Physics Procedia 62 ( 2015 ) 118 – 123

Available online at www.sciencedirect.com

1875-3892 © 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license

(http://creativecommons.org/licenses/by-nc-nd/4.0/).

Peer-review under responsibility of the ENEA Fusion Technical Unit

doi: 10.1016/j.phpro.2015.02.022

ScienceDirect

3rd International Conference Frontiers in Diagnostic Technologies, ICFDT3 2013

Performance of a Medium-Size Area nGEM Detector for Neutron

Beam Diagnostics

G. Crocia,b, , C. Cazzaniga

b,f, G. Albani

f, A. Muraro

a, G. Claps

c, M. Cavenago

d,

G. Grossoa, F. Murtas

c, R. Pasqualotto

e, E. Perelli Cippo

a, M. Rebai

f,b,

M. Tardocchia, G. Gorini

f,b

a IFP-CNR, Milano, Italy

b Sez. INFN Milano-Bicocca, Milano, Italy

c INFN-LNF, Frascati, Italy

d INFN-LNL, Legnaro, Italy

e RFX Consortium, Padova, Italy

f University of Milano-Bicocca, Milano, Italy

Corresponding author: croci@ifp.cnr.it

Abstract

Fast neutron detectors with a sub-centimetric space resolution are required in order to qualify neutron beams in

applications related to magnetically-controlled nuclear fusion plasmas and to spallation sources. Based on the results

obtained with small area prototypes, the first medium-size (20 x 35.2 cm2 active area) nGEM detector has been

realized for both the CNESM diagnostic system of the SPIDER NBI prototype for ITER and as a beam monitor for

fast neutrons beam lines at spallation sources, too. The nGEM is a Triple GEM gaseous detector equipped with

polyethylene layers used to convert fast neutrons into recoil protons through the elastic scattering process. This

paper describes the performance of the medium-size nGEM detector tested at the VESUVIO beam line of the

ISIS spallation source. Being this detector the actual largest area fast neutron detector based on the GEM technology,

particular attention was paid in the study of detector response in different points over the active area. Measurements of

GEM counting rate (both as a function of VGEM and of time) and of the capability of the detector to reconstruct the beam

in different positions are presented. This detector serves as a basis for the realization of an even larger area detector that

will be used in the MITICA NBI prototype for ITER that represents the evolution of SPIDER.

© 2014 The Authors. Published by Elsevier B.V.

Peer-review under responsibility of the ENEA Fusion Technical Unit.

Keywords: GEM (Gas Electron Multiplier), Fast neutrons, Medium-size area

© 2015 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).Peer-review under responsibility of the ENEA Fusion Technical Unit

G. Croci et al. / Physics Procedia 62 ( 2015 ) 118 – 123 119

1. Introduction

Gas Electron Multiplier (GEM) based detectors [3-7] were invented at CERN as charged particle tracking detectors

[8] but, if properly adapted, they can be used also as neutral particles detectors [6,7,9,10]. The advantage of this

technology is the good space resolution (for this detector class it can be better than 0.5 mm [11]) combined with a large

active area, that finds applications for neutron detection of fusion relevant experiments and spallation sources. In

particular, the CNESM (Close Contact Neutron Emission Surface Mapping) diagnostic system of SPIDER [1], the

prototype of the Neutral Beam Injector (NBI) for ITER, the next generation tokamak, has been designed based on

neutron GEM (nGEM) detectors [1].

Fig. 1. Assembly phases of the medium-size nGEM prototype; Top: HV test of GEM foil (a), cathode stretching (b) and GEM foil stretching (c);

Bottom: Padded readout assembly (d) and the complete detector (e).

A similar system will be used as a fast neutron beam > monitor for both the CHIPir beam line at ISIS-RAL (UK) and

the future fast neutron beam lines at the European Spallation Source (ESS). In these experiments fast neutrons are used

to study single event effects on irradiated electronic chips [2].

This paper describes the construction and test of the first full size (20 x 35.2 cm2 active area) nGEM fast neutron

detector for SPIDER. This detector has a medium size area compared to other triple GEM detectors [12] but nevertheless

it represents at the moment the largest area fast neutron detector based on Triple GEM amplification. All the tests were

performed at the ISIS-RAL (UK) VESUVIO electron Volt neutron beam line [2].

2. Full-size prototype description

Following the results obtained with the small area prototypes [13,14], the first full-size nGEM prototype was

built in May 2013. This detector was built following the construction method developed and used for the Triple

GEM of the LHCb experiment at CERN [15]. Figure 1 shows different phases of the detector assembly: the electrical

test of the GEM foils, the stretching of both GEM foils and cathode and the gluing of the padded readout-anode.

The detector has an active area of 32.5 x 20 cm2 and its cathode is composed of two layers: one 150 µm thick

polyethylene (CH2) film and one 50 µm thick Aluminum layer. The detector gap configuration (drift/transfer1/transfer2/induction gap) was 4/2/2/2 mm. The padded read-out anode is composed of 256 pads, each with an area of 22 x 13 mm

2.

The front-end chip used to read all the pads are the CARIOCA-GEM digital chips [16]. A system of patch panel

and cables carry the LVDS signals to a custom made FPGA Mother Board [17] located 1 m away from the beam

120 G. Croci et al. / Physics Procedia 62 ( 2015 ) 118 – 123

halo, that analyzes the LVDS signals coming from the chips. This configuration gives the possibility not to irradiate

the FPGA chip with neutrons, reducing the rate of soft errors; it will be used in SPIDER, too. The gas mixture

employed in all the measurements is Ar/CO2 70%/30%. A picture (rear view) of the detector installed in the

VESUVIO beam line is shown in Fig. 2. The high voltage con-figuration was generated using the HVGEM [17] NIM

module and the potentials were applied to each electrode by means of passive resistive-capacitive filters properly

designed for a Triple GEM detector.

Fig. 2. Front view of the nGEM detector installed on the Vesuvio beam line mounted on the positioner system. The alignment with the

beam was performed using the shown laser pointer (shown by P in the picture).

3. nGEM performances

3.1. The ISIS VESUVIO Facility

The measurements were performed with the medium-size nGEM placed in the neutron beam of the VESUVIO [2]

beam line with a flight path of about L=12.5 m from the target. At ISIS, neutrons are produced by a 800 MeV

proton beam with a double bunch fine structure and a repetition frequency of 50 Hz. The two proton bunches have

a duration of about 70 ns (FWHM) and are separated by 322 ns. The proton beam delivers an average current of

180 Ah on a Ta-W target yielding about 30 neutrons per incident proton. In the energy range En > 1 MeV the

neutron spectrum has approximately a 1/En behaviour.

Fig. 3. nGEM counting rate as a function of VGEM when the neutron beam is on (neutrons) and off (gamma background)

G. Croci et al. / Physics Procedia 62 ( 2015 ) 118 – 123 121

3.2. nGEM detector counting efficiency

The nGEM detector efficiency for different kind of particles was measured as a function of the effective gain by varying the sum of potential difference over the three GEM foils (Σ∆VGE M = VGE M ). Two different measurements were performed: VGE M scans with the neutron beam on and off. The former is a measurement of the neutron

detection efficiency while the latter is an indicator of sensitivity to gamma background. Fig. 3 shows the result of the measurements. The obtained result confirms what has been measured with the small area prototypes [14]: since by applying a the counting rate with the beam off is 400 times lower, it is possible to state that photons background is

completely rejected if VGE M < 900. Therefore the working point of this detector is at VGEM = 870 V.

3.3. Vesuvio beam profile measurements

Figure 4 shows the measured 2D map of the VESUVIO neutron beam profile obtained by exposing the detector

to the neutron beam. The first reconstruction of the beam width using pads of this dimensions (area 22 (x) x 13 (y) mm2

) gives σx = 14.28 mm and σy = 11.26 mm. These values are compatible both with the technical specification of ISIS

and with previous results [14].

Fig. 4. 2D map of the Vesuvio fast neutron beam fitted by a bi-gaussian function. The reconstructed σx and σy are respectively 14.28 mm and

11.26 mm . The detector parameters used are Ed (Drift Field) = 2.25 kV/cm; ET1 (Transfer 1 Field) = 1.5 kV/cm ET2 (Transfer 2 Field) = 3

kV/cm, EInd (Induction Field) = 1.5 kV/cm and VGEM = 870 V.

3.4. nGEM counting rate vs time

Figure 5 shows the measured detector counting rate as function of time during 2.5 h of irradiation and compares it to the ISIS proton current: it is possible to appreciate that the detector is able to follow the variation of the beam current. The measured statistical variation of the counting rate with time is about 7%.

Fig. 5. nGEM counting rate (bottom) compared to ISIS proton current: the detector is able to follow current variation.

122 G. Croci et al. / Physics Procedia 62 ( 2015 ) 118 – 123

3.5. Beam Position reconstruction

In order to determine the detector response over the whole active area, the detector was positioned on a X-Y

movable system so that it can be neutron-irradiated in different positions (see Figure 2). The detector was aligned to

the neutron beam by using a laser. A number of beam profiles like the one presented in Figure 4 were obtained. The

beam position was reconstructed using the so called barycentre (xb and yb ) of the counts calculated as reported in

formula (1) where xi and yi are the pad coordinates and Ci are the total counts registered by a pad in a single run.

(1)

The barycenter calculated in this way represents the centre of the reconstructed beam position considering the

counts distribution over the pads on the whole detector active area. Figure 6 shows the relation between the expected

and barycenter reconstructed beam centre both for x and y: since this relation is linear and the difference between

the nominal position and the barycenter is lower than twice the dimension of the pads (as expected rom the fact that

the read-out is discretized and not continuous), it is possible to state that the detector is able to reconstruct the beam

position in different points over the active area. The dispersion on horizontal axes for both graphs is an indication of

the uncertainty of the “real” position of the beam center. The detector resolution can be estimated to be of the order

of the pad dimension.

Fig. 6. Top: x-barycenter coordinate reconstruction and comparison with expected beam position. The origin of the coordinate system is the

one described in Figure 5. Bottom: same as top for y-barycenter coordinate. Different colours correspond to the same x (Top Graph) or y (Bottom

Graph) but to different y (Top Graph) or x (Bottom Graph).

xb =

∑256

i=1xi · Ci

∑256

i=1Ci

yb =

∑256

i=1yi · Ci

∑256

i=1Ci

G. Croci et al. / Physics Procedia 62 ( 2015 ) 118 – 123 123

4. Conclusions

The first medium-size nGEM prototype has been realized and successfully tested at ISIS and its performances are

similar to those of small area detectors. This detector represents the first full-size prototype for SPIDER and serves

as a basis for the realization of an even larger area detector that will be used in the MITICA NBI prototype for

ITER that represents the evolution of SPIDER.

Acknowledgements

This work was set up in collaboration and financial support of F4E, of INFN-Group 5 (Technology Research)

and of CNR. This work was also supported within the CNR-CCLRC Agreement No. 01/9001 concerning

collaboration in scientific research at the spallation neutron source ISIS.

References

[1] G.Croci et Al, nGEM neutron diagnostic concept for high power deuterium beams 2012 JINST 7 C03010

[2] C. Andreani et Al, Appl. Phys. Lett. 92, 11401 (2008)

[3] F. Sauli, GEM: A new concept for electron amplification in gas detectors, Nucl. Instr. and Meth. A 386 (1997) 531

[4] S. Duarte Pinto et Al, Progress on Large area GEMs, 2009 JINST 4 P12009

[5] M. Villa et Al, Progress on Large area GEMs, Nucl. Instr and Meth. A 628 (2011), 182-186

[6] M. Alexeev, et Al, Micropattern gaseous photon detectors for Cherenkov imaging counters, Nucl. Instr. and Meth A 623,(2010), 129-131

[7] M. Alexeev et Al, Development of THGEM-based photon detectors for Cherenkov Imaging Counters, 2010 JINST 5 P03009

[8] M. Alfonsi et Al, Activity of CERN and LNF groups on large area GEM detectors, Nucl. Instr. and Meth. A 617, (2010), 151-154,

[9] M. Alexeev et Al, THGEM based photon detector for Cherenkov imaging applications, Nucl Instr. and Meth A 617, (2010) 396-397

[10] G. Croci et Al, “GEM-Based thermal neutron beam monitors for spallation sources”, Nucl. Instr. and Meth A, 732, (2013), 217-220

[11] A Bressan et Al, Nucl. Instr. and Meth, 425 (1999), 262-276

[12] M. Abbaneo et Al, “An overview on the design, construction and performance of large area triple- GEM prototypes for future upgrades of the CMS forward muon system”, 2012 JINST 7 C05008

[13] F. Murtas, G. Croci et Al, Triple GEM gas detectors as real time fast neutron beam monitors for spallation neutron sources. 2012 JINST 7 P07021.

[14] G. Croci et Al, “nGEM fast neutron detectors for beam diagnostics”, Nuclear Instruments 22and Methods in Physics Research Section A, Volume 720, 21 August 2013, Pages 144-148

[15] G Bencivenni et Al, A triple GEM detector with pad readout for high rate charged particle triggering, Nucl. Instr. and Meth. A 488, (2002), 493-502

[16] W. Bonivento et Al, Development of the CARIOCA front-end chip for the LHCb muon detector, Nucl Instr. and Meth. A: 491, (2002), 233-243

[17] http://www.infn.it/csn5/joomla/GEMINI/