High resolution 3D characterization of silicon detectors using a Two Photon Absorption Transient Current Technique

Depósito de Investigación de la Universidad de Sevilla https://idus.us.es This is an Accepted Manuscript of an article published by Elsevier in Nuclear Instruments and Methods in Physics Research. Section A: Accelerators, Spectrometers, Detectors and Associated Equipment, Vol. 958, on April 2020, available at: https://doi.org/10.1016/j.nima.2019.162865 Copyright 2019 Elsevier. En idUS Licencia Creative Commons CC BY-NC-ND High resolution 3D characterization of silicon detectors using a Two Photon Absorption Transient Current Technique Marcos Fern´ andez Garc´ ıaa,b,∗, Richard Jaramillo Echeverr´ ıaa, Michael Mollb, Ra´ ul Montero Santosc, Rogelio Palomo Pintod, Iv´ an Vilaa, Moritz Wieheb aInstituto de F´ısica de Cantabria (CSIC-UC), Avda. los Castros s/n, E-39005 Santander, Spain bCERN, Organisation europ´enne pour la recherche nucl´eaire, CH-1211 Gen´eve 23, Switzerland cSGIker Laser Facility, UPV/EHU, Sarriena, s/n - 48940 Leioa -Bizkaia, Spain dDepartamento de Ingenier´ıa Electr´onica, Escuela Superior de Ingenieros Universidad de Sevilla, 41092 Spain Abstract The Two Photon Absorption Transient Current Technique (TPA-TCT) is a tool to characterize semiconductor detectors using a spatially confined laser probe. Excess charge carriers are produced by the simultaneous absorption of two sub-bandgap photons in the material. The current induced by the motion of carriers is studied using well known TCT systems. Differently to standard TCT where the energy deposition (pair creation) is continuous along the beam, TPA-TCT reduces this region to an ellipsoidal volume, achieving thus, true 3D spatial resolution. This contribution gives an overview of the technique and shows its performance in irradiated detectors, in particular diodes and High Voltage CMOS detectors. Keywords: Two Photon Absorption, Transient Current Technique, RD50, radiation hardness 2010 MSC: 00-01, 99-00 1. Introduction As feature size and thickness of silicon radiation detectors are increasingly reducing [1], testing probes of gradually higher spatial resolution are required. Thin detectors are needed for radiation tolerance (higher electric field for the same voltage,5 low leakage current) and improved timing (higher slew rate, lower arrival time fluctuations). Small pixels are essential to maintain efficiency (pileup reduction), enchance radiation tolerance (lower leakage current per pixel) and decrease readout noise. Examples of thin detectors with small pixel size are HV-10 HR/CMOS, 3D pixels for HL-LHC or thin LGADs for timing. Visible or near-infrared (NIR) lasers can be micro-focused down to the wavelength limit without diffraction effects [2], therefore providing very good transverse spatial resolution. When a detector is characterized with a beam impinging per-15 pendicular to its surface, the longitudinal resolution (along the propagation direction) is given by the absorption length (penetration depth), ranging from few microns (red laser), to hundreds of microns (NIR laser) [3]. Therefore longitudinal resolution is achieved only for short wavelengths and it is limited to20 the surface of the sample. Test beams, using particle accelerators, provide the best transverse 2D spatial resolution yielding, on top, absolute charge calibration. However the energy deposition of a particle is continuous, therefore no spatial resolution along the beam direction is obtained.25 ∗Corresponding author Email address: [email protected] (Marcos Fern´ andez Garc´ ıa) Resolution in the bulk of the detector is achieved by grazing angle [4] incidence (particle tracks) or laser edge illumination [5]. In these configurations, either the laser or the particle beam are injected from the edge of the device and depth profiling information is obtained. Yet, due to continuous charge de-30 position and reach along the beam, carriers can not be confined along the propagation direction of the probe. In the best configuration, pseudo-3D information is obtained combining normal and edge incidence. In this contribution we present, section 2, a new technique35 that yields true 3D spatial resolution at any depth in the detector and in one single measurement: the Two Photon Absorption- Transient Current Technique. Section 3 shows its performance on selected irradiated detectors. A summary concludes section 4.40 2. TPA-TCT introduction Transient Current Techniques (TCT) refer to the observation of the current induced by the transport of excess carriers generated in a semiconductor by an external ionization source [6]. The current induced by the movement of carriers is amplified45 by a current amplifier and then time resolved using an oscilloscope. From the time development of the signal one can learn about the local conditions the carriers encountered, namely electric field, weighting field and trapping. Normal incidence illumination of the top or bottom sides of a detector using red50 laser will decouple the movement of electrons from holes, while IR top or bottom illumination will induce a signal very similar to that of a charged particle. Finally, if the light is injected from Preprint submitted to Journal of L A T EX Templates March 19, 2019 Figure 1: Knife-edge scan of the TPA laser probe, with fit overimposed and inset representing the laser probe dimension at gaussian 1σlevel. the side, using a collimated beam, the performance of the detector can be also studied as a function of depth. The power of the55 laser can be adjusted to simulate a wide range of signal levels, down to a MIP. TCT has been a workhorse for the study of radiation detectors in both ROSE and RD50 collaborations [7, 8]. In standard TCT, photons ranging from visible to IR have enough energy to produce an e-h pair: this is known as Sin-60 gle Photon Absorption (SPA) TCT. No signal is produced if the energy of the photons is smaller than the difference between conduction and valence band. But if two sub-bandgap photons arrive in a very short time, a Two Step Process (TSP) transition is possible via a virtual state [9]. The existence of this vir-65 tual state is allowed during a period of time not exceeding the uncertainty principle. In order to have two almost coincident photons, pulsed femtosecond pulses are used. In such lasers photons are mode-locked into pulses typically 100 fs wide, although the TSP occurs typically within much less than 1 fs. The70 probability for such process to occur is directly proportional to the square irradiance (power per unit area) of the laser. Because of this sqaure dependence, TPA is a non-linear effect in the material. The attenuation of the beam with the distance (eq. 1) can be75 expressed as a function of the laser irradiance I(r,z) [9]: dI(r,z) dz =−αI(r,z)−βI2(r,z)−σexNI(r,z) (1) where αand βare the single and two-photon absorption coefficients, σex the absorptivity of laser-generated free carriers and (r,z) are cylindrical coordinates. The first term of equation 1 is80 the well known Beer’s law linear absorption and it is important for wavelengths above the bandgap of the material and for low laser irradiance. It leads to an exponential and continuous attenuation of the beam as the material is traversed. The integration of the quadratic term, however, yields a Lorentzian distribution85 of charges which is much more localized in space (at the focus) and leads to a negligible signal away from the focus of the beam, giving TPA-TCT a true 3D spatial resolution. TPA-TCT exploits charge confinement by using sub-bandgap pulsed lasers (α∼0, no linear term) and high enough power90 Laser power [a.u.] 0.001 0.002 0.003 0.004 0.005 0.006 Charge [a.u.] 0 10 20 30 40 50 60 2 /cm eq n 13 , 10β2 /cm eq n 15 , 10β2 /cm eq n 15 , 10α 2 /cm eq n 13 , 10α Figure 2: Collected charge as a function of laser intensity. Full (open) symbols correspond to collected charge with the focus inside (outside) the detector. to excite non-linear effects in Silicon (second term), while still preventing micro-plasmas to develop and spoil signal development (negligible third term). Figure 1 shows a knife-edge scan of the spot from which the Gaussian waist for our experiment is calculated. The in-95 set of the figure depicts the laser probe as an ellipsoid of 10 µm length and 1 µm width. The longer dimension coincides with the propagation direction of the beam. Due to the asymmetric shape of the laser probe, different spatial resolutions are obtained. The elongated dimension dominates spatial resolu-100 tion in normal (top/bottom) illumination of the sample, while the small waist is used in edge-injection measurements. 3. TPA-TCT on irradiated detectors TPA-TCT performance on non-irradiated samples was demonstrated in previous works [10, 11]. In this contribution105 TPA-TCT of both, irradiated diodes and HV-CMOS is presented. Increase of linear absorption in irradiated silicon Non-ionizing radiation in silicon creates Deep Energy Levels (DLs) in the bandgap. These DLs can trap electrons for110 much longer than the characteristic time of the TPA process. As a consequence, the linear absorption coefficient for single photons will increase, as measured in [12]. Linear absorption leads to spatially continuous generation of e-h pairs (see eq. 1), spoiling the confinement of the TPA process along the beam di-115 rection. In the presence of radiation induced linear absorption, a TPA laser will leave a signal even for out-of-focus photons. The carrier generation rate per unit of volume and photon energy ~ω, up to second order approximation, is stated in equation 2.120 dN(r,z) dt ∼αI(r,z) ~ω+βI2(r,z) 2~ω(2) where the factor of two in the denominator of the second term accounts for the generation of one electron-hole pair for every 2 Z [mm] 10.7 10.72 10.74 10.76 10.78 10.8 10.82 10.84 10.86 10.88 [a.u.] 2 Charge in 14 ns/I 0 100 200 300 400 500 600 700 800 900 3 10×1090 V 1050 V 1025 V 1000 V 1000 V 900 V 800 V 700 V 600 V 500 V 400 V 300 V 200 V 100 V Z [mm] 10.7 10.72 10.74 10.76 10.78 10.8 10.82 10.84 10.86 10.88 [a.u.] 2 Charge in 14 ns/I 0 2 4 6 8 10 12 14 16 18 20 1090 V 1050 V 1025 V 1000 V 1000 V 900 V 800 V 700 V 600 V 500 V 400 V 300 V 200 V 100 V Figure 3: Left: Collected charge as a function of focus position for an irradiated 1015 neq/cm2diode, measured at different bias voltages. Outside from the detector volume the SPA contribution is flat. Right: Invariance of SPA signal with the position of the focus is used to correct the charge distribution and improve βcontrast. two absorbed photons. Figure 2 shows the number of generated carriers, calculated as collected charge in 14 ns, as a func-125 tion of laser pulse energy. For the lowest irradiated diode (1013 neq/cm2) a clear quadratic behavior is obtained (full dots, focus inside the sample, second term of eq. 2) which implies a low linear absorption contamination (open dots, measured with the focus well outside the detector, first term in eq. 2). For a differ-130 ent sample irradiated up to 1015 neq/cm2the linear contribution is clearly seen (open triangles) and becomes an important contribution to the total charge (full triangles). TPA-TCT on irradiated diodes Figure 3 (left) shows collected charge profiles of a CNM-135 UMB diode [13] irradiated to 1015 neq/cm2, using normal TPA incidence, as a function of focus position (z coordinate), for different bias voltages. Continuous linear absorption is independent of the focus position (flat pedestal measured in fig. 3) since it only depends on the thickness of material traversed (eq. 1).140 As fig. 3 (right) shows, linear absorption can be effectively removed in normal incidence TPA-TCT. Practically, this is done at raw data level (scope waveforms) by subtracting the induced current, when the focus is out of the sensor, to all other waveforms. For this method to work it is important that the full cross145 section (coordinate r) of the beam is contained within the depletion region. This is not a strong requirement since the cross section of the beam is σ∼1µm. Therefore one can conclude that linear absorption can be subtracted in any normal incidence (top/bottom) TPA-TCT measurement.150 TPA-TCT on HV-CMOS Low resistivity HV-CMOS are challenging detectors for TCT based characterization techniques, since the depletion region can be as small as 10 µm for 100V bias. An HV-CMOS [14] fabricated in the ams H18 high voltage process [15] was suc-155 cessfully measured already in edge-TPA-TCT, see [10]. Here, a device irradiated with neutrons up to 7×1015 neq/cm2is shown in Figure 4 under edge-SPA-TCT illumination (left) and edge-TPA-TCT (right). The calculated charge in 6 ns, normalized to unity to ease comparison, of a test diode160 (DNW without NMOS or PMOS logic) is mapped as a function of the laser injection coordinates in the bulk. In both cases the depletion region grows with increasing Y coordinate. Despite linear absorption smearing of the TPA signal, substructures inside the TPA charge map are visible. In particular the165 Deep N-Well (DNW) of the detector shows up as a rectangular region of smaller charge collection, due to partial depletion. Once the implant was located, the origin of the TPA charge map was shifted to coincide with the corner of the DNW to p-bulk junction. In the SPA measurement, however, it is not possible170 to locate the position of the implant. Therefore the evaluation of the depleted thickness width from edge-SPA-TCT will overestimate the actual value since the position of the junction is not resolved. As a result only, fig. 5 shows the depletion width versus volt-175 age, for the irradiated HV-CMOS sample, and the calculated resistivity from a fit [16]. The width was calculated by Y- coordinate scans, centered in X, at different voltages. For each scan we obtained the depleted width using the collection time of the pulses as discriminating variable. The junction position and180 the end of the depleted bulk were identified for each scan. An in-depth analysis of irradiated HV-CMOS will be the subject of a following paper. 4. Summary TPA-TCT exploits a non-linear effect shown by any material185 when illuminated with a high intensity source (for instance, a laser): for certain wavelengths, light absorption (signal) happens only at the focus of the beam. No photons are absorbed out of focus. This allows to generate a laser probe confined in space, particularly along the beam direction. Our mea-190 surements employed an elliptical laser probe with measured 1 µm (10 µm) beam waist across (along) beam direction. In irradiated detectors light absorption linear coefficient increases with fluence. This leads to a degradation of longitudinal resolution. For top incidence TPA-TCT, the spatial resolution195 3 X [mm] -0.04 -0.02 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 Y [mm] -0.02 0 0.02 0.04 0.06 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 X [mm] -0.05 0 0.05 0.1 0.15 Y [mm] -0.02 0 0.02 0.04 0.06 0 0.2 0.4 0.6 0.8 1 Figure 4: Edge-TCT surface scan of the depletion region of an HV-CMOS irradiated at 7×1015 neq/cm2. Charge collected (raw data) in 6 ns as a function of laser injection point in the bulk. Left: SPA characterization (room temperature, 70 V bias). Right: TPA characterization (80 V). Despite the increase of linear absorption, features in the detector are still resolved in TPA. Bias voltage [V] 0 10 20 30 40 50 60 70 80 Depletion depth [mm] 0.015 0.02 0.025 0.03 0.035 0.04 0.045 cm] Ω [ρ 16.32± 297 cm] Ω [ρ 16.32± 297 Figure 5: Depletion wodth as a function of voltage for the 7×1015 neq/cm2ir- radiated HV-CMOS and resistivity fit overimposed can be recovered by measuring a signal with the laser focus outside of the sample. TPA-TCT measurements of irradiated HVCMOS have shown that even in raw data (no linear absorption subtraction) superior resolving power compared to edge-SPA-TCT.200 5. Acknowledgments This work was performed in the framework of the CERNRD50 collaboration under the projects RD50-2016-04 and RD50-2017-02. This activity was partially supported by the Spanish Ministry of Science under grants FPA2015-71292-C2-205 2-P and FPA2017-85155-C4-1-R; and the European Union’s Horizon 2020 Research and Innovation program under Grant Agreement no. 654168. (AIDA-2020) References [1] M. Garcia-Sciveres, N. Wermes, A review of advances in pixel detec-210 tors for experiments with high rate and radiation, Reports on Progress in Physics 81 (2018) 066101. [2] M. J. Riedl, Optical Design Fundamentals for Infrared Systems, Second Edition (SPIE Tutorial Texts in Optical Engineering Vol. TT48), SPIE Publications, 2001.215 [3] M. A. Green, Self-consistent optical parameters of intrinsic silicon at 300K including temperature coefficients, Solar Energy Materials and Solar Cells 92 (2008) 1305 – 1310. [4] B. Henrich, R. Kaufmann, Lorentz-angle in irradiated silicon, Nucl. Instr. and Meth. in Phys. Res. Section A 477 (2002) 304 – 307. 5th Int. Conf.220 on Position-Sensitive Detectors. [5] G. Kramberger, et al., Investigation of irradiated silicon detectors by Edge-TCT, IEEE Transactions on Nuclear Science 57 (2010) 2294–2302. [6] V. Eremin, N. Strokan, E. Verbitskaya, Z. Li, Development of transient current and charge techniques for the measurement of effective net con-225 centration of ionized charges (Neff) in the space charge region of p-n junction detectors, Nucl. Instr. and Meth. in Phys. Res. Section A 372 (1996) 388 – 398. [7] G. Lindstr¨ om, et al., Radiation hard silicon detectors developments by the RD48 (ROSE) collaboration, Nucl. Instr. and Meth. in Phys. Res. Section230 A 466 (2001) 308 – 326. 4th Int. Symp. on Development and Application of Semiconductor Tracking Detectors. [8] M. Moll, Displacement damage in silicon detectors for high energy physics, IEEE Transactions on Nuclear Science 65 (2018) 1561–1582. [9] M. Rumi, J. W. Perry, Two-photon absorption: an overview of measure-235 ments and principles, Adv. Opt. Photon. 2 (2010) 451–518. [10] M. Fern´ andez, et al., High-resolution three-dimensional imaging of a depleted CMOS sensor using an edge Transient Current Technique based on the two photon absorption process (TPA-eTCT), Nucl. Instr. and Meth. in Phys. Res. Section A 845 (2017) 69 – 71. Proceedings of the Vienna240 Conference on Instrumentation 2016. [11] M. Fern´ andez, et al., On the determination of the substrate effective doping concentration of irradiated HV-CMOS sensors using an edge-TCT technique based on the two-photon-absorption process, Journal of Instrumentation 12 (2017) C01038–C01038.245 [12] C. Scharf, Radiation damage of highly irradiated silicon sensors, Ph.D. thesis, Universit¨ at Hamburg, 2018. [13] CNM-IMB, Instituto de microelectr´ onica de barcelona, http://www. imb-cnm.csic.es/index.php/en/, 2019. [14] I. Peric, Active pixel sensors in high-voltage CMOS technologies for250 ATLAS, J. Inst. 7 (2012) C08002–C08002. [15] ams AG, 0.18 µm high-voltage cmos process, https://ams.com/ process-technology, 2019. [16] H. Spieler, Semiconductor Detector Systems, OUP Oxford, 2005. 4