Bled Workshops in Physics Vol. 5, No. 1
A
Proceedings of the Mini-Workshop
Quark Dynamics (p. 83)
Bled, Slovenia, July 12-19, 2004
Search for Pentaquarks at HERA-B
Tomi Zivko
for the HERA-B Collaboration
Abstract. A search for ©+ and S3/2 pentaquarks has been performed in channels p K§ and S 7tin proton - nucleus interactions at mid-rapidity and y/s = 41.6 GeV/c2 . No evidence for pentaquarks has been found in analyzed channels. Upper limits have been set on pen-taquark production cross sections.
Experimental evidence for a new hadron state at 1540 MeV/c2 decaying to nK+ was presented by experiment LEPS [1] in 2003. The particle was named ©+ (1540). Due to a quark picture of neutron and K+, the hadron 0+ must contain at least four quarks and one antiquark. After that, several other collaborations reported evidence for a peak in the invariant mass spectrum of n K+ or p K^ . The p K° peak was regarded as evidence for 0+ —> pKS on the grounds "no narrow !*+ is known around 1.5 GeV/c2 ". Currently, there are 12 experiments that have claimed evidence for decays 0+ —> nK+ or 0+ —> pK^ . The measured mass lies in the range 1521 - 1555 MeV/c2 . There is a peculiarity that p K^ experiments report smaller value of mass than n K+ ones. The measured widths have all been consistent with the experimental resolution which is typically 20 MeV/c2. The presented peaks have statistical significance of about 5 ct. In theoretical models 0+ is a member of an antidecuplet which also contains isospin 3/2 family E3/2 of doubly strange pentaquarks. Evidence for doubly charged and neutral member of the family was observed in En decay channels at mass of 1862 MeV/c2 by NA49 [2]. The statistical significance of E3/2 peak is also about 5 ct. Up to now, this has been the only evidence for E3/2. From the other side, the number of high statistics experiments reporting negative search results for 0+ and E3/2 is growing. Direct comparison of positive and negative search results is not possible because the experiments are not of the same type. However, the negative search results reported much larger yield of common particles like A (1520) and E(1530)0 , thus proving their ability to search for possible pentaquark signals in channels p K^ and E n. This short survey of experimental situation suggests that existence of pentaquarks is not proven beyond reasonable doubt. The search for 0+ and E3/2 was done also at HERA-B. The main features of the analysis are presented here, while details can be found elsewhere [3].
HERA-B is a fixed target experiment at the 920 GeV proton storage ring of DESY. It is a forward magnetic spectrometer with a high resolution vertex-ing and tracking system and good particle identification. The detector has good
84 T. Zivko
acceptance in the mid-rapidity region. The informations from the silicon vertex detector, the main tracker system, ring imaging Cherenkov (RICH) counter and the electromagnetic calorimeter (ECAL) were used in this analysis. The present study was performed on a sample of about 200 millions of minimum bias events that were taken at ^/s = 41.6 GeV/c2 using carbon, titanium and tungsten targets. Strange particles are frequent in proton - nucleus interactions at this energy, and HERA-B has reconstructed a large number of Kg —> n+n-, A —> pn- and A —> pn+ decays. A clean sample of E hyperons was obtained in decay modes E- —> An- and EE + —> An+. Background in all fore mentioned channels was efficiently reduced using decay topology, so there was no need for particle identification. Table 1 summarizes the statistics of relevant signals together with the measured mass resolutions. All measured masses are within 1 MeV/c2 compatible with the table values.
Signal	C target all targets ff [MeV/c2 ]
Ks	2.2M	4.9M	4.9
A [A]	440k[210k]	1.1M[520k]	1.6
A (1520) [A(1520)]	1.9k[1.1k]	5.1k[2.3k]	2.3
[2 + ]	4.7k [3.4k]	11.8k [8.2k]	2.6
2(1530)° [2(1530)0	] 610 [380]	1.4k [940]	2.9
Table 1. Statistics and experimental mass resolution (ff) for relevant particles are given for carbon target and for all targets.
HERA-B does not have capabilities for the identification of neutrons. Therefore, the search for 0+ was performed in the decay channel p Kg . Protons were identified requiring the proton likelihood from the RICH to be larger than 0.95. Probability that a particle which is not proton passes this cut is below 1 %. Both particles, proton and Kg had to point to the main vertex. Kg candidates had to lie in ±3ct mass window around the table mass. A clean Kg sample remained after removing particles whose mass was consistent with A or A. The invariant mass spectrum of selected p Kg pairs is shown for p+C data in Fig. 1a. The shape of background was obtained by event mixing technique and is represented by a full line. MC studies show that the mass resolution in the presented mass region is in 2.6 - 6.1 MeV/c2 range. At the 0+ mass, the resolution is 3.9 MeV/c2 . We determined the upper limit on the number of signal events in the invariant mass plot as a function the signal mass. The resulting nuclear cross section as a function of the signal mass is presented in Fig. 1b (full line).
Assuming A0 7 dependence of the nuclear cross section on the atomic number, we obtained the upper limit on Br x do/dy|y=0 < 3.7 |b/nucleon in the mid-rapidity region for 0+ mass of 1530 MeV/c2 . The upper limit varies from 3 to 22 |b in the mass region 1521 - 1555 MeV/c2 . The upper limits obtained using data from all targets are similar. We also tried with other search strategies,
Search for Pentaquarks at HERA-B
85
100
50 0
o
e
T3
oq
a) , ■ ./......	i i i i i i i i i i i p + C " ...........
■ b) - A * , , , , i , , ,	i i i i i i i i i i i . iU 11 . i .... i ... .
i i i i i i i i i i i i i i i i i i i ■	c) mult < 10 - - ^/fcA^^ ■	}t t p + CT -. +______1___________	
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 ■ d) mult < 10 + K- all targets: (^ . . +.................	
1.475	1.55
pKS mass, GeV/c
1.625
2
>
tB
5
m
p
e
■O
05
500
400
300
200
100
100
20
0 — 1.4

1.6
1.
2
2.2
mass, GeV/ c
Fig. 1. Invariant mass distributions and upper limits on nuclear cross section for channels p Ks (left) and Sn(right). Arrows denote mass region 1521 - 1555 and mass of 1530 MeV/c2 (left) and mass of 1862 MeV/c2 (right). Data were taken with carbon target. See text for details.
100
0
0
f)
50
0
0
20
like: a) requiring a low track multiplicity in an event (Fig. 1c), b) strangeness tagging, by requiring a particle with an s quark (A, K- ) in an event, c) combination (Fig. 1d) of criteria a) and b), d) relaxation of the proton identification cut. None of the attempts resulted in a significant narrow peak in the mass spectrum. We checked capabilities of the HERA-B detector by reconstruction of A(1520) —> pK-. Masses of ©+ and A (1520) are similar as well as geometrical acceptances for 0+ pK° and A(1520) pK-. Using RICH likelihood cut for both proton and K-, we obtained a clean signal for A (1520). Assuming Br(0+ —> pK^ )=1/4, we determined the UL(95%) on the particle ratio A(f52o) < 0.92% in the mid-rapidity region. This upper limit is more than one order of magnitude lower than predictions of statistical hadronization models. We also found that 1< 0.27%.
We searched for members of E3/2 family in decay channels E-n-, E-n+ and c.c. E- candidates had to lie in ±3ct mass window around the table mass. Both E- and n candidates had to point to the main vertex. Weak identification cuts with RICH and ECAL removed tracks with clear electron, kaon or proton identity from the n sample. The invariant mass spectra of En pairs obtained from
86 T. Zivko
p+C data are shown in Fig. 1e for all four charge combinations. The background shape is obtained from event mixing and is normalized to the data. The experimental resolution in the analyzed mass region is in 2.9 - 10.6 MeV/c2 range and has value of 6.6 MeV/c2 at the mass of 1862 MeV/c2. The only observed structure in the spectra are signals for E(1530)° and E(1530)0 in neutral channels. Fig. 1f gives sum of invariant mass distributions of all four charge channels after subtraction of background. Particularly, there is no enhancement in mass region around around 1862 MeV/c2, where NA49 observed E3/2 candidates. We determined UL(95%) on Br ■ da/dy|y=°, which at mass of 1862 MeV/c2 are 2.5, 2.3, 0.85 and 3.1 ^b/nucleon in E-n-, E-n+, Eand Echannels, respectively. The corresponding upper limits using all targets are 2.7,3.2,0.94 and 3.1 |xb /nucleon. We also found the UL(95%) on particle ratio Br ■ E—/E° (1530) < 4% and Br ■ E /E- < 3%. As an illustration, the UL(95%) on nuclear cross section is presented in Fig. 1g (full line) as function of E mass.
To conclude, we searched for pentaquark signals in channels p K^ and En. Having found no evidence for signals we set upper limits on production cross sections and particle ratios in mid-rapidity region. If existent, strange pentaquarks (0+ and E3/2 ) also seem to have exotic production mechanisms.
References
1.	T. Nakano et al. (LEPS Collaboration), Phys. Rev. Lett. 91, 012002 (2003), [arXiv:hep-ex/0301020].
2.	C. Alt et al. (NA49 Collaboration), Phys. Rev. Lett. 92, 042003 (2004), [arXiv:hep-ex/0310014].
3.	I. Abt et al. (HERA-B Collaboration), submitted to Phys. Rev. Lett. (2004), [arXiv:hep-ex/0408048].