- Email: [email protected]
Photoelectric X-ray Polarimetry with Gas Pixel Detectors
Nuclear Instruments and Methods in Physics Research A 720 (2013) 173–177
Contents lists available at SciVerse ScienceDirect
Nuclear Instruments and Methods in Physics Research A journal homepage: www.elsevier.com/locate/nima
Photoelectric X-ray Polarimetry with Gas Pixel Detectors Ronaldo Bellazzini a, Alessandro Brez a, Enrico Costa b, Massimo Minuti a, Fabio Muleri b, Michele Pinchera a, Alda Rubini b, Paolo Soffitta b, Gloria Spandre a,n a b
INFN Pisa Largo B. Pontecorvo 3, I-56127 Pisa, Italy INAF/IASF Rome Via del Fosso del Cavaliere 100, I-00133 Rome, Italy
a r t i c l e i n f o
abstract
Available online 16 December 2012
The Gas Pixel Detector, recently developed and continuously improved by Pisa-INFN in collaboration with IASF-Roma (INAF), can visualize the tracks produced within a low Z gas by photoelectrons of few keV. By reconstructing the impact point and the original direction of the photoelectrons, the GPD can measure the polarization plane of X-Ray photons, while preserving the information on the absorption point, the energy and the time of arrival of individual photons. Applied to X-ray Astrophysics, in the focus of grazing incidence telescopes, it can perform angular and energy resolved polarimetry with a large improvement of sensitivity, when compared with the conventional techniques of Bragg diffraction at 451 and Compton scattering around 901. This configuration has been the basis of POLARIX and HXMT, two pathfinder missions, and was included in the baseline design of IXO, the very large X-ray telescope under study by NASA, ESA and JAXA. We have recently improved the design of this low energy polarimeter (2–10 keV) by modifying the geometry of the absorption cell to minimize any systematic effect that could leave a residual polarization signal for nonpolarized source. We will report on the testing of this new concept. & 2012 Elsevier B.V. All rights reserved.
Keywords: X-ray Polarimetry Gas Pixel Detectors ASIC
1. Introduction An extensive theoretical literature predicts linearly polarized emission from most classes of X-ray sources. Yet the only successful detection has been done more than 30 years ago from the Crab Nebula, one of the brightest sources in the X-ray sky, while many of the predictions refer to much fainter sources. X-ray polarimetry can not only enlighten the nature and geometry of the emission, but can also test some quantum gravity theories on vacuum birefringence based on the Lorentz invariance violation. Unfortunately this observational technique has remained largely unexploited due to the insufficient sensitivity of the conventional astronomical polarimeters based on Bragg diffraction and Compton scattering. At X-ray energies below a few tens of keV, a very efficient mechanism on which to base sensitive X-ray polarimeters, is the photoelectric effect. A measurement of the photoelectron emission direction provides a measure of the direction of the electric field vector of the incident X-ray, namely the polarization state of the photon. The challenge is to reconstruct the very short initial part of the electron track (only some hundreds mm), before the diffusion has smeared the charge pattern, blurring in part the directional information.
n
Corresponding author. Tel.: þ39 050 2214382; fax: þ39 050 2214317. E-mail address: [email protected] (G. Spandre).
0168-9002/$ - see front matter & 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.nima.2012.12.006
In the last decade, the advances in the development of micropattern gas detectors, a very finely spaced arrays of proportional counters, have made photoelectric X-ray polarimetry a real possibility and renewed the interest of this observational strategy for scientific discovery [1–3].
2. The Gas Pixel Detector as X-ray Polarimeter The Gas Pixel Detector (GPD) developed, and continuously improved, in Italy by INFN of Pisa in collaboration with IASF/INAF of Rome [1,3,4,5] belongs to the new generation of Micro-Pattern Gas Detectors. The instrument measures the polarization of the X-ray photons by imaging the tracks of the photoelectrons created in the gas cell. A given X-ray’s polarization state is determined by the initial direction of the resulting photoelectron. When a low energy linearly polarized photon is absorbed in a low density medium, the photoelectron is emitted preferentially in the plane orthogonal to the X-ray propagation (Fig. 1) with a cos2 f modulation, where f is the azimuthal angle between the direction of emission of the photoelectron and the electric field of the X-ray (polarization vector). In a simplified quantum-mechanical treatment [6] in fact, the differential cross-section of the photoelectron emission, in the non-relativistic regime, is given by: 2 7=2 pffiffiffi 2 @s Z5 mc 4 2sin ðyÞcos2 ðjÞ ¼ r 20 4 @O hv ð1b cosðyÞÞ4 137
174
R. Bellazzini et al. / Nuclear Instruments and Methods in Physics Research A 720 (2013) 173–177
Fig. 2. Schematics of the GPD assembly.
Fig. 1. Photoelectron interaction: when incoming photons are linearly polarized, the photoelectrons are ejected preferentially in the plane orthogonal to the photon direction propagation (i.e. y 901) with a cos2 angular distribution in f).
where Z is the atomic number of the photo-absorbing atom, r0 the classical electron radius, b the photoelectron velocity given in fraction of c, y the angle between the direction of the incoming photon and the ejected electron and f the azimuthal angle defined above. The photoelectron emission is mainly at 901 with respect to the incoming photon direction. The term in the denominator accounts for a slight bending forward of the emission angle distribution with the photoelectron energy. For example, at 8 keV the photoelectrons are ejected preferentially at y ¼751. The modulation factor is a key parameter for the sensitivity of a polarimeter. It represents the response of the instrument to a 100% polarized photon beam [7]. A larger value of the modulation factor implies lower statistical fluctuations and hence the capability to detect weaker polarized signals. Schematically (see Fig. 2), a GPD is made by a gas volume enclosed by a top Berillium window, a Gas Electron Multiplier (GEM) which amplifies the charge of the electron tracks generated in the drift gap providing energy and time information and a pixellated charge collection plane directly connected to the pixellated analog read-out electronics (CMOS-VLSI). 2.1. The GPD components: GEM and VLSI The GEM consists of a 50 mm-thick polyimide sheet, sandwiched by two 5 mm-thick copper layers through which densely packed holes are drilled using the CO2 laser etching technique. Holes pitch is 50 mm. This type of GEM, designed at RIKEN [8] and fabricated by Scienergy Co. Ltd. [9], does not show any rate dependent gain instability due to the good quality of the cylindrical hole shape [10]. The GEM is the two-dimensional amplification structure of the photoelectron track. The primary ionization electrons of the track are drifted by the applied electric field ( 1 kV/cm) toward the GEM, multiplied in the high electric field region (100 kV/cm) inside the GEM holes and then collected by the pixellated read-out electrode underneath (the fine structured top metal layer of the VLSI chip).
The GEM has worked with two gas mixtures, He and Ar based (reaching large effective gain, well above 1000) at voltage difference of 450–550 V through the GEM. Typical high voltage settings with Ar 70% DME 30% gas filling are: VDRIFT¼ 3000 V, VGEMtop ¼ 950 V, VGEMbottom¼ 400 V; with He 20% DME 80%: VDRIFT¼ 2650 V, VGEMtop¼ 865 V, VGEMbottom¼ 400 V, with the read-out electrode at the charge preamplifier input voltage, i.e. 1.5 V. Fig. 3 shows the custom VLSI chip mounted on its control motherboard. The chip belongs to a third generation of ASICs of increased complexity, developed by the Pisa-INFN group in 0.18 mm CMOS technology. The chip top metal layer has been patterned with 105,600 pixels, hexagonally shaped, arranged in a 300 352 honeycomb matrix corresponding to a 15 15 mm2 active area and a pixel density of 470 pixels/mm2 (the pixel pitch being 50 mm and the fill fraction, metal/active area, 95%). A fully analog electronic chain (low noise charge preamplifier, shaping amplifier, peak&hold and input to a multiplexer stage) has been implemented beneath each pixel, in five metal and one single poly-silicon layers. The chip integrates more than 16.5 million transistors. It is subdivided into 16 identical clusters of 6600 pixels (22 rows of 300 pixels) or alternatively in 8 clusters of 13,200 pixels (44 rows) each one with an independent differential analog output buffer. The chip has customable self-trigger capability and includes a signal pre-processing function for the automatic localization of the event coordinates. Groups of 4 contiguous pixels (mini-cluster) contribute to the local trigger that has a dedicated amplifier whose shaping time is roughly a factor 2 faster ( 1.5 ms) than the shaping time of the analog signal to allow its peak detection and hold. Each noisy pixel can be masked with respect to its contribution to the trigger signal and/or its analog output. By limiting the output signal to only those pixels belonging to the region of interest, it is possible to reduce significantly the read-out time and data volume. Table 1 summarizes the main characteristics of the chip. To generate and handle the command signals to/from the chip a very compact data acquisition system has been implemented on the Altera FPGA Cyclone EPIC240 (RISC NIOS-II processor). Analog data are read, digitally converted by a Flash ADC and provisionally stored in a static RAM. In self-trigger mode just after an event acquisition it is possible to read the pixel pedestals in the triggerdefined ROI, one or more times, compute the average value and transfer the pedestal subtracted data to the offline analysis system.
R. Bellazzini et al. / Nuclear Instruments and Methods in Physics Research A 720 (2013) 173–177
The great advantage of this operation is a real time control of the data quality and the removal of drift effects of the pedestal values with temperature or any other environmental changes. 2.2. The GPD geometry The GPD is characterized by an active gas volume defined by the active area ( 2 cm2) of the charge collecting electrode (the pixellated ASIC top metal layer) and an absorption gap thickness (drift region) of 1–2 cm. In order to make more uniform the drift electric field within the absorption cell and to minimize any systematic effect that could leave a residual polarization signal for nonpolarized source, the transverse GEM dimensions are increased by several centimeters with respect to the collecting plane. The thickness of the gas gap between the GEM bottom and the ASIC is 0.5 mm. The parallelism between the GEM plane and the ASIC plane is a critical item. This requirement is satisfied if the GEM bottom side is in full contact with the package top surface that nominally has a misalignment with respect to the ASIC plane o50 mm.
175
A photo of the sealed detector fabricated in Finland in collaboration with Oxford Instruments Analytical Oy, a small company with long lasting experience in space qualified instruments, is shown in Fig. 4. The detector is manufactured with materials and components with low out-gassing rate (alumina for the spacers, kapton, and ceramic for the feedthroughs) and it is capable to withstand long-term operations in the space environment. 2.3. The photoelectron track reconstruction The photoelectron track is identified by the projected charge distribution on the pixel plane. The analysis of the track moments allows assessing some properties of this distribution such as: the barycentre, the major axis and most importantly the projection on the pixel plane of the point of absorption of the X-ray photon (see Fig. 5) used to derive the direction of ejection of the photoelectron. The amplitude and direction of the source polarization is obtained from the modulated distribution of the emission angles.
3. The results with two GPD configurations The gas mixture filling the GPD determines the working energy band of a photoelectric X-ray Polarimeter. In the 2–10 keV energy range, mixtures based on He and DME as absorber and quencher are the most suitable. In our tests the best performances in the low energy band, have obtained with 20% He–80% DME at 1-atm pressure. At higher energy, Ar based mixtures are the most effective. In the next paragraph we report the results obtained for the two baseline gas mixtures proposed respectively for the low and medium energy polarimeters at the focus of the HXMT telescope [11,12]. 3.1. He–DME (20–80, 1-atm pressure)
Fig. 3. The VLSI ASIC on its motherboard.
To test the performances of the GPD as X-ray Polarimeter in the low energy band, 95% polarized beams have been produced in the IASF-INAF laboratory in Rome [5]. The use of a mosaic graphite and flat aluminum crystals has allowed obtaining polarized photons at 2.6, 3.7, and 5.2 keV from the diffraction of unpolarized continuum or line emission. These sources have been used to measure the modulation factor of the GPD filled with the 20% He–80% DME gas mixture. As shown in Fig. 6, the modulation factor measured at these energies has resulted in good agreement with the Monte Carlo simulations. Fig. 7 represents the angular distribution that has obtained with polarized photons of 3.7 keV. A cos2 fit of the distribution gives a modulation factor of 43.10%70.84%.
Table 1 Main ASIC characteristics. Peaking time (externally adjustable) Full scale linear range Pixel noise Read-out mode Trigger mode Read-out clock Self-trigger global threshold Frame rate Parallel analog output buffer Access to pixel counter Total power dissipation Fill fraction (metal/active area)
3–10 ms 30,000 electrons 50 electrons ENC Asynchronous and synchronous Internal, external, self-trigger Up to 10 MHz 2000 electrons Up to 10 kHz (event window self trigger) 1, 8 or 16 Direct (single pixel) or serial (8–16 clusters, full matrix, region of interest) 0.5 W (self-trigger mode) 92%
176
R. Bellazzini et al. / Nuclear Instruments and Methods in Physics Research A 720 (2013) 173–177
Fig. 6. Comparison between the modulation factor expected from the Monte Carlo simulations (crosses) and the measured values at 2.6, 3.7, and 5.2 keV points with the error bars. It is assumed that the degree of polarization at 2.6 and 5.2 keV is 0.95. The dashed line is the fit to Monte Carlo values.
Fig. 4. The sealed Gas Pixel Detector. The Berillium window is capped with a titanium frame.
Fig. 7. Modulation of the reconstructed angles at 3.7 keV in He–DME (20–80) at 1atm pressure.
Fig. 5. A photoelectron track produced in He–DME mixture by a 5.9 keV photon.
A good energy resolution at 3.7 keV photon energy is obtained in the He–DME mixture, as shown in the pulse height distribution of photoelectrons in Fig.8. 3.2. Ar–DME (70–30, 2-atm pressure) To extend the polarimetric performances in the medium energy band, 6–30 keV, gas fillings with higher atomic number have to be used. Simulation results suggest argon based mixtures. A GPD prototype with a 2 cm Ar–DME (70–30) cell at 2 atm working pressure, to improve the detection efficiency, has been successfully tested. In this configuration the detector can perform
Fig. 8. Pulse height distribution of 3.7 keV photoelectrons in He–DME.
energy resolved imaging polarimetry of hard X-ray sources with high sensitivity. A measurement of the energy resolution of the GPD at 5.9 keV is shown in Fig. 9.
R. Bellazzini et al. / Nuclear Instruments and Methods in Physics Research A 720 (2013) 173–177
177
and sets the limit to the Minimum Detectable Polarization. The measured value of the residual modulation is 0.18% 70.14% (statistical error).
4. Conclusions
Fig. 9. Pulse height distribution obtained with a totally unpolarized 5.9 kev photons in a Ar 70%–DME 30% gas mixture at 2 atm pressure. The typical Ar escape peak is clearly visible This peak is due to the emission, occurring in about 15% of the photoelectric absorptions, of a fluorescence photon with energy just below the Ar K edge. This secondary photon has a very long mean free path for absorption and therefore escape the detector volume.
The Gas Pixel Detector is one of the most advanced instruments to image the photoelectrons tracks in a gas, both for performance and readiness level. The choice of a suitable gas filling allows measuring the linear polarization in the energy range 2–30 keV, and reconstructing the impact point, the energy and the time of each event. The possibility to join polarimetric information to timing, spectral and imaging capability, is a unique characteristics of this type of detector and it is particularly interesting in X-ray Astrophysics. A large literature, still without a significant experimental feedback, describes polarimetry as a fundamental tool to distinguish among competitive models, often equivalent on the basis of spectral or timing information. The GPD provides very concrete possibilities to perform polarimetric measurements at the level of accuracy of 1% and below on board of small pathfinder missions, like POLARIX or HXMT. The GPD is also included in the baseline design of the International X-ray Observatory that will definitely elevate the X-ray polarimetry to a key tool for a deeper understanding of many different classes of astrophysical objects. References [1] [2] [3] [4] [5] [6] [7] [8]
[9] Fig. 10. Residual modulation obtained with totally unpolarized 5.9 kev photons.
Fig. 10 shows the angular distribution for a totally unpolarized beam of 5.9 keV photons. A fit to the distribution allows estimating the residual modulation of the device due to systematic errors
[10] [11] [12]
E. Costa, et al., Nature 411 (2001) 662. J.K. Black, et al., Nuclear Instruments and Methods A 581 (2007) 755. R. Bellazzini, et al., Nuclear Instruments and Methods A 566 (2006) 552. R. Bellazzini, et al., Nuclear Instruments and Methods A 579 (2007) 853. F. Muleri, et al., Nuclear Instruments and Methods A 584 (2008) 149. W. Heitler, The Quantum Theory of Radiation, Oxford University Press, London, 1970. L. Pacciani, et al., Proceedings of SPIE 4843 (2006) 394. RIKEN is a Japanese Institution whose mission is to conduct comprehensive research in science and technology in a wide range of fields, including physics, chemistry, medical science, biology, and engineering, covering the entire range from basic research to practical application. SCIENERGY Co., Ltd 1-9-15-640 Kanumadai, Chuo-ku, Sagamihara, Kanagawa 252-0233, Japan /http://www.scienergy.jpS. T. Tamagawa, et al., Nuclear Instruments and Methods A 560 (2006) 418. E. Costa, et al., Proceedings of SPIE 6686 (2007) 66860Z. P. Soffitta, et al., Proceedings of SPIE 7011 (2008) 701128.
Contents lists available at SciVerse ScienceDirect
Nuclear Instruments and Methods in Physics Research A journal homepage: www.elsevier.com/locate/nima
Photoelectric X-ray Polarimetry with Gas Pixel Detectors Ronaldo Bellazzini a, Alessandro Brez a, Enrico Costa b, Massimo Minuti a, Fabio Muleri b, Michele Pinchera a, Alda Rubini b, Paolo Soffitta b, Gloria Spandre a,n a b
INFN Pisa Largo B. Pontecorvo 3, I-56127 Pisa, Italy INAF/IASF Rome Via del Fosso del Cavaliere 100, I-00133 Rome, Italy
a r t i c l e i n f o
abstract
Available online 16 December 2012
The Gas Pixel Detector, recently developed and continuously improved by Pisa-INFN in collaboration with IASF-Roma (INAF), can visualize the tracks produced within a low Z gas by photoelectrons of few keV. By reconstructing the impact point and the original direction of the photoelectrons, the GPD can measure the polarization plane of X-Ray photons, while preserving the information on the absorption point, the energy and the time of arrival of individual photons. Applied to X-ray Astrophysics, in the focus of grazing incidence telescopes, it can perform angular and energy resolved polarimetry with a large improvement of sensitivity, when compared with the conventional techniques of Bragg diffraction at 451 and Compton scattering around 901. This configuration has been the basis of POLARIX and HXMT, two pathfinder missions, and was included in the baseline design of IXO, the very large X-ray telescope under study by NASA, ESA and JAXA. We have recently improved the design of this low energy polarimeter (2–10 keV) by modifying the geometry of the absorption cell to minimize any systematic effect that could leave a residual polarization signal for nonpolarized source. We will report on the testing of this new concept. & 2012 Elsevier B.V. All rights reserved.
Keywords: X-ray Polarimetry Gas Pixel Detectors ASIC
1. Introduction An extensive theoretical literature predicts linearly polarized emission from most classes of X-ray sources. Yet the only successful detection has been done more than 30 years ago from the Crab Nebula, one of the brightest sources in the X-ray sky, while many of the predictions refer to much fainter sources. X-ray polarimetry can not only enlighten the nature and geometry of the emission, but can also test some quantum gravity theories on vacuum birefringence based on the Lorentz invariance violation. Unfortunately this observational technique has remained largely unexploited due to the insufficient sensitivity of the conventional astronomical polarimeters based on Bragg diffraction and Compton scattering. At X-ray energies below a few tens of keV, a very efficient mechanism on which to base sensitive X-ray polarimeters, is the photoelectric effect. A measurement of the photoelectron emission direction provides a measure of the direction of the electric field vector of the incident X-ray, namely the polarization state of the photon. The challenge is to reconstruct the very short initial part of the electron track (only some hundreds mm), before the diffusion has smeared the charge pattern, blurring in part the directional information.
n
Corresponding author. Tel.: þ39 050 2214382; fax: þ39 050 2214317. E-mail address: [email protected] (G. Spandre).
0168-9002/$ - see front matter & 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.nima.2012.12.006
In the last decade, the advances in the development of micropattern gas detectors, a very finely spaced arrays of proportional counters, have made photoelectric X-ray polarimetry a real possibility and renewed the interest of this observational strategy for scientific discovery [1–3].
2. The Gas Pixel Detector as X-ray Polarimeter The Gas Pixel Detector (GPD) developed, and continuously improved, in Italy by INFN of Pisa in collaboration with IASF/INAF of Rome [1,3,4,5] belongs to the new generation of Micro-Pattern Gas Detectors. The instrument measures the polarization of the X-ray photons by imaging the tracks of the photoelectrons created in the gas cell. A given X-ray’s polarization state is determined by the initial direction of the resulting photoelectron. When a low energy linearly polarized photon is absorbed in a low density medium, the photoelectron is emitted preferentially in the plane orthogonal to the X-ray propagation (Fig. 1) with a cos2 f modulation, where f is the azimuthal angle between the direction of emission of the photoelectron and the electric field of the X-ray (polarization vector). In a simplified quantum-mechanical treatment [6] in fact, the differential cross-section of the photoelectron emission, in the non-relativistic regime, is given by: 2 7=2 pffiffiffi 2 @s Z5 mc 4 2sin ðyÞcos2 ðjÞ ¼ r 20 4 @O hv ð1b cosðyÞÞ4 137
174
R. Bellazzini et al. / Nuclear Instruments and Methods in Physics Research A 720 (2013) 173–177
Fig. 2. Schematics of the GPD assembly.
Fig. 1. Photoelectron interaction: when incoming photons are linearly polarized, the photoelectrons are ejected preferentially in the plane orthogonal to the photon direction propagation (i.e. y 901) with a cos2 angular distribution in f).
where Z is the atomic number of the photo-absorbing atom, r0 the classical electron radius, b the photoelectron velocity given in fraction of c, y the angle between the direction of the incoming photon and the ejected electron and f the azimuthal angle defined above. The photoelectron emission is mainly at 901 with respect to the incoming photon direction. The term in the denominator accounts for a slight bending forward of the emission angle distribution with the photoelectron energy. For example, at 8 keV the photoelectrons are ejected preferentially at y ¼751. The modulation factor is a key parameter for the sensitivity of a polarimeter. It represents the response of the instrument to a 100% polarized photon beam [7]. A larger value of the modulation factor implies lower statistical fluctuations and hence the capability to detect weaker polarized signals. Schematically (see Fig. 2), a GPD is made by a gas volume enclosed by a top Berillium window, a Gas Electron Multiplier (GEM) which amplifies the charge of the electron tracks generated in the drift gap providing energy and time information and a pixellated charge collection plane directly connected to the pixellated analog read-out electronics (CMOS-VLSI). 2.1. The GPD components: GEM and VLSI The GEM consists of a 50 mm-thick polyimide sheet, sandwiched by two 5 mm-thick copper layers through which densely packed holes are drilled using the CO2 laser etching technique. Holes pitch is 50 mm. This type of GEM, designed at RIKEN [8] and fabricated by Scienergy Co. Ltd. [9], does not show any rate dependent gain instability due to the good quality of the cylindrical hole shape [10]. The GEM is the two-dimensional amplification structure of the photoelectron track. The primary ionization electrons of the track are drifted by the applied electric field ( 1 kV/cm) toward the GEM, multiplied in the high electric field region (100 kV/cm) inside the GEM holes and then collected by the pixellated read-out electrode underneath (the fine structured top metal layer of the VLSI chip).
The GEM has worked with two gas mixtures, He and Ar based (reaching large effective gain, well above 1000) at voltage difference of 450–550 V through the GEM. Typical high voltage settings with Ar 70% DME 30% gas filling are: VDRIFT¼ 3000 V, VGEMtop ¼ 950 V, VGEMbottom¼ 400 V; with He 20% DME 80%: VDRIFT¼ 2650 V, VGEMtop¼ 865 V, VGEMbottom¼ 400 V, with the read-out electrode at the charge preamplifier input voltage, i.e. 1.5 V. Fig. 3 shows the custom VLSI chip mounted on its control motherboard. The chip belongs to a third generation of ASICs of increased complexity, developed by the Pisa-INFN group in 0.18 mm CMOS technology. The chip top metal layer has been patterned with 105,600 pixels, hexagonally shaped, arranged in a 300 352 honeycomb matrix corresponding to a 15 15 mm2 active area and a pixel density of 470 pixels/mm2 (the pixel pitch being 50 mm and the fill fraction, metal/active area, 95%). A fully analog electronic chain (low noise charge preamplifier, shaping amplifier, peak&hold and input to a multiplexer stage) has been implemented beneath each pixel, in five metal and one single poly-silicon layers. The chip integrates more than 16.5 million transistors. It is subdivided into 16 identical clusters of 6600 pixels (22 rows of 300 pixels) or alternatively in 8 clusters of 13,200 pixels (44 rows) each one with an independent differential analog output buffer. The chip has customable self-trigger capability and includes a signal pre-processing function for the automatic localization of the event coordinates. Groups of 4 contiguous pixels (mini-cluster) contribute to the local trigger that has a dedicated amplifier whose shaping time is roughly a factor 2 faster ( 1.5 ms) than the shaping time of the analog signal to allow its peak detection and hold. Each noisy pixel can be masked with respect to its contribution to the trigger signal and/or its analog output. By limiting the output signal to only those pixels belonging to the region of interest, it is possible to reduce significantly the read-out time and data volume. Table 1 summarizes the main characteristics of the chip. To generate and handle the command signals to/from the chip a very compact data acquisition system has been implemented on the Altera FPGA Cyclone EPIC240 (RISC NIOS-II processor). Analog data are read, digitally converted by a Flash ADC and provisionally stored in a static RAM. In self-trigger mode just after an event acquisition it is possible to read the pixel pedestals in the triggerdefined ROI, one or more times, compute the average value and transfer the pedestal subtracted data to the offline analysis system.
R. Bellazzini et al. / Nuclear Instruments and Methods in Physics Research A 720 (2013) 173–177
The great advantage of this operation is a real time control of the data quality and the removal of drift effects of the pedestal values with temperature or any other environmental changes. 2.2. The GPD geometry The GPD is characterized by an active gas volume defined by the active area ( 2 cm2) of the charge collecting electrode (the pixellated ASIC top metal layer) and an absorption gap thickness (drift region) of 1–2 cm. In order to make more uniform the drift electric field within the absorption cell and to minimize any systematic effect that could leave a residual polarization signal for nonpolarized source, the transverse GEM dimensions are increased by several centimeters with respect to the collecting plane. The thickness of the gas gap between the GEM bottom and the ASIC is 0.5 mm. The parallelism between the GEM plane and the ASIC plane is a critical item. This requirement is satisfied if the GEM bottom side is in full contact with the package top surface that nominally has a misalignment with respect to the ASIC plane o50 mm.
175
A photo of the sealed detector fabricated in Finland in collaboration with Oxford Instruments Analytical Oy, a small company with long lasting experience in space qualified instruments, is shown in Fig. 4. The detector is manufactured with materials and components with low out-gassing rate (alumina for the spacers, kapton, and ceramic for the feedthroughs) and it is capable to withstand long-term operations in the space environment. 2.3. The photoelectron track reconstruction The photoelectron track is identified by the projected charge distribution on the pixel plane. The analysis of the track moments allows assessing some properties of this distribution such as: the barycentre, the major axis and most importantly the projection on the pixel plane of the point of absorption of the X-ray photon (see Fig. 5) used to derive the direction of ejection of the photoelectron. The amplitude and direction of the source polarization is obtained from the modulated distribution of the emission angles.
3. The results with two GPD configurations The gas mixture filling the GPD determines the working energy band of a photoelectric X-ray Polarimeter. In the 2–10 keV energy range, mixtures based on He and DME as absorber and quencher are the most suitable. In our tests the best performances in the low energy band, have obtained with 20% He–80% DME at 1-atm pressure. At higher energy, Ar based mixtures are the most effective. In the next paragraph we report the results obtained for the two baseline gas mixtures proposed respectively for the low and medium energy polarimeters at the focus of the HXMT telescope [11,12]. 3.1. He–DME (20–80, 1-atm pressure)
Fig. 3. The VLSI ASIC on its motherboard.
To test the performances of the GPD as X-ray Polarimeter in the low energy band, 95% polarized beams have been produced in the IASF-INAF laboratory in Rome [5]. The use of a mosaic graphite and flat aluminum crystals has allowed obtaining polarized photons at 2.6, 3.7, and 5.2 keV from the diffraction of unpolarized continuum or line emission. These sources have been used to measure the modulation factor of the GPD filled with the 20% He–80% DME gas mixture. As shown in Fig. 6, the modulation factor measured at these energies has resulted in good agreement with the Monte Carlo simulations. Fig. 7 represents the angular distribution that has obtained with polarized photons of 3.7 keV. A cos2 fit of the distribution gives a modulation factor of 43.10%70.84%.
Table 1 Main ASIC characteristics. Peaking time (externally adjustable) Full scale linear range Pixel noise Read-out mode Trigger mode Read-out clock Self-trigger global threshold Frame rate Parallel analog output buffer Access to pixel counter Total power dissipation Fill fraction (metal/active area)
3–10 ms 30,000 electrons 50 electrons ENC Asynchronous and synchronous Internal, external, self-trigger Up to 10 MHz 2000 electrons Up to 10 kHz (event window self trigger) 1, 8 or 16 Direct (single pixel) or serial (8–16 clusters, full matrix, region of interest) 0.5 W (self-trigger mode) 92%
176
R. Bellazzini et al. / Nuclear Instruments and Methods in Physics Research A 720 (2013) 173–177
Fig. 6. Comparison between the modulation factor expected from the Monte Carlo simulations (crosses) and the measured values at 2.6, 3.7, and 5.2 keV points with the error bars. It is assumed that the degree of polarization at 2.6 and 5.2 keV is 0.95. The dashed line is the fit to Monte Carlo values.
Fig. 4. The sealed Gas Pixel Detector. The Berillium window is capped with a titanium frame.
Fig. 7. Modulation of the reconstructed angles at 3.7 keV in He–DME (20–80) at 1atm pressure.
Fig. 5. A photoelectron track produced in He–DME mixture by a 5.9 keV photon.
A good energy resolution at 3.7 keV photon energy is obtained in the He–DME mixture, as shown in the pulse height distribution of photoelectrons in Fig.8. 3.2. Ar–DME (70–30, 2-atm pressure) To extend the polarimetric performances in the medium energy band, 6–30 keV, gas fillings with higher atomic number have to be used. Simulation results suggest argon based mixtures. A GPD prototype with a 2 cm Ar–DME (70–30) cell at 2 atm working pressure, to improve the detection efficiency, has been successfully tested. In this configuration the detector can perform
Fig. 8. Pulse height distribution of 3.7 keV photoelectrons in He–DME.
energy resolved imaging polarimetry of hard X-ray sources with high sensitivity. A measurement of the energy resolution of the GPD at 5.9 keV is shown in Fig. 9.
R. Bellazzini et al. / Nuclear Instruments and Methods in Physics Research A 720 (2013) 173–177
177
and sets the limit to the Minimum Detectable Polarization. The measured value of the residual modulation is 0.18% 70.14% (statistical error).
4. Conclusions
Fig. 9. Pulse height distribution obtained with a totally unpolarized 5.9 kev photons in a Ar 70%–DME 30% gas mixture at 2 atm pressure. The typical Ar escape peak is clearly visible This peak is due to the emission, occurring in about 15% of the photoelectric absorptions, of a fluorescence photon with energy just below the Ar K edge. This secondary photon has a very long mean free path for absorption and therefore escape the detector volume.
The Gas Pixel Detector is one of the most advanced instruments to image the photoelectrons tracks in a gas, both for performance and readiness level. The choice of a suitable gas filling allows measuring the linear polarization in the energy range 2–30 keV, and reconstructing the impact point, the energy and the time of each event. The possibility to join polarimetric information to timing, spectral and imaging capability, is a unique characteristics of this type of detector and it is particularly interesting in X-ray Astrophysics. A large literature, still without a significant experimental feedback, describes polarimetry as a fundamental tool to distinguish among competitive models, often equivalent on the basis of spectral or timing information. The GPD provides very concrete possibilities to perform polarimetric measurements at the level of accuracy of 1% and below on board of small pathfinder missions, like POLARIX or HXMT. The GPD is also included in the baseline design of the International X-ray Observatory that will definitely elevate the X-ray polarimetry to a key tool for a deeper understanding of many different classes of astrophysical objects. References [1] [2] [3] [4] [5] [6] [7] [8]
[9] Fig. 10. Residual modulation obtained with totally unpolarized 5.9 kev photons.
Fig. 10 shows the angular distribution for a totally unpolarized beam of 5.9 keV photons. A fit to the distribution allows estimating the residual modulation of the device due to systematic errors
[10] [11] [12]
E. Costa, et al., Nature 411 (2001) 662. J.K. Black, et al., Nuclear Instruments and Methods A 581 (2007) 755. R. Bellazzini, et al., Nuclear Instruments and Methods A 566 (2006) 552. R. Bellazzini, et al., Nuclear Instruments and Methods A 579 (2007) 853. F. Muleri, et al., Nuclear Instruments and Methods A 584 (2008) 149. W. Heitler, The Quantum Theory of Radiation, Oxford University Press, London, 1970. L. Pacciani, et al., Proceedings of SPIE 4843 (2006) 394. RIKEN is a Japanese Institution whose mission is to conduct comprehensive research in science and technology in a wide range of fields, including physics, chemistry, medical science, biology, and engineering, covering the entire range from basic research to practical application. SCIENERGY Co., Ltd 1-9-15-640 Kanumadai, Chuo-ku, Sagamihara, Kanagawa 252-0233, Japan /http://www.scienergy.jpS. T. Tamagawa, et al., Nuclear Instruments and Methods A 560 (2006) 418. E. Costa, et al., Proceedings of SPIE 6686 (2007) 66860Z. P. Soffitta, et al., Proceedings of SPIE 7011 (2008) 701128.