A 200 cm × 50 cm MRPC - based prototype for the NeuLAND detector at R 3 B ∗

A detector for high-resolution momentum measurements of neutrons in the energy range 0.2-1.0 GeV is being developed for the RB experiment at FAIR. Two solutions are currently being studied: A pure scintillator concept and an approach based on a sequence of converter material (iron) to produce secondary charged particles, and Multigap Resistive Plate Chambers (MRPC’s) to detect these particles. Here, work on the latter solution is reported. In recent years, a number of 40 cm × 20 cm MRPCbased NeuLAND prototypes been built at FZD, GSI, and SINP. Tests with the single-electron per bunch mode of the superconducting electron linac ELBE had shown that most of the prototypes fulfilled the design criterion of σt < 100 ps time resolution and η ≥ 90% efficiency, when using single-ended readout with FOPI front-end electronics and a 25 ps TDC. Recent experiments at ELBE have shown that the same is true also for differential readout, using the new PADI-3 front-end electronics. What was still missing, however, were simulations reproducing the test beam data and predicting the behavior for high-energy neutrons, and tests with a full-size prototype. Now, in simulations within the R3BRoot framework [1], a digitizer has been developed following the avalanche caused by each 31 MeV electron. The electric-fielddependent drift velocity, Townsend and attachment coefficients [2] have been modeled. The induced charge on the readout electrode is then propagated to the front-end electronics at the end of the strip. The remaining free parameters are the growth cut parameter for the space charge effect and the correlation distance parameter for several avalanches. By adjusting them, the measured efficiency curve for the 40 cm × 20 cm prototypes was reproduced (fig. 1a). These parameters were then adopted in the simulation to predict the behavior of a full MRPC-based NeuLAND detector. It shows nearly full efficiency for 400 MeV neutrons, as required. However, in the timing spectrum, in addition to a narrow peak near the ”true” time-of-flight given by the first interaction of the neutron, there is an extended tail towards later times (fig. 1b). This is due to neutrons which scatter inside the inactive converter material but do not produce a detectable signal in the MRPC structure immediately behind it. As a result, approximately 40% of the incident neutrons are not detected within the required timing window. Further simulations are in progress to explore whether this feature can be improved.