Development of a low-pressure Multi-Mesh THGEM detector for fission experiments at FRIB

A small-area imaging detector prototype for position measurement of fission fragments produced in low energy heavy-ion reactions is presented in this study. The detector readout is equipped with a 2-dimensional, position-sensitive gaseous avalanche readout based on the novel Multi-Mesh Thick-GEM (MM-THGEM), and is coupled to a delay-line board for particle localization. The prototype has an effective area of 10 × 10 cm2 and is operated in isobutane at low pressure (7–10 Torr). We present and discuss a series of systematic evaluation tests performed by irradiating the detector with α-particles and fission fragments emitted by small-rate sources (241Am, 249Cf and 252Cf). Position and time resolutions of about 0.42 mm and 1 ns were achieved, respectively. This work serves as a benchmark for the development of a large-scale array of detectors for experiments with fission and fission-like reaction products at the Facility for Rare Isotope Beams (FRIB).


Introduction
The study of fission and fission-like reactions that follow a fusion process near the Coulomb barrier provides critical insights into the mechanisms and dynamics involved in the synthesis of heavy and super-heavy elements (SHE) [1].The path to the synthesis of SHE is strongly hindered by fast fission-like processes that proceed without forming an equilibrated compound nucleus.Since the fusion cross-section is enhanced by the doubly magic nature of the 48 Ca isotope, the heaviest new elements synthesised so far have been formed by 48 Ca reactions on actinide targets.Further experimental progresses involved the use of stable heavier beams ( 48 Ti, 64 Ni) that revealed a decreasing trend of the complete fusion cross-section for the heavier projectiles [2].Advances towards neutron-rich radioactive beams (RIBs) of high intensities achievable at FRIB may lead to a breakthrough in the understanding of the mechanisms governing these fusion and fission-like processes.This is an important step in determining if successful searches for the next SHE can be undertaken with RIBs.
The separation of complete and incomplete fusion can be achieved in heavy-ion reactions by studying the properties of fission-like fragments, such as mass, energy, and angular correlations, which are derived from measurements of the velocity vectors of the binary fragments.Several detector technologies and methods can be used to measure the Time of Flight (ToF), velocity, and/or energy of fission reaction products.For instance, Multi-Wire Proportional Counters (MWPCs) have demonstrated high-gain stable operations at low pressure [3][4][5], with excellent position ( < 0.5 mm) [6] and time resolution ( < 500 ps) [7], for small/moderate detection areas.However, MWPCs also present intrinsic limitations, such as moderate spatial resolution for large detector areas (> 100 cm 2 ) and a significantly lower rate capability (up to 10 4 Hz/mm 2 ) compared to more advanced gaseous detector technologies; e.g., Micro-Pattern Gaseous Detectors (MPGDs) achieve two orders of magnitude higher rates (up to 10 6 Hz/mm 2 ) [8].
An alternative method for heavy-ion tracking is based on a position-sensitive Multi-Channel Plate (MCP), used to detect secondary electrons emitted when fast ions impinge on a thin electrode foil.The device is operated in vacuum, and thus problems associated with detector gas leaking into the environment and producing beam-induced background reactions, energy losses, and multiple -1 -scattering effects are minimized.MCP based detectors provide extremely good performance in terms of position and time resolution, down to 100 μm and 100 ps (), respectively [9].However, only small effective areas of a few cm 2 can be produced, and their sensitivity is limited.An MCP detector for time of flight can be coupled to Si-based detectors, in a configuration that provides simultaneously both ToF and energy measurements, but at a relative limited solid angle [10].
To reconstruct the kinematics of heavy-ion induced fusion-fission binary reactions, the Coincident Fission Fragment Detector (CFFD) system was developed at the National Superconductive Cyclotron Laboratory (NSCL).The operation and technical development of the CFFD continue at the Facility for Rare Isotope Beams (FRIBs).The detector system was designed for operation at FRIB/NSCL re-accelerator facilities ReA3 and ReA6 [11], which provide a unique set of isotope beams, produced by stopping RIBs in-flight and re-accelerating them in a compact linac.The initial design of the CFFD included an array of four position-sensitive Parrallel Plate Avalanche Counters (PPACs) with 30 × 40 cm 2 active area [12].The PPACs were equipped with a resistive-chain readout system capable of position resolution of around 1.75 mm (), and time resolution of 400 ps ().The large effective areas allow for large solid angle coverage, to increase the overall effective detection efficiency.
In order to improve the CFFD's performance in terms of angular and mass resolution, alternative detector technologies were explored for future applications at FRIB.In this work, we present and discuss the performance evaluation of a small-scale prototype based on a novel MPGDs, the Multi-Mesh THGEM (MM-THGEM) [13].The detector was operated in low-pressure isobutane (7-10 Torr).The MM-THGEM was coupled to a simple and economic delay-line readout for localization capability.The detector design and recent results on operation and performance are presented and discussed.A conclusion is provided regarding the evaluation of the prototype and considerations related to the implementation of MM-THGEM-based readout to large-area systems.

The MM-THGEM imaging detector prototype
The MM-THGEM-based imaging detector prototype has an effective area of 10 × 10 cm 2 and consists of several elements in cascade -see figure 1.The electron avalanche amplification is provided by a MM-THGEM directly in contact with a resistive anode.Beneath the resistive anode, a position-sensitive striped readout board, equipped with a delay-line, is used for encoding particle localization.The MM-THGEM is a novel MPGD consisting of two electrode meshes embedded within a three-layer THick Gaseous Electron Multiplier.The THGEM-like structure serves as mechanical -2 -support for the two meshes.The detector drift region, 4.2 mm wide, is defined by the MM-THGEM copper-clad top surface and an upper cathode foil.The latter comprises of a 50 μm polypropylene film coated with a 150 nm thick aluminum layer.The bottom surface of the MM-THGEM is free of copper plating and is placed directly in contact with the resistive anode in the so-called well-configuration.
The holes in the MM-THGEM electrode are displaced in a hexagonal pattern characterized by a pitch between holes of 0.7 mm.The holes have a diameter of 0.3 mm and a rim of 0.1 mm.The gaps between the upper and lower meshes, as well as the gap between the lower mesh and the resistive anode, are both 0.2 mm in width.The gap between the MM-THGEM top surface and the upper mesh is 0.4 mm.Thus, the MM-THGEM structure has a total thickness of 0.8 mm.
The resistive anode consists of a 3 mm thick Printed Circuit Board (PCB) coated with a thin layer of diamond-like carbon (DLC) film.The DLC layer is characterized by a surface resistivity of about 10 MΩ/square.The resistive anode is than placed on top of the position-sensitive striped readout.
The resistive anode technique enables the signals induced on the readout to be spread over a geometrical area larger than the effective size of the gas avalanche, leading to an optimal match with the strips pitch.Furthermore, it allows for galvanic decoupling between the multiplication stage and the readout electrode board.As a result, the resistive anode can operate at high voltage, while the readout board is maintained at ground potential.The low coupling capacitance between the resistive anode and readout board also protects the readout electronics from spurious energetic discharges.
All the elements are installed on four threaded rods at the corner of the effective area, while Teflon spacers are used to maintain the distance between electrodes -see figure 2. The electrode bias voltages of all detector electrodes were supplied and controlled remotely by high voltage power supplies (ISEG model EHS), through dedicated low-pass filters to reduce electronic noise.The segmented readout (see figure 3) consists of a series of square pads printed on both sides of the PCB, with a conceptual design similar to the one described in [14].The pads are connected by strips that run in orthogonal directions (X and Y) on both sides of the board, with a pitch of 1.5 mm.The square pads on the surface attached to the resistive anode board have an area smaller -3 - than the pads on the opposite surface to preserve equal induced charges on both coordinates.The pad areas are 0.3 mm 2 and 1 mm 2 , respectively.The delay-line was designed to have a total delay time of 150 ns and an impedance of 100 Ω. Custom-made two-stage pre-amplifiers process the signal from the four terminations of the delay-line.The low-power first stage pre-amplifiers are mounted on the PCB of the segmented readout board (see figure 3), very close to the delay-line terminations to reduce pickup noise.The second-stage pre-amplifiers are kept in air outside the detector vessel.The two stages are connected together with a twisted pair flat ribbon cable, that provides also power supply to the fist stage.Both the first-and second-stage pre-amplifiers circuits are equipped with a first-order low-pass filter with a 3 dB cutoff at about 25 MHz.-5 -

JINST 19 P05023
The position of the impinging particle is thus determined by the time difference between the time stamp of the signals recorded at the extremities of the two delay-lines, according to the following equations: where    and    are time-to-distance conversion factors,   offset and   offset are offset terms for recentering the image that take into account parasitic time delays, while

Measurements and results
In order to determine the performance of the small-area prototype in terms of gas gain, spatial/time resolutions and image quality, we have carried out a series of measurements by irradiating the detector with -particles and fission fragments, emitted by either a 241 Am, a 249 Cf or a 252 Cf source.The detector was operated in isobutane at a pressure range 7-10 Torr.The vessel was evacuated at a pressure of about 10 −5 Torr prior to the gas filling.During operation, the isobutane gas was continuously flushed through the vessel at a rate of about 5 sccm.The evacuation, the gas filling and the control of the gas flow during operation and measurement was performed through a dedicated gas handling system.

Gain measurement
Measurements of the effective gain of the MM-THGEM detector prototype, assessed for different bias configurations in isobutane at 10 Torr, are illustrated in figures 7. The gas gain is presented as a function of the reduced electric fields applied to the second multiplication stage (V/cm/Torr).In all the measurements, electric field ratio between the first and second multiplication regions was kept at a value equal to 0.6 in order to preserve a sufficient electron transparency across the meshes.The different plots in the same graph correspond to a fixed voltage applied to the MM-THGEM top electrode (yellow, blue and red graphs at 200, 250, and 300 volts, respectively).The detector was irradiated with -particles emitted by a collimated 241 Am source, and the drift field in the region above the MM-THGEM was kept at 0.6 kV/cm.Primary ionization electrons released along the -particle tracks in the drift gap migrate towards the MM-THGEM multiplier and are converged into the multiplier holes.The first stage of the MM-THGEM (collection region), namely the region between the top electrode and the first mesh, is primarily responsible for efficient collection of primary electrons into the MM-THGEM hole.A minor portion of the avalanche occurs between the two meshes (first multiplication region), while the majority occurs during the last multiplication stage (second multiplication region), between the second mesh and the resistive readout board.At around 10 Torr, the estimated amount of primaries released by the -particles in the drift region (4 mm gap) was slightly above 500 electrons.
Maximum achievable gains above 10 4 are observed for all the detector configurations.A small contribution to the total gas gain is also possible within the collection region for a higher electric field strength (see the progressive increase in gain in figure 7).Similar measurements were performed also at 7 Torr, with maximum achievable gain similar to the one obtained at 10 Torr (approaching a -6 - value of 10 5 ), but at slightly higher reduced electric fields.A more detailed study on the optimization of the MM-THGEM operation is described in [13].
Figure 7 also shows the operational conditions used to perform the imaging measurement with fission fragments (marked by the dotted lines).Note that, the MM-THGEM amplification region voltages applied for the measurement of fission fragments were significantly lower than those applied at the maximum achievable gains for  particles, due to the high ionization density of fission fragments.

Image quality
In order to establish the imaging performance of the detector system, an image of a coded-aperture brass mask with a pattern of holes of varying dimensions and pitches was analyzed (see figure 8(a)).The mask was installed inside the vessel, on top of the drift region, at a short distance (1 cm) from the cathode foil.The detector was operated at 7 Torr and exposed to about 5.5 MeV -particles from the 241 Am source over the entire effective area.
The edge-spread function analysis was used to estimate the position resolution.A portion of the calibration mask image corresponding to the rectangular aperture at the center of the mask (region within the red contour in figure 8(b)) was selected, and its projections along the coordinates was obtained figure 8(c).Assuming that the mask has sharp edges, and neglecting possible parallax effects, the profile of the aperture is essentially the response of the imaging system to a line source (line spread function -LSF).The point spread function (PSF), which defines the position resolution of the imaging system, corresponding to the response of the detector to a point source in the image plane, can be mathematically expressed as a derivative of the LSF [16].In an ideal imaging system, the LSF would be a step function and its derivative would result in two delta functions: a positive delta function at the rising edge and a negative one at the lowering edge.The PSF of a real detector, -7 - however, results in a Gaussian distribution with a non-zero spread, which represents the spatial resolution of the imaging system.As shown in figure 8(d), the average position resolution obtained from the prototype along the Y-coordinate is about 0.42 mm ().We obtained a comparable result (slightly above 0.4 mm ) for the X-coordinate.
Similar outcomes have been achieved from the analysis of the image of the holes in the central pattern of the mask (pinhole image in the red contour of figure 9).The image of the pinhole (1 mm in diameter) recorded by the imaging system arises from the convolution of the geometric projection of the extended -particle source (approximately 1 cm in diameter) through the mask hole in the detector plane and the intrinsic spatial resolution of the detector.The size of the geometrical projection on the detector plane depends on the distance between the source and the pinhole with respect to the detector plane (see red square in figure 9).The position resolution ( PSF ) can be explicitly calculated from the following expression: where  measured is the spread of the pinhole projection extracted from the image of the calibration mask (lower panel of figure 9, red circles), whereas  simulated is the size of the geometrical projection of the source that appeared through the pinhole onto the detector plane; this quantity was computed with Monte Carlo methods (lower panel of figure 9, black circles).The average  PSF calculated from the analysis of different pinholes of the mask image in figure 9 is approximately 0.41 mm, in agreement with the results obtained by edge spread function analysis.
-8 - The uniformity of the response was computed from the projection of a series of selected pinhole images along the X-coordinate (see figure 10(a)).The projection of the holes results in a sequence of peaks (see figure 10(b)) whose centroids are measured and compared to the corresponding actual position of the holes on the mask.The Integral Non-Linearity (INL) is defined as the deviation of the measured centroids from a best-fit straight line (see figure 10(c)).On average, the INL was below 0.1%.
An empty-field image obtained from the MM-THGEM detector by triggering only on fission fragments emitted by the 252 Cf source is shown in figure 11(a).The detector was operated in isobutante at 10 Torr. Figure 11(b) illustrates the projection of the number of events recorded by the MM-THGEM imaging system, projected along the Y-coordinate.The empty-field image exhibits a relatively flat profile with minor fluctuations along the coordinate axis, indicating an excellent detector efficiency and high homogeneity of images, obtained even with a relatively low gas gain (∼ 700).

Time resolution
The time resolution was measured using the setup illustrated in figure 12.The signals recorded in the MM-THGEM detector were triggered by -particles emitted by a 249 Cf source in coincidence with prompt -ray emissions.The latter were detected using a CsF scintillator assembly; the scintillation light was recorded using a Photo-Multiplier Tube (PMT), attached to the scintillator.
Signals collected from the second MM-THGEM mesh were processed by a charge-sensitive pre-amplifier and a Tennelec linear amplifier (model TC-171).The fast output of the TC-171 was -9 -    Converter (Ortec model 567) was utilized to measure the time difference between the processed signals from the CsF scintillator (start input) and the MM-THGEM (stop input).The resulting time distribution is shown in figure 13.The contribution of the CsF scintillator detector to the accuracy of time measurement is expected to be significantly smaller (about 0.1 ns) [17] than the contribution of the MM-THGEM detector, so it can be neglected.As a result, the spread of the time distribution shown in figure 13 is equivalent to the time resolution provided by the MM-THGEM based detector, corresponding to about 1 ns.

Discussion
In fission studies, the mass of the fragments emitted during the fission process, or equivalently the mass ratio of the two binary fragments produced in a single fission event, is the most critical parameters to be measured, typically correlated to the total kinetic energy or angular distribution.The mass  of the -11 -detected fission fragment can be roughly extracted from its velocity  following these simple expression: where  is the flight path, calculated from the position of arrival on the delay-line readout of the MM-THGEM detector, ToF is the measured time of flight and the kinetic energy  is calculated based on the Viola systematics [18].In a fission event, -rays are promptly emitted at the source position, so the time of flight of a fission fragment in our experimental arrangement (described in section 3.3) can be computed from the start signal recorded by the CsF scintillator detector and the stop signal provided by the MM-THGEM imaging system.Figure 14 shows the extracted mass distribution of the fission fragments emitted by the 252 Cf from the measured time of flight and position with the MM-THGEM prototype in coincidence with the CsF detector (open circles).The reconstructed mass distribution provided by he MM-THGEM prototype is compared to the data measured with an MCP-based detector system (see ref. [19]).The resolutions of both position and time have a crucial impact on the accuracy of extracting these physical quantities relevant to fission fragment experiments.As a result, both time and localization capability must be considered when evaluating the performance of the fission fragment detector system.In addition, practical and technical properties such as cost, simplicity, and mechanical stability need to be taken into account.
Table 1 shows the expected mass resolution provided by different detector systems (data taken from literature) compared to the MM-THGEM detector described in this work.A typical case of fragment masses around A = 120 has been assumed, e.g.produced in a fission process for nuclei in the actinide region.The mass resolution has been computed taking into account both the time and position resolution of the experimental systems for a fligth path of 18 cm.Based on table 1, the MM-THGEM detector offers a competitive mass resolution, although it is slightly inferior to that of other detector technologies.Nevertheless, it should be noted that the mass resolution of the detected primary fragments in a fission event is also limited by the uncertainty regarding the number of neutrons emitted either by the fissioning nucleus at the scission point or by the fragments themselves.The flight path of the fragments is not significantly affected by neutron emission.In light of this, it is expected that even if an ideal detector were to be used, a resolution of only 2-3 atomic mass units can be achieved when reconstructing the primary fragment masses from the measured velocities, since on average 4-6 neutrons are emitted following the fission of a nucleus generated by a heavy ion reaction [22].

Conclusions
We have designed and built a small-scale imaging detector prototype based on a the novel Multi-Mesh Thick GEM structure (MM-THGEM) for detecting fission fragments produced in low-energy heavy-ion reactions.The prototype has an effective area of 10 × 10 cm 2 , and provided a high-gain stable operation in low-pressure (< 10 Torr) isobutane.The detector was coupled to a delay-line based readout for encoding particle localization.
We have investigated the performance of the MM-THGEM detector prototype in terms of position and time resolution using small-rate -particle and fission sources ( 241 Am, 249 Cf, 252 Cf).Position and time resolutions of < 0.5 mm () and 1 ns () satisfy the requirements for a precise reconstruction of the velocity vectors of the fission fragments, from which the mass and total kinetic energy distributions of binary heavy-ion reactions can be extracted.This study opens the door to the development of a large-scale array of detectors, equipped with the gaseous avalanche and readout technology used in the present work.The system will be used to carry out experiments at the Facility for Rare Isotope Beams (FRIB) focused on fission and fission-like reaction products with low-energy rare isotope beams to better understand the production of super-heavy isotopes and improve cross section systematic.

Figure 2 .
Figure 2. Photograph of the components of the MM-THGEM imaging detector without the cathode foil installed on top.

Figure 3 .
Figure 3. Photograph of the striped PCB board with delay-line for encoding particle localization.

Figure 4
Figure 4  illustrates typical fast signals, from  particles emitted by a249 Cf source, that have been recorded at the terminations of the delay-line along the X-coordinate (yellow trace), processed by the two-stage pre-amplifiers.On the Y-coordinate, similar signal waveforms are recorded (blue trace).The signals exhibit a rise time of approximately 50 ns.Figure5illustrates a typical example of the coincident -particle signals measured on the delay-line (yellow trace) and the lower mesh of the MM-THGEM detector (purple trace), as well as the prompt -ray signal (blue trace) recorded by a Cesium Floride (CsF) scintillator detector.More details on the coincident measurement are given in the following sections.An overview of the Data Acquisition System (DAQ) used to process delay-line signals for particle localization is shown in figure6.The output signals from the second-stage pre-amplifiers are fed into Tennelec TC241s [15] Time Filter Amplifiers (TFA) to improve the signal-to-noise ratio through pulse shaping.The TFA outputs are then routed to a Constant Fraction Discriminator (CFD, Mesytec model MCFD-16) that generates digital time stamps for all input signals above a user-defined threshold.Finally, a Time-to-Digital Converter (TDC, Mesytec model MTDC-32) is used to generate accurate time stamps for all delay-line signals, from which the particle localization is extracted.The TDC has a 16-picosecond time resolution (RMS).

Figure 4 .
Figure 4. Typical signals recorded at the terminations of the delay-lines, processed by the two-stage pre-amplifier.The yellow and the blue traces correspond to the X and Y coordinate delay-lines, respectively.

Figure 5 .
Figure 5. Yellow trace: signal recorded at one of the termination of the delay-line.The signal was processed by the two-stage fast pre-amplifier.Blue trace: fast signal used as time reference recorded by a CsF -ray detector.Purple trace: signal recorded at the MM-THGEM lower mesh, processed by a charge sensitive pre-amplifier.

Figure 6 .
Figure 6.Schematic diagram of the DAQ used to process the delay-line signals and extract particle localization.
the arrival time of the signals on the X-coordinate (Y-coordinate).

Figure 7 .
Figure 7.The effective gain measured at 10 Torr in isobutane by irradiating the full detector area with 5.5 MeV -particle.

Figure 8 .
Figure 8. Calculation of the spatial resolution of the detector prototype by edge spread function analysis, obtained from the image (a) of a calibration mask (b).The red box in the two-dimensional histogram indicates the gate used for the one-dimensional projection.The profile of the gate along the Y coordinate (c) for which the PSFs are calculated (d).

Figure 9 .
Figure 9. (a) Schematic description of the source position with respect to the mask placed in front of the prototype detector.(b) Calculation of the spatial resolution of the detector prototype was performed by analysis of a mask pinhole image denoted by the red box indicating the gate used for the one-dimensional projection of the two-dimensional histogram.(c) The spatial resolution is extracted from the convolution of the measured pinhole image and the projection of the -particle source through the mask on the detector plane.

Figure 10 .
Figure 10.The calculation of the INL was computed by processing the position of the mask pinhole images along the coordinated ((a) and (b)) and compared the calculated centroids with the actual position on the mask (c).The red box in the two-dimensional histogram indicates the gate used for the one-dimensional projection.

Figure 11 .
Figure 11.(a) Image of an empty-field recorded with the MM-THGEM imaging system irradiated with fission fragments emitted by a 252 Cf source.(b) Projection of the empty-field image along the Y-coordinate.

Figure 12 .
Figure 12.Setup used to measure the time properties of the MM-THGEM based detector prototype.The -particles emitted by the 249 Cf source are recorded by the MM-THGEM prototype in coincidence with the -rays detected by a CsF scintillator detector.

Figure 13 .
Figure 13.Time distribution of event resulted from the coincidence measurement of -particles (MM-THGEM) and prompt gamma (CsF scintillator), emitted by the 249 Cf source.Neglecting the CsF contribution, the time resolution of the MM-THGEM detector is estimated as the spread of the time distribution (approximately 1 ns ()).

Figure 14 .
Figure 14.The derived mass distribution (open circles) of the fission fragments emitted in the spontaneous fission of 252 Cf from the time and position information measured by the MM-THGEM based detector prototype in coincidence with a CsF -ray detector, compared to the data measured in ref. [19] (red dashed line).Reprinted from [19], Copyright (2006), with permission from Elsevier.

Table 1 .
Comparison of the expected fission fragment mass resolution computed from position and time resolution for different detector systems.