Dataset for SiC and Si detectors comparison for high carbon energy spectrometry

Journal of Instrumentation PAPER • OPEN ACCESS SiC and Si detectors comparison for high carbon energy spectrometry To cite this article: L. Torrisi et al 2024 JINST 19 P07005 View the article online for updates and enhancements. You may also like On the threshold for ion track formation in CaF2 M Karluši, C Ghica, R F Negrea et al. - A didactic proposal about Rutherford backscattering spectrometry with theoretic, experimental, simulation and application activities Federico Corni and Marisa Michelini - THE XMM-NEWTON X-RAY SPECTRA OF THE MOST X-RAY LUMINOUS RADIO-QUIET ROSAT BRIGHT SURVEY-QSOs: A REFERENCE SAMPLE FOR THE INTERPRETATION OF HIGH-REDSHIFT QSO SPECTRA M. Krumpe, G. Lamer, A. Markowitz et al. - This content was downloaded from IP address 147.231.100.81 on 23/10/2025 at 10:00 2024 JINST 19 P07005 Published by IOP Publishing for Sissa Medialab Received: January 27, 2024 Revised: March 6, 2024 Accepted: June 3, 2024 Published: July 5, 2024 SiC and Si detectors comparison for high carbon energy spectrometry L. Torrisi ,𝑎,∗V. Havranek,𝑏A. Mackova,𝑏A. Torrisi 𝑐and M. Cutroneo 𝑎,𝑏 𝑎Dipartimento Scienze Fisiche-MIFT, Università di Messina, Viale F.S. D’Alcontres 31, 98166 S. Agata (ME), Italy 𝑏Nuclear Physics Institute, AS CR, v.v.i., Husinec-Řež 130, 25068 Rez, Czech Republic 𝑐Facoltà di Medicina e Chirurgia, Università Kore di Enna, 94100 Enna, Italy E-mail: [email protected] Abstract: An innovative SiC Schottky junction and a traditional p-n Si surface barrier detector have been compared to detect carbon ions with MeVs kinetic energy. To this, a comparison was performed during Rutherford backscattering spectrometry (RBS) using 2–10 MeV carbon ion beams. The energy resolution and detection efficiency for RBS analysis using the two detectors and their detection electronics are presented. The detector parameters dependencies on the surface passivating layers, ion energy and current dependence, ion penetration depth, detection efficiency, energy resolution, and others are discussed. The comparison of RBS analysis with SiC and Si is investigated highlighting the advantages and disadvantages of using SiC with respect to the traditional Si surface barrier detectors. The two detectors employed for proton, helium and carbon RBS spectrometry of different targets have been also compared on the base of the literature data. Keywords: Interaction of radiation with matter; Radiation-hard detectors; Solid state detectors ∗Corresponding author. ©2024 The Author(s). Published by IOP Publishing Ltd on behalf of Sissa Medialab. Original content from this work may be used under the terms of the Creative Commons Attribution 4.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. https://doi.org/10.1088/1748-0221/19/07/P07005 2024 JINST 19 P07005 Contents 1 Introduction 1 2 Materials and methods 2 3 Results and discussion 4 4 Conclusions 9 1 Introduction In previous experiments the SiC and Si comparison has been investigated using helium [ 1 ] and proton [ 2 ] beams applied to Rutherford backscattering spectrometry (RBS). The Si development in the last twenty years has permitted to optimize of its employment for RBS spectrometry for MeVs alpha and proton ion beams. In this paper, we want to highlight the differences in RBS spectra acquired using SiC-Schottky diode and Si p-n junction detectors of energetic carbon ions. SiC is a prototype and is not optimized for these surface analyses, but we want to verify if it has sufficient energy resolution and sensitivity to permit qualitative and quantitative analysis of the investigated targets. The used SiC detectors have a surface-active zone from 2 mm 2 up to 9 mm 2 , a maximum depletion layer of 80 microns depth and a thin film surface metallization realized with Ni 2 Si compound. They permit to detect of carbon ions from a minimum energy of about 400 keV up to a maximum value of about 400 MeV, as reported in the literature [ 3 , 4 ]. The SiC energy resolution and sensitivity for helium and proton beams are comparable but slightly worse that Si barrier detectors. With this investigation, we want to know these parameters for 2–10 MeV carbon ions in RBS configuration from calibrated targets. Has been demonstrated that Rutherford backscattering spectrometry (RBS) using 2–3 MeV helium and proton beams, normally obtained using Si detectors, can be performed also using SiC detectors with the advantage of having devices more resistant to high temperature, high radiation doses, intense visible radiation, and high fluence of heavy ion beams [ 5 – 7 ]. Generally, the surface barrier radiation detectors employed for high-resolution charge-particle spectrometry are represented by p-n junction partially depleted Si detectors and have a high sensitivity due to the low energy gap (1.1 eV) and energy requested to generate an 𝑒 - ℎ pair (3.6 eV). They have a minimum depletion depth of the order of 1000 μ m and an active surface from 25 mm 2 to more than 100 mm 2 , as well as those produced by Ortec [ 8 ]. Are polarized with a low reverse bias of the order of 50 V and their reverse current is of the order of 50 nA. Their surface is covered by about 10 nm Au and the Si dead layer thickness is of the order of 20 nm. Their energy resolution is high and, for 5 MeV alpha particles, is of the order of 0.05% [ 8 ]. The main characteristics of the SiC Schottky detector are represented by its active depletion layer of the order of 80 μ m for a reverse bias of about 500 V with a reverse current of the order of 0.1 nA, by its energy gap of 3.3 eV, and by the surface nickel silicide with a thickness between 20–200 nm and a density of 7.4 g/cm 3 [ 9 ]. –1– 2024 JINST 19 P07005 SiC detectors are able to detect UV and X-rays, energetic electrons and ions, which can be used for different applications, from plasma monitoring spectrometry to ion time-of-flight measurements, from high-power diodes and transistors to different biomedical applications [ 10 – 13 ]. In this paper, we want to compare the RBS carbon ion spectrometry performed with SiC with that obtainable using classic Si barrier surface detectors. To this we will use monoenergetic carbon beams with energy between 2 and 10 MeV incident on different solid targets placed in high vacuum, recording the Rutherford backscattered particles with SiC and Si both placed at 165 ◦ angle in high vacuum conditions. To take into consideration the different ion energy loss in SiC, based on 4H-SiC semiconductor, and Si detector, and the different parameters characterizing the two devices, some data are reported in comparison in table 1. Table 1. Some physical properties of SiC and Si detectors. Silicon Material property Carbide Silicon Detector property SiC Si (4H-SiC) (Scottky) (p-n) Band gap (eV) 3.26 1.12 Depletion layer (μm) 80 1000 Mass density 3.21 2.33 Active surface area (mm2) 9 25 (g/cm3) Mean e-h pair energy 7.78 3.63 Surface metallization (nm) Ni2Si Au (eV) (200) (10) Electrons mobility 900 1500 Reverse current (nA) 0.1 10 (cm2/V s) Holes mobility 110 550 Carbon Stopping Power 1798 1177 (cm2/V s) at 2 MeV (keV/μm) Thermal conductivity 4.0 1.5 Carbon Range at 2 MeV 1.62 2.45 (W/cm ◦C) (μm) Displacement 25 18 Carbon Energy loss in the 518.4 74.6 energy (eV) dead layers at 2 MeV (keV) Effective atomic 12.54 14 Detection efficiency 60% 100% number at 2 MeV Maximum working 600 300 Subtended solid angle 2.13 3.16 temperature (◦C) (mstr) 2 Materials and methods The carbon beams (2–10 MeV) have been obtained at the 3 MV Tandetron accelerator of the Nuclear Physics Institute (NPI) of the Academy of Science of the Czech Republique in Rez-Prague [ 14 ]. The carbon current was almost always kept at 10 nA but tests were also done with currents between 0.1 nA and 50 nA. The carbon spot was about 1 mm 2 , thus normally the current density was 1 μ A/cm 2 . For this investigation, a calibration target was prepared and analyzed by 2.0 MeV helium RBS analysis. It consists of a Si substrate (300 μ m thick) with three different covering thin films deposited –2– 2024 JINST 19 P07005 by physical vapor deposition (PVD) in high vacuum. Starting from the surface, the triple film is composed of 130 nm Au thin film, deposited on 98 nm Ag thin film, deposited on 55 nm Cu thin film, and deposited on the Si bulk substrate. The Ortec detector (A series) was a partially depleted silicon surface barrier with 25 mm 2 surface area, polarized at 50 V, with 1000 μ m active depth [ 6 ]. Its leakage reverse current was about 10–20 nA. The Si detector was placed at an 89 mm distance from the target and at a 165 ◦ backscattering angle, covering a solid angle of 3.16 mstr, The SiC Schottky detector, realized as a prototype some years ago at ST-microelectronics, in collaboration with CNR-IMM of Catania, Italy [ 15 ], was used reversely polarized at 200 V bias, at which the leakage current was 0.1 nA, as measured with a Keythely instrumentation at room temperature and in high vacuum (10 -6 mbar). It uses 4H-SiC epitaxial layers, 80 μ m thick, with 10 14 cm -3 dopant concentration onto an n-type heavily doped substrate. Ohmic contacts on the sample back side were formed by sputtering a 200 nm thick nickel film. The front contacts were obtained by sputtering deposition of a Ni thick film and performing a rapid thermal processing at 700 ◦ C producing the formation of Ni 2 Si 200 nm thick. The active surface was 3 mm × 3 mm, characterized by the 200 nm Ni 2 Si layers having a density of 7.4 g/cm 3 . The SiC detector was placed at a 65 mm distance from the target and at a 165 ◦ backscattering angle, covering a solid angle of 2.13 mstr. The low detection solid angles used in the experiment and the low carbon ion current avoid any pile-up effect in the acquired RBS spectra. Figure 1shows a photo of the used SiC (a) and Si (b) detectors. Figure 1c reports a sectional view of the device employed in this experiment. Figure 1. Photo of the SiC (a) and of the Si (b) detectors and a sectional view of the SiC device (c). –3– 2024 JINST 19 P07005 Figure 2reports the detection efficiency of the used SiC detector calculated using the ion energy loss in its assembly structure relatives to protons, helium and carbon ion beams versus the ion energy. 0.01 0.1 1 10 100 1000 0 20 40 60 80 100 Detection efficiency (%) Ion Energy (MeV) Protons Helium ions Carbon ions Figure 2. Detection efficiency of SiC detector versus proton, helium and carbon ion energy. The detection efficiency curves are agreed with the literature and indicate that carbon ions can be detected from about 400 keV, due to the energy loss in the surface metallization, up to about 400 MeV, due to the limited depletion layer of 80 microns. Both detectors have been employed using a preamplifier (Ortec mod. 142 A), followed by a linear amplifier (Ortec mod. 672) with 0.1 μ s shaping time and a very compact digital Multi-Channel Analyzer (MCA, Amptek MCA-8000D) to digitize the analogical input signal and to acquire the RBS spectra on PC. RBS analyses performed using 2–10 MeV carbon beams were performed by considering the kinematic backscattering factor depending on the target element, the backscattering detection angle, the energy of the backscattered particle with respect to that of the incident one, and by considering the differential scattering Rutherford cross sections in the laboratory system [ 16 , 17 ]. The SIMNRA code was employed to simulate the RBS spectra analysis and to evaluate qualitatively and quantitatively the amount of the elements detected in the analyzed target [ 18 ]. The energy-stopping powers, penetration depth, electronic and nuclear energy loss and energy straggling in the different materials have been calculated using the SRIM code of Ziegler [ 19 ]. 3 Results and discussion The triple film of calibration realized at NPI for this investigation, was analyzed in composition and in thickness by 2.0 MeV helium RBS spectrometry and by proton-induced X-ray fluorescence (PIXE) analysis. The calibration target corresponds to the reported thickness measurements of Au (130 nm)/Ag (98 nm)/Cu (55 nm) thin pure films deposited on the silicon substrate. A first investigation on the SiC and Si comparison has been performed by the RBS analysis of the carbon ions backscattered (165 ◦ ) by a multi-element target, the Au (130 nm)/Ag (98 nm)/Cu (55 nm) thin films deposited on a silicon substrate. –4– 2024 JINST 19 P07005 Figure 3reports the SiC (a) and Si (b) comparison obtained using 2.0 MeV, 10 nA, 1000 s detection time, and incident carbon ions. 0 100 200 300 400 500 0 2000 4000 6000 8000 10000 12000 Yield of backscattered ions (Counts) Channels 2.0 MeV C+ , #3413 target: Au/Ag/Cu/Si SiC det, 165° 7.0 keV/ch Au Ag Cu a) 0 100 200 300 400 500 0 10000 20000 30000 40000 Yield of backscattered ions (Counts) Channels 2.0 MeV C+ , #3413 target: Au/Ag/Cu/Si Si det, 165° 2.9 keV/ch Au Ag Cu Si b) Figure 3. RBS spectra at 2.0 MeV C + comparison for SiC (a) and Si (b) of C ions backscattered by the Au/Ag/Cu/Si target. The kinematic factors of the carbon ions at 165 ◦ for the target elements are 0.7866, 0.6443, 0.4712 and 0.1655 for Au, Ag, Cu and Si, respectively [ 17 ]. Thus, the backscattered carbon ions from such elements will have a maximum kinetic energy of 1.57 MeV, 1.29 MeV, 0.94 MeV and 0.33 MeV, respectively. The corresponding channel-energy calibration factor is 7.0 keV/ch and 2.9 keV/ch for the SiC and the Si recorded spectrum, respectively. Spectra show a clear higher energy resolution for carbon ion detection by Si with respect to SiC one. The different ion yields (counts) are due to the different solid angles subtended by the two detectors and to the different energy requests to produce 𝑒 - ℎ pairs in the two semiconductors. The low-energy carbon ions backscattered by the target silicon surface, corresponding to a maximum value of about 330 keV, cannot be detected because they are stopped in the surface metallization film of the SiC detector, while they are detected by the Si one having less dense and less thick surface dead layers. The Rutherford backscattering cross section increases with the square of the atomic number of the target elements, thus is expected a yield reduction (counts) with the atomic number of the element [ 16 ]. Such reduction is more evident due to the decreasing thickness of the element films from Au to Ag and Cu. The RBS spectrum acquired by Si is as expected, while the SiC one is not because the low energetic C ions backscattered from the Cu thin film are detected very near to the metal electrode of the SiC Schottky junction and are collected better than the C ions having higher penetration depth and produced far from the collection electrode. This effect is not observed in the case of Si detector due to the higher electron and hole mobilities with respect to those of SiC (see table 1). Figure 4reports the SiC (a) and Si (b) spectra comparison obtained using 3.0 MeV, 10 nA, 1000 s detection time, and incident carbon ions in the same previously described target. In this case, the backscattered carbon ions from the cited elements will have a maximum kinetic energy of 2.36 MeV, 1.93 MeV, 1.41 MeV and 0.50 MeV, respectively. The corresponding channel-energy calibration factor is 7.0 keV/ch and 3.1 keV/ch for the SiC and the Si recorded spectrum, respectively. In this case, the SiC detects the carbon ions backscattered from the silicon substrate because their energy has become higher than that absorbed in the surface metallization film. Moreover, the Si/Au –5– 2024 JINST 19 P07005 0 100 200 300 400 500 600 0 1000 2000 3000 4000 5000 6000 7000 Yield of backscattered ions (Counts) Channels Si Cu Ag Au 3.0 MeV C+, #3406 target: Au/Ag/Cu/Si SiC det, 165° 7.0 keV/ch a) 0 100 200 300 400 500 600 0 5000 10000 15000 20000 25000 Yield of backscattered ions (Counts) Channels 3.0 MeV C+ , #3406 target: Au/Ag/Cu/Si Si det, 165° 3.1 keV/ch Si Cu Ag Au b) Figure 4. RBS spectra at 3.0 MeV C + comparison for SiC (a) and Si (b) of C ions backscattered by the Au/Ag/Cu/Si target. yield ratio and the Cu/Au yield ratio are higher for SiC with respect to the Si detector, indicating that the collection of the 𝑒 - ℎ pair is good because occurs very near to the surface metallic electrode. Now we enhance the C ion energy to 5.0 MeV energy and observe the modified RBS spectra. Figure 5reports the SiC (a) and Si (b) spectra comparison obtained using 5.0 MeV, 10 nA, 1000 s acquisition time, and incident carbon ions in the same previously described calibration target. 0 100 200 300 400 500 600 700 800 900 1000 0 200 400 600 800 1000 1200 5.0 MeV C3+ , #3440 target: Au/Ag/Cu/Si SiC det, 165° 6.95 keV/ch Yield of backscattered ions (Counts) Channels Au Ag Cu Si a) 0 100 200 300 400 500 600 700 800 900 1000 0 1000 2000 3000 4000 Yield of backscattered ions (Counts) Channels 5.0 MeV, C3+ , #3440 target: Au/Ag/Cu/Si Si det, 165° 3.2 keV/ch b) Si Cu Ag Au Figure 5. RBS spectra at 5.0 MeV C + comparison for SiC (a) and Si (b) of C ions backscattered by the Au/Ag/Cu/Si target. The backscattered carbon ions from the cited elements will have a maximum kinetic energy of 3.93 MeV, 3.22 MeV, 2.36 MeV and 0.83 MeV, respectively. The corresponding channel-energy calibration factor is 6.95 keV/ch and 3.2 keV/ch for the SiC and the Si recorded spectrum, respectively. Again, the detected carbon yield for Si is higher than SiC due to the higher detection solid angle, higher detection efficiency, higher 𝑒 - ℎ mobility and minor energy to produce the 𝑒 - ℎ pairs. The best energy resolution is evident by the very distinct peaks due to Si and Cu in the case of the Si detector and, instead, near over-imposed peaks in the case of the SiC detector. In this case, the carbon ions, also at low energy, have a major penetration depth, i.e. are produced far from the surface electrode of SiC, thus the Si/Au and Cu/Au yield ratios are similar between SiC and Si detectors. –6– 2024 JINST 19 P07005 Moreover, the significant carbon ion penetration in the silicon substrate of the target indicates a not Gaussian peak of silicon, but a cassette spectrum due to its considerable thickness, as expected. Finally, a last measure was performed using 10 MeV carbon ions at the same current of 10 nA and 1000 s acquisition time on the same previously described target of reference. Figure 6reports the SiC (a) and Si (b) comparison obtained using 10.0 MeV, 10 nA, 100 s acquisition time, and incident carbon ions in the same previous reference target. 0 200 400 600 800 1000 0 200 400 600 800 1000 Yield of backscattered ions (Counts) Channels 10.0 MeV, C3+ , #3428 target: Au/Ag/Cu/Si SiC det, 165° 7.47 keV/ch Au Ag Cu Si a) 0 200 400 600 800 1000 1200 1400 1600 1800 2000 0 500 1000 1500 2000 Yield of backscattered ions (Counts) Channels 10 MeV, C3+ , #3428 target: Au/Ag/Cu/Si Si det, 165° 3.50 keV/ch Au Ag Cu Si b) Figure 6. RBS spectra at 10.0 MeV C + comparison for SiC (a) and Si (b) of C ions backscattered by the Au/Ag/Cu/Si target. In this case the backscattered carbon ions from the cited elements will have a maximum kinetic energy of 7.87 MeV, 6.44 MeV, 4.71 MeV and 1.66 MeV, respectively. The corresponding channel-energy calibration factor is 7.47 keV/ch and 3.50 keV/ch for the SiC and the Si recorded spectrum, respectively. Again, the detected carbon yield for Si is higher than SiC due to the higher detection solid angle, higher detection efficiency, higher 𝑒 - ℎ mobility and minor energy to produce the 𝑒 - ℎ pairs. The backscattered carbon ions having higher penetration depth in the SiC and Si are produced far from the surface electrode of SiC, thus the Si/Au and Cu/Au yield ratio are similar between SiC and Si detectors and the pair collection does not depend on edge effects. In this case, in fact, the significant carbon ion penetration in the silicon target substrate, of about 9.47 microns, indicates a not Gaussian peak of the thick silicon, as expected for its considerable thickness, and the 7.87 MeV carbons, scattered by Au, penetrate up to about 4.86 microns in SiC. The better energy resolution of Si compared to SiC is very evident from the sharp edge of the silicon signal compared to the very smooth edge of the signal acquired with the SiC detector. The energy resolution comparison for Si and SiC can be performed by the full width at half maximum (FWHM) for the free RBS peaks due to Au, Ag and Cu. Table 2shows the evaluated energy resolutions measured for the two detectors from the RBS peaks of the spectra reported in figure 6, relative to the backscattered 10 MeV carbon ions. At high-energy carbon ions, the SiC energy resolution appears nearly comparable to that of a silicon detector. Although the increase in energy of the carbon ions leads to a better energy resolution of the element peaks seen by the RBS via the SiC detector, the spectra of the Si detector always remain better highlighted and resolved. –7–