Transcript Slide 1
Techniques for determination of deep level trap parameters in irradiated silicon detectors AUTHOR: Irena Dolenc ADVISOR: prof. dr. Vladimir Cindro Motivation Silicon detectors usually used as inermost part of tracking system in particle physics detectors → can receive severe levels of radiation → cause of bulk damage which deteriorates detector performance Radiation damage effects caused by lattice deformation: change in effective dopant concentration which effects full depletion voltage increase of leakage current deterioration of charge collection efficiency: part of the drifting charge, created by ionizing particle, is temporarly trapped by the defects introduced with irradiation Damage mechanism: irradiation particles knock off Si atoms dislocated Si atoms (interstitials), empty lattice sites (vacancies), impurities (eg. O, C) form defect complexes → introduction of energy levels (traps) in the band gap For the developement of more radiation hard Si detectors the knowledge of defect kinetics and of correlation between microscopic defects and macroscopic properties of the detector is needed The main tools for characterization of deep level defects: Deep Level Transient Spectroscopy (DLTS) Thermally Stimulated Current (TSC) technique Operation principles of silicon detectors Neutrality of the system → N A, p x p N D ,n xn Usually p+-n → N A, p N D ,n 0 (Vbi V ) w x p xn xn e0 N D,n junction If ND,n comparable with NA,n than ND,n must be replaced with effective dopant concentration Neff =│ ND,n - NA,n│ w(V ) V N eff Desired detector operation voltage V > full depletion voltage VFD Si detector: a diode operated under reverse bias where the depleted region acts like ionization chamber N eff Capacitance C (V ) dQ 0 S dV C ( V ) Q e0 N eff Sw w V Shockley-Read-Hall statistics Occupation of a defect states wth concentration Nt , energy level Et and average occupation probability Pt density of occupied defects nt Nt Pt density of non occupied defects pt Nt nt Change of a defect occupancy possible by electron capture with rate Rn cn npt electron emission with rate Gn en nt hole capture with rate R p c p pnt hole emission with rate G p e p pt Defect parameters (besides Et ) that are determened by DLTS or TSC method: capture coefficients → cn, p vn, p n, p emission probabilities → en,p hole excess generation rate electron excess generation rat dnt The rate of change of the defect ocupancy (G p R p ) (Gn Rn ) dt Shockley-Read-Hall statistics Thermal equilibrium Pt = Fermi function steady state → dnt /dt =0 en ni cn exp Ekt TEi B no current → no net flow of electrones or e p ni c p exp Ekt TEi holes between conduction and valence band B → Rp=Gp, Rn=Gn Relations between en,p and cn,p remain valid non-equilibrium conditions → defect is fully described by Et and n,p Space charge region in steady state carrier concentration negligible: p,n ~ 0 extraction of P → E acts like t i → capture processes can be neglected E in thermal equilibirium F steady state → dnt /dt =0 electron traps generation centers (leakage current) hole traps DLTS: principle of operation Deep level transient spectroscopy (DLTS): uses capacitance transient signals resulting from relaxation processes following an abrupt change of bias voltage or light applied to the sample being investigated. 1. Electron trap → located in the upper half of the band gap: C 1 N eff w During the measurement the device must be partially depleted! DLTS: principle of operation 2. Hole trap → located in the lower half of the band gap: DLTS: determination of trap parameters Electron trap of acceptor type: 1. Capacitance after the filling pulse Effective dopant concentration after the filling pulse shalow donors N eff occupied traps (t ) N D nt (t ) Density of occupied traps after the filling pulse dnt (G p R p ) (Gn Rn ) dt in SCR : Rn , R p ~ 0 electron trap : en e p Capacitance C (t ) CR t nt (t ) nt (0)exp e emission time e 1 / en after the filling pulse n (t ) n (t ) usually N N D 1 t t CR 1 t ND 2N D DLTS: determination of trap parameters C C (t1) C (t2 ) nt (0) et1 / e et2 / e 2. DLTS spectrum High T: emission process to fast to be observed Peak observed at Tmax where emission time satisfies t1 t2 e (Tmax ) ln( t1 / t2 ) Low T: emission process to slow to be observed 3. Extraction of parameters Several temperature scans with different tW = t2-t1 2 → Arrhenius plot: ln Tmax e (Tmax ) versus 1 / Tmax Connection between en and n EC Et 2 ln e (T )T ln( A n ) k BT Nt extracted form DLTS peak since nt(0) Nt Et extracted from slope of Arrhenius plot n from the intercept of Arrhenius plot with ordinate DLTS: example of a DLTS spectrum TSC: principles of operation ♦ ♦ DLTS: observing the change in depth of SCR, due to emission of trapped charge, by measuring capacitance transient Thermally Stimulated Current technique (TSC): observing release of trapped charge directly, by measuring the current due to emission of trapped carriers Measurement process: 1. Cooling: Sample is coold to a low T. Cooling under reverse bias → traps are not filled with carriers 2. Filling: − Switching to zero bias → filling with electrons − Switching to forward bias → electrone and hole injection − Illumination with short-λ laser of n-side (filling only with electrons) or p+-side (filling only with holes) 3. Recording: Heating under reverse bias with constant heating rate. At some T emission probability is no longer negligible → trapped charge is rapidly emitted and swept out of SCR → peaks in current signal TSC: example of TSC spectrum TSC: determination of trap parameters Electron trap Density of occupied traps during the heating dnt (G p R p ) (Gn Rn ) dt in SCR : Rn , R p ~ 0 electron trap : en e p Current t nt (t ) nt (0)exp en (t )dt 0 due to emitted trapped charge during the heating ITSC (t ) e0 S w( t ) en (t )nt (t ) e p pt (t ) 2 dx − en>>ep − During whole TSC scan device fully depleted → w(t)=detector thickness D 0 defects Determination of Nt − From the area under TSC peak Q T2 T 1 − From the peak height ITSC ,max nt (0) ITSC (t ) e0 SD en (t )nt (t ) 2 dT / dt ITSC (T )dT 12 e0Dnt (0) TSC: determination of trap parameters 1. Variable heating method 4 Tmax Position of the TSC peak at Tmax which satisfies ln k E T 1 ln k AE B max B n By repeating TSC temperature scans with → ΔE=EC –Et → extracted from the 4 slope of the plot different heating rate β, a plot ln Tmax versus 1Tmax → σn → extracted from the intercept with is obtained the ordinate 2. Delayed heating method Several TSC scans performed from starting point T0 with different delay times τd between end of the filling pulse and start of the heating → plot ln(ITSC,max) versus τd nt(0) exp(- en(T0) τd ) → TSC peak decreases with increasing τd Repeating procedure at different T0 → Arrhenius plot ln T02e (T0 ) versus 1 / T0 E Connection between en and n: ln e (T )T 2 ln( A n ) k BT ITSC(t) 3. Deconvolution method: numerical fit to the whole spectrum from the slope of a plot emission probability en(T0) is obtained → ΔE=EC –Et → extracted from the slope of the plot → σn → extracted from the intercept with the ordinate Summary TSC and DLTS → techniques for determination of deep level traps parameters, based on observing reversely biased detector response to applied light or an abrupt change of biased voltage (filling of traps with holes and/or electrons) Difference between DLTS and TSC DLTS method: Capacitance transient after the filling process is measured Capacitance transient caused by the change of the width of SCR due to emission of carriers that were trapped during the filling During the measurement device must be biased with the voltage lower than full depletion voltage TSC method: Sample is cooled to a low temperature before filling Emission of trapped carriers observed directly by measuring the current while heating the sample Device is fully biased during the measurement