and photoproduction off the proton
The exclusive reactions and , leading to the p 4 final state, have been measured with a tagged photon beam for incident energies from threshold up to 2.5 GeV. The experiment has been performed at the tagged photon facility of the ELSA accelerator (Bonn). The Crystal Barrel and TAPS detectors were combined to a photon detector system of almost 4 geometrical acceptance. Differential and total cross sections are reported. At energies close to the threshold, a flat angular distribution has been observed for the reaction suggesting dominant -channel production. and higher lying hyperon states have been observed. An enhancement in the forward direction in the angular distributions of the reaction indicates a -channel exchange contribution to the reaction mechanism. The experimental data are in reasonable agreement with recent theoretical predictions.
pacs:13.60.LeMeson production and 25.20.LjPhotoproduction reactions and 14.20.JnHyperons
The internal structure of the nucleon is reflected in the rich pattern of
baryon resonances. The number of experimentally observed resonances is
much smaller than predicted from theory Capstick (). This is often
referred to as the ’missing’ resonance problem. Baryon resonances
have often large widths and overlap largely, which makes the study of the
excited states particularly difficult. It is possible to overcome this problem
by looking at specific decay channels. Up to now most existing data are based
on elastic scattering experiments. If the hypothesis is correct that
the missing states are unobserved because they couple weakly to the
decay channel, it may be possible to establish some of these missing states in
other channels. Some of the resonances are predicted to decay into final
states with strange particle pairs, coupling strongly to and
Capstick1 (). Strangeness production experiments will therefore
be an important tool to establish ’missing’ resonances and their
properties or to disprove their existence. Recently new measurements of
differential and total cross sections of hyperon photoproduction have been
reported by the SAPHIR lawall (), LEPS sumi (),
CBELSA/TAPS ralph () and CLAS brad (); brad1 () collaborations. A partial
wave analysis provided a good description of these data by introduction of a
new state with quantum numbers and two solutions for the mass -
either at 1885 MeV or at 1970 MeV anis1 (); anis2 ().
Higher mass nucleon resonances could favour decays into . Therefore vector meson photoproduction can be used to search for nucleon resonances which couple strongly to the channel, where denotes a hyperon Capstick2 (). On the other hand, photoproduction of shares elements in common with other strangeness production reactions, such as and or , which lead to or to and resonance excitations with different couplings. The investigation of higher lying hyperon resonances will provide more information about the baryon resonances in meson-hyperon decay channels and help to understand their contribution as background to photoproduction in the reaction .
The cross sections for these reaction channels are small and their experimental identification is difficult. Therefore their investigation became only feasible when high quality photon beam facilities combined with 4 high resolution detectors became available. Here we present the experimental data of (1189) photoproduction off the proton by analysing:
This reaction contains the following isobars:
where the (892) decays into and the decays into . The contribution of a non-resonant (n.r.) pair in (4) occurs together with a (1189).
Recent theoretical studies of photoproduction in the channel oset () have been performed using a chiral unitary approach for meson-baryon scattering in the energy range close to 1700 MeV, below the threshold for production. The theoretical model is based on the assumption that the (1700) is excited and decays into (1385) or (1232). It has been applied to calculate the cross sections of the reactions and oset1 (). The main conclusion is that the mechanism of both reactions is similar - going through the production of the , which is dynamically generated with strong couplings to the and . The current data will be compared with the predictions of this model for the reaction .
Quark model predictions for photoproduction via nucleon resonance excitations in the channels and were presented in Ref. zhao (). In the model, resonances are treated as genuine quark states. There are only two free parameters corresponding to the vector and tensor couplings which depend on the quark mass. Using this approach the cross sections for production have been predicted based on SU(3) symmetry and quark coupling parameters extracted from non-strange production like meson production. The assumption of t-channel exchange in this model leads to strong forward peaking of at higher energies ( 2 GeV).
Another theoretical prediction for a -channel exchange dominated reaction mechanism in photoproduction involves the assumption that the scalar (800) meson may play an important role Oh (). It was demonstrated that exchange could describe the reaction mechanism in production, but for the production the contribution from the (800) meson could be substantial. The results of the CLAS collaboration have also been compared to this theoretical prediction and are in good agreement. Nevertheless, the open question here is the controversial structure of the (800) meson. We will compare our data to this model too.
Experimentally, photoproduction has been studied with the SAPHIR detector, in the reaction wieland (). The was reconstructed from where the measured charged particles are and proton. The differential cross sections show a forward peaking of the meson. The measured total cross section is 0.35 b at 2.2 GeV incident photon energy.
In recent studies of photoproduction by the CLAS collaboration in the reaction Hleiq (), the was reconstructed from the detected particles and ; the was treated as missing particle. The angular distributions are forward peaked; good agreement with the quark model of zhao () was achieved after a slight adjustment of the vector and tensor couplings to the nucleon. The small enhancement of the cross section at backward angles has been interpreted as effect of - and -channel resonances that couple to . The production of higher hyperon resonances , decaying into or , has overlapping kinematics with production leading to a background for the channel .
Our data presented here will be compared with predictions of the available theoretical models and with the published experimental results. The reactions (1)-(4) have been identified via the neutral decay channels, which exclude the contamination from hyperon resonances subsequently decaying via production. The contribution from higher hyperon resonances, decaying into , provide an important background to the reaction (2). We will show evidence for identified higher hyperon states. For photon energies above 1850 MeV, production is dominated by reaction (2). Thus it is important to identify the contribution against the leading contribution, since they are both leading to the same final state.
This paper is organized as follows: In Section 2 we describe the experiment. Section 3 provides the analysis method and event reconstruction. Section 4 shows how to reconstruct and remove the p events which have the same final state and represent a considerable background to the reactions of interest. In Section 5 we discuss the reconstruction of , and . The differential and total cross sections are given and discussed in Section 6. The paper is summarised in Section 7.
2 The experiment
Data have been taken with the detector systems Crystal Barrel
(CB) aker92 () and TAPS Novotny1 (); Gabler1 () at the 3.5 GeV electron
stretcher facility ELSA Husmann (); Hillert (). The detector setup is shown
schematically in figure 1. Electrons extracted from ELSA with energy
hit a primary radiation target, a thin copper or diamond crystal, and
produce bremsstrahlung Elsner (). The tagging system consists of 480
scintillating fibers and 14 partly overlapping scintillator bars. It provides
the corresponding energy of the photons ()
from the deflection of the scattered electrons in a magnetic field. Photons
were tagged in the energy range from 0.5 GeV up to 2.9 GeV
for an incoming electron energy of 3.2 GeV. The total tagged photon intensity
was about in this energy range. The energy resolution varied
between 2 MeV for the high photon energies and 25 MeV for the low photon
energies at an electron beam energy of 3.2 GeV. The part of the beam that did
not produce any bremsstrahlung photons is deflected by the magnet as
well. Since the electrons have retained their full energy the curvature of
their track is smaller and they pass over the tagger into a beam dump.
At the end of the beam line a erenkov detector consisting of 9 lead glass crystals has been installed (fig. 1) which measures those photons that pass through the target without undergoing an interaction. The information provided from this detector has been used for the photon flux determination (see section 6.1).
The Crystal Barrel detector, a photon calorimeter consisting of 1290 CsI(Tl) crystals (16 radiation lengths), covered the complete azimuthal angle and the polar angle from to . The liquid hydrogen target in the center of the CB (5 cm in length, 3 cm in diameter) has been surrounded by a scintillating fibre-detector to detect charged particles suft (). The CB has been combined with a forward detector - the TAPS calorimeter - consisting of 528 hexagonal BaF crystals (12 ), covering polar angles between and and the complete 2 azimuthal angle. In front of each BaF module a 5 mm thick plastic scintillator has been mounted for the identification of charged particles. The combined CB/TAPS detector covered 99% of the full 4 solid angle. The high granularity of this system makes it very well suited for the detection of multi-photon final states.
The first level trigger was derived from TAPS, requiring either one or two hits above different thresholds. The second level trigger was based on a fast cluster recognition (FACE) logic, providing the number of clusters in the Crystal Barrel. For part of the data the minimal number of hits in FACE was one, otherwise at least two hits were requested, which did not introduce any bias for the channels analyzed here.
3 Event reconstruction and event selection
The events due to the reactions and
were reconstructed from the measured
eight photons and the proton in the final channel. Only events containig
exactly nine clusters - eight neutral and one additional charged hit - were
selected. The charged clusters were identified by using the plastic
scintillators in front of the TAPS detector and the fibre detector in the
CB. In order to reduce the background, a cut in a missing mass spectrum
derived from the identified eight photons was applied. The cut selected events
in the region of the nucleon mass; the width of the missing mass cut varied as
a function of the incident photon energy (40
MeV at = 1 GeV to 120 MeV at = 2.6 GeV). Events
which survived this cut were kinematically fitted to the hypothesis
. The procedure is
described in detail in Pee ().
Measurements of the deposited energy and the direction of the photons in the CB
and TAPS calorimeters were used in the fit. For the proton, only the two
angles of its trajectory were used, its energy was calculated, since
the fit is overconstrained. The constraints applied in
this analysis are energy and momentum conservation and the invariant masses of
The confidence level distribution of the fit is shown in fig. 3. Above 20 the distribution is flat in both data and Monte Carlo events. The combination of pairs yielding the highest confidence level was taken as the correct one. Events for which the kinematic fit yielded a confidence level of less than 10 were removed from the data. In order to eliminate time accidental background a prompt coincidence between a photon in TAPS and an electron in the tagger was required. Random time coincidences were subtracted, using events outside the prompt time coincidence window. More details of this procedure can be found in david ().
4 The reaction
An important competing channel, leading to the same final state is . Calculations oset1 () predicted 30 times larger cross section in comparison to the reactions with production. Experimentally, the cross section for reaction was determined to 4b horn (). Events with one and three ’s from -decay were selected. The two-dimensional plot of the invariant mass versus that of the 3 system as a result of the 4 possible combinations is shown in fig. 2a. A vertical band for around 547 MeV can be seen. The invariant mass distribution within the band exhibits a strong peak due to the intermediate state. The results from the analysis of the channel will be published separately. In the present analysis events due to were removed by a cut M() 600 MeV (fig. 2b). After these cuts, 9500 events remain for the analysis of strangeness production.
5 The yield of hyperon , and mesons
From the final state, the , , and , are
reconstructed. The p and 3 combinations are shown in
fig. 2b under the condition that the invariant mass of the 2
from the 3 on the x-axis should be: 470 MeV 520
MeV. This cut has been applied to reconstruct the and
which require kaons to be selected (reactions (2), (3) and (4)). A peak at
= 896 MeV and = 1189 MeV shows correlated
production. Figures 4a and 4b present the
projections on the and invariant mass axes,
respectively. The spectrum in 4a is fitted by a combination of
polynomial and Gaussian functions and shows a clear peak at 11892.0 MeV,
corresponding to . The fit yields a resolution =153.1
MeV. The invariant mass spectrum in 4b, a
projection on the x-axis of the two-dimensional plot in fig. 2b with a
cut on in 1160-1220 MeV, shows a
peak around 896 MeV, corresponding to . Also after applying the cuts a
considerable background remains. The background in this spectrum is very
complex and is discussed later in this section. The invariant
mass from six different combinations is shown in fig. 4c. All
combinations have been taken into account requiring that one of
the other two ’s and a proton have an invariant mass between 1160 and
1220 MeV, because the kaons are always produced with the . The peak
at 496 MeV corresponds to mesons from one of the reactions (2), (3) or
(4). Higher lying hyperon states could contribute to
production via: , where
could be or higher lying states
decaying into (1189). The threshold for the excitation of
(1385) is =1400 MeV. The reactions on the proton
with neutral mesons only, such as , excludes ’s as intermediate states. Requiring
the invariant mass to be close to the mass,
namely between 1160-1220 MeV, figure 5(top) shows a plot of
versus for incident photon energies of
2000-2300 MeV. The cut for this plot is important to
reconstruct the reaction (3), where decays in
. The vertical band around 500 MeV on the x-axis is from the
events (cf. figure 4c). The projection onto the y-axis, with a
cut on between 470 and 520 MeV is plotted in
fig. 5(bottom), which corresponds to a cut on the invariant mass. This cut is shown with solid
vertical lines. The non background (shaded area in
fig. 5) can be determined from side band cuts - shown with the
dashed vertical lines on figure 5(top). Left and right from the
kaon peak, in the invariant mass range 445-470 MeV and 520-545 MeV, the
y-projection of the two-dimensional plot(fig. 5) shows no
peak (shaded area on fig. 5(bottom). The full
spectrum is fitted with a polynomial background and six Breit-Wigner resonance
shapes representing the (1385), (1460), (1560),
(1620), (1660), (1670) with parameters given by
pdg (). The position and width of the resonances are taken from
pdg () and the strengths of the corresponding Breit Wigner are taken as
free parameters. As it can be seen the polynomial background (dashed line on
fig. 5(bottom)) is in good agreement with the background which we
got from the sidebands (shaded area). Apart from the (1385) no
detailed information on the other resonances can be extracted due to
overlap. The estimated production cross section of the
resonance is around 0.70.3 at 1.85 GeV, which is of the same
order as observed for the channel in wieland (); guo ().
Monte Carlo studies have been performed to understand the background
me (). We assumed that the main contribution to the background in
spectra is mainly caused by 4 sequential resonance decays
and other channels leading to the same final state. The production of higher
lying hyperon resonances is hereby of particular importance. Since the final
particles produced by and are the same, higher
lying hyperon states contribute to the spectra as
well. To investigate this contribution we have simulated the reaction and the population of higher
resonances which decay into (1189). The 3
invariant mass calculated from the decay pions and the additional pion
from the decay contributes to the background in the
spectra as shown in figure 6. The contribution of (1385)
and higher resonances to the background has been normalized to the
experimentally observed yields. The
background below the signal is composed of the decay
contributions and a 3 body phase space part. The full fit of the experimental
spectrum, shown in figure 6, has been done using
the signal including the simulated combinatorial background, the
background from , the 3-body phase space and adjusting their
The numbers of the identified , , and are listed in Table 1 for different photon energy bins. The statistical error of the cross section data has been estimated by , where S are the counts in the signal and B are the counts in the background underneath the signal.
6 Results and discussion
6.1 Absolute reaction cross sections
This section describes how the absolute reaction cross sections are
determined. The essential ingredients are the reaction yields, the detection
efficiency of the individual final states and the photon flux.
Due to the almost 4 coverage, the detection efficiency for the hyperon
final states is practically independent of the production angle. This is
illustrated for both reactions and
in fig. 7, where the total efficiency is shown, including
geometrical acceptance and the detector efficiency. The total efficiency was
determined with a GEANT-based Monte Carlo simulation. The simulated events
were evenly distributed over the available phase space and analyzed using the
same event selection criteria, kinematic fit, applied cuts, thresholds and
trigger conditions as for the experimental data. The resulting efficiency
varies slightly between 7-10% with the incoming photon energy; its angular
dependence is very small.
Uncertainties in the reconstruction of hyperons and vector mesons have been studied. By varying the fit conditions in order to achieve a consistent description of the background in different kinematical regions, an error of 3% -15% is deduced.
The photon flux through the target is determined by counting the photons reaching the intensity detector in coincidence with electrons registered in the tagger system. This provided an absolute normalization for all measurements. The main point is the accurate determination of the efficiency of the tagging system, as defined by the probability to identify the corresponding photon in the photon beam for each detected electron in the tagger. The detector has almost 100% photon detection efficiency. By comparing the number of electrons in the tagging system in coincidence with the number of counts in the detector, the tagging efficiency has been determined to vary between 64 to 74 % for different beamtimes. The systematic uncertainty in the cross sections, caused by the photon flux determination has been checked by measurements of known reactions, such as , and is estimated to be 5%-15% depending on the photon energy.
6.2 Differential cross section
The differential cross sections are calculated from the number of events identified in the respective channel using:
() are the counts of () determined in
different angle and energy bins as described in Sect. 5.
() is the efficiency determined as described in Sect. 6.1;
is the number of primary photons in the respective energy bin determined as described in Sect. 6.1;
is the target area density for the LH target used in this experiment;
is the angle bin width of the angular distributions.
() is the branching ratio of the reaction () respectively; ’s were identified via their decay into 2 which has a relative branching ratio of 98.798%.
The angular distributions for the reactions and are shown in figures 8 and 9,
respectively. The results are plotted as a function of
and as a function of
respectively. For incident photon energies higher
than 1850 MeV, above the threshold for the production, the cross
section for the has been
extracted from the difference between and yields, i.e.
In the energy range 1500-1850 MeV, below the production threshold, the differential cross section is almost flat. These measurements are in good agreement with the theoretical prediction of doering (), based on the model of oset1 (), as shown in the first picture in figure 8. The flat angular distribution indicates dominant s-channel production which is a genuine prediction of a chiral dynamical calculation based on the dominance of the in the entrance channel, plus the coupling of this resonance to . For the energy region 1500-1850 MeV, where the production is energetically not possible, the reconstructed events could come from a resonance which decays subsequently in and . For energies higher than 1850 MeV, an additional contribution from the reaction is possible. These two contributions can be separated experimentally through the mass spectrum which exhibits a sharp peak at the position of the meson (fig. 6). The observed counts in the peak at different bins of are used for the differential cross section determination (fig. 9 full circles).
The differential cross section d/d for the reaction shows a rise in the forward direction when plotted vs the production angle (fig. 9). Production of the meson via -channel exchange seems to play an important role in the reaction dynamics. These results are compared with the updated calculations from the model in zhao_up (), using the free parameters and , which describe the universal couplings for the vector and tensor part in the quark vector-meson interaction. These parameters are different from those given in zhao (), which were derived from meson photoproduction based on SU(3) symmetry. Our data are compared to the experimental data from the CLAS collaboration Hleiq () which also show an enhancement in the forward direction. At backward angles we do not observe any rise in the cross section. There is a discrepancy to the CLAS data for the forward angle bins and at incident photon energies higher than 2150 MeV. Our experimental data are consistently higher at angles with in the lower two pictures in fig. 9.
A comparison to another theoretical model assuming (800) meson exchange Oh () is also shown. The theoretical curves of the so-called model II provide a reasonable agreement with the experimental data which is not the case for model I. The main difference between models I and II of ref. Oh () is in the form and the strength of the form factor of the (800) meson.
The energy dependence of the differential cross sections for the reaction is presented in fig. 10 for six angular bins and compared with the experimental data from CLAS. The peaking in the forward direction is more pronounced than in the CLAS data as it can be seen in the lowest two pictures in figure 10.
The forward peaking of the cross section suggests that
there is a contribution to the reaction mechanism from -channel exchange. To
test this idea the differential cross section
d/d as function of was
converted to d/ as a function of .
The variable is the Mandelstam invariant that gives the 4-momentum
squared of the exchange particles. Since the momentum transfer is
limited by kinematics, lies between and , given by:
The cross section d/ is shown in fig. 11. The straight lines represent fits of in the region below 1.0 of . The slope parameter has a negative value. We plot the slope parameter as a function of the incident photon energy in fig. 12. It can be seen that the slope parameter rises with the photon energy. This is an indication for an increasing contribution to production via -channel exchange as predicted in zhao ().
6.4 Total cross section
The total cross sections for the and reactions are shown in fig. 13.
The cross sections agree within errors with those determined by integrating
over the differential cross sections.
The experimental cross section for the
channel are in reasonable agreement with theoretical predictions of
reference zhao () with the assumption of a -channel
The predicted total cross section for in oset1 () is given by a band (the shaded area in fig. 13). The good agreement between theory and experiment suggests that the dynamics used in the model is reasonable for the corresponding energy region. The cross section for rises with increasing photon energy. Above 1850 MeV the dominating channel for production is obviously . It is about factor 2 larger than the cross section of the reaction which includes excited hyperon decays into the .
We have reported measurements of differential and total cross sections for the
and reactions. The experimental data have been compared with the
available theoretical predictions. At low incident photon energies below the
production threshold ( 1850 MeV) only the
channel is energetically possible, exhibiting a flat
angular distribution, dominated by s-channel production which is the
prediction based on the dominance of the , subsequently decaying
For energies above 1850 MeV the production is mainly associated with production. The angular distributions for the reaction show a rise in the forward direction of the vector meson which indicates a -channel exchange contribution to the reaction mechanism. It is, however, not yet possible to make an explicit conclusion about the exchanged particle, due to the significant contribution to the channel from the production.
The presented data provide valuable confirmation of the theoretical predictions concerning the reaction mechanism of the and reactions. and higher resonances have been observed in the decay channel. On the basis of existing data we have estimated the production cross section of the resonance. A corresponding value for the cross section of the higher states photoproduction can not be given because of the overlap of these resonances. Further experiments are required to study the higher resonances. Polarisation experiments will be needed to clarify details of the strangeness production process.
We thank the accelerator group of ELSA as well as the technicians of the participating institutions for their outstanding support. We acknowledge illuminating discussion with Kenneth Hicks on details of the data analysis. Useful discussions with E. Oset and M. Döring and also Q. Zhao on the theoretical interpretation of the data are also highly acknowledged. This work was supported by Deutsche Forschungsgemeinschaft SFB/TR16 and Schweizerischer Nationalfond.
- (1) S. Capstick and W. Roberts, Prog. Part. Nucl. Phys. 45, 241 (2000).
- (2) S. Capstick and W. Roberts, Phys. Rev. D 57, 4301 (1998).
- (3) R. Lawall et al., Eur. Phys. J. A 24, 275 (2005).
- (4) M. Sumihama et al., LEPS Collaboration, Phys. Rev. C 73, 035214 (2006).
- (5) R. Castelijns et al., CBELSA/TAPS Collaboration, Eur. Phys. J. A 35, 39 (2008).
- (6) R. Bradford et al., CLAS Collaboration, Phys. Rev. C 73, 035202 (2007).
- (7) R. Bradford et al., CLAS Collaboration, Phys. Rev. C 75, 035205 (2007).
- (8) A. V. Sarantsev et al., Eur. Phys. J. A 25, 441 (2005).
- (9) A. V. Anisovich et al., arXiv: hep-ph/07073596
- (10) S. Capstick and W. Roberts, Phys. Rev. D 58, 074011 (1998).
- (11) M. Döring, E. Oset, D. Strottman, Phys. Rev. C 73, 045209 (2006).
- (12) M. Döring, E. Oset, D. Strottman, Phys. Lett. B 639, 59-67 (2006).
- (13) Q. Zhao, J. S. Al-Khalili, C. Bennhold, Phys. Rev. C 64, 052201 (2001).
- (14) Y. Oh and H. Kim, Phys. Rev. C. 74, 015208 (2006).
- (15) F. Wieland, Ph.D. thesis University of Bonn, Bonn, Germany (2005).
- (16) I. Hleiqawi et al., CLAS Collaboration, Phys. Rev. C 75, 042201 (2007), Erratum-ibid.C76:039905, 2007. /corrected values, see: arXiv: nucl-ex/0701036v3.
- (17) E. Aker et al., Nucl. Instr. and Methods A 321, 69 (1992).
- (18) R. Novotny et al., IEEE Trans. Nucl. Sci.38, 392 (1991).
- (19) A.R. Gabler et al., Nucl. Instr. and Meth. A 346, 168 (1994).
- (20) D. Husmann, W. J. Schwille, Phys. Bl. 44, 40 (1988).
- (21) W. Hillert, Eur. Phys. J. A 28, s01, 139 (2006).
- (22) D. Elsner et al., CBELSA/TAPS Collaboration , Eur. Phys. J. A 33, 147 (2007).
- (23) G. Suft et al., Nucl. Instr. Meth., A 538, 416 (2005).
- (24) H. van Pee et al., Eur. Phys. J. A 31, 61 (2007).
- (25) D. Trnka, Ph.D Thesis University of Giessen, Giessen, Germany (2006).
- (26) I. Horn et al., CBELSA Collaboration, arXiv: nucl-ex/07111138.
- (27) L. Guo and D. P. Weygand, CLAS Collaboration: Proceedings of the Workshop on the Physics of Excited Nucleons NSTAR 2005, Florida State University, Tallahassee, 12-15 October 2005, edited by S. Capstick et al. (World Scientific, Singapore, 2006) p. 306.
- (28) W. M. Jao et al., Particle Data Group, J. of Physics G 33, 1049 (2006).
- (29) M. Nanova for CBELSA/TAPS Collaboration: Proceedings of the Workshop on the Physics of Excited Nucleons NSTAR 2005, Florida State University, Tallahassee, 12-15 October 2005, edited by S. Capstick et al. (World Scientific, Singapore, 2006) p. 359.
- (30) M. Döring, Ph.D Thesis University of Valencia, Valencia, Spain (2006).
- (31) Q. Zhao, Numerical fit by the quark model of Ref. zhao () to the CBELSA/TAPS data, private communication.