[p]
\begin{center}
\vfill
\mbox{\epsfxsize=14truecm\epsffile{cden.ps}}
\caption{Upper window: the time evolution of central densities
of baryons (solid curve), deltas (dashed curve), and pions (dotted curve)
in Ni+Ni collisions at 1.93 AGeV and impact parameter 0 fm.
Lower window: the time evolution of the delta (dashed curve) and pion
(dotted curve) abundance.
\label{cden}}
\vfill
\end{center}
\end{figure}
\begin{figure}[p]
\begin{center}
\vfill
\mbox{\epsfxsize=14truecm\epsffile{protpidy.ps}}
\caption{The rapidity distribution of nucleons
and pions in Ni+Ni collisions at 1.93 AGeV
and impact parameter ${\rm b}\le 2$ fm.
\label{npiy} }
\vfill
\end{center}
\end{figure}
\begin{figure}[p]
\begin{center}
\vfill
\mbox{\epsfxsize=14truecm\epsffile{protpimt.ps}}
\caption{Same as Fig. \protect\ref{npiy} for the transverse mass
distribution.
\label{npimt} }
\vfill
\end{center}
\end{figure}
\begin{figure}[p]
\begin{center}
\vfill
\mbox{\epsfxsize=14truecm
\epsffile{kabnd.ps}}
\caption{The abundance of $K^+$ and $K^-$ with (solid curve) and without
(dotted curve) medium effects in the same
reaction as in Fig. \protect\ref{cden}.
\label{kabnd}}
\vfill
\end{center}
\end{figure}
\begin{figure}[p]
\begin{center}
\vfill
\mbox{\epsfxsize=14truecm\epsffile{ky.ps}}
\caption{The rapidity distribution of $K^+$ and $K^-$ in
central Ni+Ni collisions at 1.93 AGeV with (solid curve) and without
kaon (dotted curve) medium effects.
\label{ky} }
\vfill
\end{center}
\end{figure}
\begin{figure}[p]
\begin{center}
\vfill
\mbox{\epsfxsize=14truecm\epsffile{kmt.ps}}
\caption{Same as Fig. \protect\ref{ky} for the transverse mass
distribution of $K^+$ and $K^-$ with (solid curve) and without (dotted
curve) medium effects.
\label{kmt} }
\vfill
\end{center}
\end{figure}
\begin{figure}[p]
\begin{center}
\vfill
\mbox{\epsfxsize=14truecm\epsffile{phiabnd.ps}}
\caption{The phi meson yield in Ni+Ni collision
at 1.93 AGeV and impact parameter of 0 fm for three different scenarios.
\label{phiabnd}}
\vfill
\end{center}
\end{figure}
\begin{figure}[p]
\begin{center}
\vfill
\mbox{\epsfxsize=14truecm\epsffile{phipp.ps}}
\caption{The time evolution of phi meson yield
from different channels in central Ni+Ni
collisions at 1.93 AGeV without kaon and phi meson medium effects.
\label{phipp} }
\vfill
\end{center}
\end{figure}
\begin{figure}[p]
\begin{center}
\vfill
\mbox{\epsfxsize=14truecm\epsffile{phimp.ps}}
\caption{Same as Fig. \protect\ref{phipp} with kaon medium
effects but no phi meson medium effects.
\label{phimp} }
\vfill
\end{center}
\end{figure}
\begin{figure}[p]
\begin{center}
\vfill
\mbox{\epsfxsize=14truecm\epsffile{phimm.ps}}
\caption{Same as Fig. \protect\ref{phipp} with both
kaon and phi meson medium effects.
\label{phimm} }
\vfill
\end{center}
\end{figure}
\begin{figure}[p]
\begin{center}
\vfill
\mbox{\epsfxsize=14truecm\epsffile{phiimpa.ps}}
\caption{The phi meson yield in Ni+Ni collision
at 1.93 AGeV at different impact parameters for three different scenarios.
\label{phiimpa}}
\vfill
\end{center}
\end{figure}
\begin{figure}[p]
\begin{center}
\vfill
\mbox{\epsfxsize=14truecm\epsffile{phiy.ps}}
\caption{The rapidity distributions of phi
mesons in central Ni+Ni collisions at 1.93 AGeV for three different
scenarios.
\label{phiy} }
\vfill
\end{center}
\end{figure}
\begin{figure}[p]
\begin{center}
\vfill
\mbox{\epsfxsize=14truecm\epsffile{phimt.ps}}
\caption{Same as Fig. \protect\ref{phiy} for the transverse mass
distribution.
\label{phimt} }
\vfill
\end{center}
\end{figure}
\begin{figure}[p]
\begin{center}
\vfill
\mbox{\epsfxsize=14truecm\epsffile{phimtrs.ps}}
\caption{The transverse mass distribution of
phi mesons in central Ni+Ni collisions at
1.93 AGeV with (dotted curve for $\sigma_{\phi N}=0.56$ mb and
solid curve for $\sigma_{\phi N}=8.3$ mb)
and without (dashed curve) $\phi N$ scattering.
\label{phirs}}
\vfill
\end{center}
\end{figure}
\end{document}