2359-3 Joint ICTP-IAEA Workshop on Physics of Radiation Effect and its Simulation for Non-Metallic Condensed Matter 13-24 August 2012 Electrically active defects in semiconductors induced by radiation Ivana Capan Division of Materials Physics Rudjer Boskovic Institute
Electrically active defects in semiconductors induced by radiation Ivana Capan Division of Materials Physics Rudjer Boskovic Institute http://www.irb.hr/users/capan 1
Outline Radiation damage Capacitance transient techniques Experimental results: radiation induced defects in n-type Si and Ge Conclusion 2
Radiation damage Primary damage is the displacement of a lattice atom to create an interstitial and a vacancy. These can: Recombine no damage Migrate to an oxide coated surface where they may result in excess interface charge React with impurities in the silicon eg VO or displace atoms eg B i which are mobile and will react with other impurities or defects these are likely to have electrical activity The intrinsic defects may react with each other eg V 2 to generate electrically active centres or possibly inert defects. 3
Effects of particle type and flux At low levels of irradiation with electrons or gamma rays almost all the lattice damage is in the form of simple point defects which can be studied by many techniques and are relatively well understood in Si. Damage from energetic particles or heavy ions is much more complex, difficult to study and with many unknowns in terms of electrical activity and reaction energetics. This SRIM simulation of a 1MeV As ion in Ge shows high concentration clusters of primary defects which react to form stable complexes eg V 2, V 3 or larger clusters of vacancies or interstitials eg extrinsic stacking faults. Vacancy cluster in Si Main theme of our research in the last few years has been implant damage sometimes using fast neutron irradiation (at TRIGA Ljubljana) to help us to understand ion damage. 4
Factors affecting the defect species created by particle irradiation or implantation Type and concentration of impurities Nature of surface eg in Si an oxided surface sinks vacancies Particle species Particle energy Particle dose (fluence) Particles dose rate (flux) Temperature Fundamentally this is a consequence of local concentrations of intrinsic defects. 5
Impact of defects on detector performance Creation of donor and or acceptor like defects Generation current (reverse bias leakage) Recombination in neutral regions Carrier trapping The detector community has made dramatic advances in radiation hardness in the last decade both through device and system design. Is it possible to engineer defect kinetics to reduce (or eliminate) the above effects? 6
How can we engineer defect reactions to minimise the impact on electrical properties? Understand the reaction pathways for particle irradiations (essentially cluster kinetics) Use DLTS and/or Laplace DLTS to study the structure of the defects 7
Electrically active defects VO VV VP VV VSb VP E c 0.17 0.23 0.42 0.45 0.29 0.38 0.24 0.31 0.35 E v 300 C 220 C 150 C 170 C 160 C 125 C n-type Si n-type Ge 8
Deep level transient spectroscopy (DLTS) 9
DLTS τ max = ( t 1 t 2 t ) ln t 1 2 1 At typical doping levels of 10 14-10 16 cm -3 sensitivity of the technique is 10 9-10 11 cm -3 N << N t d 10
DLTS analysis 0.15-3 DLTS signal (a.u.) 0.10 0.05 ln(e/t 2 ) -4-5 -6 100 150 200 250 Temperature (K) -7 85 90 95 100 105 1/kT e = κσ ( T ) T 2 e E kt 11
DLTS analysis DLTS signal 1E-9 1E-8 1E-7 1E-6 1E-5 1E-4 1E-3 0.01 0.1 1 (a) [ 1 exp( c t )] C( t p ) = C0 p p 1E-8 1E-7 1E-6 1E-5 1E-4 1E-3 0.01 0.1 1 pulse width (s) (b) n ( t ) = σ < υ > nτn exp( qφ0 / kt )log( t / τ ) T p th T p [I. Capan et al,solar Energy Materials and Solar Cells 91 (2007) 931] 12
DLTS analysis DLTS signal 4 3 2 1 T max =89.66K; C C 0 0 70 75 80 85 90 95 100 Temperatura (K) N T N C C 0 T = 2 N D f ( W ) = ( W W 1E15 2 2 0 λ0 ) ( W1 λ1 ) / C = 2 C 10.5 MeV Si 0 N D 1 f ( W ) 2 0 Free carrier concentration (cm -3 ) 1E14 [E. V. Monakhov, Phys Rev B 65 245201] 1E13 W 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 ε 0 ε r A = C Depth (µm) ε ε ew * 2 0 N = rv 2 13
The biggest problem the lack of resolution!? [DLTS] [Laplace DLTS @ 170K] 14
Laplace DLTS f 0 st ( t) = F( s) e ds f(t) measurements F(s) Laplace DLTS signal Find F(s)! Noise!!! The number of possible solutions is.. [P. Deixler et al, Appl Phys Lett 73 (1998) 3126] 15
Laplace DLTS CONTIN [S.W.Provencher, Computer Physics Communications 27 (1982) 213] FTIKREG [J. Weese, Computer Physics Communications 69 (1992) 99] FLOG [developed for Laplace DLTS] Laplace DLTS signal Emission (s -1 ) 16
LDLTS electric-field effects Electric field Laplace DDLTS [V.P.Markevich et al, Physica B 376-377 (2006) 200] 17
LDLTS separation of capture rates DLTS signal (arb. units) 0.5 0.4 0.3 0.2 0.1 as irrad 100 125 150 175 E2 200 225 250 275 300 E8 E1 E3 E7 E4 DLTS signal (arb. units) 0.2 0.1 as irrad 100 125 150 175 200 225 250 275 300 E3 VV+O V 2 O 0.0 50 100 150 200 250 300 Temperature (K) 0.0 Temperature (K) 150 [L.Dobaczewski et al, J. Appl. Phys. 96 (2004) 4689] σ n (T)=σ n0 exp(- E nσ /kt) 18
The most used techniques are EPR, DLTS... +/- 19
How to get information on structural properties with DLTS? [Phys. Rev. Lett. 51 (1983) 1286] 20
[J.Appl.Phys. 63 (1988) 1549] 21
Uniaxial stress Laplace DLTS The symmetry of a defect centre is obtained from the intensity ratio of the split DLTS lines under uniaxial stress applied along the three major crystallographic directions, <100>, <110> and <111>. VO in n-type CZ:Si The splitting pattern establishes the orthorhombic-i symmetry of the VO complex. [O.Andersen PhD thesis] 22
Silicon Phosphorus-doped n-type CZ-grown silicon crystal with initial resistivity of (1 2) Ωcm. The accumulated doses of fast neutrons were 2 10 14 cm -2 and 2.4 10 12 cm -2 for Si:n1 and Si:n2, respectively. Au n-type Si Al 23
Si:n1 sample after neutron irradiation DLTS signal (arb. units) 0.5 0.4 0.3 0.2 0.1 Si ion implantation Neutron irradiation (/ 3.5) 4 MeV electrons (/ 4) E1 E2 0.16 ev VO 0.24 ev VV(=/-) E3 E4 (2) (1) (3) 0.42 ev VV(-/0)+? 0.10 ev IO 2 0.0 50 100 150 200 250 300 Temperature (K) (1) 4 MeV electrons (F = 1 10 15 cm -2 ) (2) 1 MeV neutrons (F = 2 10 14 cm -2 ) (3) 800 kev Si ions (F = 1 10 9 cm -2 ) [I. Kovačević et. al., J. Phys.: Condens. Matter 17 (2005) S2229] 24
Cluster related E4 = VV(-/0)+? (1) 4 MeV electrons (F = 1 10 15 cm -2 ) (2) 1 MeV neutrons (F = 2 10 14 cm -2 ) (3) 800 kev Si ions (F = 1 10 9 cm -2 ) 25
Si:n1 sample annealing study (1) as irradiated (2) 100 C (3) 150 C (4) 200 C VV(=/-) VV(-/0) 0.30 ev VOH [I. Kovačević et. al., J. Phys.: Condens. Matter 17 (2005) S2229] 26
What we can learn with Laplace? No measurements on as irradiated samples! @87K Laplace DLTS 0.15 ev 0.16 ev 10 1 10 2 emission rate (s -1 ) Laplace DLTS spectrum for the Si:n1 sample after neutron irradiation and the following annealing at 200 C. Emission rate with a slightly lower activation energy has been detected in n-type Si implanted with ions heavier than Si ions [J. H. Evans-Freeman et. al., Nucl. Instr. And Meth. In Phys. Res. B 186 (2002) 41 ]. 27
Laplace DLTS on the E4 peak (cluster related) 0.20 @ 220K LDLTS signal 0.15 0.10 0.05 0.00 1 2 10 1 10 2 10 3 10 4 Emission rate, s -1 Laplace DLTS spectra for the Si:n1 sample (1) after neutron irradiation and (2) the following annealing at 150 o C for 1 hour. 28
Same material -- lower dose of irradiation -- less damage 0.10 ev DLTS signal (arb. units) 1.0 0.8 0.6 0.4 0.2 0.0 E8 Si:n2 Si:n1 E1 E2 VO VV(=/-) E3 50 100 150 200 250 300 Temperature (K) E4 VV(-/0) 29
Si:n2 sample annealing study DLTS signal (arb. units) 0.5 0.4 0.3 0.2 0.1 as irrad 100 125 150 175 200 225 250 275 300 E8 E1 E2 T<150 C E3 E7 E4 0.0 50 100 150 200 250 300 Temperature (K) E8=>? E3=> @ T>200 C VV + O i E7=> VOH V 2 O 30
Laplace DLTS on the E4 (Si:n2:300 C) 0.008 @230K Laplace DLTS 0.004 0.000 0.45eV 100 1000 Emission rate ( s -1 ) 31
Irradiation vs. implantation n-type P-doped CZ Si 1-2 Ω-cm DLTS signal 1.2 0.6 VV(=/-) as irrad. 100 C 150 C DLTS signal 0.4 as irrad. 100 o C 150 o C 0.2 VV(-/0) 100 200 Temperature (K) 0.0 0 90 180 270 360 Temperature (K) 32
Germanium CZ-grown n-type Sb-doped Ge crystal with initial resistivity 2 Ωcm. Au n-type Ge InGa 33
Majority (electron) and minority (hole) carrier traps in nge Normalized DLTS signal (a.u.) 1.0 0.5 0.0-0.5-1.0? V-Sb 50 100 150 200 250 Temperature (K) [I. Kovacevic et al,materials Science in Semiconductor Processing 9 (2006) 606 ] 34
Annealing study C, pf 1.25 1.00 0.75 0.50 0.25 0.00 Ge:Sb Neutrons 2.4x10 12 cm -2 U b = -5.0 V U p = 0 V t p = 1 ms e n = 80 s -1 as irrad 100 o C 160 o C 200 o C 300 o C 50 100 150 200 250 C, pf 0-1 -2-3 Ge:Sb U b = -3.0 V U p = +2.0 V t p = 1 ms e n = 80 s -1 Neutrons 2.4x10 12 cm -2 as irrad 100 o C 160 o C 200 o C 260 o C 50 100 150 200 250 Temperature, K Temperature, K 35
Annealing study + C-V Normalized DLTS peak intensity (a.u.) 2.0 1.5 1.0 0.5 0.0 V-Sb??? Minority trap Carrier concentration (cm -3 ) 7x10 14 6x10 14 5x10 14 C-V @ 140K 0 50 100 150 200 250 300 Annealing temperature ( C) 0 50 100 150 200 250 300 Temperature ( C) V-Sb + Sb VSb 2 36
CZ n-type (100) Ge:Sb 5Ωcm Dose: 5 10 11 cm -2 Uniaxial stress LDLTS measurements Normalised DLTS signal 1.0 neut. elec. 0.5 V-Sb 0.0 50 100 150 200 250 Temperatue (K) 37
Point-like defects and clusters as seen by Laplace DLTS 1.2 190K 64K LDLTS signal (a.u.) 0.8 0.4 V-Sb 0.0 10 100 1000 Emission, s -1 [A. R. Peaker et al, Solid State Phenomena 131-133 (2008) 125] 38
Vacancy clusters 0.10 ev 0.04 0.03 70 K 65 K 75 K -40.4-40.6-40.8 C, pf 0.02 0.01 lnσ n -41.0-41.2-41.4 0.00 10-5 10-4 10-3 10-2 10-1 Pulse length, s -41.6 155 160 165 170 175 180 1/kT (ev -1 ) σ n (T)=σ n0 exp(- E nσ /kt) E nσ =0.04 ev; σ n0 =1 10-17 cm 2 Akceptor-like defekt Vacancy clusters!!! [I. Capan et al, Vacuum 84 (2010) 32] 39
V-SB <110> GPa 40
C 3v trigonal <100> NS <110> 1:1 <111> 1:3 C 1h - monoclinic 1:2 1:2:2:1 1:2:1 Sb-V T=195K vacancy clusters T=65K Laplace DLTS <111> <110> <100> Sb - V Laplace DLTS <111> <110> <100> 10 100 1000 Emission, s -1 100 1000 Emission, s -1 Sb-V pair C 3v Vacancy clusters C 1h??? 41
Summary and conclusions A first attempt to study neutron-irradiation induced defects by means of high-resolution Laplace DLTS in Si and Ge. Annealing study of n-type Si irradiated with fast neutrons clearly shows that the concentration of the double negative charge state of the divacancy increases with annealing temperature as vacancies and/or divacancies are released from cluster related defect, the E4 defect. The annealing behaviour of the H trap consists of negative and positive annealing stages, some of which are related to transformations of simpler radiation-induced defects to more complex structures. These changes are related to the capture of mobile Sb-V and V-V pair by Sb atoms with the creation of Sb 2 -V complexes. Trigonal symmetry for V-Sb. 42