nucl-th0306024/eta.tex
1: \documentclass[aps,prc,showpacs,amsmath,amssymb]{revtex4}
2: \usepackage{epsfig}
3: 
4: 
5: \newcommand{\acal}{{\cal A}}
6: 
7: 
8: \begin{document}
9: %\draft
10: 
11: \title{In-Medium Modifications of the $\boldsymbol{S_{11}(1535)}$ Resonance 
12:      and Eta Photoproduction}
13: 
14: \author{J. Lehr, M. Post and U. Mosel}
15: 
16: \affiliation{Institut f\"ur Theoretische Physik, Universit\"at
17: Giessen\\ D-35392 Giessen, Germany}
18: 
19: \date{\today}
20: 
21: 
22: % ----------------------------------------------------------------
23: \begin{abstract}
24: We investigate the influence of possible in-medium modifications of the
25: $S_{11}(1535)$ in eta photoproduction in nuclei.
26: Besides Fermi motion, Pauli blocking and binding effects also collisional
27: broadening is accounted for. The in-medium width is obtained from a
28: realistic resonance-hole model. Results on eta photoproduction are
29: obtained from a semi-classical BUU transport model and compared with
30: data. We find that calculations including a momentum dependent nucleon
31: and resonance potential are in agreement with the recent KEK data.
32: In contrast, collisional broadening has only little influence.
33: \end{abstract}
34: % ----------------------------------------------------------------
35: \pacs{25.20.Lj, 25.20.x}
36: \maketitle
37: 
38: 
39: 
40: 
41: \section{Introduction}
42: 
43: Photon induced reactions on nuclei provide a promising tool for the
44: investigation of in-medium properties of hadrons. However, 
45: the interpretation of data is often difficult,
46: because it is hard to disentangle contributions from different 
47: resonances and/or different decay channels as e.g. in the case of pion 
48: production in the second resonance region.
49: In this respect the study of the $S_{11}(1535)$ is very interesting because 
50: of its strong coupling to $N\eta$. 
51: In the energy regime of $E_\gamma\sim 600-900$ MeV this channel is strongly
52: dominated by the $S_{11}(1535)$ and therefore $\eta$ photoproduction gives
53: information almost solely about this resonance.
54: 
55: In the past, different experiments of $\eta$ photoproduction on several
56: nuclei (C, Al, Ca, Cu, Nb, Pb) were performed. The TAPS group covered the 
57: energy range from threshold up to photo energies of 800 MeV \cite{roebig_eta},
58: whereas measurements at KEK \cite{yorita,yamazaki} provided data up 
59: to 1 GeV.
60: Other theoretical approaches to the process discussed several 
61: in-medium effects.
62: In \cite{yorita} a QMD model was applied. Besides the trivial effects of
63: Fermi motion and Pauli blocking, the authors found a strong influence due to
64: collisional reactions of the $S_{11}$.
65: The authors of \cite{maruyama} were able to describe the KEK data on Carbon 
66: under the assumption that both scalar and vector potential of the $S_{11}$ 
67: vanish. 
68: However, the calculations were performed in nuclear matter and the $\eta$ 
69: final state interactions were modelled by using a constant absorption factor.
70: Hence, the findings of this work have to be checked by a more realistic
71: model.
72: 
73: In this work, we calculate $\eta$ photoproduction on different nuclei using
74: a semi-classical BUU transport model, which was already successfully applied
75: to the calculation of a variety of reactions (e.g. heavy-ion reactions
76: \cite{teis}, photon- and electron-induced reactions 
77: \cite{effe_pion,lehr_electro}). Within this model, $\eta$ photoproduction 
78: for energies up to 800 MeV  was already addressed in 
79: \cite{effe_pion,hombach}. In contrast, we now want to extend our study 
80: to higher energies covered by the KEK data and to discuss medium modifications
81: of the $S_{11}$. The $S_{11}$ in-medium width is obtained from self-consistent 
82: resonance-hole calculations \cite{post}.
83: 
84: We start with a brief review of the BUU model and the implementation
85: of the $\eta$ and $S_{11}$ dynamics in Sec. \ref{sec:buu_model} followed
86: by a discussion of in-medium modifications of the $S_{11}$ width in
87: Sec. \ref{sec:inmed_width}. In Sec. \ref{sec:results} we show our
88: results in comparison with experimental data.
89: 
90: 
91: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
92: 
93: 
94: \section{The BUU Model} \label{sec:buu_model}
95: 
96: For the present studies, we use the BUU model reported in 
97: \cite{lehr_electro,effe_dilep}. 
98: The model is based upon a system of coupled BUU equations describing the 
99: evolution of the spectral distribution function $F_i=\acal_i f_i$ with the
100: phase space density $f_i$ of  
101: different particle types $i=N,\pi,S_{11}(1535),\eta,...$:
102: \begin{equation}
103:   \label{eq:buu-eq}
104:   \left({\partial\over\partial t}+\vec\nabla_p H\cdot\vec\nabla_r
105:   -\vec\nabla_r H\cdot\vec\nabla_p\right)F_i(\vec r,\vec p,\mu;t)
106: =\Sigma^<_i\acal_i(1-f_i(\vec r,\vec p,\mu;t))-\Sigma^>_i\acal_i 
107:   f_i(\vec r,\vec p,\mu;t).
108: \end{equation}
109: Here  $\acal_i$ is the spectral function and $H$ is a relativistic Hamilton 
110: function 
111: \begin{equation}
112:   \label{eq:hamilton-fkt}
113:   H=\sqrt{(m+S)^2+p^2}
114: \end{equation}
115: with a scalar potential $S$ described in Eq. (\ref{eq:potential3}). 
116: 
117: 
118: $\Sigma^{\gtrless}_i$ stand for the collision rates of particle species $i$
119: and describe the gain and the loss of the distribution function at some
120: 'point' $(\vec r,\vec p,\mu,t)$ due to collisional reactions with particles
121: of the same and other types. Therefore, the BUU equations are coupled
122: via their right-hand sides. The set of 
123: equations is solved by a test particle ansatz for each density function $F_i$.
124: For model details we refer the reader to \cite{lehr_electro,effe_dilep}.
125: 
126: Besides the nucleon, our model contains 29 nucleon resonances with
127: parameters taken from the analysis of Manley and Saleski \cite{manley} and
128: all relevant mesonic degrees of freedom ($\pi,\eta,\rho,\omega$).
129: The $\eta$ couples to the resonances $S_{11}(1535)$,
130: $S_{11}(1650)$ and $F_{17}(1990)$ with pole-mass branching ratios of
131: 0.43, 0.03 and 0.94.
132: 
133: The single-particle energy of a nucleon in the local rest frame (LRF) of the
134: surrounding nuclear matter is given by
135: \begin{equation}
136:   \label{eq:potential1}
137:   \epsilon=\sqrt{(m_N-U_s)^2+p_{\textrm{LRF}}^2}+U_0\equiv
138:   \sqrt{m_N^2+p_{\textrm{LRF}}^2}+V
139: \end{equation}
140: with vector and scalar potentials $U_0$ and $U_s$.
141: For the nucleon, in the LRF the spatial components of the vector potential 
142: vanish in the mean-field approximation. In the non-relativistic limit, $V$ 
143: corresponds to the difference of scalar and vector potential.
144: For $V$ we use the density and momentum dependent non-relativistic mean-field
145: parametrization from Welke \emph{et al.} \cite{welke}
146: \begin{equation}
147:   \label{eq:potential2}
148:   V(\vec r,\vec p)=A{\rho\over\rho_0}+B\left({\rho\over\rho_0}\right)^{\tau}+
149: {8C\over\rho_0}\int{d^3p^\prime\over(2\pi)^3}{\Theta(p_F(\rho(r))-p^\prime)
150: \over 1+\left({\vec p-\vec p^\prime\over \Lambda}\right)^2}.
151: \end{equation}
152: In this equation $p_F=p_F(\rho)$ is the local Fermi momentum. We can work with
153: a choice 
154: of parameter sets $A,B,C,\tau,\Lambda$ determining the momentum dependence and 
155: stiffness of the nuclear matter equation of state (EOS) (cf. Table 
156: \ref{tab:potpar}).
157: Here, we use either a momentum independent hard EOS (H) or a momentum 
158: dependent medium EOS (M).
159: In Fig. \ref{fig:potential} we show the momentum dependence of the potential
160: $V$ for the parameter set (M) for different densities.
161: $V$ is transformed in the LRF into a scalar potential $S$ using
162: \begin{equation}
163:   \label{eq:potential3}
164:   \sqrt{(m_N+S)^2+p_{\textrm{LRF}}^2}=\sqrt{m_N^2+p_{\textrm{LRF}}^2}+V.
165: \end{equation}
166: 
167: We assume that the potential $V$ for the $S_{11}(1535)$ is also given by
168: Eq. (\ref{eq:potential2}), and the scalar resonance potential 
169: $S_R$ is obtained in the same way. The effective resonance mass is then 
170: given by
171: $$
172:   \mu_{\textrm{eff}}=\mu+S_R.
173: $$
174: Since no detailed calculations exist for the $S_{11}$, this is a reasonable
175: assumption.
176: 
177: 
178: 
179: %%%%%%%%%%%%%%%%%%%%%%%%%
180: 
181: \subsection{The elementary photon-nucleon reaction}
182: 
183: The main source of $\eta$ mesons is the elementary $\gamma N$ reaction.
184: It is known that in the threshold region the process $\gamma p\to \eta p$ 
185: is well described by the excitation of the $S_{11}$ resonance (see e.g.
186: \cite{penner}). 
187: The contribution coming from the two other resonances coupling to $N\eta$ 
188: are neglected, because the $\eta N$ branching ratio of the $S_{11}(1650)$ 
189: is very small and the $F_{17}(1990)$ is beyond the considered energy range.
190: Therefore, we use the following Breit-Wigner parametrization:
191: \begin{align}
192:   \label{eq:gamma_proton}
193:   \sigma_{\gamma p\to S_{11}\to X}&=\left({k_0\over k}\right)^2
194:    {s\Gamma_\gamma(\sqrt s)\Gamma_{S_{11}\to X}(\sqrt s)\over
195:    (s-M_{S_{11}}^2)^2+s\Gamma_{S_{11}\to X}^2(\sqrt s)} 
196:    {2m_N\over M_{S_{11}}\Gamma_0}\vert A_{1/2}^p\vert^2,\nonumber\\
197:   \sigma_{\gamma p\to S_{11}\to \eta p}&=\sigma_{\gamma p\to S_{11}}
198:      {\Gamma_{S_{11}\to \eta p}(\sqrt s)\over 
199:          \Gamma_{S_{11}\to X}(\sqrt s)}
200: \end{align}
201: with $\Gamma_\gamma=\Gamma_0\cdot k/k_0$ \cite{walker} and the
202: pole-mass decay width $\Gamma_0=0.151$ GeV.
203: The center of mass (cm) momentum $k$ of the $\gamma p$ pair
204: depends on the cm energy $\sqrt s$ (i.e. mass of the resonance), 
205: $k_0=k(M_{S_{11}})$ is the cm momentum taken at the pole mass of the $S_{11}$. 
206: The other (mass dependent) resonance widths $\Gamma_{S_{11}\to X}$ and
207: $\Gamma_{S_{11}\to\eta N}$ are parametrized as in
208: \cite{effe_dilep}. For the photocoupling 
209: helicity amplitude we use $A_{1/2}^p=0.109$ GeV$^{-1/2}$ \cite{krusche_eta}.
210: The cross section for the reaction $\gamma n\to S_{11}$ is obtained from 
211: (\ref{eq:gamma_proton}) by 
212: \begin{equation}
213:   \label{eq:gamma_neutron}
214:   \sigma_{\gamma n}={2\over 3}\sigma_{\gamma p},
215: \end{equation}
216: as suggested in \cite{krusche_deut}, corresponding to  
217: $A_{1/2}^n\sim 0.089$ GeV$^{-1/2}$. This value results from a resonance fit 
218: to eta photoproduction data on the deuteron and is connected to the value for
219: $A_{1/2}^p$ found in \cite{krusche_eta}. Therefore, we do not use the
220: (smaller) values found in other analyses (e.g. \cite{penner}).
221: 
222: In Fig. \ref{fig:gamma_nucleon} we show the cross section for the elementary
223: process $\gamma p\to\eta p$ in comparison with the data sets from 
224: \cite{krusche_eta,graal_eta}.
225: For energies up to $E_\gamma\sim 1$ GeV the agreement with the data is good.
226: 
227: Our model also contains other channels for the elementary $\gamma N$ 
228: interaction,
229: namely $\gamma N\to \pi N$, $\pi\pi N$, $P_{33}(1232)$, $D_{13}(1520)$,
230: $F_{15}(1680)$. These processes might also contribute to $\eta$ photoproduction
231: on nuclei via final state interactions (e.g. the reaction chain 
232: $\gamma N\to \pi N$, $\pi N\to S_{11}\to \eta N$). However, it was 
233: shown in \cite{effe_pion} that they are rather small.
234: The parametrization of these processes via resonance fits to experimental
235: pion photoproduction data is similar to Eq. (\ref{eq:gamma_proton}) and
236: described in Ref. \cite{effe_pion}.
237: 
238: %%%%%%%%%%%%%%%%%%%%%%%%%%
239: 
240: \subsection{Final state interactions}
241: 
242: The $\eta$ final state interactions (FSI) are assumed to be mediated by
243: the re-excitation of resonances. In the energy range under consideration the 
244: $S_{11}(1535)$ clearly dominates. The cross sections for elastic and inelastic
245: $\eta N$ reactions therefore are similar to Eq. (\ref{eq:gamma_proton}) and 
246: strongly depend on the energy of the $\eta$:
247: \begin{equation}
248:   \label{eq:eta_fsi}
249:   \sigma_{\eta N\to S_{11}\to X}={4\pi\over p_{\eta N}^2}
250:  {s\Gamma_{S_{11}\to \eta N}(\sqrt s)\Gamma_{S_{11}\to X}(\sqrt s)
251: \over (s-M_{S_{11}}^2)^2+s\Gamma_{S_{11}\to X}^2(\sqrt s)}
252: \end{equation}
253: with the $\eta N$ cm momentum $p_{\eta N}$. $X$ contains the decay channels
254: $N\pi$, $N\eta$, $N\rho$, $N\sigma$ and $P_{11}(1440)\pi$. Among these,
255: $N\pi$ and $N\eta$ are the relevant channels (branching ratio $\sim 95\%$).
256: In the medium, the $S_{11}$ also undergoes FSI via the processes 
257: \begin{eqnarray}
258:   \label{eq:s11_fsi}
259:   N S_{11}  &\leftrightarrow& NN \nonumber\\
260:  N S_{11} &\to& N S_{11} \nonumber\\
261: N S_{11} &\leftrightarrow& N R,\ R\ne S_{11},
262: \end{eqnarray}
263: which give 
264: rise to a finite collision width. This will be discussed in 
265: Sec. \ref{sec:inmed_width}. Of course, the absorption of $S_{11}$ states in 
266: such reactions also contributes to the absorption of $\eta$ mesons.
267: 
268: Finally, the coupled-channel treatment of our BUU model also allows for 
269: contributions from side-feeding reactions such as $\pi N\to R\to \eta N$.
270: As already mentioned, the $\eta$ mesons managing to escape the nucleus
271: stem to a large extent from the elementary reaction chain 
272: $\gamma N \to S_{11} \to N\eta$.
273: 
274: 
275: 
276: 
277: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
278: 
279: 
280: 
281: \section{Calculation of the $\boldsymbol{S_{11}}$ in-medium width}
282:   \label{sec:inmed_width}
283: 
284: In \cite{effe_abs} the collisional broadening of the resonances 
285: $P_{33}(1232)$, 
286: $D_{13}(1520)$, $S_{11}(1535)$ and $F_{15}(1680)$ was calculated from the 
287: collision rates resulting from processes like (\ref{eq:s11_fsi}):
288: \begin{align}
289:   \label{eq:collrate}
290:   \Sigma^>_{S_{11} N\to X}(E,p,\rho)={1\over 2E}&\intop_{}^{}
291:   {d^3 p_2\over (2\pi)^3}{d\mu_2\over 2E_2}{d^3 p_3\over (2\pi)^3}
292:   {d\mu_3\over 2E_3}{d^3 p_4\over (2\pi)^3}{d\mu_4\over 2E_4}
293:   (2\pi)^4\delta^4(p+p_2-p_3-p_4)\nonumber\\
294:   &\times \vert{\cal M}_{S_{11}N\to X}\vert^2\acal_2 f_2\acal_3 (1-f_3)\acal_4 (1-f_4)
295: \end{align}
296: with $X=NN$, $S_{11} N$ and $R N$. $f_2$ is the phase space density of the
297: incoming nucleons, $1-f_{3,4}$ are the Pauli blocking factors for the outgoing 
298: particles included in the final state $X$.
299: Note that such collision rates also appear in the collision integrals on the 
300: right-hand side of the BUU equations. Therefore, consistency between the collision 
301: widths and the processes explicitly included in the transport model is guaranteed.
302: The matrix elements appearing in the collision rates are identical to those 
303: contained
304: in the cross sections for the processes (\ref{eq:s11_fsi}).
305: Eq. (\ref{eq:collrate}) can be rearranged in the following way:
306: \begin{equation}
307:   \label{eq:ldt}
308:   \Sigma^>=\rho_N\langle v_{\textrm{rel}}^{\textrm{cm}}
309:     \sigma_{S_{11}N\to X}^{\textrm{cm}}(1-f_3)(1-f_4)\rangle_N,
310: \end{equation}
311: where the averaging is performed over the momentum distribution of the incoming
312: nucleon and $v_{\textrm{rel}}^{\textrm{cm}}$ is the relative velocity 
313: of the incoming $S_{11} N$ pair in the cm frame. We see that this expression is similar 
314: to what one obtains from the low density theorem
315: $\Gamma_{\textrm{coll}}=\rho\,v\,\sigma_{\textrm{tot}}$, leading to an increase of the
316: collision rates with momentum $p_R$, as seen in Fig. \ref{fig:gcoll}.
317: This way, for an on-shell resonance a broadening of about $35$ MeV at 
318: $\rho_0$ was found in \cite{effe_abs}. 
319: 
320: 
321: It is well known from various calculations \cite{oset_delhole,post_rho1,lutz_kaon}, 
322: that the applicability of the low density theorem may be restricted to a regime of rather 
323: small densities. In order to overcome this problem we have developed a coupled channel 
324: analysis of the properties of $\pi$, $\rho$ and $\eta$ mesons as well as baryon 
325: resonances in 
326: nuclear matter \cite{post_rho1,post_rho2,post}. The resonance parameters are taken
327:  from the 
328: analysis of Manley \emph{et al.} \cite{manley}. The in-medium propagators of the 
329: mesons are 
330: determined from the excitation of nucleon-hole and resonance-hole loops. The resonance
331: self energy in 
332: nuclear matter $\Sigma_{\textrm{med}}$ is obtained by replacing the vacuum meson 
333: propagator with 
334: the in-medium one in the meson nucleon loops, see Fig. \ref{fig:selfen}. Of course, 
335: also the 
336: correction from Pauli blocking is included. The collisional broadening is given by:
337: \begin{equation}
338:  \label{eq:coll-width}
339:         \Gamma_{\textrm{coll}}(\sqrt{s},p) = 
340:         \frac{\textrm{Im}\,\Sigma_{\textrm{med}}(\sqrt{s},p)-
341:               \textrm{Im}\,\Sigma_{\textrm{Pauli}}(\sqrt s,p)}{\sqrt{s}},
342: \end{equation}
343: where $\textrm{Im}\,\Sigma_{\textrm{Pauli}}$ is the Pauli-blocked vacuum width.
344: By coupling the meson and resonance properties a self-consistency problem arises, 
345: which we 
346: solve iteratively. The physical interpretation is that higher iterations involve
347:  reactions on 
348: more than one nucleon. For example, going to the next order effectively takes into 
349: account 
350: 3-body processes of the resonances. This is clearly beyond the low density theorem. 
351: 
352: Concerning the properties of the $S_{11}$ resonance, we find a net broadening of $35$ 
353: MeV at $\rho_0$ for an on-shell resonance. As can be seen in Fig. \ref{fig:gcoll}, 
354: in the momentum range of interest between 0.6 and 1 GeV this result is in absolute size
355: (relative to the total width at $\rho_0$) close to that of \cite{effe_abs}.
356: Effects from the dressing of $\pi$ and $\eta$ mesons are of the order of a few MeV, the 
357: only sizeable contribution to the collisional broadening coming from the $N\rho$ sector
358: (for details we refer the reader to \cite{post}). 
359: This is in qualitative agreement with the calculation of \cite{oset_eta}, where a 
360: somewhat larger value for the broadening was predicted.
361: The strong in-medium correction in the $N\rho$ channel is due to the coupling of the 
362: $D_{13}(1520)$ state to the $\rho$ meson, which moves spectral strength down to 
363: smaller invariant masses \cite{post_rho1,post_rho2} in the $\rho$ mass distribution. 
364: This way, the phase space available for the decay of a $S_{11}(1535)$ is enhanced.
365: The slightly different momentum behavior of the collision widths (\ref{eq:coll-width}) 
366: compared 
367: to the collision rates (\ref{eq:collrate}) is due to the resummation of particle-hole
368: loops in the meson propagators. This leads to a decrease of the $N\eta$ contribution and
369: an increase of the $N\rho$ contribution with increasing $p_R$, leaving the total width
370: $\Gamma_{\textrm{coll}}$ nearly constant. This behavior, however, cannot be extrapolated
371: to very small momenta $p_R<0.3$ GeV due to effects from Pauli blocking in the $N\eta$
372: channel.
373: 
374: We find that the main effect from the inclusion of higher order effects is a slight 
375: reduction 
376: of the broadening in the $N\rho$ channel. This is due to the strong broadening of the 
377: $D_{13}(1520)$ which leads to a moderate reduction of this state in the $\rho$ spectral 
378: function. Therefore, some $\rho$ spectral strength is moved up to larger invariant masses, 
379: and the decay of a $S_{11}$ to the $N\,\rho$ channel is somewhat suppressed.
380: 
381: We have also calculated the mass shift corresponding to the broadening of 
382: the $S_{11}$ by means of a dispersion analysis \cite{post}. As a result we do not 
383: observe any 
384: significant shift. This, of course, supports the hypothesis, that the mass shift as 
385: observed 
386: in the photo-nucleus data, see Fig. \ref{fig:gamma_nucleus}, is due to binding effects, 
387: see Section \ref{sec:results}. In the actual calculation, we therefore do not include
388: the dispersive mass shift.
389: 
390: The influence of the medium on the $S_{11}$ enters the calculations in two different
391: ways: The reactions displayed in (\ref{eq:s11_fsi}), which are directly 
392: connected with a collision rate via Eq. (\ref{eq:collrate}), are implemented explicitly 
393: in the model. Due to the 
394: fact that the results for the collisional width (Eq. (\ref{eq:coll-width})) are close to 
395: the collision rates, the consistency between the widths and these explicit processes is 
396: maintained. Furthermore, since the $S_{11}$ final states are not restricted
397: to the vacuum decay channels, the width $\Gamma_{S_{11}\to X}$ in Eqs.
398: (\ref{eq:gamma_proton}) and (\ref{eq:eta_fsi}) has to be substituted by the
399: full in-medium width. 
400: 
401: 
402: 
403: 
404: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
405: 
406: 
407: 
408: \section{Results} \label{sec:results}
409: 
410: In Fig. \ref{fig:gamma_nucleus} we show our results for the reaction $\gamma A\to \eta X$
411: for different nuclei in comparison with the TAPS and KEK data. The latter are obtained
412: by integration over $0^\circ\le \vartheta\le 90^\circ$, while the former data as well
413: as our calculations cover the full angular range. However, the main contribution stems from
414: angles smaller than $90^\circ$. Imposing this limitation on our calculations, deviations
415: smaller than 3\% are found.
416: 
417: The dashed and solid curves in each plot correspond to the two potentials (H) and (M) 
418: described in Sec. \ref{sec:buu_model}. The calculations include Fermi motion, Pauli blocking,
419: nuclear binding effects and FSI, but $S_{11}$ vacuum widths are used in the cross
420: sections (\ref{eq:gamma_proton}) and (\ref{eq:eta_fsi}). The dotted curves do
421: involve the full $S_{11}$ in-medium width in the cross sections and therefore show 
422: the total influence of collisional broadening.
423: 
424: The curves (M) are shifted by about 50-70 MeV compared to the curves (H).
425: This can be understood by taking a look at the elementary reaction
426: $\gamma N\to S_{11}(1535)$. The momentum independent potential (H) is 
427: identical both for incoming nucleon and outgoing resonance. 
428: The resonance mass is obtained from the cm energy $\sqrt s$ (i.e. effective resonance 
429: mass) by subtracting the resonance potential:
430: $$
431:   \mu_{\textrm{H}}=\sqrt s - S_{\textrm{H}}.
432: $$
433: In the case of the momentum dependent potential (M) we obtain a different result. For 
434: photon energies of about 800 MeV the outgoing resonance has also a momentum of
435: about 800 MeV. For such momentum values, the potential (M) is almost zero, see Fig. 
436: \ref{fig:potential}.
437: Therefore we get for the resonance mass $\mu_{\textrm{M}}=\sqrt s$, which is 
438: is smaller compared to $\mu_{\textrm{H}}$:
439: $$
440:   \mu_{\textrm{M}}=\sqrt s=\mu_{\textrm{H}}-\vert S_{\textrm{H}}\vert.
441: $$
442: Hence the peak maxima in both scenarios are shifted relatively to each other
443: by roughly $\vert S_{\textrm{H}}\vert$, which is approximately 50-70 MeV.
444: 
445: As one can see, the curves (H) overestimate the TAPS data in the threshold
446: region and exhibit a maximum shifted towards lower energies 
447: with respect to the peak maximum suggested by the KEK data. In contrast, the curves (M) only
448: slightly underestimate the TAPS data, but are in very good
449: agreement with the KEK data, especially the location of the peak maximum is well
450: reproduced. 
451: 
452: The influence of the collisional broadening of the $S_{11}$ has only little influence. As can be seen from
453: the dotted curves, which visualize the effect of the medium modification via
454: the cross sections (\ref{eq:gamma_proton}) and (\ref{eq:eta_fsi}). 
455: In order to completely analyze the influence of collisional broadening, we show in
456: Fig. \ref{fig:carbon_coll} four different szenarios:
457: The dashed curve shows the calculation without any FSI. The dashed-dotted curve includes
458: FSI except for the $S_{11}$ FSI in (\ref{eq:s11_fsi}) and therefore
459: shows the influence of the direct $\eta$ absorption via resonance re-excitation.
460: Both calculations use vacuum widths in the cross sections (\ref{eq:gamma_proton}) and
461: (\ref{eq:eta_fsi}). The solid curves include all FSI and corresponds to the  solid
462: curves in Fig. \ref{fig:gamma_nucleus}. The dotted curve also includes the 
463: full in-medium widths in the cross sections and corresponds to the dotted curves in
464: Fig. \ref{fig:gamma_nucleus}. It is seen that the main effect comes from the direct $\eta$
465: absorption via $\eta N\to R$, whereas the influence of in-medium effects concerning the 
466: $S_{11}$ is rather small. This is not surprising, because the mean free path $\lambda$ of 
467: a $S_{11}$ at $\rho_0$ is about 3 fm -- compared to $\lambda\lesssim 1$ fm for the $\eta$
468: -- whereas the RMS radius of C is about 2.5 fm. 
469: This is in contrast to the results of Yorita \emph{et al.} \cite{yorita}, who find
470: a strong effect from the $S_{11}$ FSI.
471: 
472: In Fig. \ref{fig:ca_fsi} we show the effect of the FSI in the case 
473: of Calcium. The solid line includes FSI. The dotted curve is the result 
474: without FSI divided my a constant
475: factor of 1.9. In both calculations we used the potential (M) and neglected the medium 
476: modification of the $S_{11}$ widths. It is clearly visible that the absorption depends on 
477: energy and therefore the assumption of a constant absorption factor made in 
478: \cite{maruyama} is unrealistic. This result also holds for other nuclei.
479: On the other hand, the issues raised in \cite{maruyama} concerning the
480: peak position can be verified. The claim there was that the peak position can
481: only be described by assuming vanishing scalar and vector potentials for the 
482: $S_{11}(1535)$. This scenario is close to what we get with the momentum
483: dependent potential (M) for the $S_{11}$, which -- as mentioned above -- 
484: nearly vanishes in the kinematical regime under investigation.
485: 
486: 
487: 
488: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
489: 
490: 
491: 
492: \section{Summary}
493: 
494: We have calculated $\eta$ photoproduction on several nuclei in the energy
495: region dominated by the $S_{11}(1535)$ resonance. Starting from a parametrization
496: of the data for the elementary $\gamma N\to\eta N$ reaction, we applied a BUU
497: transport model to account for final state interactions.
498: The calculations including a momentum dependent potential for the nucleons and the 
499: $S_{11}$ reproduce the available data well. In the threshold region a slight 
500: underestimation of the data is observed. The calculations involving a momentum 
501: independent potential cleary overestimate the data.
502: 
503: The collisional width of the $S_{11}$, calculated within a realistic self-consistent 
504: resonance-hole model 
505: \cite{post}, was found to be close to the collision rates from \cite{effe_abs}
506: and was included in the transport calculations.
507: The influence of collisional reactions of the $S_{11}$ on the cross section were found 
508: to be small, in contrast to the QMD calculations of Yorita \emph{et al.} \cite{yorita}.
509: The main difference between the result without FSI and the full calculation can be
510: attributed to the absorption of the $\eta$ in the FSI e.g. via 
511: $\eta N\to S_{11}\to \pi N$.
512: Finally, we have shown that the absorption reactions cannot be mimicked by applying
513: a constant factor to the results without FSI.
514: 
515: 
516: \section*{ACKNOWLEDGMENTS}
517:  
518: 
519: This work was supported by DFG.
520: 
521: 
522: 
523: % ----------------------------------------------------------------
524: \begin{thebibliography}{99}
525: \bibitem{roebig_eta} M. R\"obig-Landau \emph{et al.}, Phys. Lett. {\bf B373},
526:     45 (1996).
527: \bibitem{yorita} T. Yorita \emph{et al.}, Phys. Lett. {\bf B476}, 226 (2000).
528: \bibitem{yamazaki} H. Yamazaki \emph{et al.}, Nucl. Phys. {\bf A670}, 202c (2000).
529: \bibitem{maruyama} T. Maruyama and S. Chiba, nucl-th/0204051.
530: \bibitem{teis} S. Teis, W. Cassing, M. Effenberger, A. Hombach and U. Mosel,
531:  Z. Phys. {\bf A356}, 421 (1997).
532: \bibitem{effe_pion} M. Effenberger, A. Hombach, S. Teis and U. Mosel,
533:     Nucl. Phys. {\bf A614}, 501 (1997).
534: \bibitem{lehr_electro} J. Lehr, M. Effenberger and U. Mosel, Nucl. Phys.
535: {\bf A671}, 503 (2000).
536: \bibitem{hombach} A. Hombach, A. Engel, S. Teis, U. Mosel, Z. Phys. {\bf A 352},
537:  223 (1995).
538: \bibitem{post} M. Post, in preparation. 
539: \bibitem{effe_dilep} M. Effenberger, E. Bratkovskaya and U. Mosel, Phys. Rev.
540: {\bf C60}, 044614 (1999).
541: \bibitem{manley} D.M. Manley and E.M. Saleski, Phys. Rev. {\bf D45}, 4002  
542:    (1992).
543: \bibitem{welke} C. Gale, G. Welke, M. Prakash, S. Lee and S. Das Gupta,
544:    Phys. Rev. {\bf C41}, 1545 (1990);\\
545:  G. Welke, M. Prakash, T. Kuo and S. Das Gupta, Phys. Rev. {\bf C38}, 2101 
546:    (1988).
547: \bibitem{penner} G. Penner, U. Mosel, Phys. Rev {\bf C66}, 055212 (2002).
548: \bibitem{walker} R. L. Walker, Phys. Rev. {\bf 182}, 1729 (1969).
549: \bibitem{krusche_eta} B. Krusche \emph{et al.}, Phys. Rev. Lett. {\bf 74}, 
550:       3736 (1995).
551: \bibitem{krusche_deut} B. Krusche \emph{et al.}, Phys. Lett. {\bf B358}, 40 (1995).
552: \bibitem{graal_eta} F. Renard \emph{et al.}, Phys. Lett. {\bf B528}, 215 (2002).
553: \bibitem{effe_abs} M. Effenberger, A. Hombach, S. Teis and U. Mosel,
554:     Nucl. Phys. {\bf A613}, 353 (1997).
555: \bibitem{oset_delhole} E. Oset and L.L. Salcedo, Nucl. Phys. {\bf A468},
556:     631 (1993).
557: \bibitem{post_rho1} W. Peters, M. Post, S. Leupold, H. Lenske and U. Mosel, Nucl. Phys. {\bf A632}, 109 (1998).
558: \bibitem{lutz_kaon} M.F.M. Lutz and E.E. Kolomeitsev, Nucl. Phys. {\bf A700}, 193 (2002). 
559: \bibitem{post_rho2} M. Post, S. Leupold and U. Mosel, Nucl. Phys. {\bf A689}, 753 (2001). 
560: \bibitem{oset_eta} H. C. Chiang, E. Oset and L. C. Liu, Phys. Rev. {\bf C44}, 738 (1991) . 
561: 
562: \end{thebibliography}
563: 
564: \newpage
565: 
566: \begin{table}
567: \caption{\label{tab:potpar} Parameters for the mean-field potential $V$.}
568: \begin{ruledtabular}
569: \begin{tabular}{lcccccc}
570:    & incompressibility [MeV] & $A$ [MeV] & $B$ [MeV] & $C$ [MeV] & $\tau$ & 
571: $\Lambda$ [fm$^{-1}$] \\
572:  \hline
573:  H & 380 &-124.3 & 71.0 & 0. & 2.0 & - \\
574:  M & 290 &-29.3 & 57.2 & -63.5 & 1.76 & 2.13 \\
575: \end{tabular}
576: \end{ruledtabular}
577: \end{table}
578: 
579: 
580: %%%%%%%%%FIG 0%%%%%%%%%%%%%%%%%%%%%%
581: \begin{figure}
582: \begin{center}
583: \includegraphics[width=11cm]{fig1.eps}
584: \end{center}
585: \vspace{-0.5cm}
586: \caption{Momentum dependence of the potential $V$ with parameter set (M) given
587: in Tab. \protect\ref{tab:potpar} for different densities}. 
588:  \label{fig:potential}
589: \end{figure}
590: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
591: 
592: 
593: %%%%%%%%%FIG 1%%%%%%%%%%%%%%%%%%%%%%
594: \begin{figure}
595: \begin{center}
596: \includegraphics[width=11cm]{fig2.eps}
597: \end{center}
598: \vspace{-0.5cm}
599: \caption{Parametrization of the elementary process $\gamma p\to\eta p$ 
600: according to Eq. (\protect\ref{eq:gamma_proton}). The data are taken
601: from \protect\cite{krusche_eta} and \protect\cite{graal_eta}.} 
602:  \label{fig:gamma_nucleon}
603: \end{figure}
604: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
605: 
606: %%%%%%%%%FIG 2%%%%%%%%%%%%%%%%%%%%%%
607: \begin{figure}
608: \begin{center}
609: \includegraphics[width=14cm]{fig3.eps}
610: \end{center}
611: \vspace{-0.5cm}
612: \caption{Feynman diagram for the self energy of the $S_{11}$ resonance in 
613: vacuum (left) and in nuclear matter (right). The dashed line represents the exchange of $\pi$, $\eta$ or $\rho$ mesons. By the blob in the meson line the full in-medium propagator for the mesons, which are dressed by the excitation of nucleon-hole and resonance-hole loops, is indicated.}
614:  \label{fig:selfen}
615: \end{figure}
616: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
617: 
618: 
619: %%%%%%%%%FIG 3%%%%%%%%%%%%%%%%%%%%%%
620: \begin{figure}
621: \begin{center}
622: \includegraphics[width=11cm]{fig4.eps}
623: \end{center}
624: \vspace{-0.5cm}
625: \caption{$S_{11}$ collision width at the pole mass as a function of the 
626: resonance momentum $p_R$ for different densities (dashed and solid curves).
627: Also shown are the results of Effenberger \emph{et al.} \protect\cite{effe_abs}
628: (dotted curves).}
629:  \label{fig:gcoll}
630: \end{figure}
631: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
632: 
633: %%%%%%%%%FIG 4%%%%%%%%%%%%%%%%%%%%%%
634: \begin{figure}
635: \begin{center}
636: \includegraphics[width=11cm]{fig5.eps}
637: \end{center}
638: \vspace{-0.5cm}
639: \caption{Results of the BUU model for the reaction $\gamma A\to\eta X$ on 
640: different nuclei. The dashed and solid lines correspond to the usage of the
641: potentials (H) and (M). The dotted lines include the medium-modified
642: width for the $S_{11}(1535)$ discussed in Sec. \ref{sec:inmed_width}.
643: The data are from \cite{roebig_eta} (circles) and 
644: \cite{yorita,yamazaki} (squares).}
645:  \label{fig:gamma_nucleus}
646: \end{figure}
647: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
648: 
649: 
650: %%%%%%%%%FIG 5%%%%%%%%%%%%%%%%%%%%%%
651: \begin{figure}
652: \begin{center}
653: \includegraphics[width=11cm]{fig6.eps}
654: \end{center}
655: \vspace{-0.5cm}
656: \caption{Influence of medium modifications on $\eta$ photoproduction
657: on Carbon with potential (M). The dashed curve shows the result without
658: FSI, the dashed-dotted includes FSI except for the $S_{11}$ collision
659: reactions. The solid curve includes all FSI and vacuum widths in the 
660: cross section for the $S_{11}$ excitation
661: and corresponds to the
662: solid line in Fig. \protect\ref{fig:gamma_nucleus}. The dotted curve
663: includes full in-medium widths in these cross sections and 
664: coincides with the dotted curve in Fig. \protect\ref{fig:gamma_nucleus}.}
665:  \label{fig:carbon_coll}
666: \end{figure}
667: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
668: 
669: 
670: 
671: %%%%%%%%%FIG 6%%%%%%%%%%%%%%%%%%%%%%
672: \begin{figure}
673: \begin{center}
674: \includegraphics[width=11cm]{fig7.eps}
675: \end{center}
676: \vspace{-0.5cm}
677: \caption{Influence of the FSI on the cross section $\gamma \textrm{Ca}\to\eta X$.
678: The solid curve shows the result with FSI, while the 
679: dotted line is the result without FSI divided by a
680: factor of 1.9. All curves include the potential (M). In-medium 
681: modifications of the $S_{11}$ width are neglected.}
682:  \label{fig:ca_fsi}
683: \end{figure}
684: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
685: 
686: 
687: 
688: 
689: 
690: \end{document}
691: