20th International Workshop on Radiation Imaging Detectors

24 Jun 2018, 09:00 → 28 Jun 2018, 18:00 Europe/Zurich

Quality Hotel, Folkets Hus (Sundsvall)

Christer Fröjdh (MId Sweden University)

test

Sunday, 24 June

1 Registration

Social event: Reception

Monday, 25 June

Conference registration

Confere

Opening and welcome

Initial talk.pptx

Invited: Honorary Talk

2 Past and future of semiconductor pixel detectors

Speaker: Erik Heijne (Czech Technical University (CZ))

Oral

Convener: Prof. Christer Fröjdh (MId Sweden University)

3 First experiments with the JUNGFRAU charge integrating photon detector

The JUNGFRAU is a hybrid photon pixel detector developed at the Paul Scherrer Institut for free electron laser and synchrotron applications.

Each detector module consists of half a million pixels of $75 \; \mu \mathrm{m}$ pitch, arranged in $1024 \times 512$ arrays for a sensitive area of approximately $8\times4 \; \mathrm{cm}^2$. Eight $256 \times 256$ pixel ASICs are bump-bonded to a single $320 \; \mu \mathrm{m}$ thick silicon sensor. JUNGFRAU is a charge integrating detector and features three automatically switching gains per pixel, which adjust the amplification factor to the amount of deposited charge. This enables single photon sensitivity, while ensuring a dynamic range over four orders of magnitude [1]. Modules can be combined in various configurations to produce larger systems. A general overview of the detector and its performance will be presented, as well as details of the conversion of detector output to measured energy.

JUNGFRAU has been designed primarily for SwissFEL [2], which is now starting operation at PSI. For the pilot experiment phase, small JUNGFRAU systems were installed in the Alvra and Bernina end stations. Results from the commissioning and first user experiments will be presented.

JUNGFRAU has also been used at synchrotrons in proof-of-principle macro-molecular crystallography experiments with pink and monochromatic beam. Experience of using the JUNGFRAU away from optimal operating conditions will be shared, and the capability of JUNGFRAU for synchrotron applications will be demonstrated.

[1] A. Mozzanica et al., “Characterization results of the JUNGFRAU full scale readout ASIC” 2016 JINST 11 C02047
[2] C. Milne et al., “SwissFEL: The Swiss X-ray Free Electron Laser” 2017 Appl. Sci. 7(7), 720

Speaker: Sophie Redford (PSI - Paul Scherrer Institut)

Redford_JUNGFRAU_good.pdf

Swedish fika in the hotel lobby (2nd floor)

Invited

4 CMOS Image Sensor Technology for X-ray detection

Speaker: Renato Turchetta (STFC - Rutherford Appleton Lab. (GB))

Oral

Convener: Bernd Schmitt (Paul Scherrer Institut)

5 The Percival 2-Megapixel Monolithic Active Pixel Imager

The peak brilliance reached by today's Free-Electron Laser and Synchrotron light sources requires photon detectors matching their output intensity and other characteristics in order to fully realize the sources' potential. The Pixellated Energy Resolving CMOS Imager, Versatile And Large (Percival) is a dedicated soft X-ray imager developed for this purpose by a collaboration of DESY, Rutherford Appleton Laboratory/STFC, Elettra Sincrotrone Trieste, Diamond Light Source, and Pohang Accelerator Laboratory (Rep. of Korea).
Following several generations of prototypes, the Percival "P2M" 2-Megapixel imager -- a 4.5x5cm monolithic, stitched sensor with an uninterrupted imaging area of 4x4cm2 (1408x1484 pixels of 27x27um -- was produced and has demonstrated basic functionality with a first-light image using visible light. It is currently being brought to full operation in a front-illuminated configuration. The readout system being commissioned in parallel has been developed specifically for this imager which produces - at full frame rate of 300Hz - data at a rate of 20Gbit/s.
Currently, a first wafer with 8 Percival P2M chips is undergoing backthinning and backside processing to enable soft X-ray detection.
We will present the P2M system, the project status, and the status of the performance evaluation of the first P2M sensors under illumination with visible light.

Speaker: Cornelia Wunderer (DESY)

2018-06-25_iWoRiD_Percival_Talk_v2-noSpares.pptx.pdf

6 Single-photon imaging detector with sub-100ps and <10µm resolutions

We present the concept of a single-photon imager capable of detecting up to 10$^9$ photons per second with simultaneous measurement of position (<10µm resolution) and time (few tens of picosecond resolution) for each individual photon over an active area of 7 cm$^2$. The detector is based on a “hybrid” concept: a vacuum tube, with a transparent input window on which a suitable photocathode material is deposited, a micro-channel plate (MCP) and a pixelated read-out anode based on the Timepix4 ASIC (65nm CMOS technology) designed in the frame of the Medipix4 collaboration. A MCP with <10μm pore diameter will be used, operated at low gain (a few 10$^4$) and treated with atomic layer deposition, allowing a lifetime increase to >10 C/cm$^2$ accumulated charge. This detector will allow for the first time to fully exploit all the intrinsic excellent performance of a MCP, using a front-end electronics ASIC encapsulated in the tube with unprecedented performance. Timepix4 is an array of 512x448 pixels, 55µmx55µm each, with an active area of 28mmx25mm. It features 50-70 e- equivalent noise charge, a maximum rate of 1.2 Ghits/s, and allows to time-stamp leading-edge time and measure Time-over-Threshold (ToT) for each individual pixel. A weighted average of the cluster pixels position can be calculated using their ToT information, which allows to reach 5-10μm position resolution. The ToT information can also used to correct for the leading-edge time-walk in each pixel, and a timing resolution of few tens of picosecond is expected. The detector will be highly compact: the front-end electronics is encapsulated in the vacuum tube and allows local processing of the detector information, which are sent out of the tube in digital form. The Timepix4 architecture is data driven, producing 64 bits for each pixel hit, corresponding to a maximum data rate of 80 Gbps for 1.2 Ghits/s. A flexible design is conceived, with electro-optical transceivers linking the ASIC to a FPGA-based board, placed far from the detector, for the exchange of configuration and the collection of event data. The FPGA will perform serial decoding and send the data directly to a PC for storage using fast serial data links. These figures of merit will open many important applications allowing significant advances in particle physics, life sciences or other emerging fields where the detection of single photons with excellent timing and position resolutions are simultaneously required.
In this contribution we will present measurements performed on a prototype of such a hybrid photon detector, based on the first version of Timepix, developed by the Medipix collaboration. We will present its detailed characterization in terms of time and position resolution, measurements of its quantum efficiency using pairs of entangled photons and demonstrate its ability to reconstruct the trajectories of single high-energy electrons passing a scintillator (see Fig. 1). The challenges and prospects for an improved version of such a photon detector using Timepix4 will be presented in detail.

Fig. 1: Trajectory of a single 5 GeV electron passing a 4 mm x 4 mm x 4 mm plastic scintillator, reconstructed from single photons emitted along the trajectory which were detected at the same time from two different views with a Timepix-based hybrid photon detector. The values on the axes are given in pixels [px].

Speaker: Massimiliano Fiorini (INFN and University of Ferrara (IT))

Fiorini - iWoRiD 2018.pdf

7 Resolving Power of Pixel Detector Timepix for Wide-Range Electron, Proton and Ion Detection

We investigate the response of the Timepix detector for measurements and characterization of mixed fields such as radiation environments found in ion beam radiotherapy, Earth upper atmosphere and outer space. Of interest is the detector resolving power in terms of particle-type (species) and spectral- (energy loss) response. For this purpose we performed tests and calibration of detection response in defined fields of various radiations (electrons, protons, ions) at various energies and incident directions relative to the plane of the sensor. The signals produced (called pixel clusters) by single particles in the pixelated sensor are a convolution of (i) particle type (mass, Z), (ii) energy loss (stopping power, LET) and (iii) direction. The characteristic tracks in the form of pixel clusters are analysed at the pixel- and um-scale by so-called cluster analysis making use of high-resolution pattern recognition algorithms. Cluster parameters are analysed separately and also in correlations or ratios in the form of 2-dimensional plots (see Fig. 1). A resulting physics-based classification of event types is proposed in terms of radiation components (particle species, namely X-rays, light and charged particles), spectral/stopping power (dE/dx, LET spectra over several bins in semi-log range) and direction (limted angular resolution). Up to eight groups of events are distinguished with varying extent of range and partial overlap. Research supported by OP RDE, MEYS, Czech Republic under the project CRREAT, CZ.02.1.01/0.0/0.0/15_003/0000481. R&D of the WidePIX 3D instrument was supported by the project TACR TA04011329 of the Czech Republic.

Figure 1. (a) Two-dimensional correlated plots (event histograms) according cluster analysis parameters (shown HEA = Height to Elevation Angle, and LET = Linear Energy Transfer) for (a) electrons, (b) protons and ions: (c) 4He and (d) 12C at indicated energies and incident angles.

Speaker: Carlos Granja (Advacam)

11.40 Carlos Granja + TPX + IWORID + june2018 .pdf

Lunch

Invited

8 In-Vivo Monitoring in Proton Radiotherapy with Prompt-Gamma Multi-Slat Imaging: a Realistic Monte Carlo Study

Speaker: Paulo Crespo

Oral

Convener: Roelof de Vries (Malvern-PANalytical)

9 Status and calibration concept for the AGIPD detector

AGIPD (Adaptive Gain Integrating Pixel Detector) is a 2D hybrid pixel detector system designed and developed for the European XFEL (Eu.XFEL) [1]. At the Eu.XFEL photons will arrive in pulse trains every 100 ms (or at the fundamental repetition frequency of 10 Hz). Each pulse train consists of up to 2700 pulses that arrive within 600 µs (i.e. a bunch spacing of 222 ns, meaning 4.5 MHz frame rate) followed by 99.4 ms between trains. Each single pulse consists of 10$^{12}$ X-ray photons in less than 100 fs and with an energy tunable between 250 eV up to 25 keV. The challenges for the detector at the Eu.XFEL include a dynamic range up to 10$^{4}$ photons per pixel, single photon sensitivity, a frame rate up to 4.5 MHz, as well as high radiation tolerance. In order to cope with the large dynamic range, the first stage of each pixel of the AGIPD ASIC [2] is a charge sensitive preamplifier with three different gains that are dynamically switched during the charge integration. Dynamic gain switching enables single photon resolution (with a measured S/N ratio of 11 at 12.4 keV) in high gain and a dynamic range of 10$^{4}$ x 12.4 keV photons in low gain. The time structure of the beam of the Eu.XFEL does not allow a continuous readout of the single frames during the bunch train, thus each pixel of the AGIPD ASIC (area 200 x 200 µm$^{2}$) is equipped with an in-pixel memory which consists of 2 storage cell matrices of 352 storage cells each onto which the frames are stored before being read out in the gap between trains. One of the most critical aspects of this detector concerns its calibration [3] and many aspects have to be taken into account such as linearity, calibration time, radiation damage and high dynamic range to probe. A high dynamic range test suite is needed which must be linear to the required degree over the entire dynamic range. Due to the high brilliance the radiation damage expected at the Eu.XFEL is significant and may cause a variation of many parameters of the detector such as noise, baseline and gain. The detector will therefore need a periodic re-calibration, taking into account that the calibration suite itself may be affected by radiation damage. This is especially true for the on-chip sources. This contribution will be focused on the calibration concept of the AGIPD detector. Different calibration techniques will be shown and compared such as internal current source, backside pulsing, IR pulsed laser, LED light and mono-energetic protons (see figure). Moreover an update of the status of the AGIPD detector will be given and some relevant results of the first experiments will be shown.

[1] M. Altarelli et al., European XFEL technical design report, (2006), ISBN 978-3-935702-17-1.
[2] X. Shi et al., Challenges in chip design for the AGIPD detector, Nucl. Instrum. Meth. A 624 (2010) 387.
[3] D. Mezza et al., 2016 New calibration circuitry and concept for AGIPD JINST 11 C11019.

Speaker: Dr Davide Mezza (Paul Scherrer Institut)

iWoRiD2018_DaM_v8.pptx

10 First X-ray measurements with microstrip LGAD sensors

Although planar silicon sensors are commonly used for X-ray detection, they currently cannot be used by single photon counters at low energies due to their electronic noise being too high. A 1 keV-photon for instance, creates less than 300 electron-hole pairs. In this case, single photon counting detectors would require a noise lower than 30 electrons, which is still unachievable, and are therefore unusable in the low energy range. For soft X-rays, Low Gain Avalanche Detectors (LGAD), which amplify the signal in the sensor, are suitable to overcome this limitation.

Microstrip LGAD sensors, fabricated at FBK, are under investigation with X-rays of energies down to a few keV using both the Gotthard (charge-integrating) and the Mythen (single-photon-counting) readout chips. The LGAD sensors are 50 µm thick and have a strip-pitch of 150 µm, which is however not optimized for applications in photon science. Instead of a homogeneous layer, the charge multiplication region (also known as gain region) of the sensor is only below the implantation of each strip, with electrons being collected by the readout strips.

We found that X-ray photons with energies below 3 keV can be detected using the LGAD microstrips, suggesting a significant improvement compared to the minimum detectable energy of 8 keV of a planar silicon sensor of the same not optimized geometry.
The gain, electronic noise, signal-to-noise ratio as well as the fill factor of the sensor were characterized as a function of the bias voltage of the sensor and the results agree with simulation and design parameters.
The measurements show that LGADs are a promising technology for photon-counting detectors and fast charge-integrating detectors with the capability to decrease the minimum detectable photon energy.

We will present the first results of the characterisation of the LGAD sensors with X-rays and discuss the future developments based on the study of this work.

Speaker: Marie Andrä (PSI - Paul Scherrer Institut)

20180624_iWoRiD_MAndrae_LGADs_noAnimation.pdf

11 Development of full-field XRF imaging system for non-invasive investigation of artworks

Analysis of the spatial distribution of elements in the heritage objects, using X-ray fluorescence (XRF) spectroscopy, has become a very important support for conservators in recent years. Elemental mapping offers possibility to study spatial distributions of pigments on the surface of artworks and in the hidden painting layers.

In this work we present progress in the design and performance of a full-field XRF imaging system being developed in collaboration between the Faculty of Physics and Applied Computer Science of the AGH University of Science and Technology and the Laboratory of Analysis and Non-Destructive Investigation of Heritage Objects of the National Museum in Krakow. Compared to commercially available scanning macro-XRF systems using detectors with no position sensitivity, the full-field XRF technique is based on a pinhole camera and a large area position-sensitive and energy dispersive detector. Thus, XRF radiation from large area of an investigated object is simultaneously projected on the detector and such a technique has a number of advantages: (i) it offers shorter measurement times, (ii) the pinhole camera can be adjusted to optimise the measurement procedure, making a compromise the spatial resolution and the measurement time, (iii) fast moving measuring head is eliminated, which is an important aspect for safety of the investigated objects, (iv) the infinite depth of field of the pinhole camera allows one to investigate non-flat objects.

Our system is equipped with two Varian (Varex) VF-50J low power air-cooled X-ray tubes and employs 10 cm by 10 cm Gas Electron Multiplier (GEM) detector assembled with custom designed readout and data acquisition system based on Application Specific Integrated Circuit [1]. Moderate energy resolution of GEM detectors limits elemental selectivity of the system, however, by optimization of the front-end electronics and detector readout significant improvement of the energy resolution has been achieved compared to typical values reported in the literature [2].

Performance of the imaging system and results of elemental mapping obtained with the developed apparatus are presented. The imaging measurements have been performed on several phantoms, painted with different pigment compositions. Selected imaging results for non-flat objects are also presented and discussed.

References
[1] W. Dąbrowski, T. Fiutowski, P. Frączek, S. Koperny, M. Lankosz, A. Mendys, B. Mindur, K. Świentek, P. Wiącek, P.M. Wróbel, Application of GEM-based detectors in full-field XRF imaging, JINST 11, 2016, C12025.
[2] T. Fiutowski, S. Koperny, B. Łach, B. Mindur, K. Świentek, P. Wiącek, W. Dąbrowski, ARTROC - a readout ASIC for GEM-based full-field XRF imaging system, JINST 12, 2017, C12016

Acknowledgments
The authors would like to acknowledge The National Centre for Research and Development for financial support within Applied Research Programme (project no. PBS3/A9/29/2015).

Speaker: Dr Tomasz Fiutowski (AGH University of Science and Technology, Faculty of Physics and Applied Computer Science)

fiutowski_XRF_IWORID_2018.pdf

12 Characterisation of FBK small-pitch 3D diodes after neutron irradiation up to 3.5x10ˆ16$ neq cmˆ-2

Owing to their radiation hardness, 3D pixel sensors are the most promising option for the innermost layers of tracking detectors at the High Luminosity LHC (HL-LHC). This application requires very high hit-rate capabilities, increased pixel granularity, extreme radiation hardness, and low material budget. As a result, future 3D pixels will have all geometries significantly downscaled as compared to the existing ones, and in particular, a smaller pitch (e.g., 50$\times$50 or 25$\times$100 $\mu$mˆ2) and reduced active thickness (~100 $\mu$m). To this purpose, a new generation of 3D pixel sensors has been developed in the past few years within the “Phase2” INFN-FBK collaboration, using a single-sided technology on Si-Si Direct Wafer Bonded 6” substrates [1]. Beam test results of the first prototypes are encouraging, with hit efficiency values of ~99% (before irradiation) and ~97% (after irradiation to ~1 $\times$10$ˆ{16}$ n$_{eq}$ cm$ˆ{-2}$).
The sensor performance at larger fluences of interest for the HL-LHC could not be measured yet due to the limited radiation tolerance of the available read-out chips (FEI4 and PSI46dig). Nevertheless, TCAD simulations incorporating advanced radiation damage models anticipate that charge trapping effects are strongly mitigated in these small-pitch 3D sensors, so that the charge collection efficiency can be still very good (~70%) after an irradiation fluence of 2$\times$10$ˆ{16}$ n$_{eq}$ cm$ˆ{-2}$. In order to confirm the simulation results, we used 3D diodes, e.g., small (~2 mmˆ2) arrays of basic 3D cells with all columns of the same doping type shorted together, reproducing the same layout details of their parent pixel sensors. Sensors were irradiated with neutrons at the TRIGA Mark II reactor at JSI (Ljubljana, Slovenia) to three different fluences: 1$\times$10$ˆ{16}$, 2$\times$10$ˆ{16}$, and 3.5$\times$10$ˆ{16}$ n$_{eq}$ cm$ˆ{-2}$. The breakdown voltage of the irradiated samples, generally higher than 200 V, ensures a wide bias voltage range that is essential to obtain high efficiency [2]. For functional tests, we have used a vacuum chamber allowing to operate the samples at -15°C. The samples are contacted from the back side by placing the silicon dice on a metallic thermal finger covered with a conductive paste that improves the electrical contact. The signals are read-out from the front side by contacting the probing pads by microneedles. The read-out chain includes a low-noise charge amplifier and a shaper. By using a position resolved pulsed laser (wavelength 1055 nm, nominal pulse width 40 ps), we have collected two-dimensional maps of the relative signal intensity across a 3D basic cell at different bias voltages. Data were then normalised to the signals acquired on non-irradiated samples. The relative efficiency reaches ~75% for sensors irradiated to the smallest fluence and ~55% for those irradiated to the highest one. The 2-d signal maps were compared to those calculated according to Ramo’s theorem with input data obtained from TCAD simulations, and the agreement is satisfactory.
A comprehensive overview of the experimental and simulation results will be reported at the conference.

REFERENCES
[1] D.M.S. Sultan, et al., First production of new thin 3D sensors for HL-LHC at FBK, Journal of Instrumentation, JINST 12, C01022, 2017.
[2] G.-F. Dalla Betta, et al., Electrical characterization of FBK small-pitch 3D sensors after $\gamma$-ray, neutron and proton irradiations, Journal of Instrumentation, JINST 12, C11028, 2017.

Speaker: Gian-Franco Dalla Betta (INFN and University of Trento)

13 MARS spectral CT imaging using CZT Medipix3RX

In this presentation, we will report on the use of the MARS scanner equipped with the Cadmium Zinc Telluride (CZT) assembled Medipix3RX detector in Charge Summing Mode (CSM). We have looked at methods for optimising scanning protocols in a preclinical small animal spectral scanner that will translate to human scanning in the diagnostic energy range (20 – 140 keV).

We have used a MARS small bore scanner incorporating a CZT Medipix3RX camera and a 0.375mm brass or 2mm aluminium filtered poly-energetic x-ray source to simulate the expected absorption during human scanning. Different energy band selections were trialled for different applications in imaging. We have imaged gold nanoparticles targeted to ovarian and breast cancer cells. We have distinguished gout and compared DECT and spectral CT images, and various calcium based crystals found in common joint diseases. We have measured lipid content in fatty liver and quantify gadolinium uptake in human knee samples as a measure of cartilage health. We have also used MARS to measure bone density and structure of normal and osteoporotic human femora comparing DEXA, DECT, pQCT, and spectral CT. A post material decomposition assessment method based on the sensitivity (true positive rate) and specificity (true negative rate) of an individual voxel was used to generate a quantitative metric to characterize material misclassification.

Spectral CT is likely to play a major role in identifying and directing treatment for a range of inflammatory diseases. Metrics and methods for assessing errors in material identification and quantification are currently being validated using imaging protocols designed to be translated to human imaging when it becomes available.

Speaker: Lieza Vanden Broeke (MARS Bioimaging Limited, Christchurch, New Zealand)

Swedish fika in the hotel lobby (2nd floor)

Poster session: part 1

Convener: Val O'Shea (University of Glasgow (GB))

14 Absorbed dose distributions in intensity-modulated dental computed tomography

Dedicated cone-beam computed tomography (CBCT) for dental diagnostics has become popular and nowadays it is not difficult to find CBCT scanners even in private dental clinics. On the other hand, cancer risk with the increasing use of CT has been a great issue in diagnostic radiology [1]. If there was the slice war in the CT industries, the dose war has begun. The tube-current modulation technique in medical CT, which was developed for low-dose imaging while minimizing the loss of image quality, may also be applied to dental CT.
This study investigates the feasibility of beam-intensity (kVp and/or mAs) modulation (BIM) technique in dental CBCT. It is a motivation of this study that the BIM technique may avoid severe radiation dose to critical organs, such as eye lens, thereby reducing the effective dose. Various BIM scenarios for a single circular scanning are designed accounting for the cervical spine through which x-ray beam attenuates largely. Using the Monte Carlo (MC) technique, we obtain absorbed dose distributions in a numerical anthropomorphic head/neck phantom for the designed BIM scenarios, and compare the results with that obtained for the conventional scan, as demonstrated in Fig. 1. We implement the BIM scenarios to our laboratory bench-top CBCT system, and investigate image quality (e.g., image noise) of the reconstructed images. This study will determine which BIM scenario can provide less patient dose with image quality comparable to the conventional dental CBCT.

This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIP) (No. 2017M2A2A6A01071267).

Speaker: Jinwoo Kim (Pusan National University)

15 Characteristics of a Flexible Radiation Detector Fabricated with Non-Fullerene Acceptor for Indirect-type X-ray Imaging

With noticeable improvements in synthesis and processing of organic semiconductor material, flexible devices are widely studied in various areas such as organic light emitting diode (OLED), organic solar cell, and radiation detector. Among radiation detector applications, dentistry, mammography, and non-destructive inspection areas have a significant interest in flexible and conformable imager. In this study, we proposed the indirect-type and flexible X-ray detector fabricated on polyethylene naphthalate (PEN) substrate with non-fullerene acceptor material. In order to evaluate the performance of the organic detector (Figure 1a) fabricated on glass and PEN substrate, the sensitivity1) of the detector having P3HT2) and PC70BM3) blended active layer was measured. The sensitivity of the detector fabricated on PEN substrate was 6% lower than that of the detector made on glass substrate due to relatively low transmittance and relatively high contact resistance. Fullerene-based acceptors, such as PC60BM and PC70BM, have issues with high synthetic cost, limited spectral absorption and morphological instability. In order to improve the thermal and mechanical stability of the detector fabricated on PEN substrate, a non-fullerene acceptor, O-IDTBR4), was applied instead of PC70BM acceptor. In addition, absorption wavelength of O-IDTBR (Figure 1b) was well-matched with emission wavelength of CsI(Tl) scintillator. Various curvature bending tests (Figure 1c) were carried out to verify its suitability as a flexible detector. After undergoing 30 bend cycles of different curvatures, the detector with the P3HT:PC70BM active layer showed poor stability than the detector with the P3HT:O-IDTBR active layer.

Speaker: Kang Jungwon (Dankook University)

16 Comparison between bulk semi-insulating GaAs-based and Si-based sensor for Timepix hybrid pixel radiation detector

Monocrystalline silicon is a well-established and widely used material in the field of X-ray detection and imaging, however, its low sensitivity to the photons with energy above 20 keV motivates the development of sensors made of high Z materials, like CdTe or GaAs. Semi-insulating (SI) GaAs is one of the promising candidates due to its advantageous properties. In comparison to silicon, it has about 20 times higher sensitivity for 12-60 keV X-rays and 3-5 times faster charge collection. In contrast to CdTe, it has no or minimal polarization effect and internal fluorescence photons have energies below range important for medical soft tissue imaging such as mammography. In this paper, we continue our study of 350 µm thick SI-GaAs sensor bump-bonded to the Timepix readout chip [1, 2]. Standard 300 µm thick silicon-based sensor was chosen as a reference for our test measurements. Firstly, homogeneity in terms of open beam image and flat-field corrected image was investigated. Secondly, spectrometric properties in counting (thresholded) mode as well as in energy (Time over Threshold) mode were studied. As a source of radiation, we used gamma rays from Am-241 (59.54 keV) and In α and β fluorescence photons (24.14 and 27.27 keV) in both counting and energy mode plus Cu, Mo, Cd fluorescence photons (8.04, 17.44 and 23.10 keV, respectively) in ToT mode. Thirdly, imaging properties were examined using X-ray source with micro-focal spot size of approx. 2 µm. Slanted edge method was used to determine the modulation transfer function (MTF) and the spatial resolution of the sensors. Also, images of several objects were taken for different energy thresholds, namely default sensor threshold (~3 keV for Si and ~6,5 keV for GaAs), 15 keV and 25 keV. In last measurement we tested contrast resolution limits of Timepix cameras with our fabricated testing object using nano-machining technique.

[1] Llopart, X., et al.: Timepix, a 65k programmable pixel readout chip for arrival time, energy and/or photon counting measurements, NIM in Phys. Res. A 581 (2007) 485-494.
[2] Zaťko, B., et al.: Imaging performance of Timepix detector based on semi-insulating GaAs. In Journal of Instrumentation, 2018, vol. 13, C01034.

Speaker: Mr Dávid Kubanda (IEE, Slovak Academy of Sciences)

17 Development and Evaluation of Parylene C-type Additive detector for High Resolution in High Energy X-ray CT

Development and Evaluation of Parylene C-type Additive detector for High Resolution in High Energy X-ray CT

M.J. Han^a, Y.H. Shin^a, K.T. Kim^b, Y.J. Heo^c, K.M. Oh^d, E.T. Park^c, J.Y. Kim^e, H.L. Cho^c, S.K. Park^c*

^aDepartment of Radiation Oncology, Collage of Medicine, Inje University
^bDepartment of Radiological science, International University of Korea
^cDepartment of Radiation Oncology, Busan Paik Hospital, Inje University
^dRadiation Equipment Research Division Korea Atomic Energy Research Institute
^eDepartment of Radiation Oncology, Haeundae Paik Hospital, Inje University

For the high-energy fan beam CT scanning applied to nondestructive inspections on heavy equipment, the one dimensional line detector that maximizes the shield of radiation scattering is used. This, while increasing the image quality, is notified to consume a great deal of time for using a large scaled scan feature on a limited radiation area causing issues of activation due to the irradiation for a long period of time. Therefore, for the industrial nondestructive inspections, the necessity of the high-energy cone beam CT scanning that is available for validation on large amount of component is emerging and researches on the large-area two dimensional detector for the image are actively performed. Currently, researches on detectors of indirect detection based on detection material of CdWO_4 are actively performed for such applications while Si-Diode is utilized for the circuit board to receive the produced radiation. However, the diode shows mechanical vulnerabilities so that the issue of performance change due to the radiation damage has been notified. Therefore, researches on photoconductors for direct detection are actively performed to replace the diodes. On the other hands, among photoconductors, HgI_2 shows relatively higher actual atomic no (Z_eff : 186) and electronic density so that researches on this material for highenergy X ray detections are actively performed. However, HgI_2 has issues that the leakage current exponentially increases following the increase of applied voltage producing unstable factors to implement high-resolution images. In the meanwhile, researches on parylene C layer as an insulating material have been conducted to improve the performance and the drive stability of FET structure in the semiconductor field.^[2] The chemical activity is low and the pinhole-free coating is available so that the evaporation characteristic with a low concentration of charge trap is identified. Therefore, this study aims to make improvements for issue of the unstable leakage current by using the parylene C on the HgI_2 based sensor as the basic work to produce a high resolution large-area detector and to evaluate the response characteristics of parylene C on the evaporation thickness. The sensor is produced in the size of 1 x 1 cm^2 with the thickness of 150 μm considering the dielectric breakdown when the driving voltage of 1 V/μm is applied. Then, using the CVD evaporation method, Parylene C is applied in the thickness of 10 to 20 μm on the photoconductor. The reproducibility and the linearity of the produced sensor are evaluated with the photon energies of 2 MeV and 5 MeV and the Photo/Dark current ratios are presented to evaluate the resolution. For the reproducibility, with the dose rate of 5 Gy/min, the radiation dose of 1 Gy is applied 10 times repeatedly and the evaluation is performed by calculating the standard error(SE) based on the confidence interval of 95%. Also, for the linearity, the radiation doses of 0.001 to 1 Gy with the dose rates of 1 to 5 Gy/min are applied and the evaluation is performed based on the determination coefficient(R-sq) acquired from the linear regression analysis. The experiment results show that the reproducibility of non-parylene with 2 MeV and 5 MeV are 1.20% and 0.94% to be in the confidence interval of 1.5%. On the other hands, the Parylene applied in the thickness of 18 μm shows the reproducibility of 1.01% and 0.72% respectively with 2 MeV and 5 MeV remarking improvements against the non-parylene case. In case of the linearity, R-sq value of the nonparylene shows to be 0.9931 and 0.9855 respectively with 2 MeV and 5 MeV. And, the 18 μm parylene showing the best reproducibility also shows relatively excellent linearity with 0.9984 and 0.9993. Also, the 18 μm parylene shows improvements of photo/dark current ratio by 5.4%p and 7.1%p respectively with 2 MeV and 5 MeV against that of the non-parylene. This study results are considered to be applicable as basic data to develop a highresolution image detector using parylene C for the insulation as well as for the sensor areas showing various detection structures.

[1] N. Uhlmann et al., Metrology, Applications and Methods with High Energy CT Systems, 2014 ResearchGate.
[2] Jan Jakabovic et al., Preparation and properties of thin parylene layers as the gate dielectric for organic field effect transistor, Microelectronics Journal 40, 3

Speakers: Mr YOHAN SHIN (Department of Radiation Oncology, Busan Paik Hospital, Inje University), Prof. SUNGKWANG PARK (Department of Radiation Oncology, Busan Paik Hospital, Inje University), Mr MOOJAE HAN (Dept. of Radiation Oncology, Collage of Medicine, Inje University)

18 Development of a dual-sided readout DOI-PET detector using 500-µm-pitch 64ch SiPMs

In this study, we used a novel type of 500um-pitch SiPM to develop a high resolution small animal DOI-PET detector. This SiPM has a very fast recovery time (~8ns) and high gain (10^6). The signal was read from both ends of the GAGG scintillator (0.4mm0.4mm20mm) to obtain the DOI information. According to the simulation based on GEANT4, the spatial resolution of the small animal DOI-PET system can be improved to ~0.64mm.

1.Introduction
Positron emission tomography (PET) is a nuclear imaging technique that can be used to visualize metabolic processes in the body by measuring the concentration of molecular probes labelled with positron-emitting radionuclides.At present, the resolution of PET imaging is sub-millimeter. Pixel size and parallax error are considered as two of the factors that limit spatial resolution. By using smaller photosensors and applying DOI (depth of interaction) estimation, effects of these factors can be reduced. Dual-sided readout detectors is a common approach for DOI measurement. The DOI information of a dual-sided readout detector is a function of the ratio of energy deposition in one SiPM array (E1) and the total energy deposition in both of the SiPM arrays (Etotal=E1+E2).

2.Materials and Methods
2.1.GEANT4 Simulation
We simulated single pixel to find the relationship between DOI and energy deposition ratio (E1/Etotal). The scintillator was 0.4mm×0.4mm×20mm Gd3Al2Ga3O12 (GAGG) wrapped by reflector (BaSO4). And each end of the crystal had a SiPM. We let the gamma rays irradiate the crystal at different DOI and record energy deposition in the SiPMs. We also simulated the whole ring system to evaluate its spatial resolution. The PET system consisted of 8 detector modules and each module had 80×80 pixels. The inner diameter of the detector ring was ~98mm.

2.2.Experiment
A novel type of SiPM [1] which is very compact (500um pitch), has very fast recovery time (~8ns) and high gain (10^6) was used to develop the detector. A 0.4mm×0.4mm×20mm GAGG scintillator was placed between two SiPM and fixed by a supporter. After filtering and amplifying, signals were readout by a time-over-threshold (ToT) ASIC.

3.Result
The simulation result showed that the relationship between DOI and E1/Etotal is linear and DOI resolution at the center of the crystal (10mm) was ~0.76mm. The spatial resolution of the PET system can be improved to ~0.64mm. In the real experiment, a very small peak of 511keV can be seen from the spectra.

4.Future Work
In the future, we will calibrate DOI in the real experiment, and evaluate DOI resolution.

Reference:
[1] A. Koyama. Fabrication and Characterization of a 500-µm-Pitch 64-Channel Silicon Photomultiplier Prototype[J]. 2017.

Speaker: Ms Laiwen Zheng (the University of Tokyo)

19 Development of a large-area multilayer detector for single-shot dual-energy imaging

Contrast in x-ray radiography is mainly determined by the difference between linear attenuation coefficients of target and surrounding background. Dual-energy (DE) x-ray imaging typically assumes that an object to be imaged consists of two materials. By arithmetically controlling the magnitudes of attenuation coefficients measured at two different energies, DE imaging can enhance one material intensity while suppressing the other one. Therefore, DE imaging is capable of material-of-interest imaging without background clutter. Instead of using dual shots of different x-ray energies, DE imaging is also possible with a multilayer detector at a single shot accounting for the average energy increase of x-ray photons at the rear detector layer, which are survived from the transport through the front detector layer. Although a prototype multilayer detector was limited to a small size less than a mouse, the single-shot DE technique has been successfully applied to radiography [1] and tomography [2].
In this study, we introduce a large-area multilayer detector for the single-shot DE imaging and present its initial performance. The detector consists of two phosphor-coupled CMOS photodiode arrays. To support the front detector layer as well as to include readout electronics, flexible printed circuit boards are used as a substrate. We describe development of the multilayer detector and report quantitative performance, including some demonstration images.

Speaker: Mr Ho Kyung Kim (School of Mechanical Engineering, Pusan National University)

20 Development of prompt-gamma and positron emission tomography (PG-PET) system for in vivo dose verification in proton therapy: a Monte Carlo study

The interest in proton therapy is increasing worldwide thanks to its unique dose delivery characteristics, which can deliver conformal dose distributions to target volumes while minimizing damage to normal tissue. However, the proton therapy has a quite important issue for the treatment verification; in vivo real-time proton beam range verification. As the representative in vivo range verification systems, there are prompt-gamma (PG) imaging system and positron emission tomography (PET) system. The PG imaging system has high accuracy for measurement of the proton beam range, while the sensitivity is low. The PET system is able to estimate three-dimensional (3-D) dose distribution, while the accuracy of range determination is low. Therefore, we proposed a PG-PET system that combines the advantages of the two systems and optimized the PG-PET system using Monte Carlo method (figure 1).
The PG-PET system was designed in Geant4.10.00 with a 2-D scintillator array, two silicon photomultiplier (SiPM) arrays for a dual-ended system, and a parallel-hole collimator covering 64 x 64 mm2 area. The geometry of the collimator and the scintillator array, scintillation material, energy window (EW), and time-of-flight (TOF) window were optimized based on the radiation field generated by the interaction between 150 MeV proton beam and 30 x 20 x 20 cm3 water phantom. As an additional background-reduction technique, a depth-of-interaction technique (DOI) was proposed and assessed.
As a result of the optimization study based on the sensitivity of the system and the capability of measuring proton range, the thickness, the pitch and the hole sizes of the collimator were 200 mm, 8 mm, and 7 mm, respectively. The thickness and the pitch size of crystal bar were 30 mm and 4 mm, respectively. The LaBr3 and GAGG scintillator showed good performance considering light yield of the crystal and a figure-of-merit that we made in this study. Among 12 energy windows, 3-5 MeV energy window was taken to be the optimal energy window reducing the portion of background radiation 14.88% compared to the not applied case. In the case applied DOI and TOF techniques, the background was reduced by 17.38% and 57.36%, respectively.
In this study, we present the capability of proton beam range verification according to the change of geometrical parameters of the PG-PET system and background reduction techniques and offer the possibility of integrating PG imaging system and PET system.

Speaker: Mr HyunJoon Choi (Yonsei Univ.)

21 Dose distribution visualization by multi-pixel photon detectors containing lead(II) iodide for the development of radiation safety management system of radioactive source

Safety regulation of technologies like this has increasingly strengthened as national safety management systems have globally taken roots for the safe use of radiation. Significant amounts of time and resources have therefore been invested in developing a radiation safety management system to prevent radiation disasters in NDT; In particular, to perform gamma radiography safely, it is essential to determine the radioactive source location.[1,2] In the current industry field, to monitor the location of radioactive source, the research on gamma camera composed of scintillator that converts radiation to visible ray and photo-diode that converts visible ray to electrical signal has been performed variously. However, as compton recoil electron occurred by high energy radiation gets damage, photo-diode based on silicon might show single event latch-up and it would ultimately increase the possibility of electrical error.[3] Therefore, the development of a new detector which is suitable for NDT field that uses high energy radiation is necessary. In this paper, visualizes the dose distribution of radioactive source in real time, the research was performed with a purpose of developing MPPD(multi-pixel photon detectors) based on PbI2 (Lead iodide) in order to construct the safety management system of radiation that can prevent potential accidents in advance. To decide on the distance and CF (calibration factor) between pixels of MPPD, basic evaluation on the pixel detector was performed. Response characteristic and directional dependence based on the distance between radioactive source and pixel detector was each evaluated by PID (Percentage Interval Distance) and CV (coefficient of variation). Additionally, active area of pixel detector was produced with 1 x 1 cm2 and gamma ray source used Ir-192(effective energy is approximately 300keV). As a result of experiment on the response characteristic based on distance of pixel detector, PID value was shown as a curve form of exponential decay and maximum readable range was analysed to be 11.3 cm; considering this result, the distance between pixels of MPPD was decided to be 7cm. Also, as a result of experiment on directional dependence about effective angle of 8° ≤ θ ≤ 172°, difference of less than 1% was analysed; In order to visualize more accurate dose distribution, CF was drawn based on this. After that, in order to visualize the dose distribution, MPPD based on 7 pixel was produced and reproducibility and linearity on radioactive source were each evaluated by CV and R-sq(coefficient of determination). As a result of reproducibility experiment, the credibility of MPPD was verified as CV was analysed in less than 0.5%. Also, as a result of linearity experiment, all pixels was analysed to R-sq more than 0.99. Finally, conformity of dose distribution was verified based on the relationship between planned irradiation time and signal measure at MPPD. As a result of experiment on conformity of dose distribution, it corresponds well with the time rate which was planned by using a remote controller. Based on the research results, we verified MPPD, through visualization, based on PbI2 to confirm the dose distribution on radioactive source in real time. Also, It suggested the possibility as a safety management system of radiation that is suitable to international trend to use the NDT apparatus.

REFERENCES
[1] E. Massoud, Dose assessment for some industrial gamma sources with an application to a radiation accident, Open J. Model. Simul. 2 (2014) 4.
[2] G.T. Joo et al., Development of automatic remote exposure controller for gamma radiography, J. Korean Soc. Nondestruct. Test. 22 (2002) 490.
[3] Z. Li, Radiation damage effects in Si materials and detectors and rad-hard Si detectors for SLHC, JINST 4 (2009) P03011

Speaker: Mr YO HAN SHIN (Dept. of Radiation Oncology, Collage of Medicine, Inje University)

22 Double-side pixelated X-ray detector based on metamorphic InGaAs/InAlAs quantum well

Modern flat-panel x-ray imaging detectors have played an important role in the transition from analog to digital x-ray imaging. They capture an x-ray image electronically and hence enable a clinical transition to digital radiography. Research over the past two decades has indicated that the most practical flat-panel medical digital radiographic systems are based on a large-area integrated circuit or array of many single pixels. For both the indirect and direct conversion approaches, the latent image is a charge distribution residing on the array’s pixels. In other hands, current pixel detector technologies for X-rays at synchrotron radiation sources and Free electron laser are building a hybrid technology consisting of a passive sensor bump-bonded to an active read-out chip. Due to the high atomic number, low-energy band gap and high electron mobility, the opportunity to use of a metamorphic InGaAs/InAlAs quantum well-based devices was proposed for the fast pixelated photon detectors. In this work, we are presenting a double side-segmented QW device under discussion of spatial resolution and a cross-talk between pixels. The devices were tested with 10 keV, 11 keV, 14 keV and 20 keV photon energies with a beam size of 90×90µ2. The position of the synchrotron radiation was estimated with 1.3 µm precision. The charge spread in the material and related crosstalk function between pixels were extracted from the position estimation measurements of the photon beams. The results show that the cross-talk between pixels is actually responsible for the different resolutions obtained, independently of the experimental conditions, pointing out the importance of the geometry of the fabricated devices. The cross talk between the pixels is minimized when the QW is segmented.

Speaker: Tamiraa Ganbold

23 Energy dependence of two-dimensional Al2O3 thermoluminescence dosimeter in robotic radiosurgery

For the dose distribution verification in robotic radiosurgery which is one of stereotactic radiosurgery (SRS), X-ray films are used due to their high sensitivity uniformity and resolution. Two-dimensional (2-D) Al₂O₃ thermoluminescence dosimeters (TLDs) are reusable 2-D passive detectors with good TL properties [1]. Previous study has suggested that the 2-D Al₂O₃ TLDs had high applicability for quality assurance of robotic radiosurgery [2]. However, the influences of energy dependences for the Al₂O₃ TLDs in robotic radiosurgery are unknown. In robotic radiosurgery, energy spectrums of photons vary in accordance with the field size, source-to-axis distance (SAD) and depth in water. TL efficiencies also vary in accordance with energy spectrums of photons. For this reason, it is important to characterize energy dependences of the Al₂O₃ TLD for the verification of the 2-D dose distribution in robotic radiosurgery. In this study, we investigated the energy dependences on the field size, SAD and depth in water in robotic radiosurgery.
The 2-D Al₂O₃ TLDs were composed of Al₂O₃ (>99.5 wt%) doped with 0.05 wt% Cr₂O₃. The effective atomic number and bulk density was 11.14 and 3.7 g/cm³ respectively. The 63×63×0.7 mm³ Al₂O₃ TLDs were irradiated with 6 MV X-ray beams from CyberKnife®, which is an only robotic radiosurgery system, and then read out with a dedicated 2-D TL reader. The TL reader consists of a hot plate and a sensitive complementary metal-oxide-semiconductor (CMOS) camera with a heat absorption filter.
For the field size dependence, an Al₂O₃ TLD was set at a depth of 15 mm and with three constant SADs. The diameters of irradiation fields were 5, 10, 20, 40 and 60 mm. This geometry was the same as the measurement geometry of output factors. The SADs were 650, 800 and 1000 mm. For the depth dependence, an Al₂O₃ TLD was set at six depths with a constant SAD of 800 mm. The depths in water were 0, 15, 50, 100, 200 and 300 mm. The diameters of irradiation fields were same as them in the measurement for the field size dependence.
The TL response to the reference value by the semiconductor detector decreases as the field size decreases. This is because the mean energy of photon increases as the field size decreases. At the field size of 5 mm in diameter, the field size dependence was some 5%. There are small changes of field size dependences by different SADs. The TL response to the reference value decreases as the depth in water increases because of the beam hardening effect. Therefore, when applying the 2-D Al₂O₃ TLD to the absolute dose verification, we should use care about the energy dependences of it.

Speaker: Shin Yanagisawa (Tokyo Metropolitan University)

24 EVAULATIONS OF COMMON SOLID-STATE DETECTORS WITH MULTI-CRITERIA DECISION-MAKING THEORY FUZZY PROMETHEE

Solid-state detectors are radiation detectors which employ semiconductor crystals as the detecting medium. These detectors produce a pulse of electric current, by means of pairs of charge carriers, electrons and holes generated when the detectors come in contact with ionizing radiation. The type of semiconducting material used in the detectors determines the the number of paired charge carriers generated, which is a very important parameter that affects the performance of the device. Other parameters to be considered in determining the performance of the solid-state detectors include energy resolution, atomic number, density, and electron/hole mobility. The trend in development of new cameras for nuclear medicine applications is the principal reason behind the extensive research into detectors with superb spatial resolution and energy resolution. Semiconductor based detectors have received much attention due to their high energy and intrinsic spatial resolution. Although various types of such detectors exist, to our knowledge no study has categorically provided the best amongst the existing detectors. Fuzzy PROMETHEE is a multi-criterion decision-making method (MCDM) that has been applied in many fields to solve selection problems involving multiple criteria. We applied the MCDM technique in order to evaluate the various semiconductor materials using their distinct physical parameters including density, bandgap, energy resolution and electron/hole mobility. Evaluation results showed that Germanium is the best detector, mainly due to relatively very high electron/hole mobility, for use in fast nuclear medicine applications, on the ranking followed by Gallium Arsenide, then Cadmium Telluride and Cadmium Zinc Telluride based on the selected weights. Results also showed that Silicon and Thallium Bromide came last on the ranking, having a low performance value from the most important parameters. By using fuzzy PROMETHEE, one can determine the most suitable crystal based on the selected criteria and defined weight according to desired application.

Speaker: Dr Ilker Ozsahin (Department of Biomedical Engineering, Near East University)

25 Feasibility Study on Current-mode SPECT for B-10 Concentration Estimation in Boron Neutron Capture Therapy

Boron neutron capture therapy (BNCT) is a promising method to treat invasive cancers. In BNCT, medicine including boron-10 (B-10) which will be delivered to cancer tissue is injected into blood vessel. With the neutron absorption reaction of B-10, two charged particles, i.e., alpha particle and Li-7, are emitted. Because the ranges of these charged particles are similar to the size of a cell, canter cells are killed by neutron irradiation.
For the estimation of B-10 concentration, prompt gamma rays with 0.478 MeV in energy emitted by Li-7 are utilized. The background gamma rays, however, prevent from measuring 0.478 MeV gamma rays at the BNCT treatment facility: high flux neutrons create 2.2 MeV gamma rays by H(n, gamma)D reactions and accompanied 0.511 MeV annihilation gamma rays. For the measurement of 0.478 MeV gamma rays by photon counting method, a very thick collimator is necessary as demonstrated in Ref. [1].
For overcoming high rate background gamma rays and obtaining B-10 concentration image, we have been studying the feasibility of current-mode single-photon emission computed tomography (SPECT) with applying the principle of a “transXend” detector, which we have invented for energy-resolved X-ray computed tomography [2]. In this method, all the gamma rays are measured as electric current. After unfolding analysis, a gamma ray energy spectrum is obtained.
Because the number of annihilation gamma rays is proportional to the one of 2.2 MeV gamma rays, total number of 0.478 and 0.511 MeV gamma rays is obtained first, and the number of 0.478 MeV gamma rays, which is proportional to the B-10 concentration, is estimated.
A phantom used in the calculation was 18 cm diameter 20 cm high water with 5 cm diameter cancer with various B-10 concentrations. In the simulation study, the gamma ray energy spectrum induced by neutron reactions was calculated by PHITS code [3]. The transXend detector consisted of four TlBr segmented detectors with 5×5×10 mm3. The electric currents measured by each segmented detector were estimated. With unfolding analysis and applying a neural network method, images of B-10 concentration distribution in the phantom were shown.
[1] T. Kobayashi, et al., Med. Phys., 27, 2124 (2000).
[2] I. Kanno, et al., J. Nucl. Sci. Technol., 45, 1165 (2008).
[3] T. Sato, et al., J. Nucl. Sci. Technol., 50, 913 (2013).

Speaker: Prof. Ikuo Kanno (Kyoto University)

26 Improvement of the sensitivity of organic polymer-based photodetector fabricated with p-type conjugated polymers for indirect X-ray detection

In this study, an organic polymer-based photodetector was investigated as a candidate for the indirect-type X-ray detector. Organic photodetectors fabricated with polymers have the advantage of being able to apply flexible substrates and simple processes while reducing manufacturing costs. In order to enhance the performance of the photodetector, conjugated polymer PBDB-T1) was selected as a p-type material instead of P3HT2) commonly used as a p-type material to form a bulk-heterojunction with n-type PC70BM3) (Fig. 1a). Compare with P3HT, PBDB-T has more alkyl chain and quinoid character. The alkyl chain facilitates the intermolecular covalent bonds and helps to form the percolation paths with n-type material. Quinoid helps electrons migrate from p-type to n-type and leads to high short circuit current in the photodetector. Absorption spectra of the thin film with 1:1 blending ratio of P3HT: PC70BM and different blending ratios of PBDB-T:PC70BM were measured (Fig. 1b). PBDB-T:PC70BM-blended thin film showed higher absorption in the entire visible-light range and the absorption spectrum of PBDB-T:PC70BM was well-matched to CsI(Tl) emission spectrum. The organic photodetectors were fabricated with different blending ratios and spin-rates of PBDB-T:PC70BM. The detector fabricated at the condition of PBDB-T:PC70BM = 2:3 blending ratio and spin rate of 1,100 rpm showed the highest collected current density (CCD) and highest sensitivity which were 434.14 nA/cm2 and 3.25 mA/Gy*cm2, respectively (Fig. 1c). This is 152% enhancement of CCD and 189% enhancement of sensitivity than the common P3HT:PC70BM photodetector.

Speaker: Mr Daehee Ban (Dankook University)

27 Investigation of Low Temperature Bonding Process using Indium Bumps

Hybrid pixel detector modules are state of the art in vertex detectors of the high-energy physics as well as the x-ray cameras in synchrotron radiation experiments. Each module of such a detector consists of a sensor chip and one or more electronic readout chips. In order to connect every pixel on the sensor with an electronic readout cell both parts are bump bonded together. In comparison to the reflow soldering process with In or InSn thermo-compression bonding enables an even lower temperature to form the interconnection. Such a low temperature bonding process is required especially for high-Z sensor materials like Germanium, Cadmium-Telluride, Cadmium-Zinc-Telluride and MCT-sensors due to their limited temperature budget. The investigation presented here focus on a compression bonding process that will be applicable for such sensors. This low temperature compression bonding process is based on the use of a ductile bump material like indium and a dedicated UBM surface that is able to form intermetallic compounds with the bump metal.
A special test chip design was developed in order to investigate the influence of bonding parameters on the quality of the interconnection. The chip size of the test chip is adapted to the size of MEDIPIX3 readout chip with an identical number of bumps arranged in a 256 x 257 bump matrix with 55µm pitch in X- and Y-direction. When flip bonding this chip onto the sensor-test-chip, daisy chain test structures are formed on relevant positions of the resulting test module. In addition, four-point-kelvin-test structures have been designed to measure the electrical resistance of individual bumps. Other test features deal with the investigation of bonding pressure related to the risk to create shorts between adjacent bumps.
In order to achieve comparable results between daisy chain test chips and functional MEDIPIX readout chips and sensor chips a similar bonding metal layer stack was chosen [1]. Indium bumps were deposited by electroplating on the readout-test-chip side. Two different types of sensor-test-chips were designed. One pad defined type has a NiAu pad layer stack on top of a passivation layer. A second type of sensor-test-chip was designed with a passivation opening defined bonding area which means an Au-finished metal pad circumferential covered by a SiO2-passivation.
The investigation of the bonding process includes the variation of the bonding pressure, the bonding temperature and the dwell time of bonding pressure. A good quality of the bonding process was defined by a minimum value of the interconnection resistance and a closed daisy chain of bump interconnects. After electrical characterization pull tests were performed to investigate the corresponding bonding area. Good bonding results were achieved at temperatures of 80°C and 100°C. A minimum bonding pressure of 3.2 MPa is necessary to get a reliable bonding result. Daisy chain failures occur especially in the corner of the chips. Comparing the two different UBM pad structures we found a narrow distribution with less variation of all resistance values for the NiAu pads on top of the passivation layer. Pull test results show a larger bonding area for this UBM pad in comparison to the passivation opening defined bonding pad.
This project was performed in cooperation with the Photon Science Detector group of DESY, Hamburg. The author would like to thank Prof. Dr. Heinz Graafsma and his group for their support.

[1] F. Broennimann and et. al.: Development of an indium bump bond process for silicon pixel detectors at psi. Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment, 565(1):303–308, 2006.

Speaker: Thomas Fritzsch (Fraunhofer IZM)

28 Linear approximation of single-shot dual-energy computed tomography with a multilayer detector

Stacking two imaging detectors in tandem configures a multilayer detector, and the front detector measures relatively low-energy photons whereas the rear detector measures relatively high-energy photons survived from their transports through the front detector layer. This energy separation in the multilayer detector facilitates dual-energy (DE) imaging. Different phosphor thicknesses between the two detector layers provides a unique characteristic, known as unsharp masking-like effect, in the resultant DE images as shown in Fig. 1. This characteristic results from the weighted combination of different modulation-transfer functions (MTFs) of the corresponding two detector layers [1].
We are developing a single-shot DE computed tomography (CT) technique [2]. For a better design of the multilayer detector and its application to DE-CT, it is important to understand the effect of the subtraction DE algorithm on the resultant DE images. In this study, we develop a simple signal-formation model in DE-CT incorporating filtered backprojection and weight-log subtraction for image and DE reconstructions, respectively. We validate the developed model in comparisons with the measured MTF results obtained for operation parameters such as subtraction weights and reconstruction filters. This study will be helpful to understand the principle and physics underlying the single-shot DE-CT based on the multilayer detector, and from which we may elaborate the system design and find appropriate applications.

Speaker: Prof. Ho Kyung Kim (School of Mechanical Engineering, Pusan National University)

29 Noise spectral density calculation of different CdTe Metal-semiconductor-metal contacts

The Cadmium telluride Schottky diodes are widely used for several applications [1], specifically as spectroscopic detectors. The actually CdTe Schottky devices are composed of thick semiconductor (S) layer (in mm) coated by two metals (M) contacts, therefore they have M–S–M common structure form. The CdTe M-S-M structure presents a high signal-noise ratio depends to the intrinsic noise level. Particularly, the fluctuations of the electric field at Schottky contacts lead to the fluctuations of leakage current [2] where their properties change by the choice of metal contact. An additional component of noise becomes from the fluctuations of free carrier inside the whole Schottky barrier region, which depend to the perturbation of the electric field inside the whole structure.
This contribution presents an analytical approach for the current noise calculations of Schottky CdTe for three different contacts (Au, Al, and In). The calculations taking into account the fluctuations of the electric field, of the leakage current through Schottky barrier at the M-S interface, and of the surface charge density current. The fluctuations of electric field in different CdTe Schottky contacts exhibit a sharp resonance followed by decreasing in gigahertz and terahertz frequencies, respectively (see Fig. 1).

Speaker: Prof. Franc J (3Faculty of Mathematics and Physics, Institute of Physics, Charles University)

30 Noise-reduction approach to single-shot dual-energy imaging with a multilayer detector

Dual-energy (DE) x-ray imaging can improve lesion conspicuity by suppressing anatomical background noise. Compared to the conventional radiography, however, DE imaging results in relatively higher image noise because subtraction of two energy images increases the resultant DE image noise, hence lowering signal-to-noise performance. Degradation of noise performance becomes severer in single-shot DE imaging, which uses a multilayer detector, because of high quantum noise in high-energy image due to the lower number of photons attenuated through the front detector.
We are developing the single-shot DE technique based on the multilayer detector. For more practical use of this technique, we apply noise-reduction strategy to high-energy images obtained from the rear detector layer before DE subtraction [1]. Figure 1 compares images obtained for a post-mortem mouse using a multilayer detector, and the Gaussian noise-reduction operation to the high-energy image shows a better DE image quality. However, the noise-reduction operation may yield an adverse effect in high-energy and DE images, for example, image blur. In this study, we investigate quantitatively the effects of various noise-reduction algorithms on the DE image quality. This study may find a best noise-reduction algorithm with less image distortion for the successful single-shot DE technique.

Speaker: Mr Dong Woon Kim (School of Mechanical Engineering, Pusan National University)

31 Numerical simulations of CMOS sensors for high-precision silicon trackers

The tracking system of the Circular Electron Position Collider (CEPC) [1] proposed by the Chinese particle physics community includes a high-precision silicon tracker. In order to fulfill the high precision track reconstruction, the single point resolution should be < 7 µm, the material budget < 0.65% X$_{0}$ and the radiation hardness > 1 MRad for this tracker.

A new CMOS pixel sensor based on the rolling-shutter readout is under development for the CEPC tracker by using a 180 nm CMOS imaging sensor technology [2]. The prototypes were fabricated with two different wafers in terms of the resistivity and the growing method. One is a high resistivity epitaxial layer grew on a low resistivity substrates, and another is a purely high resistivity substrate. This high resistivity based technology could result in better charge collection efficiency and improved radiation hardness. The prototypes are aimed at the investigation of the charge collection efficiency depending on various pixel geometries and pitches.

In order to explore the electric properties and optimize the sensor charge collection efficiency for this CMOS technology, the two dimensional and three dimensional Technology Computer Aided Design (TCAD) simulations have been performed carefully. The simulations have been carried out to investigate the depletion, the breakdown voltage, the leakage current, the capacitance as well as the charge collection behaviors for the different two kinds of wafers. Both the Non Ionizing Energy Loss (NIEL) and the Total Ionizing Dose (TID) effects are introduced into the simulation in order to investigate the sensor property degradation after irradiation. The above mentioned simulation results will be shown in details. Some comparisons between the simulations and the latest measurements will be addressed as well.

REFERENCES:
[1] The CEPC-SPPC study group, CEPC-SPPC preliminary conceptual design report -
volume 1: physics and detectors, (2015) [http://cepc.ihep.ac.cn/preCDR/main_preCDR.pdf ].
[2] L. Zhang, et al, Investigation of CMOS pixel sensor with 0.18 µm CMOS technology for high-precision tracking detector, 2017 JINST 12 C01011

Speaker: Dr Jian LIU (Shandong University)

32 PVT based Neutron Monitor for Monitoring of Nuclear Smuggling in Cargo Container

ABSTRACT
Generally, nuclear material monitoring in container security uses He-3 to measure neutrons generated from nuclear materials. But due to the decrease in the supply of He-3, the development of nuclear material monitors using alternative materials has been performed[1]. Generally, in order to measure neutron, secondary products (electrons, charged particles, and light) which is generated by the reaction of a neutron with a sensor should be measured. Gas-filled proportional counter using a BF3 or 10B thin film has been used to monitor nuclear material in large size container because it is easy to manufacture with large-area. However, BF3 is a toxic gas and has limitations in increasing the gas pressure[2]. And in the case of using 10B thin film, due to the low absorption efficiency, it is necessary to construct a multi-layered thin film of several m thick[3]. For this reason, it is necessary to fabricate large-area sensor at low price in order to develop high efficiency neutron monitoring system. For this purpose, we have conducted a study on a hybrid system that combines PVT (polyvinyltoluene) to measure  -ray and sensors that react with neutrons to generate  -ray. The neutron reactors used in this study are boron, Cd and Gd, and the size of PVT is 1001005cm3. The detecting efficiency of nuclear materials in container using each materials was measured by MCNP. To calculate the detection efficiency, neutron (Cf-252, surrounded by 1-inch-thick HDPE) were generated at 2 m from the detector surface. As shown in the below figure, efficiency of Cd was the lowest and efficiency of Gd was the highest. As a result, it was found that 2 to 3 cps for 1 ng Cf-252 source, which is the efficiency of a general vehicle neutron detector, was satisfied, and the system can be configured using RPM and it also will be advantageous in terms of price.

ACKNOWLEDGMENTS
This research was supported by a grant from National R&D Project of “Research on Fundamental Core Technology for Ubiquitous Shipping and Logistics” funded by Korea Institute of Marine Science and Technology Promotion(PMS3791).
REFERENCES
[1] Timothy M. Persons, Gene Aloise, Neutron Detectors Alternative to using helium-3, Report to Congressional Requesters, GAO-11-753, pp. 1-48
[2] A. Lintereur, K. Conlin, J. Ely, L. Erikson, R. Kouzes, E. Siciliano, D. Stromswold, M. Woodring, 3He and BF3 neutron detector pressure effect and model comparison, Nuclear Instruments and Methods in Physics Research A, vol. 652, pp. 347-350.
[3] C. Höglund, J. Birch, K. Andersen, T. Bigault, J-C. Buffet, J. Correa, P. van Esch, B. Guerard, R. Hall-Wilton, J. Jensen, A. Khaplanov, F. Piscitelli, C. Vettier, W. Vollenberg, and L. Hultman, B4C thin films for neutron detection, Journal of Applied Physics, vol. 111, pp. 104908-1-8.

Speaker: Mr Kim Ik-Hyun (KRISO)

33 Response function of GaAs:Cr X-ray sensors

In this work an investigation of current-voltage dependencies on temperature and time were carried out in order to analyze relaxation time of electric field distribution in Me-GaAs:Cr-Me structures. To make conclusions of the possibility of deep levels recharging the charge collection efficiency dependency on the temperature was measured when irradiating with gamma radiation source and IR pulses. Slight increase of electron lifetime was observed with the increase of the temperature which corresponds to the improvement of the response function of GaAs:Cr sensors shown before [1, 2]. For the tests pad GaAs:Cr samples with active area size 3×3 mm^2 and thickness 500 μm were used.
The calculation of photocurrent dependency on the dose of irradiation was done. Saturation of the current is observed with the increase of the dose. This can be explained with the change of electric field distribution over a time under the generation of big charge – when increasing the applied voltage the dependency becomes linear.
The research was supported by The Tomsk State University competitiveness improvement programme.

1. A.N. Zarubin et al., Evaluation of suitability of GaAs:Cr sensors for X-ray transmission technology of diamond-bearing ore enrichment, Nuclear Science Symposium, Medical Imaging Conference and Room-Temperature Semiconductor Detector Workshop (NSS/MIC/RTSD), 2016, DOI: 10.1109/NSSMIC.2016.8069941
2. J. Becker et al., Characterization of chromium compensated GaAs as an X-ray sensor material for charge-integrating pixel array detectors, 2018 JINST 13 P01007, https://doi.org/10.1088/1748-0221/13/01/P01007

Speaker: Ms Anastasia Lozinskaya (Tomsk State University)

34 Simulation of Particle Tracks in a Timepix Detector

This presentation demonstrates a method for the simulation of particle tracks in Timepix hybrid pixel detectors using Geant4.

NASA uses Timepix hybrid pixel detectors for several applications including measurement of radiation on the International Space Station and the Orion Spacecraft. An important goal of these measurements is the development of algorithms for particle charge and energy ID. An important tool for this development is software that can produce accurate simulations of detector response for particles of arbitrary charge, energy, and incidence angle. To do this we use a Geant4 Monte Carlo simulation coupled with a model of the detector response.

The typical approach to this problem seen in the literature is to use a Monte Carlo simulation coupled with a drift/diffusion charge transport model and some simulation of detector characteristics such as pixel noise and threshold effects. These methods work extremely well for particles with low stopping power such as protons, often producing results that are excellent matches for experimental data. However they do not accurately reproduce the shapes of tracks for heavy ions. These tracks feature a characteristic ‘core’ region corresponding to the primary track with far more charge sharing than would be expected from drift/diffusion and an extensive `skirt’ region from charge induction effects.

Our approach uses a semi-empirical charge sharing model based on a large database of beam measurements with Timepix sensors. Simple models are developed for both the charge sharing in the core and induction of the skirt. The free parameters in these models are then minimized against experimental data for several different datasets. It is shown that these parameters are largely governed by the particle stopping power. The volcano effect (a saturation effect caused by the behavior of the Timepix front end) can also be simulated resulting in accurate predictions of stopping power for a variety of heavy ions.

This approach differs considerably from a first principles approach in that the model is specifically tuned to 500 μm silicon Timepix detectors. However, it allows rapid simulation speed and accurate reproduction of clusters for all tested species. Particular emphasis is placed on the computational reproduction of experimental observable quantities such as the stopping power, delta electron distributions and reproduction of the volcano effect.

Speaker: Dr Stuart George (University of Houston, NASA Space Radiation Analysis Group)

35 Sparse Compton X-ray imaging using MAPS technology

X-ray crystallography is one of the well-known inventions of the 20th century. It is even more impressive considering the facts that both the X-ray sources and detectors were primitive compared with what is available more than one century later. Cast in the context of recent advances in sparse imaging theory and practice, scene redundancy, or atomic periodicity in the case of crystallography, is the key physical basis for the success of X-ray crystallography. Opportunities exist to further apply the concept of redundancy reduction in imaging, including X-ray imaging. Here we summarize the current status of X-ray imaging enhanced through sparse methods, followed by new analysis, simulations, and initial experimental data on using hard X-rays for sparse Compton X-ray imaging. The imaging hardware is based on Monolithic Active Pixel Sensors (MAPS), a technology being developed for high-luminosity colliders. Some of the attractive features of MAPS include high-spatial resolution, small material footprint, low cost, small dynamic-range pixels, high-speed (for the next generation in particular) and radiation hardness, which make them one of the best choices for a sparse Compton X-ray imaging demonstrator. Sparse Compton X-ray imaging can be used to enhance the performance in many traditional areas of Compton imaging.

Speaker: Dr Zhehui Wang (Los Alamos National Lab)

36 Study on photon counting detector using CZT and CdTe

The energy integrating method X-ray imaging devices have made many advances in medical diagnostics, nondestructive and security detection. In order to solve the problem of excessive X-ray exposure to the object which is the biggest disadvantage of the X-ray measurement by the integrating method. Researches on photon counting method using CdTe and CZT in the low energy X-ray imaging device have been actively conducted recently. The photon counting method imaging device can be applied to industrial and medical applications because of the sensitivity of the detector material is higher than that of typical scintillators for X-ray detection and this method is capable of distinguishing X-ray energy.
In this study, we developed a prototype detector module for image acquisition using analog signal processing ASIC chip developed to apply CZT and CdTe (4X4 and 1X16) detector for linear scanner. We tested the performance of energy resolution, detector response as like count rate and validate imaging etc.
The most important feature of the analog signal processing circuit used in this study for developing the low energy X-ray linear scanner is that it can output 5 digital values according to the charge pulse size from the readout ASIC.

This research was funded by the MSIP ICT R&D Program 2016 (2016M2A2A4A04913449), National Research Foundation of Korea (NRF-2015M3A7B7045525) and KI institution-specific project (N10180002) of 2018 KAIST's own research projects.

Speaker: Prof. Gyuseong Cho (Dept. of Nuclear & Quantum Engineering, KAIST)

37 Temperature Sensor based on PEDOT:PSS Ink-Jet printed on Paper Substrate

Printed electronics is rapidly developing, where more and more components are printable. High speed roll-to-roll processes are preferred for low cost production of flexible electronics. Often, the substrates used for printed electronics are some type of plastic such as PET or Kapton. An alternative to plastic is to use paper substrate that has the benefits of being environmentally friendly, recyclable and renewable. Paper is also the material of choice for packages of various goods.
In this work we have developed an ink-jet printable temperature sensor, a thermistor, consisting of poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate)( PEDOT:PSS) on paper substrate. The starting material is a commercial PEDOT:PSS ink-jet ink (Orgacon IJ-1005, Agfa). This ink was then modified to increase the thermal sensitivity by addition of the co-solvents Dimethyl sulfoxide (DMSO) and Polyethylene glycol (PEG) in different quantities. DMSO has been shown to increase the conductance by arranging the PEDOT into more conductive pathways and by removing PSS [1] and PEG to increase the carrier density and mobility [2].
The sensors consisted of modified PEDOT:PSS lines printed on photo-paper substrate between contacts printed using silver nano-particle ink. The line widths were varied from one pixel, corresponding to one pass of one nozzle up-to 20 pixels. The linewidths were then measured to be from 45 µm up-to 450 µm. The thickness of the sensor was also varied as one, two or three printed layers. The characterization as a temperature sensor was performed by using a setup consisting of a Peltier cooler and a heating element to step the temperature. As a reference a PT-100 element fixed to the surface of the cooler/heater was used.
An increase in resistance from 30.5 MΩ to 85 MΩ, corresponding to a change of 2.8 times, were measured when the temperature were changed from 22 °C to -12 °C as can be seen in the figure. This gives a ΔR/ΔT of 0.093.

Such a printed sensor can be used for applications where a low cost, printable solution is needed, such as printed directly on packages, for environmental monitoring and similar.

[1] C.S. Pathak, J.P. Singh, R. Singh, Effect of dimethyl sulfoxide on the electrical properties of PEDOT:PSS/ n-Si heterojunction diodes, Current Applied Physics, 15, (2015), 528-534
[2] Yow-Jon Lin, Wei-Shih Ni and Jhe-You Lee, Effect of incorporation of ethylene glycol into PEDOT:PSS on electron phonon coupling and conductivity, Journal of Applied Physics 117, (2015), 215501

Speaker: Mr Mašlĺk Jan (Tomas Bata University in Zlin)

38 The Adaptive Gain Integrating Pixel Detector for fast experiments at the European XFEL

Adaptive Gain Integrating Pixel Detector (AGIPD) is designed in a collaboration between Deutsches Elektronen-Synchrotron (DESY), Paul-Scherrer-Institut (PSI) and the Universities of Hamburg and Bonn. It is a hybrid pixel X-ray detector developed for the new cutting-edge science experiments at the European XFEL, whose key features are the high brilliance coherent pulses with a specific bunch structure [1]. The detector should be able to capture multiple images at a 4.5 MHz burst mode frame rate with a dynamic range from 1 to 10⁴ 12.5 keV photons. In order to fulfill the mentioned requirements an Application Specific Integrated Circuit (ASIC) was designed [2]. The ASIC has 64x64=4096 pixels in total, each containing an analogue signal pipeline, driven by the digital periphery block. The input interface are three CML signals and the output are 4 analog differential lines.

A hybrid of front-end module, consisting of a silicon sensor bump-bonded to 16 ASICs, contains 128x512 pixels, 200x200 µm² each, and is served by back-end electronics including external ADC and FPGA boards. The 1M pixel detector incorporates four quadrants of 4 modules each, operating in vacuum. The output of the detector via 16x10G optical links makes the data handling a challenge. Behind the successful start of operation of Eu-XFEL in the September 2017 there were a lot of efforts for different parts involved. From the detector side it was a big challenge for both Eu-XFEL and DESY teams to commission the AGIPD at the SPB instrument in a very short time before the first experiments started, although some precision work is still ongoing. The commissioning itself included mechanical, electrical and infrastructure integration of AGIPD into the Eu-XFEL, low level tuning and optimization of the detector operation for a given beam structure. A big task is also to perform an on-a-beamline calibration of the detector including the processing of Terabytes of calibration data to enable the correction of scientific data. It is particularly difficult to calibrate the detector in the Eu-XFEL-specific run mode and to estimate the precision of data post-correction.

One system was installed and commissioned (see Figure 1 for an example of commissioning activity) at the SPB experimental station of the European XFEL. A second 1M system is ready for the delivery to the MID station. A 4M version for the SFX experimental station is now in development, some of the system's building blocks are already under test. Assembly, commissioning and applied calibration efforts will be presented.

References:

[1] M. Altarelli et al., European X-ray Free Electron Laser. Technical design report, (2006) [ISBN:978-3-935702-17-1].

[2] A. Klyuev et al., Front end ASIC for AGIPD, a high dynamic range fast detector for the European XFEL. Journal of Instrumentation 11 · January 2016. DOI: 10.1088/1748- 0221/11/01/C01057

Speaker: Alexander Klyuev (Deutsches Elektronen-Synchrotron)

39 The influence of contact material and its fabrication on X-ray GaAs:Cr sensor noise characteristics

In the report the investigation results of GaAs:Cr sensors noise characteristics by means of amplitude spectrum analysis are presented. Test sensors were 3×3 mm^2 “Me-GaAs:Cr-Me” structures and thickness was 500 μm. Metal contacts were made by means of electron beam deposition, magnetron sputtering and electro-chemical deposition of various metals such as Ni, Ni/Au, Cr, Cr/Ni, Cr/Al, NiV/Au, NiV/Al.
Noise characteristics dependencies on applied bias in the range from 0 to 500 V were investigated at different shaping times (50-500 ns). Energy resolution (FWHM) of the sensors was studied during electron-hole pairs generation using IR (830 nm) laser with the pulse width less than 1 ns. For these measurements sensors with mesh contact were fabricated.
The nature of dominant noise, optimal shaping time and extreme energy resolution are determined by analyzing obtained results. The conclusion of contacts fabrication method characterized with lowest noise is drawn.
The research was supported by The Tomsk State University competitiveness improvement programme.

Speaker: Mr Ivan Chsherbakov (Tomsk State University)

40 Virtual monochromatic computed tomography based on compressed-sensing theory for improving image characteristics

Since the introduction of dual-source computed tomography (CT) scanners, dual-energy CT (DECT) has been popularly used in many clinical applications for material decomposition, contrast enhancement, and artifact reduction in conventional CT images. In addition, DECT may synthesize monochromatic CT images at different x-ray energies, which can be used for routine diagnosis similar to conventional polychromatic single-energy images. These images utilize all the radiation dose in the dual-energy acquisition, and are typically reconstructed using a common filtered-backprojection-based algorithm. However, the quality of the resultant monochromatic CT images is often impaired; artifacts by scatter and noise or photon starvation still may exist in the reconstructed images. In this study, we investigated a recently popular compressed-sensing (CS)-based algorithm to produce virtual monochromatic images of less artifacts and thus to improve the image quality. CS is an innovative signal processing technique used in data acquisition and image reconstruction by finding solutions to underdetermined linear systems. A systematic simulation and experiment was performed to evaluate the effectiveness of the proposed approach. Polychromatic projections were obtained at two tube potentials of 80 kVp and 140 kVp and monochromatic CT images were reconstructed using the CS-based algorithm after applying projection-domain dual-energy processing. We compared the quality of virtual monochromatic CT images to that of polychromatic single-energy images, acquired at different tube potentials and the same radiation dose.

Speaker: Ms Jeongeun Park (Yonsei University)

Social event: Guided tour Sundsvall Stonetown

Social event: Archipelago Cruise M/S Medvind

Tuesday, 26 June

Invited: CANCELLED

41 Latest development of high effective X-ray detectors using high-z materials

Speaker: Michael Fiederle (University Freiburg)

Oral

Convener: Heinz Graafsma

42 Evaluation of a Compton Camera Concept Using the 3D CdZnTe Drift Strip Detectors

At DTU Space an early version of a 3D CZT drift strip detector prototype has successfully been developed and fabricated [1]. The design of the prototype detector (20 mm x 20 mm x 5 mm) is based on the CZT Drift Strip detector principle [2]. It contains 12 drift cells, each comprising one collecting anode strip with 4 drift strips, biased such that the electrons are focused and collected by the anode strips. The electrode geometry of these prototype detectors was optimized for having anode pitch of 1.6 mm and cathode pitch of 2 mm. A theoretical 3D CZT detector model has been developed using COMSOL for electrostatic calculation and IDL for charge collection and signal formation to display the expected pulse shapes for given interactions position. Furthermore, the model also calculates the electron and hole trapping during the charge movements. A readout technology and algorithm were developed, resulting in excellent position determination in 3D, and energy determination for high energies (>50 keV - few MeV, the developed algorithm is tested with a narrow beam at 662 keV). However, until this point the algorithm only comprised position and energy determination of single hit events.
The signal formation for all electrodes provides 3D position information of the interaction together with enabling a 3D correction of residual charge trapping. It also allows for detection and characterization of multiple interaction events occurring in the same or multiple detector cells, e.g. a Compton interaction with subsequent photoelectric absorption. In this project, the position determination of events with multiple hits were analysed and an algorithm developed for multiple interaction events. Compton imaging capability and achievable angular resolution of the single 3D CZT detector is evaluated with a point source illumination using Cs-137 source. The simulation platform MEGAlib [3] is used for image reconstruction from position and energy information provided by the developed algorithm. Furthermore, a 3D CZT Compton camera instrument model is evaluated with MEGAlib simulations of realistic source illumination scenarios to provide spatial and spectral performance figure of merit.

Speaker: Mrs Selina Howalt Owe (DTU Space, Technical University of Denmark)

iworid2018_presentation_selinahowaltowe.pdf

43 Intrinsic XRF correction in CdTe Timepix3 spectral detectors

ADVAPIX is an energy dispersive hybrid pixel detector based on a semiconductor sensor (Si/CdTE/CZT/GaAS) and developed for X-ray spectral imaging (3-200 keV). The sensor is read out by the recently developed Timepix3 ASIC. Timepix3 is formed of 256 x 256 pixels with a pixel pitch of 55 μm developed by the Medipix consortium. Each Timepix3 pixel is capable of recording simultaneous Time-Over-Threshold (Energy) and fast-Time-of-Arrival (fTOA, resolution: 1.56 ns) up to 0.4 Mcts/s/mm2. The fTOA information is used to recover the pixel clusters corresponding to a single photon interaction, which significantly suppresses the charge sharing artifacts in the energy spectrum. The fTOA information can also be used to identify simultaneous photon interactions in separated parts of the detector sensor. By combining fTOA and TOT (Energy) information, it is thus possible to identify intrinsic XRF photons from the detector sensor as well as the photons with reduced energies, and correct for them in the spectrum and spectral images. In this work, we present a new energy dispersive, 1 mm thick, CdTe detector (ADVAPIX-Timepix3) and we use simultaneous TOT and fTOA information to correct for artifacts caused by the intrinsic XRF photons from the CdTe sensor.

The attached figure shows an Americium 241 spectrum with and without the XRF correction. The original spectrum shows the peak close to 59 keV from the Americium 241 decay lines. It also shows the intrinsic XRF emission from the CdTe sensor close to 23 keV as well as the escape peak. The corrected spectrum shows significant suppression of the XRF peak and the escape peak as well as an increase in the photon counts in the emission peak.

Speaker: Dr Mohamad Khalil (Technical University of Denmark)

180626_Intrinsic XRF correction in CdTe Timepix3 spectral detectors_iWoRiD20.pdf

180626_Intrinsic XRF correction in CdTe Timepix3 spectral detectors_iWoRiD20.pptx

44 GaAs sensors in combination with charge-integrating, pixelated FEL detectors

Abstract: The Detector Group of the Swiss Light Source (SLS) at the Paul-Scherrer-Institute (PSI) has a long history of developing detectors for the needs of synchrotron sources. The advent of XFELs (like SwissFEL, which is constructed at PSI) has triggered the development of low noise, charge-integrating detectors with a high dynamic range of up to 104 . 12.4 keV photons per pixel. This high dynamic range per pixel has to be achieved without sacrificing single photon resolution in case of few incoming photons. Additionally the noise performance has to be better than the Poissonian fluctuations of the source at each point of the dynamic range to ensure the same data quality as single photon counting detectors. The SLS Detector Group has developed a novel method called dynamic gain switching, which is adapting the gain of each pixel to the incoming photon flux by dynamically adding additional capacitors into the feedback loop of the charge sensitive preamplifier.
JUNGFRAU is a charge-integrating pixel detector with a pixel pitch of 75 x 75 µm2, employing the aforementioned dynamic gain switching method to extend the dynamic range. Various high-Z sensors have been (GaAs and CdTe) has been investigated in order to extend the usable range of photon energies of our detectors. As JUNGFRAU has been developed for silicon sensors, dynamic gain switching is only working in hole collection mode. Therefore in case of using high-Z sensors which are preferable operated in electron collection mode, the gain stage (high, medium or low) has to be chosen in advance (fixed mode). When operated in fixed high gain mode the noise performance is 82 e- ENC and the available dynamic range is 71 keV.
Charge-integrating detectors offer interesting insights into the sensor properties as each pixel provides a direct measure of the collected charge of a well-defined area as output. On the other hand, each pixel is very sensitive to temporal as well as spatial sensor effects, which are affecting the charge collection (e.g. fluorescence, dislocation lines, detrapping of charge carriers etc.)
Over the last year several beamtimes have been performed at the Swiss Light Source (SLS) and the ESRF to test the usability of high-Z sensor materials (GaAs and CdTe) in combination with the JUNGFRAU readout chip. This presentation will focus on the results obtained with the GaAs-JUNGFRAU system. Amongst others the effect of the short hole lifetime of the GaAs sensors will be discussed and quantified.

Speaker: Dominic Greiffenberg (PSI - Paul Scherrer Institute)

iWoRID_Greiffenberg_GaAs_23Jun2018.pptx

Swedish fika in the hotel lobby (2nd floor)

Oral

Convener: Ralf Hendrik Menk (Elettra Sincrotrone Trieste)

45 Dual-energy and energy-dispersive X-ray diffractions using CdTe pixel detector

We have been developing CdTe pixel detectors combined with an aluminum Schottky diode sensor and photon-counting ASICs [1-3]. The hybrid pixel detector was designed with a pixel size of 0.2 mm by 0.2 mm and an area of 19.0 mm by 20.0 mm or 38.2 mm by 40.2 mm. We have investigated Pt/CdTe/Al configuration by using high resistivity p-type single-crystal CdTe wafers of 750 μm in thickness. The front side (X-ray irradiation side) was processed with a single platinum-electrode. The back side was deposited with a 95 × 100 or 190 × 200 matrix of aluminum-electrodes. This electrode configuration of Pt/CdTe/Al-pixel has an advantage that a high Schottky barrier formed on the Al/CdTe interface leads us to operate the CdTe as an electron-collecting diode pixel detector. The photon-counting ASIC, SP8-04F10K, has a preamplifier, a shaper, 3-level window-type comparators and a 24-bits counter in each pixel.
The synchrotron radiation facility provides monochromatic X-rays but the beams have higher harmonics contamination. In the standard experiments higher energy contamination is regarded as background and eliminated by optical mirrors and/or an energy window of detector. However, high energy contribution is still usable as real signals from the sample. Therefore, we came up with a novel method measuring multi-energy diffractions at the same time. In this presentation we will describe simultaneous X-ray diffraction measurements with 25 keV (primary) and 75 keV (third order) X-ray beams performed by the dual window comparator mode.
We have performed a feasibility study for lattice spacing measurement by performing energy-dispersive Laue diffractions. White X-ray microbeams were irradiated to a polycrystalline copper alloy sample. Figure shows the results for a Laue diffraction pattern, energy plots measured by the threshold scans and the lattice spacing plots. X-ray energies were distributed between 60 and 120 keV in one diffraction spot, but the lattice spacings were almost constant. These were classified to three values. The averaged lattice spacings were calculated to be 1.294 (Å), 1.830 (Å) and 2.106 (Å). These correspond to the Cu(220), Cu(200) and Cu(111) surfaces. The present values are fairly in agreement with the standard values.
References
[1] H. Toyokawa et al., Nucl. Instr. and Meth. A636, S218-S221 (2011).
[2] H. Toyokawa et al., Journal of Physics: Conference Series, 425, 62014 (2013).
[3] H. Toyokawa et al., Journal of Instrumentation, 12, C01044 (2017).

Speaker: Hidenori Toyokawa (Japan Synchrotron Radiation Research Institute)

46 IBEX based CdTe Detectors for Spectral Applications at High X-ray Energies

IBEX is a versatile readout ASIC developed at DECTRIS as successor of PILATUS and EIGER [1] that can be operated in electron collection mode to count the number of X-ray photons absorbed in pixelated CdTe sensors. Since it also implements multiple energy thresholds, it is very well suited for spectral applications at X-ray energies up to 160 keV. To quantify how well photons of different energies can be separated, we recently introduced a quantity called Spectral Efficiency [2]. This figure of merit is defined as the fraction of the incoming photons that is measured within a certain energy window around the incident photon energy. The Spectral Efficiency of IBEX based detectors with different pixel sizes was measured at the BAM beamline at BESSY as a function of the photon energy and flux. The results show very good agreement with the outcome of detailed Monte-Carlo simulations. The Spectral Efficiency turns out to be a very useful tool to quantitatively optimize the pixel size for spectral applications. For a given photon flux it allows finding the ideal pixel size that simultaneously minimizes the effects of charge sharing and pulse pile-up.

Some novel medical applications, such as molecular imaging using gold nanoparticles, require the energy threshold to be set above 75 keV. For a uniform response of all pixels, an accurate equalization of their energy thresholds is essential. Since fluorescence radiation above the Kα doublet of Pb is not practically available, an alternative energy reference point like the end-point of an X-ray tube spectrum has to be used. Due to the finite energy resolution of the ASIC, the measurement of the position of the end-point is not straightforward. However, by using a very simple model to describe the spectrum, we found a reliable way to use the end-point for accurate threshold equalization. The procedure has been validated with 140.5 keV photons from a 99mTc source at the Nuclear Medicine Department of the Kantonsspital Baden. The energy threshold is found to only deviate by 0.5 keV from the nominal value and the response is very uniform with a threshold dispersion as low as 1.7 keV rms.

[1] Bochenek, M. et al. (2018), IBEX: Versatile Readout ASIC with Spectral Imaging Capability and High Count Rate Capability. Submitted to Transaction on Nuclear Science
[2] Trueb, P., Zambon, P. and Broennimann, C. (2017), Assessment of the spectral performance of hybrid photon counting x‐ray detectors. Med. Phys., 44: e207-e214. doi:10.1002/mp.12323

Speaker: Dr Peter Trueb (Dectris)

IBEX_CdTe.pdf

IBEX_CdTe.pptx

47 Large high-Z pixel detectors with LAMBDA readout

Hybrid pixel detectors using high-Z semiconductors can provide a combination of efficient hard X-ray detection, low noise, fast readout and additional features such as energy binning. This is particularly important for hard X-ray experiments at synchrotron beamlines, where the high brightness can allow fast time-resolved measurements or rapid scanning of a sample. LAMBDA [1] is a photon-counting X-ray detector system based on the Medipix3 chip that was designed to provide large, tileable modules, high speed readout at 2000 fps, and compatibility with high-Z materials.

Using newly-developed 4” Cr-compensated GaAs wafers with 500 µm thickness, Tomsk State University produced a GaAs pixel array with an area of 84 mm x 28 mm, corresponding to a full LAMBDA module. The pixel size was 55µm, giving a total of 1536 x 512 pixels, and ohmic contacts were used. The sensor was bonded to 6 by 2 Medipix3 readout chips, and integrated into the readout system. Flatfield X-ray tests showed a high bump yield. Like previously-produced Cr-compensated GaAs sensors, the material shows nonuniformities with a granular structure. However, these are stable over time, and can be effectively corrected with flat-field correction to achieve a good image quality.

Large high-Z detector systems have also been built by tiling together multiple GaAs or CdTe sensors manufactured from 3” wafers, each with a 42 mm x 28 mm size. The largest systems built so far are composed of 6 GaAs sensors, giving a total active area of 84 mm x 84 mm (with gaps between modules) and a pixel count of 2.3 megapixels. These are read out by 3 sets of readout boards to allow 2000 frames per second operation. These have been used in experiments at PETRA-III synchrotron beamlines, and have demonstrated sub-ms time resolution in hard X-ray diffraction experiments.

[1] D. Pennicard, S. Smoljanin, B. Struth, H. Hirsemann, A. Fauler, M. Fiederle et al., “The LAMBDA photon-counting pixel detector and high-Z sensor development”, JINST 9 (2014), C12026

Speaker: David Pennicard (DESY)

DPennicard_Lambda_IWorid2018_v5.pdf

DPennicard_Lambda_IWorid2018_v5.pptx

48 Development of Hafnium-Halide-Based Scintillator for Radiation monitor

A novel halide scintillator with low hygroscopic nature, Cs₂HfCl₆ (CHC), reaching high light output of 54,000 photons/MeV and excellent energy resolution of 3.4% at 662 keV (FWHM) using a photo multiplier tube (PMT) was reported [1]. However, its decay is as long as around 4 µs. In order to accelerate the decay time, we investigated optical and scintillation properties of Hafnium-halide-based materials such as Tl-doped Cs₂HfCl₆ (Tl:CHC), Cs₂HfX₆ single crystals (X = Cl, Br, I) grown by the vertical Bridgman technique.We succeeded in growing several crystals such as Tl:CHC and Cs₂HfI₂ (CHI). CHI had red-emission wavelength around 650 nm, and the CHI emission would be originating from self-trapped exciton or defect luminescence, as well as CHC. Moreover, CHI had a fast decay time of ~2.5 µs, and high scintillation light output up to ~70,000 photons/MeV
In this presentation, we report several novel scintillators with faster decay time than CHC using a photo-multiplier tube, Si-avalanche photo-diode and Si- photo-multiplier (MPPC) due to green and red emission wavelength.

Speaker: Prof. Kurosawa Shunsuke (Tohoku University, Yamagata University)

In memory of Sherwood Parker

Lunch

Invited

49 K-edge imaging with photon counting detectors for medical applications.

Speaker: Pasquale Delogu

Oral

Conveners: Prof. Cinzia Da Via (University of Manchester), Cinzia Da Via (University of Manchester (GB))

50 Timepix3 sub-pixel resolution of 200 and 300 keV electron tracks by machine learning

Detectors which are a valid choice for imaging applications are not necessarily a good choice for diffraction applications. The Chemistry Nobel Prize 2017 went to high-resolution imaging of biomolecules in solution by cryo Electron Microscopy (cryo-EM). Three major technology breakthroughs were celebrated: 1) Computation methods for data analysis; 2) Sample preparation and 3) understanding the limitations of the techniques and the importance of excellent near-ideal detectors.

Current state-of-the-art detectors can, while very powerful for imaging applications, not be used for diffraction experiments. For imaging application, current detectors reach their best and DQE and MTF responses only at the higher electron energies of 300 keV [1]. In the past Medipix, EIGER and Timepix detectors have been used successfully for imaging applications (< 120 keV) and diffraction (< 200 keV) experiments, but not at higher energies as their MTF and DQE collapses at higher energies. With the Timepix 3 event-based Time-over-Threshold and Time-of-Arrival information can be obtained [2], which opens up new possibilities where every single electron track can be analyzed. This allows determining the point of impact of every electron a wide range of energies.

Here, we present an initial characterization of the Timepix3 for electron microscopy. Two detector quad assemblies are packaged in vacuum camera housings and mounted on a 200kEV FEI Tecnai Arctica as well as a 300keV FEI Tecnai Polara electron microscopes. The simulation package GEANT4Medipix, developed and validated by Krapohl et al. [3] for X-rays, is able to simulate interactions of the incident electron track in the sensor layer, drifting of hole pairs, and electronics responsible for digitizing the signal of the induced charge separation. The package was used to generate thousands of electron tracks and their simulated pixel responses. The output from this simulation was used to train a convolutional neural network to predict the incident position of the corresponding electron within a pixel cluster. Using a secondary set, the neural network was able to locate the known point of impact with an accuracy better than half a pixel. To compensate for non-uniform response of the chip, an algorithm was developed which uses the deposited energy spectrum of many tracks. This method is invariant to changes in the feedback of the pixel over time and removes the need to calibrate each pixel by the use of radiation sources outside of the microscope. We ascertained that the neural network out-performed other computational methods for improving the DQE and MTF of the Timepix3.

[1] G. McMullan, et al. Ultramicroscopy 147 (2014): 156-163
[2] T. Poikela, et al. Journal of instrumentation 9.05 (2014): C05013
[3] D. Krapohl, et al. IEEE Transactions on Nuclear Science 63.3 (2016): 1874-1881

Speaker: Dr Eric van Genderen (PSI - Paul Scherrer Institut)

IWORID2018-EricvanGenderen.pptx

51 Mapping of the XRF data onto surface of the reconstructed artifact volume

It is advantageous to combine information about geometry and the inner structure of historical artifacts with information about the elemental composition of decorative layers (polychromy), typically covering historical wooden sculptures. X-ray computed tomography describing artifact structure is quite common and easy. Standard X-ray fluorescence (XRF) analysis of decorative layers is typically done for several selected spots of the artifact’s surface utilizing single pad detector. XRF imaging fully describing the surface’s elemental composition is commonly done for flat objects, while time consuming XRF tomography is applied to relatively small objects. It will be shown in this work that an effective fusion combination of XRF imaging and X-ray tomography describing the whole object can be realized even when using a limited number of XRF images.
Sculpture polychromy is typically composed from several layers as for instance basic white color layer containing zinc or led which is silvered or golden. Metal foil is usually covered by several color layers which can contain pigments based on other metals like iron, mercury, copper etc. The polychromy is relatively thin in comparison with sculpture dimensions, therefore XRF signal can be expected from the sculpture surface only as it is extremely low probability that XRF photons will transmit through the investigated body.
The gamma/XRF camera (XRF camera hereafter) used in this work is employed for XRF imaging [1] - XRF photons are detected by the XRF camera pixelated detector through a pinhole collimator. XRF camera is equipped with two stacked Timepix detectors [2]. Detectors are operated in counting or time-over-threshold mode (ToT), which enables measuring the energy of each incident photon utilizing single event analysis. Front detector of the XRF camera used has a 300 um thick silicon sensor, and the back detector has a 1 mm thick CdTe sensor. XRF images are projected through pinhole with 100 um diameter. The XRF camera communicates via a USB 2.0 interface allowing frame rates up to 30 fps.
The Twinned Orthogonal Adjustable Tomograph (TORATOM) was used for CT and XRF data acquisition. TORATOM consists of two independent X-ray imaging lines in an orthogonal arrangement, with a shared rotational stage. TORATOM is equipped with 160 kV and 240 kV tubes (X-RAY WorX GmbH). CT data are recorded utilizing 160 kV tube and flat panel detector Perkin Elmer (400x400mm, 0.2 mm pixel pitch). XRF photons are induced by the 240 kV tube.
Mapping of the XRF data onto surface of the topographically reconstruction volume is done as follows [3]. The reconstructed volume (3D matrix) is binarized, i.e. a value of 1 is written into voxels which have a value higher than the given threshold (corresponding to air) and a value of 0 is written into voxels with a value lower than the same threshold. The coordination system of the volume is rotated to have vertical slices parallel with the XRF camera detector. The XRF image is multiplied with vertical slices in successive steps from the slice nearest to the XRF camera up to the furthest slice. When the first voxel with a value of 1 is met, the value of the related pixel of the XRF image is written into this voxel. The same pixel of the XRF image is then set to zero. Therefore, such an XRF pixel is not written into other voxels laying behind the first detected material. Obviously, the XRF image has to be gradually enlarged due to the subsequently increasing distance between the vertical slices and the camera. An example of the XRF data mapped onto surface of the reconstructed baroque Pieta is depicted in the figure bellow. XRF image coded in the RGB representation: R 3-7 keV, G 8-16keV, B 18-28 keV. The mixture of materials is expressed by the resultant color.
It was proven that the fusion of XRF imaging and CT helps to identify the distribution of the various elements and relationships between the surface’s material composition and the geometry of the reconstructed volume. It is relatively easy to implement descibed method of the XRF mapping into surface layer of the CT volume. Analysis of the Middle age wooden sculpture from the National Gallery in Prague will be presented as example.

REFERENCES
[1] Zemlicka, J., et al., Energy and position sensitive pixel detector Timepix for X-ray fluorescence imaging, Nucl. Instrum. Meth. A 607 (2009) 202.
[2] Llopart X. et al., Timepix, a 56k programmable pixel readout chip for arrival time, energy and/or photon counting measurements, Nucl. Instrum. Meth. A 581 (2007) 485.
[3] D. Vavrik, J. Jakubek, J. Lauterkranc, I. Kumpova, M. Vopalensky, J. Zemlicka, Multimodal analysis of cultural heritage artefacts utilizing computed tomography and x-ray fluorescence imaging, IEEE Nuclear Science Symposium Conference Record, 2016, DOI: 10.1109/NSSMIC.2016.8069960

Speaker: Daniel Vavrik (Czech Academy of Sciences)

52 Optimization of spatial resolution and detection efficiency for photon/electron/neutron/ion counting detectors with Microchannel Plates and Quad Timepix readout

Particle counting detectors with Microchannel Plates (MCPs) have been successfully used in various applications ranging from astrophysical imaging, where incoming fluxes are relatively low, to synchrotron instrumentation and neutron imaging with input fluxes exceeding 10^8 cm^-2s^-1. Among the attractive capabilities of MCPs is their ability to convert an incoming particle into 10^3-10^7 electrons within single or several adjacent pores, enabling imaging with spatial resolution as high as several micrometers through event centroiding. In such devices the center of gravity is calculated for each electron cluster registered by a certain readout, providing individual events are separated spatially or temporally. However, the detection efficiency in such devices is reduced compared to non-centroiding configurations as some events are rejected by the processing electronics (e.g. events with footprint being too narrow). That reduction of detection efficiency is strongest for applications where incoming particles are absorbed within the bulk of the MCP glass (such as gamma and neutron detection), not just its front surface, leading to substantial variation of gain between different events.

In this paper we demonstrate the possibility to compromise between the detection efficiency and the spatial resolution by acquiring several images simultaneously: one with the best spatial resolution, but reduced statistics, and the others with more events accepted, but a lower spatial resolution, with the lowest resolution defined by a single pixel/strip width of the readout. The results presented in the this paper are obtained with a quad Timepix parallel readout with 512x512 pixels encoding events produced by a chevron stack of MCPs. The highest spatial resolution image (down to single MCP pore spacing of 7.2 µm) is acquired simultaneously with the image where resolution is defined by the readout pixel size of 55 µm, where no events are rejected. The challenge of real time data analysis is addressed by the implementation of multicore parallel processing in order to cope with multi-MHz event rates. The ability to acquire high resolution and high statistics images simultaneously increases the flexibility of data analysis and may be very attractive for applications where a compromise between spatial resolution and detection efficiency must be determined before the experiment.

Speaker: Anton Tremsin

53 Advances in Pixel Cluster Analysis in Hybrid Pixel Detectors

This presentation outlines several advances in cluster analysis for hybrid pixel detectors. Hybrid pixel detectors are used by NASA for the radiation monitoring in a variety of space applications and different missions – namely deployment of several sensors on the International Space Station, as a sensor for radiobiology on the BioSentinel satellite and the radiation detector on the Orion Spacecraft. NASA’s monitors are based on the Timepix detector produced by the Medipix2 collaboration.

NASA has historically required sensors to measure radiation dose levels and particle stopping power. Recently there is a need to gather more detailed information about the radiation field, especially for future Exploration Missions which will carry humans outside of Low Earth Orbit for the first time in 50 years.

We have previously shown that hybrid pixel detectors can be utilized as a ‘single layer particle telescope’, where the long tracks extending over many pixels are selected for more detailed analysis. In this case, such a detector can be thought of as many independent subunits, each separately sampling the stopping power of a traversing particle.

This approach allows for more detailed analysis to be carried out on long clusters. For example, theoretical slowing curves can be compared to the measured using subunits. This enables unambiguous identification of the energy and charge of particles with energies less than 100 MeV/A. Similar methods can be used to identify nuclear interactions in the silicon sensor.

More information can also be gathered about the energy of fast particles using maximum likelihood reconstruction of the stopping power measurement in each subunit. In this case the energy of interacting particles can be reconstructed far more accurately than via a single stopping power measurement of a whole cluster.

Furthermore delta electrons can be used as a veto on the maximum particle energy, as well as in a probabilistic manor to estimate particle charge and energy. A combination of these tests allows for charge and energy binning using a single hybrid pixel detector as shown in figure 1.

Figure (1) – Example of testing of measured slowing curves (red) against theoretical cases (blue) and identification of delta electron. Due to the presence of the energetic delta-ray, this particle was identified as a carbon ion. Example performance shown on right.

Speaker: Dr Stuart George (University of Houston, NASA Space Radiation Analysis Group)

54 SIMULATION AND EVAULATION OF HIGH PERFORMANCE COST-EFFECTIVE POSITRON EMISSION MAMMOGRAPHY (PEM) SCANNER WITH LASER INDUCED OPTICAL BARRIER TECHNIQUES

Breast cancer is the main cause of tumor deaths in women. Early diagnosis of the disease is widely approved as being essential for effective treatment. An estimate of about 246,660 new breast cancer cases was done in 2017, and the death rate rises to about 40,450 in that year [1]. Due to these reasons, breast cancer diagnosis at the very early stage is important in order to reduce the incidences and mortality rates. Positron Emission Mammography (PEM) is a breast dedicated imaging device which uses pair of annihilation gamma photons to detect cancerous breast tissues. The PEM device is compact in nature with a reduced field of view to cover the entire breast region, and it employs few detector modules which makes it cost-efficient. To effectively diagnose breast cancer at a very early stage, a device with high spatial resolution and sensitivity is required. PEM detectors based on semiconductor materials are characterized by an excellent intrinsic system spatial resolution but are not cost-effective [2], whereas detectors based on scintillator crystals are cost-effective but have limited intrinsic resolution to detect small breast lesions. This study focuses on improving the resolution of scintillator detectors by simulating a PEM scanner employing 1 × 1× 10 mm3 laser processed [3] scintillator crystal such as Lutetium-Yttrium Oxyorthosilicate (LYSO). The simulation was done with Geant4 application for emission tomography (GATE) software, and performance evaluation was done with National Electrical Manufacturers Association (NEMA) phantom studies using Maximum Likelihood Expectation Maximization (MLEM) technique. The scanner geometry has 110 mm trans-axial field of view (FOV) and 128 mm axial FOV. Evaluation result showed that the scanner has 7.6% system sensitivity and 1 mm system spatial resolution.

Keywords: PEM; LYSO, GATE; Semiconductor materials; Scintillator crystals; laser induced optical barrier

REFERENCES
1. Li, L., et al., (2015). Performance Evaluation and Initial Clinical Test of the Positron Emission Mammography System (PEMi). IEEE Transactions on Nuclear Science, 62(5), 2048-2056.
2. D Uzun., et al., (2014) Simulation and evaluation of a high resolution VIP PEM system with a dedicated LM-OSEM algorithm, Journal of Instrumentation, Volume 9 C05011
3. Sabet, et al., (2016). Novel laser‐processed CsI: Tl detector for SPECT. Medical physics, 43(5), 2630-2638.

Speaker: Dr Dilber Uzun Ozsahin (Near East University)

Swedish fika in the hotel lobby (2nd floor)

Poster session: part 2

Convener: Thilo Michel (Max-Planck-Institute for the Science of Light, Erlangen)

55 A system modification to a single-grid phase-contrast x-ray imaging (PCXI) system for improving image quality

In this study, we propose a modification to a single-grid phase-contrast x-ray imaging (PCXI) system using a Fourier domain analysis technique to extract absorption, scattering, and differential phase-contrast images. The proposed modification is to rotate the x-ray grid in the image plane to achieve spectral separation between the desired information and the moiré artifact, which is introduced by the superposition of the periodic image of the grid shadow and the periodic sampling by the detector. In addition, we performed some system optimization by adjusting distances between source, object, grid, and detector to further improve image quality. This optimization aimed to increase the spectral spacing between the primary spectrum (lower frequency) and the harmonics of the spectrum (higher frequency) used to extract the various image contrasts. The table-top setup used in the experiment consisted of a focused-linear grid with a 200-lines/inch strip density, a microfocus x-ray tube with a 55-um focal spot size, and a CMOS flat-panel detector with a 49.5-um pixel size. The x-ray grid was rotated at 27.8 degrees with respect to the detector and the sample was placed as close as possible to the x-ray tube. Our results indicated that the proposed method effectively eliminated the PCXI artifacts, thus improving image quality.

Speaker: Prof. H. S. Cho (Yonsei University)

56 A USB 3.0 readout system for Timepix3 detectors with on-board processing capabilities

Timepix3 is a high speed hybrid pixel detector consisting of a 256x256 pixel matrix with a maximum data rate of up to 5.12 Gbps (80 MHit/s). The ASIC is equipped with eight data channels that are data driven and zero suppressed which makes it suitable for particle tracking and spectral imaging.

In this paper, we present a programmable USB 3.0-based readout system with online pre-processing capabilities. USB 3.0 is present on all modern computers and can, under real-world conditions, achieve up to 370 MB/s, which allows up to 40 MHit/s of raw pixel data. With on-line processing, the proposed readout system is capable of achieving higher transfer rate (approaching Medipix4) and thus hit rate since only information rather than raw data can be transmitted.

The system is based on an Opal Kelly development board with a Spartan 6 FPGA providing a USB 3.0 interface between FPGA and PC. It connects to a standard CERN Timepix3 chipboard with a VHDCI connector via a custom designed adapter card. The firmware is structured into blocks such as detector interface, USB interface, system control, and an interface for data pre-processing. On the PC side, a Qt/C++ multi platform software library is implemented to control the readout system, providing access to detector functions and handling high speed USB 3.0 streaming of data from the detector.

We demonstrate equalisation, calibration and data acquisition using a Cadmium Telluride sensor and optimise imaging data using simultaneous ToT (Time-over-Threshold) and ToA (Time-of-Arrival) information.

Speaker: Mr Till Dreier (Mid Sweden University)

57 Absorption grating based X-ray phase contrast CT using Timepix detector

Single absorption grating based X-ray phase contrast imaging (XPCI) with a hybrid pixel detector Timepix is a promising technique in many application fields (material sciences, biology, medicine etc.) [1]. Conventional grating interferometry uses two absorption gratings where one of them is moved in steps (so-called phase stepping). In contrast to that, the absorption, phase contrast and dark field images can be obtained in one acquisition using the proposed approach. On top of that, the single exposure method decreases the delivered dose and it is easier to maintain system stability without phase stepping.

Standard laboratory X-ray imaging setup equipped with Timepix detector and nano-focus X-ray tube was utilized for the measurements. The maximal source to detector distance (SDD) of the setup is roughly 90 cm, which provides long enough phase propagation path for 55 um large detector pixels. The system enables insertion and positioning of both 1D and 2D absorption gratings producing micro-line or micro-pencil beams, respectively.
In this system series of acquired phase contrast images taken under different angles can be used as input projections for computed tomography (CT). CT reconstructions were performed by both in-house and commercial software packages. Results of pilot CT measurements for phantoms and complex objects will be presented in the contribution.

[1] F. Krejčí, J. Jakůbek, M. Kroupa, "Single grating method for low dose 1-D and 2-D phase contrast X-ray imaging", Journal of Instrumentation 6 C01073 (2011)

Speaker: Dr Jan Žemlička (IEAP, Czech Technical University in Prague)

58 Characterization of CdTe Diode Detector with Depletion Layer Modulation

Introduction

CdTe diode detector is an appealing sensor device for medical imaging applications, such as X-ray radiography, CT and so on. Particularly, room-temperature operation availability and high detection efficiency across the wide energy range, due to the large band-gap energy (1.44 eV) and the high atomic number (Cd:48, Te:52), are remarked for integrated sensors, i.e., line sensor and panel sensor [1]. While photon counting signal processing enables to measure the energy spectrum of photon of irradiated X-ray, it is difficult to implement large-size two-dimensional array of photon counting sensors because huge data traffic is required between sensors and computers.

In this study, we propose a novel method for realizing X-ray energy discrimination with accumulation-type CdTe flat panel detector (FPD). It utilizes the difference of the depletion layer thickness caused by the modulation of bias voltage applied on the diode detector. Because the thickness of depletion layer is proportional to the root of bias voltage, if the higher bias voltage is applied, the depletion layer gets thicker. The probability of interaction between the detector and X-ray increases as the thickness of the depletion layer increases, especially for higher energy X-ray. When the bias voltage is modulated per frames, we can obtain the images containing different energy information for each frame. This enables energy discrimination imaging like dual-energy CT [2] with a single typical X-ray source.

The characteristics of CdTe Schottky diode radiation detector at high bias voltage have already been investigated [3]. However, the characteristics at low bias voltage is not yet known well because very low pulse height from very slow carrier mobility. To make it clear, our group have investigated the characteristics of a single CdTe diode detector at low bias voltage, both theoretically and experimentally.

Analysis and experiment

Figure 1(a) shows a p-type CdTe Schottky diode with thickness $W$ , applied bias voltage $V$. Assuming injected X-ray with energy $h\nu$ is absorbed at distance $x$ from injection surface, charge collection efficiency $\eta$, in other words, the sensitivity of the detector to the radiation, is formulated as

$\eta(V, h\nu ) = \int_0^W \eta_H (V, x) \alpha_\gamma (h\nu) \exp(-\alpha_\gamma (h\nu) x) dx $
$\eta_H(V,x) = \frac{\lambda_n}{W} [ 1 - \exp (-\frac{x}{\lambda_n})] + \frac{\lambda_p}{W} [ 1 - \exp (-\frac{W-x}{\lambda_p})] $

where $\lambda_n$ and $\lambda_p$ are the mean free path of electrons and holes in the detector, respectively [3]. Figure 1(b) shows that the difference of the efficiency is significant where injection energy is greater than 100[keV], while less difference at 30 - 60[keV]. By comparing the efficiency ratio at 72[V] over 36[V], the ratio at 60[keV] is 1.16 while the ratio at 120[keV] is 1.36.

We have performed the spectrum measurement using multi-channel analyzer. Am241 and Eu152 are used as RI. Peak count around 59.5[keV] (Am241) and 122[keV] (Eu152) are measured for bias voltage 36[V] and 72[V]. While the peak count ratio around 59.5[keV] = 1.15 were in good agreement with the analysis result, the peak count ratio around 122[keV] was 1.69, which is larger than the analysis result. Compton scattering and interaction with electrodes should be considered for more precise analysis.

Reference

[1] Wenjuan. Zou, et al., Jpn. J. Appl. Phys., 2008, 47(9R), 7317-7327
[2] Johnson, et al. Eur Radiol (2007) 17: 1510.
[3] Kosyachenko et al, J. Appl. Phys. 113, 054504 (2013)

Speaker: Mr Tsuyoshi Terao (Graduate School of Science and Technology, Shizuoka University)

59 Characterization of LVDS drivers for MAPS detector designed in 180 nm SOI CMOS technology

The Low-Voltage Differential Signaling (LVDS) drivers for the X-ray imaging pixel detector X-CHIP-03 have been developed using a 180 nm deep submicron Silicon On Insulator (SOI) CMOS commercial technology. The X-CHIP-03 is a Monolithic Active Pixel Sensor (MAPS) and its architecture is based on the first prototype X-CHIP-02 with hit counting. The advantages, design, features, test structures and measurements of the first prototype X-CHIP-02 fabricated in SOI CMOS technology have already been published [1, 2].
The applicability of the X-CHIP-02 has been found for many commercial applications, for example the imaging detectors in medicine industry, detectors for dosimetry and spectroscopy, tracking detectors, etc. However, all of them require to improve the speed of data acquisition from the detector. The communication speed and its radiation resistance is the one of the most important parameters of the detector. The X-CHIP-02 communication consists of long shift register daisy chaining all pixels. This architecture limits the maximum data rate below 50 kbps. The LVDS transceiver and receiver allow to increase the speed of communication to 500 Mbps, Figure 1, with respect to low power consumption (6 mW) and radiation resistance. The motivation for development the LVDS IP block is the utilization of the detector in the commercial application. The LVDS receiver and transmitter are included in the X-CHIP-03 detector. The design of the LVDS IP blocks, as well as circuit simulation, laboratory measurement technology are described. The experimental results are of great importance for further development of the LVDS communication with radiation tolerance design which is intended for MAPS sensors in SOI technology.

[1] T. Benka, M. Havranek et al., 2018 Characterization of pixel sensor designed in 180 nm SOI CMOS technology JINST 13 C01025
[2] M. Havranek et al., 2017 MAPS sensor for radiation imaging designed in 180 nm SOI CMOS technology, proceeding from IWORID 2017, submitted to JINST

Speaker: Mr Tomas Benka (1) Faculty of Electrical Engineering, Czech Technical University in Prague; 2) Faculty of Nuclear Sciences and Physical Engineering, Czech Technical University in Prague)

60 Compact four-layered Timepix-based particle tracker WidePIX 3D: The first applications in X-ray CT imaging

WidePIX 3D is a compact Timepix-based ionizing radiation detector with 3D position sensitivity [1]. The device consists of four Timepix chips with 300 µm silicon sensors tightly stacked together. The insensitive gap between the subsequent layers is 250 µm only, since the read-out chips were thinned down to 100 µm. The sensor volume is, therefore, divided into a 3D array of 256×256×4 voxels, 55×55×300 µm each.
Although the detector was originally designed as a wide field-of-view particle tracker its construction is advantageous for X-ray imaging as well. Four layers of the device can be either used for improvement of quantum efficiency of silicon sensors (as the total thickness of four sensors is 1.2 mm) while keeping the spatial resolution of a thin sensor or for multi-channel energy sensitive imaging. Similar approach is known from a single-shot dual-energy X-ray computed radiography utilizing a dual-layer detector [2]. WidePIX 3D provides up to four sample images produced by different photon spectra from a single exposure. Furthermore, the detection threshold of each layer can be individually adjusted to optimize the resolving performance for elements expected within the scanned sample.
In this contribution we present the first results of using the WidePIX 3D detector as an imaging camera for energy-sensitive X-ray computed radiography and CT. The procedure of precise threshold-energy calibration of the detector using X-ray fluorescence will be described and material-resolving radiographies and CT data will be demonstrated.

[1] ADVACAM s.r.o., WidePIX 3D, Compact Large Field of View Particle Tracker (2018) http://advacam.com/camera/widepix-3d
[2] Rassouli N. et al, Detector/based spectral CT with a novel dual-layer technology: principles and applications; in: Insights Imaging, 8(6):589–598 (2017), DOI: 10.1007/s13244-017-0571-4.

Speaker: Dr Jan Žemlička (IEAP, Czech Technical University in Prague)

61 Comparative analysis of the readout front-end electronics implemented in deep submicron technologies.

There are two principal approaches for the detector signal processing in the front-end electronics: a charge integration (used in CCDs, CMOS imagers, etc.) and a single photon counting (SPC). Nowadays, hybrid pixel detectors with fast readout electronics operating in the SPC mode are becoming increasingly popular in the development of X-ray imaging systems. This is mainly because of their advantages like a very high dynamic range (compared to the integration type detectors), noiseless imaging (a discriminator with properly set threshold cuts off the noise and counts only the hits whenever pulse amplitude is sufficiently high) and the possibility of counting photons only within a given energy window (in systems with two or more discriminators). However, the main disadvantages of the SPC operating detectors in comparison to charge integration type systems are their worse spatial resolution and worse capability of coping with high photon fluxes. Though, many prominent groups worldwide put much effort to minimize gap among SPC and charge integration systems as it may highly improve current X-ray imaging systems performance. The possible solution to increase both spatial resolution and count-rate performance of the SPC systems is to use modern nanometer processes for integrated electronics fabrication. It is however known that mainly the digital part of the pixel can be efficiently downscaled with a possible small risk of their parameters degradation. Analogue blocks may often be scaled down only at the expense of their parameters.
Therefore, based on our experience [1, 2] and currently realized project, we will present comparative analysis of two modern deep submicron CMOS processes, i.e. 40 nm and 28 nm in terms of design of the front-end electronics dedicated to detector signal fast processing. Authors will present considerations about the design of the readout front-end electronics dedicated for high-count rate applications and their limitations, with the emphasis on:
- input noise of the system,
- power consumption,
- speed of input pulses counting,
- area occupied by a front-end electronics.

Speaker: Piotr Kmon (AGH UST Krakow)

62 Compressed-sensing based digital tomosynthesis (DTS) using a moving multi-slit collimator for low-dose x-ray imaging

Conventional digital tomosynthesis (DTS) reconstruction based on the filtered-backprojection (FBP) method requires a full field-of-view scan and relatively dense projections, which results in still high x-ray dose especially for medical imaging purposes. In this work, to reduce the x-ray dose delivered to the patient in DTS examinations, we propose a new type of DTS scan in which the x-ray span is blocked by a moving multi-slit collimator having a duty cycle of 0.5 during the projection data acquisition. We considered a compressed-sensing (CS)-based algorithm for more accurate DTS reconstruction in the proposed scan geometry. To validate the proposed method, we performed a systematic simulation and investigated the image characteristics. In the simulation, several types of multi-slit collimators having a closing size equal to the widths in the range of 14-60 pixels were tested. All projections were taken at a tomographic angle range of ±50o and an angle step of 2o. We successfully reconstructed collimated DTS images of high accuracy and no truncation-related artifacts in the proposed scan. We expect that the proposed approach can considerably reduce the x-ray dose in the present DTS examinations, if the moving collimation is realized in real DTS systems.

Speaker: Soyoung Park (Yonsei University)

63 DAQ software for GEM-based imaging system

ABSTRACT

In the paper we report on development of a dedicated data acquisition (DAQ) software being part of a full-filed X-ray fluorescence spectroscopy (XRF) imaging system equipped with a standard 3-stage Gas Electron Multiplier (GEM) detector of 10 cm by 10 cm [1].

Analysis of the spatial distribution of elements on the heritage objects, using XRF, nowadays is a very important support for conservators. Elemental mapping offers possibility to study spatial distribution of pigments of artworks. Presented system is able to perform a fast imaging of the spatial distribution of the elements over the large area of an object. Moreover, thanks to the infinite depth of field of the pinhole camera the system is able to investigate the non-flat surfaces.

Simulations measurement of a large surface introduces additional constraints and requirements (in comparison to e.g. macro-XRF) for the front-end electronics, data acquisition system and the software which are necessary for efficient system operation. The details on the front-end electronics and the data acquisition system have already been reported elsewhere [2, 3]. Therefore, in the paper we focus mostly on the software part of the developed system.

The software suite consists of a few independent components, each responsible for well-defined task. The components are interacting all together. A crucial part of the software, in a form of a graphical user interface (GUI) application written in modern C++ language with help of Qt framework and Boost libraries, is responsible for system configuration, monitoring and online data reconstruction. The GUI software is very helpful as convenient diagnostic tool during setting up and debugging the measurement campaign. However, when the system is in an optimum configuration and ready for data-taking one needs a different approach. In this case an automated system is utilized for batch recording of the measurement data, event reconstruction and processing with results visualization. That part of the software has been decomposed into: (a) dedicated raw data recorder, (b) non-GUI event reconstruction component (C++ based) able to use most of the workstation resources (mainly CPUs time) and (c) independent, flexible offline visualization component for plots or histograms preparation (Python based). All the components of the system are completely independent from each other. Therefore, distribution of the tasks to many workstation PCs is straightforward and provides flexibility along with data processing acceleration.

The software is running in the Linux environment (with possibility to be deployed to the Microsoft Windows operating system, if needed). It is also equipped with CMake build recipes in order to make building, testing and deployment of the system simpler and faster.

This work was supported by the Polish National Centre for Research and Development, grant no. PBS3/A9/29/2015.

REFERENCES

[1] A. Zielińska et al., X-ray fluorescence imaging system for fast mapping of pigment distributions in cultural heritage paintings, JINST 8 (2013) P10011.
[2] B. Mindur et al., A compact system for two-dimensional readout of Gas Electron Multiplier detectors, JINST 8 (2013) T01005.
[3] T. Fiutowski et al., ARTROC - a readout ASIC for GEM-based full-field XRF imaging system, JINST 12 (2017) C12016.

Speaker: Dr Bartosz Mindur (AGH University of Science and Technology, Faculty of Physics and Applied Computer Science)

64 Deformation Analysis of the Spongious Sample in Simulated Physiological Conditions Based on In-situ Compression, 4D Computed Tomography and Photon-Counting Detector

Spongious materials, such as trabecular bone, are currently broadly investigated structures as they represent a key element for proper operation of bone in terms of its load bearing function and permeability. Advanced mechanical analysis of the spongious materials has to be carried out volumetrically in micro-scale as the cellular structure of the material is very complex and has fundamental effects on the macro-scale properties [1]. Moreover, reliable analysis requires environment conditions equivalent to the physiological conditions in real body. In this work, an in-house designed table top loading device equipped with a bioreactor is used for in-situ compression of the spongious sample in simulated physiological conditions. 4D computed tomography is used as a tool for an advanced volumetric analysis of the deforming microstructure of the specimen.
The compact in-house designed table top loading device was used for mechanical compression of the spongious specimen during the experiments.
The device has loading capacity 3 kN with 1 micrometer position tracking accuracy and sub-micron position sensitivity. The device is equipped with load-cell and position encoders for real-time closed loop control of the experiment in both displacement and force driven mode. The bioreactor with closed circulation of the fluid is a modular optional part of the loading device. It is equipped with a simulated body fluid reservoir with a heating device, a fluid pump, and a thermometer for closed-loop control of the temperature and flow of the fluid. The specimen was placed in a loading chamber consisting of a thin carbon-fibre composite tube with nominal shell thickness 0.45 mm ensuring low attenuation of X-rays. It was submerged in a simulated body fluid from the bioreactor. The loading device with the bioreactor was placed directly on the rotation stage of the modular X-ray imaging device TORATOM [2]. As the loading device is equipped with slip-ring cable system, it can perform an unlimited number of revolutions during on-the-fly scanning procedure.
In this study, two types of X-ray imaging detectors were used: i) scintillating large-area flat panel detector Perkin Elmer XRD1622 (resolution 2048×2048 px, physical dimensions 410 x 410 mm) and ii) large area single photon counting detector (LAD), the 10 x 10 TimePix assembly with resolution 2560×2560 px (single TimePix detector - resolution 256 x 256 px, pixel size 55 x 55 µm). As the low-attenuating specimen was submerged in the fluid with considerable attenuation of X-rays, image quality and contrast of the both types of detectors were evaluated and analyzed. The specimen was compressed with low loading velocity until collapse of the structure. Continuous on-the-fly tomography capability of both detectors was evaluated and is discussed in the study.
A set of volumetric data capturing deformation of the specimen during the experiment was prepared from the images acquired by both detectors. Digital image correlation (DVC) algorithm was used for evaluation of the volumetric strain fields in the specimen. To conclude, a sophisticated experimental procedure for an advanced analysis of the spongious sample based on the loading device with bioreactor, 4D computed tomography and photon counting detector was introduced.

ACKNOWLEDGEMENT
The research has been supported by the Interreg V-A Austria - Czech Republic programme in frame of the project Com3D-XCT (ATCZ38) and Operational Programme Research, Development and Education in project INAFYM (CZ.02.1.01/0.0/0.0/16_019/0000766).

REFERENCES
1. D. Kytyr, P. Zlamal, P. Koudelka, T. Fila, N. Krcmarova, I. Kumpova, D. Vavrik, A. Gantar, S. Novak, Deformation analysis of gellan-gum based bone scaffold using on-the-fly tomography, Materials and Design 134 (2017) 400–417.
2. I. Kumpova , D. Vavrik, T. Fila, P. Koudelka, I. Jandejsek, J. Jakubek, D. Kytyr, P. Zlamal, M. Vopalensky, A. Gantar, High resolution micro-CT of low attenuating organic materials using large area photon-counting detector, J. Instrum. 11 (2016).

Speaker: Tomas Fila (Czech Academy of Sciences, Institute of Theoretical and Applied Mechanics)

65 Design of high resolution gamma camera for radioactive waste monitor

The hot spot, radioactivity, and discrimination of radionuclides must investigate in order to determine the radiological characterization during decommissioning. It is widely known how the radiological characterization evaluates at each stages [1]. Decontamination is essential process to reduce radioactive waste because we need large area to store a lot of the waste. The scintillation detector is widely used for radioactive waste monitor. Many scintillation detectors are needed to identify the position of the radioactive contamination. There is, therefore, a limitation to detect the hot spot accurately. This problem can overcome by gamma camera which detects and visualizes of the radioactive sources. The high resolution gamma camera is useful for effective decontamination and cost reduction. The purpose of this study is to design of high resolution gamma camera for radioactive waste monitor. The gamma camera consisted of a pinhole collimator, NaI (Tl) crystal, position sensitive photo-multiplier (PSPMT), signal processing board, and data acquisition board (DAQ). The acceptance angle and source-to-collimator distance were around 37° and 70 cm, respectively. The channel-height pinhole collimator was optimized for high resolution gamma camera. For 2° resolution, the optimized diameter and height of pinhole collimator by Monte Carlo simulation were 1 mm and 4 mm, respectively. The scintillation crystal was determined as cylinder shaped NaI (Tl) crystal, diameter of 10 cm and height of 7.6 cm, which has 7 % energy resolution. The PSPMT (Mo. R3292-02) which is manufactured by Hamamatsu Photonics was used. The experimental measurements were evaluated using Tc-99m, Cs-137, Co-60, and K-40 sources. The information of position and energy were calculated by anger logic. The discrimination of radionuclides and radioactivity also were investigated.

Speaker: Ms Minju LEE (Korea Advanced Institute of Science and Technology)

66 Dual asymmetric and high resolution detection heads of a novel compact Molecular Breast Imaging system for small breast cancer diagnosis

Breast cancer is the most common cancer in women. Early diagnosis is a crucial point to achieve an efficient therapy.
In order to identify small-size tumours at the early stage, mammography is the most used technique for screening having showed to be able to decrease mortality from the breast cancer.
Nevertheless it showed reduced performance in case of dense breast [1].
Magnetic Resonance Imaging (MRI), Ultrasound (US) and Molecular Breast Imaging (MBI) techniques have been proposed as complementary to mammography.
The most promising among them is the MBI based on dedicated gamma camera, which provides functional, specific, information, particularly appropriated to dense breast.
It showed better sensitivity and specificity than mammography in case of dense breast.
A new compact system, consisting of a two asymmetric (different geometries and collimations) detectors, has been developed [2].
The two detector heads face each other in anti-parallel viewing direction, compressing, appropriately, the breast between them and allowing Limited-Angle Tomography.
The detectors provide somehow complementary planar images that shall be properly combined (fused) to get enhanced, diagnostic information with high specificity and sensitivity.
A full scale prototype based on matrices of Position Sensitive Photo-Multiplier Tube (PSPMT), coupled to segmented NaI(Tl) scintillators with parallel and pin holes optics has been constructed to test different design solutions, and evaluate the expected performances.
We will present the results of the measurements, in different modalities, including the Limited-Angle Tomography, with 99mTc on a rather complex perspex phantom simulating a woman breast with up to four spheric tumour lesions of different sizes, positions and uptakes.
First results on performance will be reported.
[1] Hruska C. et al., Medical Physics, June 2012, 39(6)
[2] Garibaldi F. et al., Nuclear Instruments and Methods in Physics Research A 617, 2010, 227–229

Speaker: Gaetano Elio Poma (INFN - National Institute for Nuclear Physics)

67 Evaluating the effectiveness of a mixed lead(Ⅱ) iodide and oxide semiconductor-based sensor for the development of a position sensitive detector for tracking a radioactive source

Evaluating the effectiveness of a mixed lead(Ⅱ) iodide and oxide semiconductor-based sensor for the development of a position sensitive detector for tracking a radioactive source

Ye Ji Heo1, Kyo Tae Kim2, Moo Jae Han3, Yo Han Shin3, Ji Koon Park2 and Sung Kwang Park1

1. Department of Radiation Oncology, Busan Paik Hospital, Inje University, Republic of Korea
2. Department of Radiological science, International University of Korea, Republic of Korea
3. Department of Radiation Oncology, Collage of medicine, Inje university, Republic of Korea

Radiation sources are being constantly used around the world, not only in the field of industry but also in medical radiation therapy. The radiation sources present a high risk of exposure due to their high emission energy. Therefore, a position-sensitive detection technology capable of accurate measurement is required, by increasing the sensitivity to the point of the radiation source. Conventional detectors are mostly prepared with PIN type silicon detectors, where P-type and S-type substances are doped on intrinsic silicon wafers. These silicon-based detectors cause defects in the silicon lattice when a radiation energy of 1 MeV or more is incident, thus degrading the detection efficiency.[1] Also, the lowered detection efficiency due to the silicon lattice defect can lead to potential radiation accidents, since it is impossible to accurately locate the radiation source. Hence, in this study, a Lead(Ⅱ) iodide (PbI_2) semiconductor, which exhibits excellent characteristics against the detection of gamma-ray and high-energy X-ray, was combined with Lead mono oxide (PbO), which exhibits sensitivity for high-energy radiation and low drift current characteristics, to manufacture a sensor using a screen printing method.[2,3] Also, a Silicon dioxide (SiO_2) deposition was performed to reduce the leakage current, which is caused by the movement of carrier inside the manufactured sensor, while improving the stability of signal detection. For the evaluation of the sensor developed in this study, iridium 192 (IR-192), a gamma-ray radiation source, was used. First, a morphological evaluation of the sensor was conducted. Second, an leakage current according to driving voltage change (I-V) test at the driving voltage ranging from -1000V to 1000V, and a photon current according to driving voltage change (V-S) test according to the changes in the driving voltage were conducted to obtain the signal to noise ratio (SNR). Based on the results, an effective driving voltage, at which the SNR is maximized, was derived to select an optimum driving voltage for the evaluation of energy characteristics. Third, as the items for the evaluation of detector characteristics with respect to the energy being emitted, reproducibility, linearity of the dosimeter response, and percentage internal distance (PID) were assessed to evaluate the effectiveness of the detector.
First, the result of the I-V test was 0.09±0.19 nA, and the result of the V-S test was -0.10±12.47 nA. The result of SNR analysis based on the above results was @ - 400, which was thus selected as the optimum driving voltage for the evaluation of energy characteristics. The result of reproducibility evaluation, conducted based on the selected driving voltage, was 0.15%, as a result of a standard deviation analysis of the signal currents. The linearity of the dosimeter response was analyzed by the R-Sq value of the linear regression function, and the value was 0.98 or above, indicating an excellent linearity. Finally, in the evaluation of the PID, based on the result of a linear attenuation coefficient analysis and the result of I-V test, the effective detection distance was derived as 10 cm. Therefore, the sensor developed in this study was found to have characteristics suitable for monitoring the position of radionuclides used in medical radiology. Also, further studies are required to enhance the sensitivity by increasing the effective detection distance.

References
[1] Z. Li, “Radiation damage effects in Si materials and detectors and rad-hard Si detectors for SLHC”, JINST 4 03011, 2009.
[2] J. F Condeles, T. M. Marins, et al., Fabrication and characteriazation of thin films of PbI2 for medical imaging, crystalline solids, 2004, Vol 338-340.
[3] K. T Kim, M. J. Han et al, Feasibility study of a lead monoxide based dosimeter for quality assurance in radiotherapy, JINST 11 P11006, 2016.

Speaker: Mr MOO JAE HAN (Department of Radiation Oncology, Collage of medicine, Inje university)

68 Feasibility Evaluation of Mercury(Ⅱ) Iodide based Curved Dosimeter for measurement of Entrance Surface Dose in Radiotherapy

Feasibility Evaluation of Mercury(Ⅱ) Iodide based Curved Dosimeter for measurement of Entrance Surface Dose in Radiotherapy

M.J. Han^a, Y.H. Shin^a, K.T. Kim^b, Y.J. Heo^c, K.M. Oh^d, D.H. Lee^c, J.Y. Kim^e, H.L. Cho^c, S.K. Park^c*

^aDepartment of Radiation Oncology, Collage of Medicine, Inje University
^bDepartment of Radiological science, International University of Korea
^cDepartment of Radiation Oncology, Haeundae Paik Hospital, Inje University
^dRadiation Equipment Research Division Korea Atomic Energy Research Institute
^eDepartment of Radiation Oncology, Busan Paik Hospital, Inje University

Today, in the radiotherapy for skin, the criticality of skin dose detection is increasing to prevent unnecessary exposures. For the phase of skin coordination to prevent the accumulating skin dose during treatment, Glass Rod Dosimeter(GRD) and Optically Stimulated Luminescent Dosimeter (OSLD) are the representative applications while the point measurement that reads by attaching many dosimeters is utilized. However, in clinics, the complexity and the time consumption in reading on the integrating analog detection are pointed as issues. Also, the inaccuracy of the dosimeter attachment point visually determined on the local area leads the difficulty of accurate dose analysis. ^[1-2] In the meanwhile, a photoconductor HgI_2 shows advantages beneficial for minimization since its actual atomic number and density are higher than those of GRD or OSLD. Also, the photoconductor in powder form, when mixed with a polymer binder, makes production if an array dosimeter in a curved structure. ^[3] Therefore, in this study, as a basic research to develop a real-time skin dose monitoring technology for the computerized treatment system, a unit cell dosimeter and a film based line type array dosimeter are produced. The unit cell dosimeter is produced in the size of 1 x 1 cm^2 with the thickness of 150 μm utilizing the Particle In Binder evaporation method and evaluated on its reproducibility and linearity to verify the performance. For the reproducibility, the same dose amount of 100 MU with the dose rate of 500 MU/min is examined 10 times. Then, Standard Errors(SE) are calculated on all the acquired measurements and evaluated based on the number less than 1.5% satisfying the 95% confidence interval. For the linearity, against the dose rates of 100 to 600 MU/min, the dose is gradually increased from 1 MU to 1,000 MU. Then, the measured data is evaluated based on the determination coefficient given from the linear regression analysis. Later, to verify the applicability of curve measurement, a 7-Pixel Line Array Dosimeter showing the pixel size of 1 x 1 cm^2 and the pixel pitch of 3 cm is produced. The irradiation condition of the line array sensor is the 10 irradiations of the dose of 100 MU with the dose rate of 500 MU/min. To present the dose distribution exposed to the skin, the average error rate is acquired from the dose distribution measured from the 4 pixels located at the iso-centers of the flat and curved circuit board (Radius: 20.5 cm). According to the results of this trial, the reproducibility evaluation of the unit cell dosimeter shows that the cases of 6 MV and 15 MV are 0.613% and 1.465% respectively to be in the 1.5% confidence interval. According to the linearity evaluation, the R-sq values of 6 MV and 15 MV are 0.9999 and 0.9981 respectively showing superb linearity close of the determination coefficient of 1. The line array dosimeter produced based on these results shows the average error rates are 23.337% and 12.264% with 6 MV on the flat and curved area respectively while being 11.691% and 8.372% with 15 MV. Then, the curved area measurements show reduced error rates reduced 11.073%p with 6 MV and 3.318%p with 15 MV from the cases of flat area demonstrating its applicability. Therefore, if additional studies are performed on the dose distribution by the photon energy, the application of curved dosimeter would become possible to analyze the dose distribution on the geometrical structure of human body.

[1] W. J. Yoo et al., Measurement of Entrance Surface Dose on an Anthropomorphic Thorax Phantom Using a Miniature Fiber-Optic Dosimeter, 2014 Sensors 14, 6305-6316.
[2] McCabe, B.P. et al., Calibration of GafChromic XR-RV3 radiochromic film for skin dose measurement using standardized x-ray spectra and a commercial flatbed scanner, 2011 Med. Phys 38, 1919–1930.
[3] K. M. Oh et al., Flexible X-ray Detector for Automatic Exposure Control in Digital Radiography, 2016 JNN 16, 11473-11476.

Speakers: Mr MOOJAE HAN (Department of Radiation Oncology, Busan Paik Hospital, Inje University), Mr YOHAN SHIN (Department of Radiation Oncology, Busan Paik Hospital, Inje University), Prof. SUNGKWANG PARK (Radiation Oncology, Busan paik hospital, Inje University)

69 Geant4 Simulations of a Scintillator and SiPM System for Medical Applications

A detector system based on the use of plastic scintillators and silicon photomultipliers (SiPMs) can be mounted externally to a standart medical LINAC head in order to get the E-LINACs real-time 3D dose profile. To design such a system, the effect of plastic scintillators on the linac dose profile should be investigated. In this work, the effect of the dose distribution on the water phantom of 10 MeV electron beams were simulated using different thicknesses of plastic scintillators for a medical linac. For these simulations the Geant4 Monte Carlo simulation package was utilized together with Root6 toolkits and Paraview software.The simulation includes the major components of the linear accelerator (LINAC) with multi-leaf collimator and a homogeneous water phantom. Calculations were performed for the electron beam with treatment field sizes ranging from 5 cm × 5 cm at 100 cm distance from the source.

This simulation package has also the ability to simulate an intensity modulated radiotherapy scenario comparison with real time detector data. SiPM based and, LYSO scintillator coupled fiber grid system has been developed in the simulation as the detector layer where the beam pattern converted into two dimensional flux and energy based contour image to construct real time dose profile. In the light of this study, a design of a new three dimensional real time dosimeter system is undergoing, including relevant data collection systems and software.

Speaker: Prof. Tuba Conka (Turkish German University)

70 High resolution X-ray micro-CT imaging of fibrin scaffold using large area photon counting detector

Cross-linked fibrin based biomaterial allows one to design scaffolds with controlled stiffness, strength, and permeability for broad use in tissue engineering applications. To be able to assess their microstructural properties reliably, it is necessary to study the material’s internal structure on a volumetric basis and in high detail [1]. In this work, we demonstrate X-ray micro-CT imaging of the synthesized artificial tissue material.
We performed the micro-CT measurement of a fibrin bone scaffold using the in-house designed modular X-ray scanner TORATOM with the dual-source/dual-energy capability, consisting of two perpendicularly mounted imaging chains, each with one X-ray source/detector pair. In the presented study, we used two different types of X-ray imaging detectors to highlight influence of instrumentation on acquired results. The results from the large-area flat panel detector Perkin Elmer XRD1622 (pixel resolution 2048×2048 px, active area 410×410 mm, Gadox scintillator) were compared against data from large area device (LAD) composed of a matrix of 10 × 10 tiles of silicon pixel detectors Timepix (each of 256 × 256 pixels with pitch of 55 um) having fully sensitive area of 14.3 × 14.3 cm2 without any gaps between the tiles. Superior quality of the LAD for imaging of the low attenuation objects was demonstrated in [2].
The radiograms acquired during imaging of the specimen using either detector were used to develop detailed geometrical models of its microstructure. The morphological conformity between the investigated material and the reconstructed topographical data is presented for each considered detector type. We show that organic base material used for production of the relatively small investigated specimens yielding generally low achievable signal-to-noise ratio (SNR) in the radiograms. However, performance of the LAD was superior due to its sensitivity for low energy photons. Nevertheless, to obtain the highest achievable resolution using the LAD detector, it was necessary to introduce correction procedure to compensate drift of the tube focal spot due to relatively long total measurement time necessary to acquire whole CT data set together with relatively high geometrical magnification 34 x. Resultant reconstruction has 1 um voxel size (slight oversampling during tomographic reconstruction). Other corrections were necessary due to LAD imperfections: Edgeless Timepix detector used for LAD assembling has larger border pixels with different response than others; LAD detector is not fully flat. Besides of spot shift compensation and other post-processing corrections, most important improvement was reached thanks to slight tilting of the detector.

Acknowledgment

The research has been supported by the European Regional Development Fund in frame of the project Kompetenzzentrum MechanoBiologie (ATCZ133) in the Interreg V-A Austria - Czech Republic programme and project INAFYM (CZ.02.1.01/0.0/0.0/16_019/0000766).

References
1. D. Kytyr, P. Zlamal, P. Koudelka, T. Fila, N. Krcmarova, I. Kumpova, D. Vavrik, A. Gantar, S. Novak, Deformation analysis of gellan-gum based bone scaffold using on-the-fly tomography, Materials and Design 134 (2017) 400–417.
2. I. Kumpova , D. Vavrik, T. Fila, P. Koudelka, I. Jandejsek, J. Jakubek, D. Kytyr, P. Zlamal, M. Vopalensky, A. Gantar, High resolution micro-CT of low attenuating organic materials using large area photon-counting detector, J. Instrum. 11 (2016).
3. X. Llopart, M. Campbell, R. Dinapoli, D. San Segundo, E. Pernigotti, Medipix2: a 64-k pixel readout chip with 55lmsquare elements working in single photon counting mode, IEEE Trans. Nucl. Sci. 49 I (2002) 2279–2283, 2014.

Speaker: Petr Koudelka (Czech Academy of Sciences, Institute of Theoretical and Applied Mechanics)

71 In-beam PET treatment monitoring including short and intermediate lifetime isotopes

Background: The quality of charged particle therapy treatments depends on the ability to predict and achieve a given particle range in the patient. Non-invasive moni- toring of the particle range can be performed by detecting with a dedicated PET system β+ emitters, produced in the patient as a result of nuclear interactions of charged hadrons with tissue, with a dedicated PET system. The correctness of the dose delivery can be verified by comparing measured and pre-calculated Monte Carlo space and time distributions. The reliability of Monte Carlo (MC) predictions is hereby a key issue. Most studies performed so far focus on long time intervals, where the signal is dominated by the β+ decays of 11C and 15O, and typically do not include β+ isotopes with shorter life-times.

Goal: In this work, we investigate the reliability of MC predictions of space and time (decay rate) profiles in various time intervals relevant for in-beam PET monitoring up to a few minutes after treatment. Moreover, we show how the decay rates can give an indication about the elements present in the phantom.

Methods and Materials: Various phantoms were irradiated in clinical and near- clinical conditions at the Cyclotron Centre of the Bronowice proton therapy centre. PET data were acquired with the DoPET system, a planar 16x16 cm2 PET system based on LYSO crystals and Hamamatsu H8500 position sensitive photo-multipliers. Monte Carlo simulations of particle interactions and photon propagation in the phantoms were performed using the FLUKA code. The analysis included a comparison between experimental data and MC simulations of space and time profiles, as well as a fitting procedure to obtain the various isotope contributions in the phantoms.

Results and conclusions: We show comparisons of the space and time distributions between data and Monte Carlo, including examples where the Monte Carlo and data are in slight disagreement (mostly in short-time intervals from the beginning of irradiation, i.e.,¡20 s). Moreover, we show how to separate the different isotope contributions including 11C, 15O, 10C and 8B in the various time intervals. Our results demonstrate that it is possible to obtain useful information about the treatment by using not only space but also time profiles.

Speakers: Aafke Kraan, Prof. Valeria Rosso (INFN and University of Pisa)

72 Investigation of radiation hardness in lateral position sensitive detector irradiated with 13.5 nm EUV

The development of lateral position sensitive detector for EUV irradiation is important for use in the lithographical equipment for 13.5 nm, which is the new standard in semiconductor industries.Using a scintillator (CsI(Tl)) at 13.5 nm wavelength produces ~6 light photons which results in 6 eh-pairs in the detector. Direct conversion produces 26 eh-pairs in the silicon detector, which is more than 4 times as efficient. Producing 13.5 nm needs hot dense plasma or a synchrotron source. By filtering hot dense plasma by using multilayer Mo/Si mirrors, a beam of a wavelength 13.5 nm is generated. Several mirrors can be involved which makes a measurement of the beam position necessary for optimal beam performance, homogeneous wave front and increased brightness. Therefore a position sensitive direct conversion detector for 13.5 nm is needed.

Two different types of detector are tested, n in p-substrate and p in n-substrate, silicon lateral position sensitive detectors. The detectors are manufactured by Sitek electro optics. The passivation of surface is by oxidation, oxynitride, and titanium annealed at 400 °C for 1 hour (forming gas annealing). The layer thicknesses (~10 nm) have been simulated by using MCNP to keep control of the attenuation. Stress measurement have been done at Elettra Sincrotrone Trieste using the beam line for Circular Polarization

Speaker: Dr Göran Thungström (Mid Sweden University)

73 K-edge sensitive X-Ray Radiography for Small Animal using fully spectral Timepix3 detector

Tissue type sensitive imaging with photon counting detector of Medipix family has been demonstrated recently. This capability stands promising for diagnosis of different diseases and pathological changes. In comparison methods with contrast agents provide functional information. Usage of contrast agents brings however complexity and drawbacks such toxicity. Most contrast agents used in clinical practice are iodine-based compounds. Iodine contrast agent in high concentration that is required for high contrast can lead to fatal effects. The main goal of researchers is to develop a new type of contrast agents with low toxicity or decrease the volume of the contrast agent injected to specimen. To achieve this the new more sensitive imaging methods are needed. One of these methods is K-edge imaging. K-edge imaging is a unique method that provides detection and quantification of given element thanks to element specific increase in x-ray absorption occurring at a specific energy. For this purpose, a high resolution energy and high spectral sensitive imaging detector such as Timepix3 is needed. Spectral measurements were performed using new developed Timepix3 and Medipix3 detectors. Spectral performance and properties of 300 um thick silicon and 1000 um CdTe were evaluated. Commonly used contrast egents such as iodine resp. gold based compounds with K-edges at 28.1 keV (Ka) resp. 68.8 keV (Ka) were used as suitable candidates.
Thanks to revealing of contrast agents based on their spectral responses gain K-edge imaging method a high potential for preclinical diagnostic.

Speaker: Mrs Eliska Trojanova (1)

74 Material Discrimination Using Dual Energy X-ray Beam for Cargo Inspection system

ABSTRACT
Generally, the x-ray cargo inspection systems are used to smuggled and dangerous goods detection inside of the container. When scanning with dual energy x-ray beams, different materials can be identified by comparing the attenuations of the transmitted dual energy x-ray beam. In this paper, we present a method to improve detection efficiency by using dual-energy x-ray. we use a new logarithmic curve for material discrimination rather than the conventional curves such as banana curve, alpha curve, R curve, and H-L curve. It is because the logarithmic curve has a better linearity in shape so as to make it easy. For each material, probability distribution along the curve is estimated with jig material images acquired by our dual energy system. The distribution is used as a weight of each material for material discrimination. We measured seven materials for different thicknesses using the dual x-ray energy of 6MeV and 9MeV. Experimental results show that our method based on the proposed weights gives accurate material discrimination. We will test the various combinations of materials in our test-bed for detection of each materials. In addition we will continue to study on improving detection efficiency for overlapped objects based on experimental data.

ACKNOWLEDGMENTS
This research was supported by a grant from National R&D Project of “Research on Fundamental Core Technology for Ubiquitous Shipping and Logistics” funded by Korea Institute of Marine Science and Technology Promotion(PMS3791).

REFERENCE
[1] V. L. Novikov, S. A. Ogorodnikov, and V. I. Petrunin. Dual energy method of material recognition in high energy introscopy systems. Questions of Atomic Science and Technology [translated from Russian], 4(2):93–95, 1999.

Speaker: Mr JeongHee Lee (Korea Research Institute of Ship & Ocean Engineering)

75 Pattern Recognition Algorithm for Charge Sharing Compensation in Single Photon Counting Pixel Detectors

Thanks to the detector technology development based on high Z materials (GaAs, CdTe, CZT, etc.), the hybrid pixel detectors with direct photon–to–charge conversion become more and more popular, even in medical applications. Single photon counting systems aim at good position resolution and operation with high X-ray flux, so making a pixel size smaller is a general tendency in such systems. However for the detector pitch of about 100 m and smaller the effect of charge division is present in about 20-30% cases of all incoming photons, while for the smaller pixel size this effect is even more pronounced. In this presentation we analyze different algorithms implemented in integrated circuits of a pixel architecture and we propose a new algorithm to eliminate the effect of charge sharing called Multithreshold Pattern Recognition algorithm (see Fig. 1). The algorithm is extensively tested for X-ray energy range 20-160 keV and finally implemented in the design of readout chip with pixel pitch of 100 m in CMOS 130 nm process. Operation at four different energy threshold allows a photon counting in selected energy windows and fast hit allocation.

Fig. 1. Three approaches examples of choosing a proper energy threshold for hit allocation.

This work has been supported by the National Science Center, Poland under Contract No. UMO-2016/21/B/ST7/02228.

[1] R. Ballabriga, et al., "The Medipix3RX: a high resolution, zero dead-time pixel detector readout chip allowing spectroscopic imaging", 2013 JINST, 8, C022016, pp 1-15.
[2] A. Krzyzanowska, G. W. Deptuch, P. Maj, P. Grybos, and R. Szczygiel, “Characterization of the Photon Counting CHASE Jr., Chip Built in a 40-nm CMOS Process with a Charge Sharing Correction Algorithm Using a Collimated X-Ray Beam,” IEEE Trans. Nucl. Sci., vol. 64, no. 9, 2017.
[3] P. Otfinowski et al., “Asynchronous Approximation of a Center of Gravity for Pixel Detectors' Readout Circuits,” IEEE Journal of Solid State Circuits, in print.

Speaker: Dr Piotr Otfinowski (AGH University of Science and Technology)

76 Radiation damage effects caused by soft X-rays in silicon pad detectors

Detailed study of energy resolution and charge division effects in silicon pad detectors performed recently show that these are suitable devices as high energy resolution detectors offering high count rate capability and moderate spatial resolution [1]. An energy resolution better than 220 eV FWHM at 5.9 keV at room temperature has been demonstrated using a low noise front-end electronics optimised for this application. Such detectors can be used in laboratory instruments using X-ray tubes. Given that the noise performance of the sensor and front-end electronics has been pushed to the limits such devices are potentially susceptible to radiation damage effects caused by soft X-ray even if the total doses are relatively low compared to synchrotron applications, for example.
In the paper we present the results of a systematic investigation of X-ray induced radiation damage in silicon pad sensors for total ionising doses up to 300 Gy. The radiation induced degradation of energy spectra as well as of critical sensor parameters, capacitance and leakage current, have been measured vs total dose. Detailed analysis of the measured energy resolution shows that degradation of energy resolution can be explained by degradation of the sensor parameters. Additional radiation effects related to charge division has been observed and will be discussed in the paper.
References
[1] P. Wiącek et al., Limitations on energy resolution of segmented silicon detectors, 2018 JINST 13 P04003

Acknowledgement
This work was partially supported by the AGH UST, statutory tasks No. 11.11.220.01/4 within subsidy of the Ministry of Science and Higher Education.

Speaker: Mr Piotr Wiacek (AGH University of Science and Technology, Faculty of Physics and Applied Computer Science)

77 Sensor and module developments for the ATLAS ITK pixel system

The entire tracking system of the ATLAS experiment will be replaced during the LHC Phase II shutdown (foreseen to take place around 2025) by an all-silicon detector called the “ITk” (Inner Tracker). The pixel detector will comprise the five innermost layers and it will be built of new sensor and readout electronics technologies to improve the tracking performance and cope with the severe HL-LHC environment in terms of occupancy and radiation.
A new on-detector readout chip has been designed and produced in the context of the RD53 collaboration in 65 nm CMOS technology. This paper will present the on-going R&D within the ATLAS ITK project towards the new pixel modules.
Planar and 3D sensors have been re-designed with cell sizes of 50x50 or 25x100 μm2, compatible with the RD53 chip. A sensor thickness equal or less than 150-200 μm is foreseen for the outer layers, where high yield and low costs are required, and 100-150 μm in the two innermost ones where radiation hardness is the major concern. Several prototyping sensor productions are being carried out at the moment in collaboration with commercial vendors to identify the best technology for the different pixel layers.
A particularly challenging aspect of the ITK pixel module assembly is represented by the sensor-chip interconnection given the large area to be covered and the compressed production schedule. An intense R&D is being carried out to satisfy the yield requirements when using 100-150 m thin chips with the higher bump density and the larger chip wafer size (12” for the 65 nm CMOS technology versus the 8” of the 130 nm CMOS).
After extensive characterization of the sensors in the laboratory, their charge collection properties and hit efficiency are measured in common test-beam campaigns, which provide valuable feedback for improvements of the layout. Testbeam measurements of the final prototypes will be used for the decision of which sensor types will be installed in ITk.
The setups used in the ITk Pixel testbeam campaigns will be presented and results from the latest measurements will be shown, highlighting some of the developments and challenges for the ITk Pixel sensors.

Speaker: Anna Macchiolo (Max-Planck-Institut fur Physik (DE))

78 Spectral analysis of 3D and planar sensors with Timepix3

Abstract

Timepix3 is a hybrid pixel detector readout chip that features an event driven mode. Upon incident particle hits, it sends processed pixel coordinate, time-over-threshold and time-of-arrival to the readout [1].

Timepix3 modules with planar and 3D sensors where studied with different x-ray sources, low and high-energy hadrons at test beams. This paper will report on the comparative spectral analysis of the two detectors and discuss their performances.

Summary

3D sensors compatible with Timepix3 where fabricated at CNM Barcelona, Spain and bump-bonded at Advacam, Finland. The structure of 3D sensors, where electrodes are processed inside the silicon bulk is known to affect the spectral response of x-ray and charged hadrons due to the presence of the electrodes themselves and the special electric field layout. In particular 3D are shown to be less prone to charge sharing than planar sensors. In this study we will use the Timepix3 features to study the detailed spectral response of the 3D sensors at different energies and particle incident angles to understand the spatial response of the sensor when compared to a planar one.

Comparisons studies of 3D and planar sensors include cluster size, as can be seen in Figure 1, cluster energy, pixel energy versus induced charge, and effects of 3D electrodes on the overall detection efficiency. The studied angles with respect to the incident beam are 0, 45, 60, and 70 degrees. Data were collected at the SPS facility of CERN, with 40GeV/c pion beams and at the Van Der Graaf facility in Prague with low energy proton and with radioactive sources at Manchester and Prague. For test beams the detectors were configured in a telescope arrangement of 2 synchronized planar and 3D-Timepix3 detectors and Readout with the Katherine software.

As an example of the results that will be shown, in Figure 2 top the Landau distribution of 40 GeV pions at zero degrees incident angle collected with the 3D sensors (main figure cluster size>2, insert cluster size>=1). At the bottom, images of the corresponding event distributions on the sensor for the low energy peaks and for the main landau peak for cluster size greater than 2. We will analyse these finding and report correlations with the geometrical response of the sensor.

[1] T Poikela et al 2014 JINST 9 C05013

Speaker: Tianqi Gao (University of Machester)

79 Stack of Timepix-based detectors with Si, GaAs:Cr and CdTe sensors with optimized thickness for spectral micro-CT

Hybrid pixel detectors based on Medipix chips have proven themselves as a good tool for spectral X-ray imaging. An important advantage of such detectors is the possibility to use different sensor materials (Si, GaAs:Cr, CdTe etc.). Each of these materials has own advantages and disadvantages. The higher Z materials provide the higher photon registration efficiency, and the greater energy of the fluorescent photons distorting the detector's energy response. The thickness of the sensor affects the same way: the thicker sensor means the higher photon registration efficiency, but the worse energy resolution. At the same time, photons that have passed through the detector without interacting in it can be registered by the next detector. Combining several detectors with different sensor materials, allows increasing the overall photon registration efficiency. In this case, each detector will operate in the optimal energy range. In this paper, we consider a system based on three Timepix detectors with sensors made from Si, GaAs:Cr, CdTe. The Monte Carlo simulation was used for the sensors thickness optimization. The results of measuring spatial resolution, energy resolution, photon registration efficiency in each layer are presented. The possibility and advantages of using such imaging system for spectral micro-CT are demonstrated.

Speaker: Mr Danila Kozhevnikov (Joint Institute for Nuclear Research)

80 Time of flight measurement by the PH32 chip

The frontend readout chip PH32 suitable for measurement of X-rays, beta radiation and ions first presented in [1] is primarily dedicated to the dose rate measurement and basic spectroscopy. Present article is focused on the time of flight functionality which was implemented in the new version of the PH32 chip and can be used in various applications as particle tracking or ion mass spectroscopy [2]. The PH32 chip was manufactured using a commercial 180 nm CMOS process and can operate in two operation modes. The first mode, the High Gain Mode (HGM), is suitable for the measurement of the generated charge in the range from about 5 ke- to 70 ke- for soft X-rays and beta radiation. The second, the Low Gain Mode (LGM), is used in the range from about 500 ke- to 7 Me- suitable for alpha particles. The PH32 chip contains 32 identical channels operating individually, which can be connected to the silicon strip sensor by the wire bonding. The chip is optimized to the strip sensor capacitance of 8 pF with the AC coupling. The electronic noise is about 1100 e- for the chip calibration charge of 10 ke- at HGM and about 2300 e- for the chip calibration charge of 2 Me- at LGM. The measurements presented in this paper are focused on the channel response to the injected charge including channel dispersion and time-walk measured with variety of injected charge. Time of flight is derived from an internal oscillator which can be set by internal data to analog converter up to around 333 MHz. Time-stamp is stored in a 16 bit asynchronous counter for every channel separately, which can be read as a shift register from all channels after exposition. The internal oscillator is a significant source of dispersion between channels and between hit occurrences. The variety of sampling frequencies and their influence on the measurement is discussed in this article as well as the effect of equalization of the chip which has to be performed for the proper chip functioning. The PH32 chip contains also a 40 bit synchronous counter designed in STSCL logic which can be used for time of flight measurement triggered by an internal trigger derived from all channels. This method of time of flight measurement can be efficient only for low dose rate, because only the first hit from all channels is detected. However, the chip using this functionality can operate in hit energy measuring mode [1] and time-walk can be corrected. This article presents correlation between time of flight measured from all channels by one 40 bit synchronous counter and energy measurement for every channel separately. Internal trigger used for triggering STSCL logic is also propagated to output for triggering external hardware. The output is realized as proprietary differential signaling to avoid crosstalk between digital output at PCB and wirebonding between sensor and chip. This article compares the results provided by the 40 bit synchronous counter with trigger output and comparison between time of flight measured by individual channel by the 16 bit asynchronous counter with trigger output. The article describes a newly developed frontend readout chip suitable for particle rate measurement, measurement of deposited charge and time of flight.

[1] Janoska, Zdenko, et al. "Measurement of ionizing particles by the PH32 chip." Nuclear Science Symposium and Medical Imaging Conference (NSS/MIC), 2015 IEEE. IEEE, 2015.
APA
[2] Nomerotski, Andrei, et al. "Characterization of TimepixCam, a fast imager for the time-stamping of optical photons." Journal of Instrumentation 12.01 (2017): C01017.

Speaker: Zdenko Janoska (Czech Technical University (CZ))

81 Transient phenomena in Schottky Au / CdZnTe interface

CdTe optoelectronic based devices has been extensively studied during last decade due to their multiple applications. The time change in the physical properties due to the charge dynamics in these devices is the main reason of the degradation of their performances.
In this contribution, we present a study of the current transients and polarization phenomena [1, 2] in n-type CdZnTe material with Au contacts that showed upwards bending of the bands near the metal-semiconductor (M-S) interface. The transient of the electric field along the sample due to the accumulation of positive space charge below the cathode at the M-S interface was studied by means of the Pockels effect. The analysis of the time and temperature variation of the electric field values at the M-S interface has provided the parameters of the deep donor responsible for the polarization mechanisms in the n-type CdZnTe. We explain the change of the current with time as a result of changing electric field at the interface. The field dependence of barrier height due to the formation of interfacial layer at the M-S interface is the mechanism responsible for the current transient with no more assumption. Model to explain the transient phenomena in the studied structure has been proposed based on our findings, Fig.1.

Fig. 1. Illustrates the energy bands bending of Au-CZT at 200V. De- trapping of electrons from deep donor level and transitions from the metal to CZT conduction band are shown based on values extracted from our experimental data.

[1] H Elhadidy, V Dedic, J Franc and R Grill, J. Phys. D: Appl. Phys. 47 (2014) 055104.
[2] V. Dědič, M. Rejhon, J. Franc, A. Musiienko, and R. Grill, Appl. Phys. Lett. 111, 102104 (2017)

Speaker: Dr Hassan Elhadidy (, Institute of Physics of Materials, Academy of Sciences of the Czech Republic)

82 XRF Methods for Online Measurement of Chromium, Molybdenum and Cadmium Contents in Fly Ash from Municipal Solid Waste Incineration

Incineration of the municipal solid waste (MSW) is the most efficient way in waste management, because of the large reduction in volume and high energy recovery. According to European Commission, already in 2006, the EU-15 countries were treating approximately 20 - 25% of their MSW by incineration, with an anticipation of a massive expansion over the following years. After the incineration process; ash is produced along with flue gas and heat. The ash contains mostly inorganic components and, can be spread into the atmosphere by the flue gas. At the final stage the ash is dumped into landfills, and its contents can be washed away and carried to our natural water reservoir by the rain water. Some of the contents in the ash are hazardous to health and the environment. Therefore, it is important to routinely monitor the constituents in the ash, and clean it before abandonment.

In this work, X-ray fluorescence (XRF) measurements have been conducted in order to measure Chromium (Cr), Molybdenum (Mo) and Cadmium (Cd) in ash, collected from the local incineration plant in Sundsvall. These elements have been on the focus of the study because of their carcinogenic properties and other severe effects on health. The concentrations of these elements in the ash are relatively lower than other elements, such as Iron. As a result, the fluorescence signal from these elements are very low in the spectra. For this reason, it is crucial to have good signal-noise-ratio (SNR) in the detection system in order to detect them with certainty. Due to the large range of fluorescence energies of the materials of interest, several filters and tube voltages have been used to optimize the performance for different elements. By doing this, a significant increase in the SNR has been observed, which led to more distinct detection of the aforementioned elements.

Speaker: Dr Salim Reza (Mid Sweden University)

Wednesday, 27 June

Invited

83 Direct Electron Detectors for Electron Cryomicroscopy

Speaker: Dr Wasi Faruqi (MRC Laboratory of Molecular Biology)

Oral

Convener: Renata Longo (UNIVERSITY OF TRIESTE & INFN)

84 Low dose single shot two-dimensional phase contrast imaging

X-ray phase contrast imaging provides a method to distinguish materials with similar density and effective atomic number, which otherwise would be difficult using conventional X-ray absorption [1]. In the recent decade, multiple methods have been developed to acquire X-ray phase contrast images using incoherent laboratory sources. One of these is the Double Masked Edge Illumination (DM-EI) method. DM-EI enables a large field of view and makes it possible to utilize the full energy range of conventional X-ray laboratory sources [2]. However, the method needs a minimum of two image acquisitions to do a full phase retrieval, and it requires relatively complicated mask apertures or even additional sample exposures to make a two-dimensional phase contrast image.

We have developed, and experimentally tested, a low dose, single mask, edge illumination method that enables two-dimensional phase contrast images in a single shot (Figure 1). The method is based on the technique originally proposed by F. Krejci et al. [3], and uses a micro focus source, a two-dimensional mask, and an Advapix detector, which contains the Timepix3 chip developed by the Medipix consortium. By using a highly absorbing, pre sample tungsten mask, our single shot method is capable of producing phase contrast images at low sample radiation dose and is applicable for both hard and soft X-rays. In addition to this, the method shows great potential for further development by using the Timepix3’s fast time of arrival information.

Speaker: Mr Erik Schou Dreier (Niels Bohr Institute, University of Copenhagen)

ErikSchouDreier.pdf

85 Active personal dosimetry with Dosepix, a 16 energy channel hybrid pixel detector

Dosepix is a hybrid pixel detector with 16x16 pixels at a pitch of 220µm based on the time-over-threshold measurement principle of Timepix. It has been developed for dosimetry of ionizing radiation, especially for personal dosimetry in pulsed X-ray radiation fields as available active personal dosimeters can fail at very high dose rates or short pulse durations. The pixel size is a compromise between the necessary small size of the pixel for high dose rates, reasonable energy resolution and the area needed for the electronics in the pixel cell. The pixel cell circuitry features a charge sensitive preamplifier which is connected to a leading edge discriminator. Dosepix features a register for analog threshold adjustments among pixels. The length of the digital output of the discriminator is the time--over--threshold (ToT). A clock signal for ToT measurement is generated in the ASIC by a phase-locked-loop. The ToT is determined for each event in a binary counter. The content of the ToT counter is copied to the ToT latch for energy channel assignment. The ToT counter is then reset to zero so that the next arriving pulse can be processed. The energy channel to which the ToT value has to be assigned is identified by the binning-state machine. It sequentially compares 16 stored digital thresholds with the measured ToT value. The 16 pre-defined time-over-threshold (ToT) values in each pixel serve as borders of the 15 energy channels (ToT channels) and as low threshold of the 16th energy channel. The 16 thresholds can be defined individually for every pixel. If the measured ToT value is above or equal to a certain digital threshold and if it is below the threshold of the next energy channel, the event is assigned to a 16-bit binary counter (energy bin register) associated with the energy channel. This way, a histogram of the number of detected events with energy depositions in the predefined energy intervals (ToT intervals) is accumulated in each pixel. Readout proceeds one column at a time (while the others remain sensitive) to avoid dead-time for the whole detector. As the spectra of deposited energies are very similar for different high photon energies, several detectors with adapted filters have to be used in order to determine personal dose equivalents correctly up to 1.5 MeV. Personal dose equivalents can be calculated from the measured energy histograms as a linear combination of the number of counts with conversion factors determined with a Monte-Carlo-simulation. First measurements at the German Metrology Institute PTB indicate, that personal dose equivalents can indeed be measured accurately with photon counting pixel detectors such as Dosepix, even at very high dose rates. Figure 1 shows the deep dose response (measured/true dose) as function of the average photon energy for perpendicular irradiation and +-60 degrees angle of incidence. One can see that Dosepix measures the doses with the required accuracy (green band).

In this talk we will review the functionality of Dosepix and present first measurements of personal dose equivalents with a setup with 3 Dosepix detectors.

Speaker: Patrick Hufschmidt

86 Molecular additives for high pressure Xe TPCs to improve the topological signature

Gaseous time projection chambers (TPC) is a very attractive detector technology for particle tracking. Efficient discrimination of the events through pattern recognition of the topology of primary ionisation tracks is a major motivation for such type of TPCs. High-pressure xenon (HPXe) time projection chambers (TPCs) gained increasing importance during the last years due to their application to rare-event detection such as double-beta decay (DBD), double-electron capture (DEC) and directional dark matter (DDM). Gaseous xenon TPCs offer important advantages when compared to liquid xenon and double phase xenon TPCs: interaction of rare events in the gas allows to determine the event topological signature, as demonstrated for DBD and DEC detection in contrast to interaction in the liquid, where the extremely reduced dimensions of the primary ionisation trail rules out any possible topology-based pattern recognition. In addition, the achieved energy resolution in interactions in the gas is much better than that achieved in the liquid, having also a positive impact on the background reduction. Along this line, optical TPCs based on electroluminescence (EL) amplification of the primary ionisation signal present much better energy resolution than what is achieved in TPCs based on charge avalanche readout being, therefore, the most competitive alternative. The Next-White (NEW) detector is currently the World’s largest radio-pure high pressure xenon gas time projection chamber with electroluminescence readout. NEW has been operating at Canfranc Underground Laboratory (LSC) since December 2016.
However, drift velocity and diffusion play a major role in the TPC tracking performance, being the large electron diffusion in pure Xe a limiting factor for the capabilities of background discrimination based on the topological signature. In the HPXe TPC used by the NEXT collaboration, electron transverse diffusion may be as high as 10 mm/sqrt(m): after one meter of drift, a point-like ionisation deposit becomes a cloud distributed as a Gaussian of 10 mm sigma in the direction perpendicular to the electric field. This situation is far from ideal and can be largely improved by adding molecular electron coolants to the main gas. By adding a molecular gas to pure xenon, new molecular degrees of freedom from vibrational and rotational states are made available for electron energy transfer in inelastic collisions and the energy distribution of the ionisation electron cloud in the drift region tends to build up around the energy of the first vibrational level, typically ~ 0.1 eV, even in the presence of sub-percent concentrations of the additive. The thermal diffusion limit gives a diffusion factor of ~ 1.5 mm/ sqrt(m) for a drift field of 250 V/cm.
The presence of molecular species in a noble gas was believed to dramatically reduce the EL yield. If an electron has a significant probability of colliding with a molecular impurity before it obtains from the electric field sufficient energy to excite a noble gas atom, it may lose part of its energy without producing EL photons. Besides electron cooling, additional losses of scintillation originate from excimer quenching, photo-absorption and electron attachment, with impact on the statistical fluctuations of the EL. The reduction of the EL yield and respective statistical fluctuations depend on the amount and type of impurity present in the gas.
In this talk, we present an overview of the NEW TPC (10 bar Xe, with a length of 664.5 mm and a diameter of 522 mm) and its performance characteristics. We also present the results from a campaign designed to systematically add several molecular gases (CO2, CH4 and CF4) to Xe, at minute concentration levels to find a suitable mixture able to reduce diffusion and improve the topological discrimination of the events, without compromising the performance of the TPC significantly in terms of EL yield and energy resolution.

Speaker: Dr Cristina M. B. Monteiro (LIBPhys, Physics Department, University of Coimbra, P-3004-516 Coimbra, Portugal)

Swedish fika in the hotel lobby (2nd floor)

Oral

Convener: Seppo Nennonen

87 TCAD simulations of pixel sensors for the ATLAS ITk upgrade and performance of annealed planar pixel modules

In view of the high luminosity phase of the LHC (HL-LHC) to start operation around 2026, a major upgrade of the ATLAS inner tracker system is in preparation. Thanks to their reduced power dissipation and high charge collection efficiency after irradiation, thin planar pixel modules are the baseline option to instrument all, but the innermost layer of the pixel system.
To optimise the sensor layout for a pixel cell size of 50$\times$50 $\mu$m$^2$, TCAD simulations are being performed. Charge collection efficiency, electronic noise and electrical field properties are investigated. Two radiation damage models are employed to estimate the performance before and after irradiation.
The effects of storage time at room temperature for the ITk pixel detector during maintenance periods are reproduced using sensors irradiated up to a fluence of 1$\times$10$^{16}$ n$_\text{eq}$/cm$^2$. Pixel sensors of 100-150 $\mu$m thickness, interconnected to FE-I4 read-out chips with pixel dimensions of 50x250 $\mu$m$^2$, are characterised using the test-beam facilities at the CERN-SPS and DESY. The charge collection and hit efficiencies are compared before and after annealing at room temperature up to one year.

Speaker: Julien-Christopher Beyer (Max-Planck-Institut fur Physik (DE))

jbeyer_iworid2018.pdf

88 The silicon strips Inner Tracker (ITk) of the ATLAS Phase-II upgrade detector

The inner detector of the present ATLAS detector has been designed and developed to function in the environment of the present Large Hadron Collider (LHC). At the next-generation tracking detector proposed for the High Luminosity LHC (HL-LHC), the so-called ATLAS Phase-II Upgrade, the particle densities and radiation levels will be higher by as much as a factor of ten. The new detectors must be faster, they need to be more highly segmented, and covering more area. They also need to be more resistant to radiation, and they require much greater power delivery to the front-end systems. At the same time, they cannot introduce excess material which could undermine performance. For those reasons, the inner tracker of the ATLAS detector must be redesigned and rebuilt completely.
The inner detector of the current detector will be replaced by the Inner Tracker (ITk). It consists of an innermost pixel detector and an outer strips tracker. This contribution focuses on the strips tracker. The basic detection unit of the strips tracker is a silicon microstrips “module”. It consists of a ~ 10 x 10 cm2 n-in-p silicon microstrip sensor fabricated in a float-zone (FZ) substrate, with its associated power, control and readout electronics directly glued on top of it. There are eight flavors of modules in total with slightly different geometries and strip lengths, two for the barrel central region and six for the end-cap forward regions. The barrel modules contain rectangular, 10x10 cm2, sensors, and the end-cap modules have wedge-shaped sensors of different geometries. The modules use binary readout architecture by means of differential SLVS data signaling. The readout electronics consists of readout chips manufactured in 130 nm CMOS technology, mounted on top of polyimide-copper Printed Circuit Boards (PCBs) called “hybrids” and connected to the microstrips via wirebonds. A power board contains the power and control electronics, based on Low Voltage (LV) DC-DC power conversion, a High Voltage (HV) multiplexer, and an Autonomous Monitoring And Control ASIC (AMAC). The modules are glued on both sides of lightweight carbon fiber-based support structures, called “staves” for the barrel region and “petals” for the end-cap regions. The petals and staves consists of a carbon fiber sandwich structure with embedded titanium cooling pipes, hosting 28 modules in the staves and 18 modules in the petals. Dual-phase, evaporative CO2 is used as a coolant. Connection to the modules is achieved via thin, polyimide-based circuit boards, called “bus tapes”, which include low impedance differential data lines, control lines, LV and HV power rails and shielding. The bus tapes are co-cured directly with the carbon fiber facesheets, on the skins of the staves and petals. Additionally, a data concentrator board called “End of Substructure” (EoS) is mounted on each stave/petal side. The EoS board interacts with the modules and with the outside world, hosting the components and connectors that multiplex the module data and sends it in and out the detector via high-speed optical links. The support structures are then mounted into lightweight carbon fiber-based global structures (the barrel and the end-caps), providing mechanical support for the staves and petals and the electrical, optical, and cooling services. The barrel and the end-caps are integrated together into the ITk tracker, inside the Outer Cylinder (OC), which, along with the pixel tracker, also integrated in the OC, constitute the full ITk.
This contribution will focus on the latest R&D activities performed by ITk strips community with respect to the assembly and test of the strip modules and the stave and petal structures, and their integration into the global structures. The overall plans for the production phase of the experiment will also be detailed.

Speaker: Dr Mercedes Minano Moya (Instituto de Física Corpuscular (CSIC-UV))

89 Lab and Test Beam Results of Irradiated Silicon Sensors with Modified ATLAS Pixel Implantations

The inner detector of the ATLAS experiment consists of a pixel detector, a strip detector and the transition radiation tracker. During the Phase-0 upgrade in 2013 the Insertable B-Layer (IBL) with a new beam pipe was added as the innermost layer. The planar pixel sensors of the IBL are produced in n$^+$-in-n technology and have a pitch of $250\,\mu$m $\times$ $50\,\mu$m. The pixel design of these sensors with a rectangular n$^+$-implant are the baseline for modified pixel designs which were developed in Dortmund.
Different implantation shapes are designed to cause electrical field strength maxima to achieve an increased charge collection after irradiation and thus higher particle detection efficiencies at lower operation voltages.
To test and compare the different pixel designs, all six modified pixel designs and the standard IBL design are placed on one sensor. With individual guard rings and separated high voltage pads on the p-side of the sensor, any pixel design matrix can be depleted separately. These sensors can be read out with FE-I4 read out chips.
Several sensors have been irradiated with protons or neutrons in irradiation test facilities to simulate the radiation damage. In this way the influence of the pixel design on the sensor performance after irradiation can be investigated.
Test beam measurements provide an environment close to the real operating conditions in the detector because the sensor detects charged particles with high energies. With the results of test beam measurements advanced sensor properties like charge collection efficiency and track recognition efficiency can be examined. The tracks are measured with a Mimosa26 telescope and reconstructed with the EUTelescope software. The sensor properties are analyzed with TBMon2.
Basic sensor characterization like IV measurements and source scans are performed in the lab.
In this talk, lab measurements of these prototype sensors are presented. Results of test beam measurements complete the characterization of the different pixel designs.

Speaker: Mareike Weers

90 Timepix3 detector network at ATLAS experiment

Detectors of Medipix/Timepix family are usable in a wide range of applications. The original purpose of an x-ray imagining was broadened to many areas, where mainly Timepix detectors became very often used as a powerful tool in a particle and an experimental physics. The latest member of this class – the Timepix3 ASIC (read-out chip) – allows measuring the Time-over-Threshold (ToT) and the Time-of-Arrival (ToA) values of hits at the same time. It also yields a time resolution of 1.56ns, a high hit-rate and event data-driven readout mode.
Nowadays, on the market and in the scientific community there are several available readout devices and instruments (Katherine readout from UWB, CTU; AdvaDAQ from Advacam or SPIDR from Nikhef) that make it possible to utilize the detectors easily. However for specific applications, as measuring at ATLAS cavern (experiment area) at CERN is, there was a need to design an enhanced readout system meeting requirements of a time synchronization (detectors among each other and with a trigger), a long distance among sensors and readouts, a radiation hardness, a cooling, a power supply etc.

Attached figure
Figure 1. Radiation during collisions (10s)

In this contribution, authors describe the design of the detector network based on the Timepix3 and the Katherine readout (with Gigabit Ethernet interface). The main emphasis is put on the technical solution and the concept of the project. The designed system, which was already installed in January 2018 at ATLAS experiment at CERN, consists of four Timepix3 detectors (in a configuration of a two-layer telescopes) and readout system. For reason of expected harsh radiation field, only the detector units (sensors with power supplies) of the system are placed in the cavern, all readout electronics is far outside of the radiation field – localized in a rack room. Actually, the distance between detector and readout device is 80m (up to 120m was tested) in the installed system. Time-stamping of all detectors in the network work synchronously to each other; it means user can observe events in various parts of the cavern (generally place with rad. field) with uniform time stamps. Additionally, this time-stamping is also in exact relation to the orbit clock generated by LHC accelerator.
Authors also present first results of the design system, they are focused mainly on the time-stamping performance and the stability of distributed detector network.

Speaker: Dr Petr Burian (Czech Technical University (CZ))

burian_talk.pptx

91 Characterization of the ePix100a and the FastCCD semiconductor detectors at European XFEL

ABSTRACT

Intense X-ray FEL beams allow research using experimental techniques: e.g. record chemical reactions on timescales previously not accessible (femto-second experiments), resolve atomic details of viruses and study new material structures. The European X-ray Free Electron Laser (XFEL.EU) is a research facility providing spatially coherent X-ray flashes in the energy range from 0.25$\,$keV to 25$\,$keV of unprecedented brilliance and with unique time structure: X-ray pulses with a 4.5$\,$MHz repetition rate arranged in trains with 2700 pulses every 10$\,$ms [1]. There are three beamlines (SASE1, SASE2 and SASE3), each hosting two scientific experiments. In order to gain information from the experimental data, properly characterized and calibrated X-ray detectors are required. Supplementing high repetition rate detectors at MHz speeds, detectors at 10$\,$Hz, matching the train rate of 10$\,$Hz, such as the ePix100a and the FastCCD will be used at the facility. These 2D silicon pixelized detectors use column-wise readout as CCDs or hybrid pixel detectors, with columns defined either on the sensor- (CCD) or ASIC-level (ePix). Characterization and analysis approaches of the FastCCD [2] and ePix100a [3] detectors are discussed and the performance of the detectors is evaluated using appropriate state-of-the-art analysis techniques. Figure 1 shows an Fe-55 single events spectrum measured with a FastCCD comparing uncorrected and corrected data, as analyzed with the XFEL.EU calibration tools. ![Figure 1. Fe-55 single events spectrum of the FastCCD comparing uncorrected and corrected data. Improvement of energy resolution after applying corrections is observable. FWHM is changing after corrections from ca. 573.1$\,$eV to ca. 422.6$\,$eV at 5.9$\,$keV. Red and green vertical lines show K$_{\alpha}$, respectively K$_{\beta}$ peak positions before (dotted) and after (dashed) corrections.] At XFEL.EU the FastCCD and ePix100a detectors are two of multiple detector designs with similar readout architectures in use, allowing us to compare analysis approaches and improve accuracy and efficiency of calibration and characterization across different charge shifting detectors technologies.

REFERENCES

[1] Altarelli, Massimo, et al. "The European x-ray free-electron laser." Technical Design Report, DESY 97 (2006): 1-26.
[2] Januschek, Friederike, et al. "Performance of the LBNL FastCCD for the European XFEL." arXiv preprint arXiv:1612.03605 (2016).
[3] Blaj, Gabriel, et al. "X-ray imaging with ePix100a: a high-Speed, high-resolution, low-noise camera." SPIE Optical Engineering+ Applications. International Society for Optics and Photonics, 2016.

(1): https://drive.google.com/open?id=12XdqAU9FfsFBj4eE-3zTUmvjpIPccSIO

Speaker: Ivana Klačková (a European X-ray Free Electron Laser GmbH, Holzkoppel 4, 22869 Schenefeld, Germany; Faculty of Electrical Engineering and Information Technology, Slovak University of Technology, Ilkovičova 3, 81219 Bratislava, Slovakia )

Lunch

Invited

92 Journeys in particle and detector physics

Speaker: Ken Smith

Oral

Convener: Valeria Rosso (INFN and University of Pisa)

93 A pixel detector readout circuit with digital charge sharing compensation hit allocation algorithm

This paper presents a design of a pixel readout circuit implementing an charge sharing compensation method based on an algorithm for asynchronous approximation of a center of gravity – Cogito [1]. In contrary to the former published methods, developed by Medipix Colaboration [2] and C8P1 algorithm by AGH-Fermilab [3], which rely on analog properties of the observed signals, the presented solution bases only on the shape of the area of an array which has detected the fractional charge. The hit allocation process is carried out entirely in the digital domain, while the analog front-end is responsible only for signal summation and amplitude discrimination.
The digital part, responsible solely for the algorithm operation, has been experimentally verified [1]. The measurement results showed that the hit allocation process can be completed in time on the order of tens of nanoseconds. This work presents the design of a pixel readout circuit implementing the Cogito algorithm, complemented by the analog front-end block. The input charge sensitive and shaping amplifier’s cores are based on a modified inverter circuit. The feedback circuits of both amplifiers have been optimized for linearity, which is required to correctly reconstruct a full charge of a charge-shared event. The design is implemented in 40 nm CMOS process. Its details, backed-up by comprehensive simulation results will be presented during the conference.

Fig. 1. Schematic of a hybrid pixel detector with the COGITO algorithm [1].

This work was supported by National Science Center, under contract no. DEC- 2014/13/B/ST7/01168.

[1] P. Otfinowski et al., “Asynchronous Approximation of a Center of Gravity for Pixel Detectors' Readout Circuits,” IEEE Journal of Solid State Circuits, in print.
[2] R. Ballabriga, et al., "The Medipix3RX: a high resolution, zero dead-time pixel detector readout chip allowing spectroscopic imaging", 2013 JINST, 8, C022016, pp 1-15.
[3] G. W. Deptuch, et al. “An Algorithm of an X-ray Hit Allocation to a Single Pixel in a Cluster and Its Test-Circuit Implementation,” IEEE Trans. Circuits Sys. I, Reg. Papers, vol. 65, no. 1, 2018, pp. 185-197.

Speaker: Dr Piotr Otfinowski (AGH University of Science and Technology)

Otfinowski_IWORID_2018.pdf

94 A cost and energy efficient data acquisition architecture for pixel detectors in mobile applications using embedded multi-core processors

Versatile pixel detector readout chips like the Timepix3[1] offer interesting features for various physics related applications. Since these fast chips record hit rates of up to 80 Mhits/s, available readout systems focus on ensuring data throughput at highest possible rates [2]. Typical data acquisition electronics for modern pixel detectors are therefore centred around field-programmable gate arrays (FPGA) consuming power in the order of 10 Watt and more.
For mobile applications like educational settings [3], low power consumption, a streamlined user experience and compact size is more important than the ability to detect large amounts of particle flux. This contribution elaborates on an alternative data acquisition architecture based on embedded multi-core processors, emphasizing size, cost and energy efficiency with a power consumption of 1-2 Watt on average. A silicon sensor bias supply design meeting these requirements including safety is featured as well.
The developed electronics and software is geared towards battery-powered use of Timepix3 in inquiry-based learning environments but can be also used in other mobile applications such as dosimetry. Based solely on embedded processors, this novel approach enables faster development cycles and an overall simplified system design compared to a traditional FPGA-based solution. Additional functionality such as on-board processing of pixel data can be handled by separate processor cores leading to a scalable solution. The presented architecture can be adapted for other pixel detectors as long their serial or parallel output lines can be configured to operate at speeds of 100 Mbit/s or less. Data output options are either wired or wireless Ethernet and local storage.

[1] T. Poikela et al (2014) Timepix3: a 65K channel hybrid pixel readout chip with simultaneous ToA/ToT and sparse readout, JINST 9 C05013
[2] J. Visser et al (2015) SPIDR: a read-out system for Medipix3 & Timepix3, JINST 10 C12028
[3] O. Keller et al (2016) iPadPix - A novel educational tool to visualise radioactivity measured by a hybrid pixel detector, JINST 11 C11032

Speaker: Oliver Keller (CERN / University of Geneva)

Oliver_Keller_efficient_multi-core_architecture_for_pixels_IWORID_2018.pdf

95 Experimental Performance of a Highly-Innovative Low-Noise Charge-Sensitive Preamplifier With Integrated Range-Booster

Integrated charge-sensitive preamplifiers suffer from a reduced available dynamic range respect to discrete-type equivalents. This is due to the limits on maximum supply voltages that modern scaled technologies can tolerate. In this work we present a low-noise low-power integrated charge-sensitive preamplifier (CSP) for solid-state detectors. This device is equipped with an integrated range-booster that can enhance the spectroscopic range of the preamplifier by more than one order of magnitude, enabling high-resolution spectroscopy even if the preamplifier is in deep saturation condition.
If the input signals from the detector are under the natural saturation threshold (40 MeV), the preamplifier works in an usual linear way, producing at the output the typical exponential signals. With proper filtering a resolution of approximately 1 keV is achievable. When a large signal from the detector saturates the preamplifier, a sensing circuit detects the saturation and switches the operation mode of the CSP to the “fast-reset mode”. In this mode a constant and controlled current generator discharges the input node of the preamplifier until the normal operating point is reached. Meanwhile an auxiliary circuit similar to a TAC (Time-to-Amplitude converter) retrieves the energy of the signal that caused saturation. Although the natural dynamic range of the CSP is 40 MeV, the fast-reset mode enables for high-resolution spectroscopy (under 0.2% FWHM) up to several hundreds of MeV (700_MeV typically).
One issue in this kind of circuits is the dependence of the energy measured with the TAC circuit on the baseline value of the CSP before the “fast-reset event” [1]. As a solution to this problem we propose a correction algorithm implemented inside the TAC block in the form of an analog circuit.
On a test-bench a series of large 3 pC charge signals is injected in the input node of the preamplifier through a test capacitor. Before these events, residual charges ranging from 0 to 0.56 pC produce non-zero baseline voltages at the output of the CSP. The TAC with correction not only retrieves correctly the energy of the main event, but also rejects the baseline voltages, leaving the energy measurement unaffected. The fluctuations of the flat-top voltage in the signals produced by the auxiliary TAC circuit due to the different baseline voltages are under the 0.6% of the total signal amplitude.

[1] A. Pullia, F. Zocca, G. Pascovici, and D. Bazzacco, “Extending the dynamic range of nuclear pulse spectrometers,” Review of Scientific Instruments, vol. 79, no. 3, p. 36105, Mar. 2008.

Speaker: Stefano Capra (University of Milan, INFN (Milan))

Stefano_Capra_27_June_iWoRiD.pdf

96 Characterization, Optimization and Implementation of a Hyperspectral X-ray Camera control and processing software chain

There is an increasing research interest in combining X-ray spectroscopy with imaging. The availability of full-field energy resolving detectors is enabling the use of true spectroscopic X-ray imaging techniques at small-scale X-ray microscopy facilities, such as the Ghent University Centre for X-ray Tomography (UGCT). At UGCT we operate the SLcam [1], a full-field hyperspectral camera prototype. This system uses a 450 µm thick silicon CCD sensor with 264 by 264 pixels measuring 48² µm². The camera operates in photon counting mode where incident X-rays generate deposited energy-dependent charge clusters spread over one or more pixels. These charges are amplified, digitized, and recombined into photon events with an associated energy and “centre-of-mass” location, resulting in a developer specified energy resolution of 152 eV or better at the manganese Kα line.

We present an optimized software chain for camera control and image processing. Given the SLcam’s original low-flux applications, working in the higher flux environments of transmission imaging poses significant challenges. To this end the processing chain takes advantage of improved calibration and characterization of the SLcam. By quantifying and analysing a large number of photon events of different energies and origins, detector performance is significantly improved for our workloads, and the energy resolution further refined. This work was performed on data from radioactive sources, X-ray tube spectra using different target materials, and data from a recent beamtime at I13-2 [2], Diamond Light Source, Didcot, UK. The latter has allowed to discriminate double-counted photons against real photons at very similar energies, as the harmonic of the used monochromator is slightly below two times the fundamental energy, as shown in Fig. 1.

Speaker: Mr Frederic Van Assche (Ghent University)

iWoRiD2018-FredericVanAssche.pdf

97 Pulse Pileup Models for Spectral X-ray Imaging

This paper reports on the development and testing of analytical models for simulating the pileup effects on photon-processing detectors. With the Medipix3RX detector and other energy-resolving photon-counting detectors, pulse pileup distorts the measurement of the photon energy. This is particularly important in photon-processing detectors that use charge summing and inter-pixel communication such as the Medipix3RX detector used in MARS scanners.

The processing time for shaping and amplifying analogue pulses within the ASIC limits the count rate capability of photon-processing detectors when performed under high photon flux rates. Pileup effects also occur in semiconductor sensors arising from the overlap of the charge clouds from different photon events. When two or more photon events occur within this processing time, they are registered as a single event with an energy different from those of original photons.

The properties of GaAs and CZT sensors bump-bonded to Medipix3RX are investigated. Measurements of the resolving time and true interaction rate are presented. Several distribution functions are used to quantify the probability of counts that are registered at different energies due to pileup effects. The predicted spectrum from our model is compared with the measurements within the energy range of 20-130 keV. There is an excellent agreement between the measurements and our model for flux rates up to ten times higher than the typical operational rate. The results show that our pileup model is capable of simulating the measurements, even when up to 37.2% of counts were lost due to pileup.

Knowing the contribution of various pileup effects to the detector measurements allows compensation for energy distortion at higher flux rates. Integrating our pileup model into the MARS material reconstruction algorithm will result in improved spectral CT image quality and classification of materials particularly when using shorter scan times.

Speaker: Dr Ali Atharifard (Mars Bioimaging Limited)

Swedish fika in the hotel lobby (2nd floor)

Poster session: part 3

Convener: David Krapohl (Mid Sweden University)

98 A Comparative Study of the TID and SEE Radiation Effects on ASICs Manufactured in 180 nm Commercial Technologies

Radiation damage is a widely studied topic for its effects on semiconductor sensors and front-end electronics used in medical and aerospace applications. Radiation tolerance and stability of the detector properties are critical issues for many applications.

Presented study compares the effects of radiation on circuit structures manufactured in a standard 180 nm CMOS [1] and, in addition, on a 180 nm SOI [2] technology. These ASICs are ongoing R&D in the field of radiation detectors and apart from transistor testing matrices for TID measurements, they contain analog and digital circuits with different transistor sizes useful for SEE measurements.

The TID study was performed using a high flux Co-60 source in charged particle equilibrium conditions, where test structures on ASICs were irradiated up to a total absorbed dose of 100 Mrad.
The test structures in both technologies consist of transistors of different width to length ratios, polarities and placement in deep insulation wells, as shown in Figure 1. The irradiation was performed in several steps and, after each step, the IV characteristics of the transistors in test structures have been measured. After irradiation, test structures were left to anneal at room temperature and their parameters have been monitored periodically. The measurement results are expressed in the form of transistor IV characteristics, threshold voltage shifts and a relative leakage current increase. In addition, the integral power consumption of the whole analog and digital part was measured and compared to the transistor level measurements.

Concerning the SEE measurements, the digital registers of the ASICs were used to study the bit flip cross sections, measured for different LET values by beams of protons, He-4 and C-12 ions at several energies delivered by a combination of cyclotron a tandetron accelerators.

References
[1] Z. Janoska et al., The PH32 Readout Integrated Circuit, in Proceedings of the 10th International Conference on Measurement, Bratislava: VEDA, 2015, pp. 207-212. ISBN 978-80-969672-9-2

[2] M. Havranek et al., 2017 MAPS sensor for radiation imaging designed in 180 nm SOI CMOS technology, proceeding from IWORID 2017, submitted to JINST

Speaker: Maria Marcisovska (Acad. of Sciences of the Czech Rep. (CZ))

99 A synchronization and data acquisition system for silicon detectors

A dedicated synchronization bus has been developed and its support has been integrated into the FITPix COMBO device. It can be used as Timepix read-out (involving back-side-pulse acquisition) or as a simple spectrometer device (when an external single pad sensor is connected, e.g. ΔE detector). The synchronization bus allows to build-up a system comprising up to 32 separated devices running in clock locked mode while the absolute value of the timestamp is distributed to all involved devices. Any combination of Timepix or spectrometer devices (up to 32, i.e. max number) can be connected-up to create a final measurement set-up. The synchronization bus was also designed regarding the controlling of a trigger signal and busy signal to allow effective filtration of unpaired events when coincidence measurement is performed. The system has been tested with a ΔE−E telescope consisting of a thin detector and Timepix. The thin detector has been used for ΔE and the Timepix detector for E measurements. The ΔE detector has an area of 10×10 mm2 and a thickness of 12μm with non-uniformity of 8%. An area and thickness of Timepix is 14*14 mm2 and 300 μm, respectively. The detection system can provide simultaneous information about position, energy, time and type of registered particles with high synchronization accuracy. Some measurements have been carried out with alpha particle sources (U-233, U-235, and Pu-239) in a vacuum and obtained results are presented.

Speakers: Gadir Ahmadov, Michael Holik (Czech Technical University (CTU))

100 Absolute measurement of Kr Electroluminescence Yield

Gaseous detectors present advantages over solid-state competitors in applications where room temperature operation, large detection areas or volumes and low-cost are important assets. On the other hand, optical gaseous detectors based on electroluminescence (EL) amplification of the primary ionisation signal are highly competitive alternatives to those based on charge avalanche amplification, presenting much better energy resolution and much larger amplitude signal output, due to the additional amplification in the photosensor used for the EL readout. The electrons produced by the radiation interaction are driven towards a region where the electric field is large enough to promote signal amplification. The applied electric field in the scintillation region is only high enough to excite but not ionise the noble gas atoms by electron impact, producing a scintillation-pulse through atom de-excitation, the so-called electroluminescence, which is proportional to the number of electrons produced in the foremost interaction. The statistical fluctuations inherent to the EL processes are much less than those associated to the avalanche ionisation processes, and even fewer than those associated to the charge produced by the radiation interaction. In addition, EL readout through a photosensor has the advantage of electrically decoupling the amplification region from the photosensor, rendering more immunity to electronic noise, radio-frequency pickup and high voltage discharges. For instance, high-pressure or dual-phase time projection chambers (TPCs) based on electroluminescence have gained increasing importance during the last decade due to their application to rare-event detection such as dark matter, double-beta decay and double-electron capture, being xenon and argon the most studied detection media.
Krypton has a radioactive isotope, which disfavours its use in rare-event applications. Nevertheless, Kr is much less expensive, is denser than Ar, and presents even the highest absorption cross section for x-rays in the 14–34 keV energy range, when compared to Xe. These are advantages when large detection volumes and/or high-pressure are requirements and in specific applications where its natural radioactive background will not seriously affect its operation due to the high intensity of incident radiation, like in some x- and gamma-ray spectrometry applications, or the possibility of efficient background discrimination in rare-event detection. Kr detectors have been already proposed for double-beta decay and double electron-capture detection.
The El yield of Xe and Ar have been determined both experimentally and by simulation while for Kr only simulation results are present in the literature. In this work, we present absolute measurements for the EL yield in Kr, using the same methodology as used for Xe and Ar. The simulation results are in agreement with the obtained experimental results.

Speaker: L.M.P. Fernandes (LIBPhys – Physics Department, University of Coimbra, Rua Larga, 3004-516 Coimbra)

101 Addressing of imperfection of hybrid pixel sensors for X-ray detection through corrections of gains and DC offsets mismatch in a circuit for charge sharing cancellation.

A developing trend in the design of event-driven sensors for X-ray detection are hybrid pixel detectors working in the single photon counting mode. The factors limiting detector performance, especially for small pixel sizes, include among others, mismatch of circuit components due to process variations, electronic noise and charge sharing. Charge sharing in a hybrid pixel detector occur when the charge generated in X-ray photon interaction with a sensor is collected by more than one pixel. Charge sharing effect may significantly impair the detector energy resolution and result in counting extra events or missing some of the events. The impact of the charge sharing increases with a decrease of the pixel sizes. The small pixel size is a desired feature of a novel hybrid X-ray detector, as it allows better spatial resolution and helps overcoming the high count rate limits by accepting more photons per unit area. Therefore, the charge sharing effect must be dealt with by a dedicated readout IC or processed off-chip.
Minimization of the analog parameters dispersion in the pixel matrix is crucial for the circuits designed for charge sharing cancellation. The DC offsets and gains spread can be optimized using correction circuits and dedicated trimming algortihms. The correction algortihms implemented in a readout integrated circuit with an inter-pixel communication for charge sharing cancellation are presented. The chip’s sensitivity to the analog parameters spread and mitigation of the performance achieved with the digital correction blocks are studied. Detection efficiency in case of charge sharing for the corrected and uncorrected pixel matrix is addressed. The simulation results as well as pencil beam measurements using synchrotron radiation are presented.

Speaker: Mrs Aleksandra Krzyżanowska (AGH University of Science and Technology)

Poster-Iworid-2018-Krzyzanowska_A0.pdf

102 Calibrating MARS Cameras using X-ray Fluorescence

ABSTRACT
We develop a XRF (x-ray fluorescence) method to characterise the energy response of a multi-element camera in a MARS spectral scanner. Due to the low emission and detection efficiency of the XRF signal it is challenging to measure a well-defined XRF peak. Nearly monochromatic XRF sources are created in the MARS small bore spectral scanner by placing metallic foils between the x-ray source and the camera. The metallic foils are permanently mounted on the filter bar (in addition to a range of filters) inside the MARS scanner, thereby providing an automatable method for generating the XRF photons.

The experimental parameters that affect the strength of the XRF signal are discussed, followed by experimental results of a geometrical setup that measures oblique (off-axis) fluorescence by minimising contamination from the primary beam. The mathematical model developed in [1] is used to obtain the optimum scanner parameters for this oblique fluorescence setup. We present results where each element in the camera is Medipix3RX bump bonded to a 2mm thick CZT or CdTe sensor.

The XRF photons can be used to characterize the energy response of each of the pixels in the camera. This work improves the quality of the 2D spectroscopic data set obtained with the MARS small bore scanners, hence makes the material analysis more accurate, and in turn makes our MARS images more medically useful.

REFERENCES
[1] L. Vanden Broeke, A. Atharifard, and others, Oblique fluorescence in a MARS scanner with a CdTe-Medipix3RX, Journal of Instrumentation, vol. 11, December 2016.

Speaker: Ms Lieza Vanden Broeke (University of Canterbury, New Zealand)

103 Characterization of noise and quantum efficiency of the Pixirad-1/Pixie-III CdTe X-ray imaging detector

Characterization of noise and quantum efficiency of the Pixirad-1/Pixie-III CdTe X-ray imaging detector

V. Di Trapani a, F. Brun b , D. Dreossi c, R. Longo d, L. Rigon d, P. Delogu a
aDipartimento di Scienze Fisiche, della Terra e dell’Ambiente, Università di Siena and INFN sez. di Pisa, Italy
bINFN, sez. di Trieste, Italy
cElettra-Sincrotrone Trieste S.C.p.A, Basovizza, Trieste, Italy
dDipartimento di Fisica, Università di Trieste and INFN, sez. di Trieste, Italy

In this work we present a study of the noise and quantum efficiency of a Pixirad detection system based on a CdTe Schottky sensor and PIXIE-III readout system [1]. In particular we compare the Signal to Noise Ratio (SNR) and the detective quantum efficiency (DQE) in different operation modes, when considering monochromatic radiation.
The Pixirad imaging system (PANalitycal) is based on a hybrid architecture were the sensor is bonded to the readout electronics. This system has square pixels with 62 μm pitch arranged in a matrix of 512×402 elements, thus covering an area of 31.7×25.0 mm2.
The sensor is made of CdTe with a thickness of 650 μm. Due to the high thickness-to-pixel pitch ratio, charge sharing between adjacent pixels is present. This effect can degrade both spatial and energy resolution [2] [3]. Another issue of this crystal is the production of fluorescence photons when high energy photons (above Cd or Te K-shells) are employed.
The readout implements, at single pixel level, two 15 bit counters fed by two independently settable energy discriminators. Moreover the chip can be configured in three different operation modes:
1. Pixel Mode (PM): each pixel counts independently from the others.
2. Neighbor Pixel Inhibit mode (NPI): only one pixel per event is allowed to count.
3. Neighbor Pixel Inhibit and Pixel Summing Mode (NPISUM): the signals of 4 neighbor pixels are summed together to correctly evaluate the total energy of any event involving up to 4 pixels; only one pixel per event is allowed to count.
For a monochromatic X-ray beam of energy E below Cd K-shell (26.7 KeV), and threshold below E/2, almost all the interacting photons are registered. In PM mode multiple counts from a single interacting photon can be registered, and the number of counts can exceed the number of interacting photons. Moreover multiple counts induce statistical correlation between adjacent pixels reducing the noise and increasing the SNR. In NPI and NPISUM modes multiple counts are avoided and the number of counts correspond to the number of interacting photons.
For threshold above E/2, in PM and NPI modes some events may be not recorded, thus decreasing the SNR and the detection efficiency. On the other hand, in NPISUM mode all the interacting photons are recorded.
For photons of energy above Cd k-shell, fluorescence photons can interact far from the point in which the primary interaction occurs. In this case, if the threshold is not set to a sufficiently high value, all the considered acquisition modes can register multiple counts due to both the primary interacting photon and the fluorescence photon.
To quantify this behavior, we used monochromatic synchrotron radiation of two different energies (26 KeV and 33 KeV), below and above the K-shells of the sensor material.
In particular we acquired sets of images of the beam at SYRMEP beamline (Elettra, Trieste) in all the different operating modes and varying the discriminator threshold. On these images we measured the SNR, compared it to the pure Poisson case and finally calculated the DQE. Results are in good agreement with the expected behavior and a constant SNR is observed in NPISUM without dependence by the threshold values. This feature candidates the NPISUM mode as the best choice in spectral imaging with polychromatic beam.

References
[1] R. Bellazzini et al., 'Pixie-III: a very large area photon-counting CMOS pixel ASIC for sharp X-ray spectral imaging', JINST 10 C01032 (2015)

[2] P. Delogu et al., 'Characterization of Pixirad-1 photon counting detector for X-ray imaging' JINST 11 P01015 (2016 )

[3] A. Vincenzi, P.L. de Ruvo, P. Delogu et al., ' Energy characterization of Pixirad-1 photon counting detector system', JINST 10 C04010 (2015)

Speaker: Vittorio Di Trapani (Dipartimento di Scienze Fisiche, della Terra e dell’Ambiente, Università di Siena and INFN sez. di Pisa, Italy)

104 Characterization of silicon sensors for the soft X-ray energy range

The MÖNCH [1] detector is a low-noise charge-integrating hybrid pixel detector with 25 μm pixel pitch currently being developed at the Paul Scherrer Institut (PSI) optimized to be used in soft X-ray applications at synchrotrons and X-ray free electron lasers (XFELs). Suitable instruments for experiments in the soft X-ray energy range (250–2000 eV) present several technological challenges. Detectors should provide high Detective Quantum Efficiency (DQE), high Signal-to-Noise Ratio (SNR), high resolution, large dynamic range and need to handle frame rates higher than 1kHz. Additionally, they need to be operational under high vacuum conditions.

MÖNCH 03 [2], a prototype readout chip of the detector, meets most of the requirements to be used in soft X-ray experiments due to its remarkable low noise performance of 36 e- ENC and single photon resolution. But some optimizations are still needed such as reducing the thickness of the sensor backside implant for an improved quantum efficiency below 4 keV. Therefore, the aim of this work is to optimize the DQE of the MÖNCH detector with a thin entrance window to increase the detection efficiency in the soft X-ray regime.

Four silicon sensors with different entrance windows were produced at FBK and bump-bonded to the MÖNCH readout chip. The four prototypes of 160x160 pixels with 25 μm pitch (4 x 4 mm$^{2}$ of active area) were investigated using X-ray photons with energies between 400-2000 eV at the SIM beamline of the Swiss Light Source at PSI. The quantum efficiency of the four sensors has been compared and the thickness of the entrance window has been estimated. Additionally, the noise of the detection system has been measured as a function of the integration time and of the applied bias voltage. In this contribution, the first results of the DQE measured with the four sensors will be presented and discussed.

[1] R. Dinapoli et al, J. Instrum. 9 (2014) p. C050115.
[2] M. Ramilli et al, J. Instrum. 12 (2017) p. C01071.

Speaker: Sabina Chiriotti Alvarez (PSI - Paul Scherrer Institut)

105 Design and Characterization of a Preamplifier for Photodiodes With Separate Gain Channels for High-Frequency and DC Signal Components

In this work a preamplifier for photodiodes is presented, with a very peculiar feature: the circuit provides two physically separated gain channels for the DC and the high-frequency components of the spectrum. This is achieved with an active transconductor as feedback network of the first gain stage. Even if this one is DC coupled to the photodiode, the low-frequency component of the signal is collected from the active transconductor and separately amplified. In this way the preamplifier behaves like if it was AC coupled to the photodiode. This enables to adopt different gains for the DC and the high-frequency channels.
This circuit was originally designed to be inserted inside a Pound-Drever-Hall (PDH) stabilization loop for a laser optical cavity. The mechanical vibrations of an optical cavity (from DC to some KHz) continuously bring it out of tune respect to the laser frequency. For this reason, an active loop is required to correct the position of the mirrors in order to keep the cavity in tune. In order to do so an “error function” is required. The reflectivity of a cavity is not a good parameter to be used as error function since it is “even” around the resonance. The PDH technique is used to produce an error signal which is an “odd” function around the resonance frequency. An electro-optic modulator is used to produce two sidebands in the laser’s spectrum with 10-100 MHz frequency difference with the carrier. Due to these sidebands the reflected beam from the cavity shows an amplitude modulation that is an “odd” error function respect to the difference between the laser’s frequency and the resonance of the cavity. This signal can be de-modulated with a signal mixer, sent to a PID (Proportional, Integrative, Derivative) module and used to make the cavity track the laser’s frequency.
The circuit was designed for a modulation frequency between 80 MHz and 130 MHz. The high-frequency power fluctuations of the reflected beam are generally two order of magnitude lower than the average power. Thus if the photodiode DC current in typical experimental conditions is around 200 μA, the 100MHz signal is only 1-2 μA. The problem of a small interesting signal surrounded by a huge low-frequency background is a common problem in very different applications involving photosensors [1,2]. The possibility to have two different gains enables to amplify the 100 MHz signal without saturations due to an extremely high DC voltage level.
The circuit characterization was performed with an Agilent 4395A network-spectrum-impedance analyzer, a Tektronix AFG3252 function generator and an Agilent 54832D oscilloscope. The circuit is based on a three-stage gain circuitry involving only discrete components to ensure high gain and wide bandwidth. The circuit was designed with wide-availability components. The chosen gain transistors are BFR82 while for active loads and buffers MMBTH81 and MMBTH10 were chosen. The gain of the circuit at 100 MHz is 3.79 mV/uA with 15 pF of photodiode capacitance. The test-bench characterization was performed injecting an AC current to the input node through a test capacitor of 1 pF. The circuit shows significant improvements respect to previous adopted solutions both in terms of S/N ratio and absolute signal amplitude.

[1] A. Pullia, T. Sanvito, A. Potenza e F. Zocca. «A low-noise large dynamic-range readout suitable for laser spectroscopy with photodiodes». In: Review of Scientific Instruments 83.104704 (2012).
[2] N. A. Lockerbie e K. V. Tokmakov. «A low-noise transimpedance amplifier for the detection of "Violin-Mode" resonances in advanced Laser Interferometer Gravitational wave Observatory suspensions». In: Review of Scientific Instruments 85.104705 (2014).

Speaker: Stefano Capra (University of Milan, INFN (Milan))

106 Gain and Noise in GaAs/AlGaAs Avalanche Photodiodes with Thin Multiplication Regions

In this presentation avalanche photo diodes based on GaAs/AlGaAs and separated absorption and multiplication regions (SAM – APD) will be discussed. The structures have been fabricated by molecular beam epitaxy utilizing a δ p-doped layer of carbon to separate the multiplication region and the absorption region. The layer thickness of the latter defines the quantum efficiency and subsequently the time resolution of the structure, which in turn allows tailoring the device for specific scientific applications. Within the multiplication region a periodic modulation of the band gap is obtained by growing alternating nanometric layers of AlGaAs and GaAs with increasing Al content, which enables to tune the band gap and subsequently provides a well-defined charge multiplication. The use of such staircase hetero-junctions enhances electron multiplication and conversely reduces – at least in principle - the impact of the noise associated to the hole amplification, which should result in a decreased overall noise, when compared to p-i-n diodes. Since the doping density of the δ layer controls the punch-through of photo induced primary charges from the absorption to the multiplication layer, several devices with different carbon concentrations have been fabricated and have been characterized. In this presentation gain and noise measurements, which have been carried out on these devices utilizing photons from visible light to hard X-ray, will be discussed and will be compared to the results of a nonlocal history-dependent model specifically developed for this kind of staircase APDs.

Speaker: Ralf Hendrik Menk (Elettra Sincrotrone Trieste)

107 Gas gain and secondary processes in Ne – N2 mixtures – concentration and pressure scaling.

In the present work, the gas gain curves have been measured in the range from the ionization chamber regime to the breakdown limit in Ne – N2 gas compositions (2 – 20% N2, also in pure Ne) at various mixtures pressures (50 – 1800 hPa).
The measured gas gain curves have been fitted to the Diethorn, Williams & Sara and of Aoyama gain models to determine the characteristic mixture constants like effective ionization potential or mean ionization free path. The possible relation between these parameters and the mixture pressure and N2 – concentration will also be discussed. The secondary processes related to electron avalanche development have been also investigated. The measured values of gas gains will be compared with those calculated from the MAGBOLTZ software.

Speaker: Dr Tadeusz Kowalski (AGH University of Science and Technology, Fac. of Physics and Applied Computer Science)

108 Implementation of large area CdTe Hybrid Pixel detector for purpose of X-ray spectroscopic imaging.

S.Hasna,b , M. Pichotkaa,c, D. Vavrika,c
a Institute of Experimental and Applied Physics, CTU in Prague, Horska 3a/22, 128 00 Prague 2, Czech Republic. b Faculty of Electrical Engineering, Czech Technical University in Prague, Technicka 2, 166 27 Prague, Czech Republic.
c Institute of Theoretical and Applied Mechanics AS CS, v. v. i., Prosecka 76, 190 00 Prague 9, Czech Republic.
E-mails: salman.hasn@cvut.cz, vavrik@itam.cas.cz, Martin.Pichotka@utef.cvut.cz

ABSTRACT

In this contribution the performance of a CdTe direct converting photon counting detector in spectroscopic imaging is investigated. The tiled detector, which is based on the Timepix readout ASIC, features a frame-rate of 40fps and edgeless sensors.
In comparison to Si, the main advantage of using high-efficiency sensor materials like CdTe is the significantly higher stopping power for high energy photons (figure 1).
This kind of large area detectors represents a suitable tool for the purpose of X-ray spectroscopic imaging of large samples, employing single event analysis and THL(energy threshold ) scanning respectively.
The measurements based on this detector will carry additional spectroscopic information about the materials within the sample, which will help us for the separation process of the materials based on their attenuations characteristics.

Figure. 1. X_ray attenuation in Sillicon and CdTe sensors

Speaker: Mr Salman Hasn

109 Multilayer Compton Camera based on Timepix3 Technology

The Compton camera concept is based on reconstruction of recorded Compton scattering events of incoming gamma rays. The scattering of primary gamma ray occurs in the first detector (called scattering detector – usually thin) recording position and energy of recoiled electron. The scattered gamma quantum continues towards the second detector (called absorption detector - usually thick) where it is absorbed. The second detector records the energy and position of this scattered gamma. Using Compton scattering equation it is possible to determine the scattering angle and estimate possible directions of the original gamma ray as a surface of the cone. When the Compton camera records number of such events the location and shape of the gamma source can be reconstructed.

Timepix3, a hybrid single photon counting pixel detector, is perfect device for creation of a compact Compton camera. Timepix3 is an event based readout chip (every hit pixel is immediately sent to a readout) and can record time-of-arrival (ToA) and energy of incident gamma simultaneously in each pixel. The chip offers high energy resolution (1 keV at 60 keV, 7 keV at 356 keV) as well as time resolution (1.6ns). The Timepix3 readout chip can be combined with different sensor materials (Si, CdTe, CZT).

In this contribution we present a very compact detector system for imaging with gamma-rays using Compton camera principle. The system consists of at least two layers of hybrid pixel detectors Timepix3 with sensors optimized for gamma-ray tracking. The front detector layers (scattering) are made of silicon of various thickness (up to 1 mm) while the last layers (absorbing) are equipped with thick CdTe or CZT sensors up to 2 mm in thickness. The total absorption of the whole detector can be very high if several CdTe or CZT layers are used. The maximal number of layers is not limited but the practical evaluation was performed with 2 layers. Thanks to Timepix3 simultaneous measurement of ToA and energy it is possible to precisely detect coincidence events in the detector layers. Based on the energy and position of these events it is possible to estimate the possible direction of the original gamma. The angular resolution of the presented Compton camera depends on the detected energy and reaches units of degrees.

Speaker: Mr Daniel Turecek (ADVACAM s.r.o.)

110 Photodetector of multilayer exfoliated MoS2 deposited on polyimide films

We fabricated a photodetector based on multilayer molybdenum disulfide ($MoS_2$) by micromechanical cleavage of a molybdenite crystal using a polyimide film. We deposited 40nm of gold by vacuum sputtering and copper tape was used for the contacts. Without any surface treatment, we achieved high responsivity at different incident optical power. The calculated responsivity are 2$3\mu AW^{-1}$ of incident optical power in the range between 400 and 800nm. For the responsivity measurement it was estimated that $MoS_2$ have a bandgap of 1.6eV, which lies between monolayer and multilayer films. The thickness of the $MoS_2$ thin film was determined by Raman spectroscopy evaluating the difference between the in plane $E^1_2g$ and out of plane $A_1g$ Raman modes. The measurement of IV curves indicated Ohmic contacts in respect to the Au regardless of the incident optical power. Our device fabrication was much simpler than previous reported devices and can be used to test the light absorption and luminescence capabilities of exfoliated $MoS_2$.

Speaker: Viviane Forsberg (Mid Sweden University)

111 Prototype of detector of reactor antineutrinos for reactor monitoring

In recent years, much attention has been focused on neutrino research because it can shed new light on greatest mysteries in physics. A new experiment devoted do detection and investigation of reactor antineutrinos is being performed at the IEAP CTU Prague and JINR Dubna using highly segmented scintillating detector $S^3$. This paper describes present status of polystyrene based setup (detector part, front-end electronics, data acquisition system) which does not contain any dangerous or flammable materials and is absolutely safe to be place in the close vicinity from the reactor. Close vicinity from the reactor core enables study of neutrino properties with higher efficiency, for example investigation of short-range neutrino oscillations and verification of sterile neutrino hypothesis because of its short oscillation length. if it is possible to measure the antineutrino energy spectrum, the operational status, thermal power of the reactor and isotopic composition of the reactor fuel can be also determined. As a result, it will be possible to prevent illegal production and extraction of $^{239}Pu$, which is an essential part of nuclear weapons. Therefore, this research has applications in physics beyond the Standard Model as well as find practical applications in reactor physics.

Speaker: Maria Savorova (Institute of Experimental and Applied Physics CTU in Prague, Horska 3A/22 Czech Republic)

112 Rospix: modular software tool for automated data acquisitions of Timepix detectors on Robot Operating System

We present a new software package for interfacing FitPix and USB Lite compatible Timepix readout electronics using the Robot Operating System (ROS) [1]. ROS is a widely adopted middleware for integration of sensors, processing algorithms and logic to autonomous systems such as robots, unmanned helicopters, and robotic payloads. Thanks to ROS, Timepix detectors can be set for automated experiments on platforms spanning from traditional desktop computers to small ARM devices such as Raspberry Pi and Odroid. Acquisition and detector settings can be controlled in Linux shell allowing native deployment on screen-less devices. Using ROS networking capabilities, the measured data such as captured frames and detector controls can be transmitted via a network, which allows building distributed systems. The proposed software is a lightweight package, easily connectable to existing visualization, logging, and processing software in ROS. It offers simple bindings to custom Python and C++ programs for real-time control of the acquisition or processing of the captured frames. Rospix was deployed on a Timepix equipped payload on board a NASA suborbital rocket, which was successfully launched on April 4, 2018, from Kwajalein Atoll by Pennsylvania State University. The REX (Rocket Experiment) utilized Odroix-XU4 computers recording data from two Timepix detectors, which served as focal plane imagers for an onboard 2-dimensional X-ray telescope for the observation in open space of stellar X-Ray objects such as the Vela supernova remnant.

We release Rospix to the community using the GitHub platform [2] and plan to provide ROS modules for cluster analysis and real-time feedback-based acquisition control. Research carried out in frame of the Medipix collaboration through the Institute of Experimental and Applied Physics, CTU in Prague.

[1] M. Quigley, K. Conley, B. Gerkey, J. Faust, T. Foote, J. Leibs, R. Wheeler, AY. Ng, "ROS: an open-source Robot Operating System", ICRA workshop on open source software 2009 May 12 (Vol. 3, No. 3.2, p. 5).
[2] T. Baca, "Rospix on GitHub", accessed 09/04/20018, URL: https://github.com/klaxalk/rospix

Speaker: Tomas Baca (Czech Technical University in Prague)

113 Simulation of New Scintillation Crystals for Brain PET

ABSTRACT

In the last decade, different Positron Emission Tomography (PET) crystals have been proposed for brain PET detectors [1]. Brain PET cameras with restricted field of view (FOV) to a smaller size to cover the brain can both exhibit much higher performance and lower the cost when comparing to conventional whole-body PET scanners. All the present designs based on scintillating crystals employ most commonly known materials such as Sodium Iodide (NaI), Lutetium Yttrium Orthosilicate (LYSO), Bismuth Germanate Oxide (BGO) etc. This is due to their relatively good parameters such as light yield, density, and especially low cost comparing to solid-state crystals. The only drawback of using scintillation crystals is having a modest energy resolution. Besides, crystal based detectors have intrinsic uncertainty because of the depth of interaction (DOI) which is not accurate due to thick scintillation materials. This uncertainty causes parallax error which is a major problem in PET.
In this study, both conventional crystals used in commercialized scanners and the recently proposed scintillation crystals with promising physical properties were analyzed. In total, 13 crystals were simulated using Geant4 Application for Tomographic Emission [2] (GATE) toolkit and the performance in terms of sensitivity of each of the crystals was evaluated according to NEMA standards [3]. New scintillation crystals were added to the GATE material database.
The performance evaluation results showed that Hafnium dioxide (HfO2) is a very promising material with an outstanding sensitivity when compared to the remaining crystals and followed by BGO which is the most common type of crystal in use. Lutetium Fine Silicate (LFS), a newly proposed crystal also shows the potentials of being a promising detector material for PET imaging, while Lutetium Aluminum Perovskite (LuAP), Lutetium Oxyortho-Silicate (LSO) and Lutetium Gadolinium Oxyortho-Silicate (LGSO) crystals show high sensitivity which categorizes them among the mostly used materials for medical imaging. Those crystals whose sensitivities are low, Cesium Hafnium Chloride (CHC), Strontium Iodide (SrI2), and Cs2LiLaBr6:Ce (CLLB) can be considered for either high spatial resolution or very fast cameras.

Keywords: PET, scintillation crystals, GATE simulation, sensitivity, HfO2, LFS

REFERENCES
1. Paul Lecoq, "Development of new scintillators for medical applications", Nuclear Instruments and Methods in Physics Research Section A: Accelerators, Spectrometers, Detectors and Associated Equipment, 809, 130-139, 2016.
2. Jan, S. et al., "GATE: A simulation toolkit for PET and SPECT", Physics in Medicine and Biology, 49, 4543-4561, 2004.
3. NEMA NU-4 2008 Standards Publication, "Performance Measurements of Small Animal Positron Emission Tomographs", Rosslyn, VA: National Electrical Manufacturers Association, 2008.

Speaker: Mr Musa Sani Musa (NEAR EAST UNIVERSITY)

114 Study of electromigration phenomena in Au/P-type CdTe of two Schottky contacts

In spite of the importance of understanding and controlling the electromigration (drift and diffusion of the intrinsic ion defects in external bias) of defects in CdTe based optoelectronic devices, there is, according to our knowledge, no systematic study about the phenomena in the literatures. The reasons for that are the complexity to prepare samples with the same distributions of defects and the interpretation of measured data.
According to our model [1] the electric field confined in the depletion region below the Schottky M-S interface causes electromigration of donor defects in the semiconductor bulk, which leads to time change in the resistance of the depletion region values. The immigrated ions showed no tendency to revert to their initial distribution even after longtime, which is necessary to obtain comparative measuring data at different T. Such limitation of measurements results in a lack of understanding of the electromigration phenomena, and limitation of extracted parameters.
We present in this work transient measurements of Au/CdTe/Au structure in the temperature range of 323K-363K. Removal of anode contact and of a thin layer below was necessary after each measurement to have similar distribution of the donor ions in depletion region for each new measurement. The diffusion coefficients at different T have been determined and the activation energy of the electromigration process was extracted - Fig 1. Current – Voltage measurements have been carried out to determine the reverse current transport mechanisms, and the concentration of the acceptors ions was determined from the resistivity measurements.

Fig. 1. Arrhenius plot of the diffusion coefficients of the donor ions based on values extracted from our experimental data.
[1] H. Elhadidy, R. Grill, J.Franc, O.Šik, P.Moravec, O.Schneeweiss, Solid State Ionics 278 (2015) 20–25.

Speaker: Dr Hassan Elhadidy (, Institute of Physics of Materials, Academy of Sciences of the Czech Republic)

115 Study of low power front-ends for hybrid pixel detectors

The classical pulse processing chain for a radiation detector consists of a charge sensitive amplifier followed by a shaping filter that optimizes the signal to noise ratio. Hybrid pixel detector readout circuits usually omit the band-pass filter stage because of area and power consumption constraints. Moreover, when digitizing the energy information using the Time-Over-Threshold principle, a charge sensitive amplifier whose feedback capacitor is reset at constant current provides superior linearity and dynamic range than the circuit including the shaper.
In this paper we analyze the behavior of a classical charge sensitive amplifier with constant current discharge of the feedback capacitor with the input transistor current from circuit basic principles and compare the results with measured results in the Timepix3 chip. The expression for the Equivalent Noise Charge due to the preamplifier input transistor thermal noise is:
$$ ENC^2_s = \frac{K_BTN_{\gamma}\frac{C_F(C_F+C_I)^2}{(C_FC_I+C_FC_O+C_IC_O)}}{q^2\alpha^2}$$ Where $K_B$ is the Boltzmann constant, T is the temperature in degrees Kelvin, N is the excess noise factor, $\gamma$ is a parameter measuring the degree of inversion of the transistor, $C_I$, $C_F$ and $C_O$ are the input, feedback and output capacitances respectively q is the electron charge and $$ \alpha = ma x( e^{\frac{1}{\tau_{fall}}} -e^{\frac{1}{\tau_{rise}}} ), t>0$$ $\alpha$ is a parameter accounting for the separation of the time constants defining the preamplifier output rise and fall times. An important conclusion is that, provided the rise time and the fall time constants are far apart, the noise due to the input transistor is independent of its transconductance and as a consequence, independent of the input transistor power consumption. The reason is that when increasing the current in the input transistor, its noise decreases but the bandwidth over which the noise is integrated increases by the same amount, the two effects cancelling each other. If the device operates in a low count rate environment substantial reductions in power consumption can be obtained with little or no noise penalty. The measurements below show 241Am source spectra measured with Timepix3 [1] at preamplifier bias currents of $2.4\mu A$ (left) and 150nA (right). In the paper we will report on full circuit simulations helping to identify optimized power consumption depending on the count rate environment including any trade-offs between power consumption, maximum count rate, noise and minimum operating threshold.

[1] T. Poikela et al. 2014 JINST 9 C05013

Speaker: Xavi Llopart Cudie (CERN)

116 Study of power consumption of Timepix3 detector

Timepix3 readout chip – the latest member of Medipix family of a hybrid detectors – brought several great functionalities in comparison with older Timepix, i.e. a high hit-rate, a time resolution of 1.56ns, event data-driven readout mode, and mainly the capability to measure the Time-over-Threshold (ToT) and the Time-of-Arrival (ToA) values at the same time.

However, there are postulates that Timepix3 is not suitable in low power applications and in applications where heating and cooling are crucial. As it has been hinted in [1], these expectations may not be valid. Authors of this contribution deal with the dependency of various settings of an analogue and a digital part of the detector on a power consumption. For this measurement, Timepix3 chipboard and firmware in Katherine readout were modified so that user can measure power consumption of an analogue and a digital part “on-line” and directly in a control software.

In the contribution it is shown that a standard power consumption is approximately 1.5W. However, just by a simple change of several internal DACs, the consumption can be reduced by a half. Other reductions can be achieved by the change of clock management of a digital part of the Timepix3. Mainly a decrease of frequency of a matrix clock can yield significant reduction of power consumption. This fact is important not only from viewpoint of a power consumption, but it means also benefit, for example, for measurement in a vacuum chamber, where heating of the detector is relevant issue. The reduction of the power consumption and heating of detector could open possibilities of usage of Timepix3 in space applications.

References
[1] Llopart, X.: Timepix3 in low power mode. Contribution in Medipix meeting 3/2018. CERN

Speaker: Pavel Broulim (University of West Bohemia (CZ))

117 The SpacePix-D Radiation Monitor Technology demonstrator

We present a newly developed radiation detector based on ASIC manufactured in a 180nm SoI technology [1]. It serves as a capability demonstration device for the SpacePix ASIC suite, currently under development. The active part of the presented demonstrator SpacePix-D is a monolithic radiation detection ASIC X-Chip03 [2] with an active area of 3.8 ´ 3.8 mm2 made of an array of 64 ´ 64 pixels, each with the pulse height measurement capability with a fast 10-bit ADC. The power consumption of the ASIC is 20 mW and it has O(100) Hz imaging capability.
The SpacePix ASICs are novel stand-alone monolithic semiconductor pixel detectors with a non-linear amplifier, allowing them to detect and measure energy deposition of a wide range of ionizing particles. The SoI technology ensures a high degree of SEU and TID tolerance, which makes them attractive for use in demanding environments such as orbital or interplanetary radiation field.
The current SpacePix-D demonstrator is equipped with accumulator, LCD for visualization of hits and cluster analysis as well as with temperature sensor, accelerometer and wireless data connection. With its total power consumption of 1.2 W, the device can be operated for up to 8 hours without the need of an external power supply, thus making it practical for simple measurements and it can be even used as a teaching aid.
One of its main features is easy operation and real-time imaging of ionizing radiation. For a more advanced usage, it can be connected via either USB or Bluetooth to the PC. The operating software, ASPIRE, enables visualization and acquisition of data which can be later processed.
The device has already been successfully tested with various sources of ionizing radiation – ranging from different table-top nuclide sources (55Fe, 241Am) to high energy protons.

[1] M. Havranek et al., MAPS sensor for radiation imaging designed in 180 nm SOI CMOS technology, proceeding from IWORID 2017, submitted to JINST
[2] M. Havranek et al., X-CHIP-03 – SOI MAPS sensor with hit counting and ADC mode, manuscript in preparation

Speaker: V. Vrba (FNSPE CTU, Brehova 7, Prague, Czech. Rep.)

118 TOF studies for dedicated PET with open geometries

Currently, there are three trends in the design of human Positron Emission Tomograph (PET) scanners. The conventional ones are based on detector rings of about 70 cm transaxial Field Of View (FOV) and about 20 cm axial FOV. The other two approaches in the research arena are the newest total body PET that aims at completely exploring the subject with 200 cm axial length, or the use of less-bulky and cost-effective, organ-dedicated systems. Cylindrical PET geometries in the XY plane favor the system design, providing an optimal performance as in the case of sensitivity. However, most suitable geometries for dedicated PET scanners exhibited a limited angular tomography that varies depending on the organ and application. That supposes a handicap, since the sensitivity is reduced and the lost of information in the opening axis causes serious artifacts. One way to overcome these artifacts is to use the annihilation photons Time Of Flight (TOF) information in the image reconstruction process. This allows reducing the uncertainty in the emission probability of the voxels in the corresponding Lines Of Responses (LORs).
Our group has a large experience working in the development of dedicated PET systems. Recently, two novel PET scanners have been developed with open geometries: breast- and prostate-dedicated (see figure 1). The breast-dedicated system comprises two detector rings of twelve modules with a field of view of 170mm x 170mm x 94mm [1]. Each module consists of a continuous trapezoidal LYSO crystal and a PSPMT. The system has the capability to vary the rings opening to 60 mm in order to allow the insertion of a needle to perform a biopsy procedure. The prostate system has an open geometry consisting on two parallel plates separated 28 cm. One panel includes 18 detectors organized in 6x3 matrix while the second one comprises 6 detectors organized in a 3x2 matrix. All the detectors consist in a continuous LYSO crystal of 50mm x 50mm x15mm and a SiPM array of 12x12 individual photo-detectors. The system geometry is asymmetric maximizing the sensitivity of the system at the prostate location, located at about 2/3 from the abdomen-anus distance.
In this work we present a TOF resolution study for open geometries. With this aim, our dedicated systems have been simulated using different TOF resolution in order to determine the image quality improvements obtained with the existing TOF electronics in the market. The images have been reconstructed using a modification of the TOR projector with LMOS algorithm [2] which includes the TOF modeling in the calculation of voxel-LOR emission probabilities [3].

Figure 1. Open geometries scanners developed.
References:
[1] L. Moliner et al. “Performance characteristics of the MAMMOCARE PET system based on NEMA standard”, JINST Vol 12, 2017
[2] L. Moliner, at al. “Implementation and analysis of list mode algorithm using tubes of response on a dedicated brain and breast PET” NIMA, Vol. 702, 129-132 (2013)
[3] Spanoudaki VC, Levin⋆ CS. Photo-Detectors for Time of Flight Positron Emission Tomography (ToF-PET). Sensors (Basel, Switzerland). 2010;10(11):10484-10505. doi:10.3390/s101110484.

Speaker: Dr Laura Moliner (Institute of Instrumentation for Molecular Imaging (I3M))

119 Using Timepix pixel detectors in educational physics experiments about natural radioactivity and cosmic rays

Within the education context, radiation imaging devices like pixel detectors based on the Timepix chip offer advantages compared to traditional detectors for ionising radiation [1]. In contrast to simple Geiger counters, the energy of single and low-energetic particles or photons can be revealed. Different types of ionising radiation are distinguished based on the shapes of recorded pixel clusters similar to traces in a cloud chamber. Pixel detectors are sensitive enough to measure low-intensity environmental radiation whereby the usage of problematic radioactive sources can be avoided. This is an important benefit for learning places.
We show examples of exploring natural radioactivity as well as cosmic particles in educational physics experiments using Timepix. With the help of static electricity and filter paper, radon progeny is collected from drinking water, air and cigarettes providing harmless samples of alpha radiation [2]. These sources of heavy ionising particles are contrasted with the measurement of cosmic rays. After appropriate filtering and analysis of pixel clusters, general properties of minimum ionising particles from atmospheric muons are revealed.

[1] T. Whyntie et al (2015) CERN@school: demonstrating physics with the Timepix detector, Contemporary Physics, 56:4, 451-467
[2] M. Pohl, H. von Philipsborn (1996) Recent Progress in Sampling and Measurement of Radon and Thoron decay products, proceedings of IRPA9, Volume 2

Speaker: Oliver Keller (CERN / University of Geneva)

120 X-ray Transmission Imaging Performance with a Sealed MPGD

A sealed MicroPatterned Gaseous Detector (MPGD) developed with low outgassing materials and with a gas purification system was applied for X-ray transmission imaging. The purification system is based on getters that remove the gas impurities and keep the detector efficiency stable along time.
The filling gas (pure Kr) allows for high detection efficiency [1], expected good spatial resolution [2] and high gains [3] for 1 – 30 keV, the photon energy range suitable for breast and small animal imaging.
In order to characterize the detector performance for spectral X-ray imaging, studies of energy resolution, uniformity, signal-to-noise ratio and spatial resolution are being performed. Images to evaluate the detector performance for X-ray imaging were corrected with a flat field acquisition in an attempt to uniform the image intensity and performance in the full detector active area. First results show an energy resolution of about 17% and a spatial resolution
below 500 µm. Results of uniformity, signal-to-noise ratio and spatial resolution as a function of photon energy will be presented.

REFERENCES
[1] M. J. Berger, J. H. Hubbell, S. M. Seltzer, J. Chang, J. S. Coursey, R. Sukumar, D. S. Zucker and K. Olsen, XCOM: Photon Cross Sections Database, (2011)http://www.nist.gov/pml/data/xcom/index.cfm.
[2] C. D. R. Azevedo, S. Biagi, R. Veenhof, P. M. Correia, A. L. M. Silva, L. F. N. D. Carramate and J. F. C. A. Veloso, Position resolution limits in pure noble gaseous detectors for X-ray energies from 1 to 60 keV, Phys. Lett. B 741, (2015) 272.
[3] L. F. N. D. Carramate, A. L. M Silva, C. D. R. Azevedo, I. Fortes, S. G. Monteiro, S. Sousa, F. M. Ribeiro, S. De Francesco, D.S. Covita and J. F. C. A. Veloso, THCOBRA X-ray imaging detector operating in pure Kr, submitted to JINST; Preprint JINST_061P_0317.

Speakers: Lara Filipa Das Neves Dias Carramate (University of Aveiro (PT)), Joaquim Marques Ferreira dos Santos (University of Coimbra)

Social event: Transport to Hotell Södra Berget - Conference dinner

Social event: Transport to Hotell Södra Berget - Conference dinner

Social event: Conference dinner

Social event: Transport back to Quality Hotel

Social event: Transport back to Quality Hotel

Thursday, 28 June

Invited

121 A look back on development of Medipix/Timepix detectors for noiseless tracking of neutrons and charged particles

Speaker: Stanislav Pospisil (Czech Technical University (CZ))

Oral

Convener: Joaquim Marques Ferreira dos Santos (University of Coimbra)

122 Radiation image display on a simple virtual reality system

Please see the attached file in the abstract.

Speaker: Dr Yuki Sato (Japan Atomic Energy Agency)

123 First Use of Timepix3 Hybrid Pixel Detectors in Ultra-High Vacuum for Beam Profile Measurements

The Timepix3 hybrid pixel detector’s precise time resolution of 1.5625 ns is of particular interest for new beam instrumentation applications since it enables time-resolved bunch-by-bunch measurement of the beam [1]. A transverse beam profile monitor is currently under development for the CERN Proton Synchrotron (CPS) to provide non-destructive continuous measurements during a beam cycle. As protons travel through the vacuum of the accelerator residual gas is ionized; the liberated ionization electrons are accelerated by a strong electric field towards the Timepix3 hybrid pixel detectors where they are detected. The beam profile is inferred from the distribution of the detected ionization electrons.

Figure 1. Top: 3D Rendering of the beam profile instrument. Bottom: Example of data recorded from the Timepix3 hybrid pixel detectors showing the ionisation electrons from the beam.

In early 2017 a prototype instrument was installed and used successfully during the year to measure the time evolution of the transverse beam profile. This demonstrated for the first time at CERN the potential of using hybrid pixel detectors directly inside the accelerator beam pipe for a beam instrumentation application. Due to the highly radioactive environment inside the CPS a radiation hardened readout system has been designed to connect to the pixel detectors inside the vacuum. The electrical data signals from the Timepix3 are brought to an FPGA where packets of data are routed onto optical fibers using components from the GBT and Versatile Link projects. A first version of the readout system has successfully been used during 2017 and a new system is currently being developed to enable the full data rate of the Timepix3 detectors.

An overview of the instrument will be presented with particular emphasis on the radiation hardened readout electronics for the Timepix3 detectors. Measurements from the 2017 run demonstrating the potential of hybrid pixel detectors for beam instrumentation will be shown.

REFERENCES
[1] POIKELA, T., et al. Timepix3: a 65K channel hybrid pixel readout chip with simultaneous ToA/ToT and sparse readout. Journal of instrumentation, 2014, 9.05: C05013.B.

Speaker: Hampus Sandberg (University of Manchester (GB))

2018-06-28-IWoRiD-First Use of Timepix3 Hybrid Pixel Detectors in Ultra-High Vacuum for Beam Profile Measurements.pdf

figure.png

124 Calibration and first user experience with the Large Pixel Detector at the European XFEL

The Large Pixel Detector (LPD) was in development since 2009 at the Rutherford Appleton Laboratory, Science and Technology Facilities Council in Oxford, UK, for the European XFEL. While smaller prototype systems existed much earlier[1], the first Megapixel LPD detector was delivered in early 2017 to the European XFEL and successfully went into operation at the Femtosecond X-ray Experiments (FXE) instrument in late summer that year. The detector enables time-resolved studies of structural dynamics in solids, liquids and gases by means of X-ray scattering on ultrafast timescales down to 100 fs with 4.5 MHz pulse repetition rate.
LPD is a hybrid pixel detector based on a Silicon sensor with 500$\mu$m $\times$ 500$\mu$m sized pixels with 512 analogue memory cells each, allowing the detector to take more than 5000 frames per second at the 4.5 MHz repetition rate of the European XFEL. Its high dynamic range of 105 12 keV photons is realized by three parallel gain stages and two preamplifier configurations. Simultaneously, it allows single photon sensitivity down to energies of about 14 keV, while the 500 $\mu$m-thick sensor provides good detection efficiency also for higher energy photons up to 20 keV.
Characterization and calibration of LPD is challenging due to the high number of pixels and memory cells in combination with the very short integration time of about 100 ns, yielding billions of calibration parameters, totalling tens of gigabytes of data. The calibration strategy and results ranging from making use of calibration sources implemented in the front end ASIC and external sources like continuous X-ray tube irradiation to the final implementation currently in use at the FXE instrument will be shown.
Calibration parameters are then applied online within the Karabo framework [2] used at the European XFEL, enabling users to see already pre-calibrated data in a near real time online preview; and offline, using dedicated python libraries [3]. First results about the performance of the detector during the early user operation phase of the European XFEL will be presented as well.

[1] M. Hart, C. Angelsen. S. Burge, J. Coughlan, R. Halsall, A. Koch, M. Kuster, T. Nicholls, M. Prydderch, P. Seller, S. Thomas, A. Blue, A. Joy, V. O'Shea, M. Wing (2012). Development of the LPD, a high dynamic range pixel detector for the European XFEL. Nuclear Science Symposium and Medical Imaging Conference (NSS/MIC), 2012 IEEE pp.534-537, DOI: 10.1109/NSSMIC.2012.6551165.

[2] B.C. Heisen et al., Karabo: an integrated software framework combining control, data management and scientific computing tasks, 14th International Conference on Accelerator & Large Experimental Physics Control Systems, Proceedings of ICALEPCS 2013, San Francisco, U.S.A.

[3] M. Kuster, D. Boukhelef, M. Donato, J.-S. Dambietz, S. Hauf, L. Maia, N. Raab, J. Szuba, M. Turcato, K. Wrona & C. Youngman (2014) Detectors and Calibration Concept for the European XFEL, Synchrotron Radiation News, 27:4, 35-38, DOI: 10.1080/08940886.2014.930809

Speaker: Dr Philipp-M. Lang (European XFEL)

Calibration and first user experience with the LPD.pdf

Swedish fika in the hotel lobby (2nd floor)

Invited

125 20 Years of IWoRID

Speaker: Jan Visschers (NIKHEF (NL))

Invited

126 New Generation of Neutron Detectors at the European Spallation Source

Speaker: Ramsey Al Jebali

Oral

Convener: Wasi Faruqi (MRC Laboratory of Molecular Biology)

127 First results from the LUCID-Timepix spacecraft payload onboard the TechDemoSat-1 satellite in LEO orbit

The Langton Ultimate Cosmic ray Intensity Detector (LUCID) is a payload on the satellite TechDemoSat-1 (TDS-1), used to study the radiation environment in low Earth orbit. TDS-1, placed in a 635 km, 98.4° Sun-synchronous orbit, is a technology demonstration satellite developed by Surrey Satellite Technology Ltd as a platform for testing novel technologies from UK academia and industry. The detector operated successfully from late 2014 to mid 2017, capturing over 1.3 million frames of radiation data from its five Timepix detectors on board. LUCID (Fig.1) is the second use of the Timepix detector technology in open space, the first with a 3D configuration of detectors and the first on a commercial platform - providing a valuable testbed for the performance of this technology in a new environment.

Before launch, LUCID was extensively simulated and modelled using GEANT4 [1]. Since launch, the data from LUCID has been analysed with computing resources provided by GridPP, using a deep learning machine learning algorithm to classify particle tracks. This algorithm provides a new approach to processing Medipix data, using a training set of volunteer labelled tracks, combining machine and human classifications. We present some early science applications from these particle track classifications, in particular constructing radiation maps for different particle types, tracking the South Atlantic Anomaly, comparing the day and nightside magnetosphere, and identifying heavy ions. We also compare to measurements from the Timepix detectors on the International Space Station [2], and the Timepix-based SATRAM detector on Proba-V [3].

For managing, accessing, and visualising the LUCID data, we have developed an online, open access, platform called Timepix Analysis Platform at School (TAPAS). This provides a swift and simple way for users to analyse data that they collect using Timepix detectors from both LUCID and other other Timepix experiments. We also also discuss LUCID's link to a system of ground-based Timepix detectors in UK secondary schools, that enables students from more than 80 schools across the country to use Timepix detectors for their own research projects and experiments. Students have taken radiation measurements from soil samples across the country, up mountains, from Unmanned Surface Vessels (`AutoNauts’) at sea, as well as contributing to analysing Medipix data from the MoEDAL experiment at the LHC and the IceCube Neutrino Observatory.

References:

[1] T Whyntie and M A Harrison, Full simulation of the LUCID experiment in the Low Earth Orbit radiation environment, Journal of Instrumentation, Volume 10, March 2015
[2] N Stoffle et al., Timepix-based radiation environment monitor measurements aboard the International Space Station, Nucl. Instr. Meth. Phys. Res., Volume 782, May 2015
[3] C Granja et al., The SATRAM Timepix spacecraft payload in open space on board the Proba-V satellite for wide range radiation monitoring in LEO orbit, Planetary and Space Science, Volume 125, June 2016

Fig.1
LUCID mid assembly. The four orthogonally positioned detectors are visible on the right hand side of the instrument.

Speaker: Dr Peter Hatfield (Institute for Research in Schools; University of Oxford)

hatfield_talk_IWORID_2018.pdf

128 Timepix in LEO orbit on board VZLUSAT-1 nanosatellite: 1-year of Space Radiation dosimetry measurements

The VZLUSAT-1 satellite, the first Czech CubeSat, was successfully launched on June 23, 2017, to a 510 km Sun-synchronous low-Earth orbit. It carries several scientific payloads [1] including a Timepix detector as focal plane imager for the X-Ray telescope onboard. The Timepix payload, already presented in [2], contributes significantly to the satellite data collection, with more than 20 000 frame acquisitions in the first year of deployment. Despite limitations of the satellite attitude control system, necessary for capturing X-Ray images of the Sun, the Timepix detector allows measuring the space radiation environment along the satellite low-Earth orbit. As of April 2018, we conducted 28 whole-Earth mappings, recording radiation doses around the planet with radiation maps comparable to those from the SATRAM - Timepix payload [3]. Further, we show data from scans of the South Atlantic Anomaly and polar radiation horns, where the location and acquisition time were tailored to minimize event pile-up and particle track overlap. Since October 2017, the optics segment of the onboard X-Ray telescope was deployed, which exposed the Timepix unshielded to free open space. This produced entirely new observations namely of low energy (< 10 keV) X-Ray and charged particles as well as a significant increase of measured particle flux. We also registered effects of exposing the sensor to direct intense Sunlight. We will summarize on the actual performance of the custom readout interface, which exceeds expectations in the constrained environment of the low-cost and low-powered CubeSat nanosatellite.

Image: "Radiation map (dose rate) along the orbit of VZLUSAT-1 registered by the Timepix on board. Data collected from 3 months."

[1] M. Urban, O. Nentvich, V. Stehlikova, T. Baca, V. Daniel and R. Hudec, "VZLUSAT-1: Nanosatellite with miniature lobster eye X-ray telescope and qualification of the radiation shielding composite for space application", Acta Astronautica vol. 140, 96-104, 2017.
[2] T. Baca, M. Platkevic, J. Jakubek, A. Inneman, V. Stehlikova, M. Urban, O. Nentvich, M. Blazek, R. McEntaffer and V. Daniel, "Miniaturized X-ray telescope for VZLUSAT-1 nanosatellite with Timepix detector", Journal of Instrumentation 11(10):C10007, 2016.
[3] C. Granja, S. Polansky, Z. Vykydal, S. Pospisil, A. Owens, Z. Kozacek, K. Mellab, and M. Simcak, "The SATRAM Timepix spacecraft payload in open space on board the Proba-V satellite for wide range radiation monitoring in LEO orbit", Planetary and Space Science, 125, pp. 114-129, 2016.

Speaker: Mr Tomas Baca (Czech Technical University in Prague)

Oral: Concluding remarks

Lunch