G. Croci 1,2, C. Cazzaniga 3, G. Claps 4, M. Cavenago 5, G. Grosso 1, F. Murtas 4,6, S. Puddu 6, A. Muraro 1, E. Perelli Cippo 1, M. Rebai 2,3, R. Pasqualotto 7, M. Tardocchi 1 and G. Gorini 2,3 Development of GEM-based neutron beam monitors 1 Istituto di Fisica del Plasma, IFP-CNR - Milano (IT) 2 INFN, Sezione di Milano-Bicocca (IT) 3 Dipartimento di Fisica, Università di Milano-Bicocca (IT) 4 INFN – LNF - Frascati (IT) 5 INFN – LNL - Legnaro(IT) 6 CERN – Geneva (CH) 7 Consorzio RFX – Padova (IT)

OUTLINE Why and how to use GEM-based detectors to detect neutrons FAST NEUTRON DETECTORS Mainframe projects Prototypes construction Performances on neutron beams Large area detector (35 x 20 cm 2 ) THERMAL NEUTRON DETECTORS Mainframe projects Prototypes construction Performances on neutron beams Conclusions and Future Perspectives 2

WHY AND HOW TO USE GEMS TO DETECT NEUTRONS GEMs offer the following advantages Very high rate capability (MHz/mm 2 ) suitable for high flux neutron beams like at ESS Submillimetric space resolution (suited to experiment requirements) Time resolution from 5 ns (gas mixture dependent) Possibility to be realized in large areas and in different shapes Radiation hardness Low sensitivity to gamma rays (with appropriate gain) GEM detectors born for tracking and triggering applications (detection of charged particles) In order to detect neutral particles you need a converter Fast Neutrons: Polyethylene converter + Aluminium Neutrons are converted in protons through elastic scattering on hydrogen Thermal Neutrons: 10 Boron converter Neutrons are detected using the productus (alpha,Li) from nuclear reaction 10 B(n,alpha)7Li 3

FAST NEUTRON BEAM MONITORS Details about triple GEM detector, HV-GEM Power Supply, CARIOCA chips and FPGA-Board have been already shown by G. Claps talk 4

Complete GEM detector system HVGEM HV Filters 3 GEM detector with padded anode 3 GEM detector with padded anode FPGA Board LNF 128 ch FPGA Board LNF 128 ch DAQ PC 12 V PS Charged particles X Ray GammasNeutrons Current Monitor 2D monitor with pads readout Possibility to set time slices from 5 ns up to 1 s 5

Mainframe Projects CNSEM (Close Contact Neutron Surface Emission Mapping) diagnostic for ITER NBI Prototypes (SPIDER & MITICA) Beam monitor for ISIS and ESS E d =100keV nGEM neutron Detector Aim: Reconstruct Deuterium beam profile from neutron beam profile. Angular resolution and directionality property needed ChipIr CAD model at ISIS-TS2 ESS Model Aim: Construct large area, real-time and high rate beam monitors for fast neutron lines Deuterium Beam (100 Kev) Neutron Flux n/cm 2 s Deuterium Beam composition: 5x16 beamlets See G.Gorini Talk 6

nGEM (fast neutrons GEM) prototypes 1 «Analogue» Prototype (nGEM-S-1) 100 cm 2 active area Cathode: Aluminium (40 μm) + Polyethylene (60 μm) 2 Small area Digital Prototypes (10x10 cm 2 – nGEM-S-2/3) nGEM-S-2 Cathode: Aluminium (40 μm) + Polyethylene (60 μm) Gas Ar/CO 2 & Ar/CO 2 /CF 4 nGEM-S-3 (same cathode as full size prototype) Cathode: Aluminium (50 μm) + Polyethylene (100 μm) 1 Full-Size SPIDER prototype (nGEM-FS-1) Cathode: Aluminium (50 μm) + Polyethylene (100 μm) 20 x 35 cm 2 active area 4 Prototypes of nGEM have been built and tested so far with Gas Mixture Ar/CO 2 & Ar/CO 2 /CF 4 7

Neutron Facilities Directionality Property  nGEM-S-1 (Analogue) High Voltage Scan (efficiency scan)  All prototypes Linearity w.r.t neutron flux  nGEM-S-2 Beam Profile Measurements  All Digital prototypes Gamma Background sensitivity  All prototypes Fast neutron time-line (ISIS beam time profile reconstruction)  nGEM-S-2 Counting stability  All digital prototypes Imaging  nGEM-S-2/3 FNG Enea Frascati (Italy) 2.5 MeV neutrons 14 Mev neutrons Max Flux: n/s (14 MeV) 10 9 n/s (2.5 MeV) ISIS – Rutherford Appleton Laboratory Didcot (Uk) Spectrum from Thermal to 800 MeV Flux: Thermal (<100 meV): 7*10 5 Fast (> 1MeV): 6*10 5 n/cm 2 s nTOF – CERN Geneva (Ch) Spectrum from a few meV to several GeV Flux 10 5 n/cm 2 /pulse 8

2.5 MeV neutron Test at FNG (Frascati Neutron Generator – ENEA) Deuterium beam Deuterium target nGEM detector Analog Prototype nGEM-S-2 See P. Valente Talk 9

Directionality Property 10 Neutron Flux ≃ 10 8 n/cm 2 s (measured by in-site NE213 scintillator). The optimized aluminium thickness that allows to discard protons emitted at an angle > 45°is 40 μm (determined by MCNP Simulations) Each pulse height spectrum was normalized considering the total number of neutrons generated by the neutron gun measured by the NE213 scintillator. n p p Al gas CH 2 n pp G. Croci et Al, JINST C Results confirm that nGEM is fully able to discard protons emitted at θ>45°. 10

Neutron flux Linearity nGEM-S-2 Very important feature for a beam monitor Neutron Flux up to 10 8 n/cm 2 /s Counts over the full area scales linearly with neutron flux Efficiency 2.5 MeV) = 2*10 -5 ΔV GEM = 1020 V 2.5 MeV neutrons (Ar/CO2/CF4 gas mixture) Detector working point and gamma rays background rejection Counting rate Vs chamber gain: up to 890 V the chamber is sensitive to fast neutron but not to gamma rays (Ar/Co2 70%/30% gas mixture) ISIS FNG 11

Real-time 2D beam map measurements Monitor for a fast neutron beam with energies ranging from a few meV to 800 MeV Tested at neutron beam of the Vesuvio facility at RAL- ISIS 2D Beam profiles and intensity in real time Neutron beam monitorig during the shutter opening nGEM-S-2 12

Vesuvio Beam 2D Measurement Y direction cut X direction cut 2D Fast Neutron Intensity Map FWHM = 34 mm FWHM = 36 mm G. Croci et Al, NIM A 720, OFFLINE Analysis 13

Detector Counting Rate Stability in time Counting stability Neutron flux = 10 5 /n/cm2 nGEM counting rate exactly follows the ISIS beam Measured% of counting rate variation with time = 4.7 % Stability is a very important feature for a beam monitor G. Croci et Al, NIM A 720,

Fast Neutron time line Rate measurement scan on time delay from beam T 0 using GEM detector with 100 ns gate. Comparison with proton ISIS current impinging on the target (double structure)  nGEM is able to see the double proton structure E n >2MeVE n <2MeV G. Croci et Al, JINST P

Delay 2000 ns, HV 870 V, gate 10 ms Two different intensity beams arrive to the facility Instantaneous counts Cumulative counts 16 y y x x Constant:515 ± 15 s Mean1: 6.0 ± 1.8 cm Mean2: 5.4 ± 1.7 cm Constant: 3640 ± 40 Mean1: 5.9 ± 1.8 cm Mean2: 5.5 ± 1.8 cm nTOF Online 2D Beam Measurement

17 The FPGA can detect neutrons vs a delay in time allowing to make a time (i.e. Neutron energy) scan that allows the efficiency vs energy to be measured (uncertainty ~1% ). 100 keV 2 e-4 3 MeV 10 MeV Scan in energy at nTOF

First nGEM full size prototype for SPIDER GEM Foil HV TestCathode Stretching and Framing GEM Stretching and Framing 35 cm 20 cm Assembly 256 Pads At the moment it is the largest area GEM- based fast neutron detector!!!! 18

First ISIS Vesuvio In this case the MBFPGA is put outside of the neutron beam using flat LVDS cables to carry the signal out. This decreases neutron induced soft errors in the FPGA. Using this setup the prototype run for several days without any inconvenience 19

Preliminary results ISIS beam 2D profile normalized to current: 12x22 mm 2 pad area; half detector shown Data analysis in progress 20

THERMAL NEUTRON BEAM MONITORS (for ISIS and ESS) 21

n α Low efficiency detector  1% is sufficient since the neutron flux is very high (>10 6 n/cm 2 s) bGEM prototype of thermal neutron beam monitor Triple GEM detector equipped with an aluminum cathode coated with 1μm of B 4 C: first bGEM prototype Exploit the 10 B(n,α) 7 Li reaction in order to detect thermal neutrons B 4 C coated aluminium cathode mounted on its support B 4 C coated aluminium cathode assembled inside the bGEM chamber layout Detector Schematics 22

e CN CH N CH 3 N2N2 CH 2 e Atoms, Radicals Molecules, Ions and Electrons powered electrode grounded electrode Time-average voltage profile across electrodes in rf discharge Plasma deposition area B 4 C target RF plasma sputtering system for B 4 C coating at IFP-CNR (Milano,Italy) Gas Injection Courtesy of E. Vassallo (IFP-CNR ) 23

Thermal neutron measurements as a function of detector gain (wp and γ-rejec) at ISIS-Vesuvio A wide plateau is present for 820 V <V GEM < 910 V  This confirms that the detector is revealing all the alpha particles emitted from the 10 B(n,α) 7 Li reaction. The detector reached its maximum efficiency. The detector is gamma-background free with V GEM < 900 V At V GEM =870V corresponding to a GEM effective gain of 100, the measured efficiency is about (0.95±0.08)%, very similar to the expected one 0.86% G. Croci et Al, NIMA (2013), In Press bGEM 24

Measurement of ISIS-vesuvio 2D thermal neutron beam profile The measured FWHM is around 3 cm compatible with ISIS-Vesuvio data 3 cm G. Croci et Al, NIMA (2013), In Press 25

Detector Counting Rate Stability in time bGEM counting rate exactly follows the ISIS beam Measured% of counting rate variation with time = 3.5 % Stability is a very important feature for a beam monitor G. Croci et Al, NIMA (2013), In Press 26

Thermal neutrons time of flight spectrum The TOF spectrum was measured using the multi-gate property of FPGA- MB. The spectrum is compatible with TOF spectrum measured by standard Vesuvio beam monitors G. Croci et Al, NIMA (2013), In Press Performed using the FPGA-MB 27

Conclusions GEM-based fast and thermal neutron beam monitors have been successfully realized and tested. They provide: Real-time neutron beam profile with a portable system (HV System + CARIOCAS & MBFPGA LNF) Measurements with the necessary space resolution (pad dimension) Time resolution of 100 ns for fast neutron measurements TOF thermal neutron spectrum Complete Gamma ray background rejection Stability in time First «large area» detector built and results are under study 28

Future Perspectives A new larger area nGEM neutron detector for MITICA (the evolution of SPIDER) is under design and will be developed next year A new high efficiency (>50%) thermal neutrons GEM- based detector - based on a 3D cathode of thin lamellas - for future spallation neutron sources has been designed and is currently been built. Results will be presented in the next months. This detector can represent a valid alternative to 3 He detectors We are working on a new GEMINI chip which will be able to increase the number of channels. The new chip can manage 32 channels, in comparison to the 8 channels of CARIOCA. This new GEMINI chip will be used to upgrade all these detectors 29

Relationship with the industry HVGEM : MPElettronica – Rome (Italy) CARIOCA Chips: Artel SRL – Florence (Italy) MB-FPGA: Athenatek – Rome (Italy) GEM FRAMES: Meroni & Longoni – Milan (Italy) GEM Foils: CERN Detector construction: LNF-INFN (Frascati) and IFP-CNR (Milano) 30

Spare Slides 31

Filters in the beam line: effect on nGEM counting rate MaterialCountrate (Hz) %Expected if fast neutrons (6 MeV) Expected if thermal neutrons Expected if gamma rays No Material // Lead (5 cm) %15 %7.3 % Cadmium (1 mm) %0%0%97% Polyethylene (15 cm) %9%0%29% Aluminium (2.5 cm) %79%75% Lead: the observed decrease is compatible with the hypothesis that the fast neutron beam is scattered by the lead block and that the detector is non sensitive to gammas Cd: the observed decrease is compatible with the thesis that we are not detecting thermal neutrons CH 2 : the observed decrease is compatible with the fact that we are detecting fast neutrons G. Croci et Al, NIM A 720,

Material Filters in the beam (Imaging with bGEM) CH 2 L- Shaped Cd Neutrons are scattered Neutrons are absorbed 33