Measurement of antiproton annihilation on Cu, Ag and Au with emulsion films

The characteristics of low energy antiproton annihilations on nuclei (e.g. hadronization and product multiplicities) are not well known, and Monte Carlo simulation packages that use different models provide different descriptions of the annihilation events. In this study, we measured the particle multiplicities resulting from antiproton annihilations on nuclei. The results were compared with predictions obtained using different models in the simulation tools GEANT4 and FLUKA. For this study, we exposed thin targets (Cu, Ag and Au) to a very low energy antiproton beam from CERN's Antiproton Decelerator, exploiting the secondary beamline available in the AEgIS experimental zone. The antiproton annihilation products were detected using emulsion films developed at the Laboratory of High Energy Physics in Bern, where they were analysed at the automatic microscope facility. The fragment multiplicity measured in this study is in good agreement with results obtained with FLUKA simulations for both minimally and heavily ionizing particles.


A
: The characteristics of low energy antiproton annihilations on nuclei (e.g. hadronization and product multiplicities) are not well known, and Monte Carlo simulation packages that use different models provide different descriptions of the annihilation events. In this study, we measured the particle multiplicities resulting from antiproton annihilations on nuclei. The results were compared with predictions obtained using different models in the simulation tools GEANT4 and FLUKA. For this study, we exposed thin targets (Cu, Ag and Au) to a very low energy antiproton beam from CERN's Antiproton Decelerator, exploiting the secondary beamline available in the AEgIS experimental zone. The antiproton annihilation products were detected using emulsion films developed at the Laboratory of High Energy Physics in Bern, where they were analysed at the automatic microscope facility. The fragment multiplicity measured in this study is in good agreement with results obtained with FLUKA simulations for both minimally and heavily ionizing particles. K : Detector modelling and simulations I (interaction of radiation with matter, interaction of photons with matter, interaction of hadrons with matter, etc); Particle tracking detectors (Solidstate detectors)

Introduction
Emulsion films have recently been considered as possible position detectors for low-energy antimatter studies. These studies include the AEgIS (AD6) experiment at CERN [1][2][3][4], whose goal is the measurement of the Earth's gravitational acceleration on antihydrogen atoms. Another collaboration proposed emulsions for their studies on positrons, as described in [5]. In particular, in the case of the AEgIS experiment, the position-sensitive detector must have a micrometer-level resolution to allow the required sensitivity of ∼1% for the gravitational acceleration measurement. Spatial resolutions of ∼1-2 µm can be achieved with emulsion films [6], and they have been exploited before for the reconstruction of antihydrogen impact points from annihilation products [3]. Films with this resolution, combined with a time of flight detector, could allow the experimental goal to be achieved. In the same paper, a preliminary study of antiproton-nuclei annihilations was also reported. That study assessed particle multiplicities resulting from antiproton annihilations on emulsion films and aluminium. Recently, again within the context of the AEgIS experiment, similar measurements were performed by means of a silicon detector also acting as an annihilation target [7]. Apart from the obvious applications in nuclear physics, measuring the decay products of low-energy antiproton annihilation in different materials provides a useful check of the ability of standard Monte Carlo packages to reproduce fragment multiplicities, type and energy distributions stemming from antiproton (or antineutron) annihilations on nuclei at rest. Although measurements of the multiplicities of pions and other charged particles with energies higher than ∼ 50 MeV are available in the literature [8,9], the production of highly ionizing nuclear fragments with short range has not been studied sufficiently. A measurement of the multiplicities of charged products in antiprotonaluminium annihilations was reported in [3], although only 43 events were analysed in this study, and the tracking efficiency of the detector was limited to 80%. In this paper, we present the results of a study of the multiplicities of charged annihilation products on different target materials, namely copper, silver and gold, using emulsion detectors at the Antiproton Decelerator (AD [10]) at CERN.

Experimental setup
Emulsion detectors were used to study the antiproton annihilation products generated in different materials. Before reaching the targets, the 5.3 MeV antiprotons from the AD (3×10 7 p/shot every 100 s) -1 -  were slowed down using several different titanium and aluminium foils with variable thicknesses. Finally, the beam was collimated in a vacuum test chamber after crossing a titanium vacuum separation window with a thickness of 12 µm. The emulsion detector was situated at the downstream end of the vacuum chamber (∼1 m in length), where it could be reached by a defocused beam of low-energy antiprotons (∼100 keV). This distance from the degrading layer was necessary to reduce the background due to annihilations taking place at the moderator. A sketch of the experimental assembly is shown in figure 1. The emulsion detector was operated under ordinary vacuum conditions (10 −5 − 10 −6 mbar). The antiproton intensity measured by the detector was approximately 150/cm 2 per shot.
For this study emulsion detectors were produced at the Laboratory for High Energy Physics (LHEP) of the University of Bern by pouring the emulsion gel with a thickness of ∼100 µm, provided by Nagoya University (Japan), on a glass plate (for a review on the emulsion technology see [6]). Glycerin was added to so that the emulsion could operate in vacuum [3]. This emulsion features a very low background with approximately 1-2 thermally induced grains per 1000 µm 3 [5].
Foils of copper, silver and gold, each having a thickness of 10 µm, were placed as targets at the end of the vacuum chamber, in front of the emulsion detectors. Figure 2 shows the targets (2 × 2 cm 2 each) fixed to the emulsion film and an example of antiproton annihilating on the emulsion surface. At the antiproton energies obtained after degrading, all annihilations are expected to take place within a few µm of the target surface. During data taking, we collected approximately 1500 antiprotons per cm 2 in about 10 AD shots.

Data analysis and results
Data recorded by the emulsion detectors were scanned by an automatic optical microscope and then analysed by exploiting a recently developed fast 3D tracking algorithm [11]. The measured tracking efficiency of our detector was approximately 99% for minimally ionizing particles over a wide angular range, as reported in [11]. Figure 3 shows the profile of detected tracks (xy positions of tracks at the emulsion layer) in one of the assemblies. Among the reconstructed tracks we only considered those that were longer than 30 µm to avoid considering tracks that were due to the background. An angular cut of 0.4<tanθ<2.0 (22 • <θ<63 • ) was applied, where θ is the track angle with respect to the beam direction. To reconstruct a vertex, at least two three-dimensional tracks were required. The efficiency of vertex reconstruction was estimated by applying the criteria given above to the output of the FLUKA simulation. It was found to be 22% for copper, 24% for silver, and 18% for gold. Figure 4 (left) shows the distribution of the reconstructed vertex position perpendicular to the film surface (z-direction) in a subarea of the copper target. The peak in z is a measure of the target foil position (the gap between the emulsion surface and the target foil). This measurement was performed by segmenting the analysed area into smaller areas since the target foils were neither flat nor in contact with the film surface. The surface topography obtained from the reconstructed vertices is shown in figure 5 for copper, silver and gold targets. In our analysis, we only considered regions of the emulsion film surface with the z value smaller than 100 µm because the vertex reconstruction efficiency was uniform within a few percent in this region. The analysed fiducial area was 1.68 cm 2 for the copper target, 1.96 cm 2 for silver and 0.80 cm 2 for gold. The fraction of signal vertices became dominant by requiring vertex reconstruction at the position of the target foil.
The nearly flat distribution in figure 4 (right) is due to combinatorial background in the vertex reconstruction. The main source of the background was due to accidental combinations of tracks  coming from annihilations taking place upstream in the apparatus, which were not completely excluded by the angular cut due to the broad angular distribution. The fraction of tracks from signal vertices to all detected tracks was estimated using the vertex finding efficiency described above and found to be 9% for copper, 9% for silver, and 6% for gold. The number of background vertices was estimated using all the detected tracks in the analysed area by subtracting the above signal track fraction, randomizing positions and slopes of the remaining tracks, reconstructing the vertices and counting the number of vertices that mimicked annihilations in the target. The background estimated from measured data, which depends on the number of tracks in the event, is shown in the left panes of figure 6, while the right panes show the multiplicity distributions after subtraction of the background, compared with the Monte Carlo predictions based on the CHIPS [12][13][14][15] and FTFP (FTFP_BERT_TRV) [16,17] models in the GEANT4 (4.9.5.p02) and FLUKA (2011.2c) [18,19] frameworks. A total of 654 signal annihilation vertices were reconstructed for copper, 941 for silver and 233 for gold. Gold.   -6 -We were also able to discriminate between heavily ionizing particles (HIPs) such as protons and nuclear fragments and minimally ionizing particles (MIPs), namely pions. Continuous dense tracks correspond to HIPs, while faint tracks are produced by MIPs, since the aligned grains of these last tracks are separated. The local energy deposition (dE/dx) of each track can then be assessed in terms of signal density (S.D.) along the reconstructed tracks, using Here, x, y and z are the coordinates of voxels in the 3D image data. C is a group of voxels in a cylinder along the track, and S xyz is an 8-bit grey-scale signal of the voxel. L is the length of a reconstructed track in the 3D image data. The S.D. is proportional to the dE/dx of the particle and does not depend on the angle. However, there is a saturation effect for higher values of dE/dx. Figure 7 shows that the S.D. distribution of tracks revealed a peak at 3000 for MIPs. As the simulated dE/dx distribution of MIPs peaked at 1.2 MeV·g·cm −2 , we define particles with dE/dx smaller than 2.4 MeV·g·cm −2 , corresponding to an S.D. below 6000 µm −1 , as being MIPs. The complementary particles are defined as HIPs. Figure 8 shows the track multiplicity distributions for MIPs and HIPs. The errors on the data are statistical. The histograms represent the Monte Carlo predictions by CHIPS, FTFP and FLUKA. The error bars of the Monte Carlo predictions account for uncertainties in the dE/dx classification. These uncertainties were estimated using simulations, which smeared the threshold for assigning tracks to either class of ionizing particles by 20% and checked for effects on the multiplicity distributions for MIPs and HIPs. The statistical errors for the simulations were 0.01-0.02, which are significantly smaller than the errors reported above. The average measured multiplicities are summarized in table 1. Both CHIPS and FLUKA are in good agreement for copper, particularly in the case of MIPs. Neither CHIPS nor FTFP accurately describe particle multiplicities for annihilations on silver and gold nuclei, while FLUKA more closely reproduce the data than the other models.
The mean values of particle multiplicities measured for the three target materials are shown in figure 9 as a function of atomic number along with the simulation outcome. Results obtained for MIPs with the FTFP model do not agree with our experimental data for any material, while those obtained with both CHIPS and FLUKA are in fair agreement as far as copper is concerned, although only FLUKA reproduces the higher atomic number behavior. Good agreement with CHIPS was also found for annihilation on bare emulsions and for aluminium [3]. Multiplicities related to HIPs are well described by the FLUKA simulation, while the CHIPS and FTFP models clearly underestimate the number of particles produced by antiproton annihilation.

Conclusions
The goal of the study presented in this paper was to measure the products of low-energy antiproton annihilation on different materials, utilizing a secondary beam line of CERN's Antiproton Decelerator in the AEgIS experimental area. The characteristics (e.g. hadronization and fragmentation multiplicities) of low-energy antiprotons annihilating on nuclei are not well known, and experimental data are needed to validate models used by simulation packages such as GEANT4 and FLUKA. We exposed several thin targets (Cu, Ag and Au) to the antiproton beam and measured fragment tracks using emulsion detectors with a vertex position resolution at the level of a few micrometers, which allowed the separation between minimally and highly ionizing particles. The fragment multiplicities we measured were not well reproduced by the different models used in Monte Carlo simulation with the exception of FLUKA, which is in good agreement with the particle multiplicities for both minimally and heavily ionizing particles. Future measurements with more materials are needed to gain a better understanding of antinucleon annihilations also on low-Z materials, and to obtain a full description in terms of particle types, multiplicities, as well as energy, for a more complete benchmarking of Monte Carlo simulations.