Preliminary Results of the CiS-HH SRD Project G. Lindström*, E. Fretwurst*, I. Pintilie, R. Röder, J. Stahl*

Transcription

1 Preliminary Results of the CiS-HH SRD Project G. Lindström*, E. Fretwurst*, I. Pintilie, R. Röder, J. Stahl* * Uni-Hamburg, guest from NIMP, Bucharest, CiS Institute for Microsensors, Erfurt Optimization of DOFZ for manufacturing process Cost effectiveness, how much [O] is enough? <111> vs. <1>, influence on surface properties Hidden parameters affecting radiation hardness (example: HCl vs. TCA) Czochralski silicon, a possible alternative? Material: FZ: Wacker, Cz: Sumitomo Processing: CiS Erfurt, process very similar to ATLAS wedge Test diodes multi g.r.(hh) + test fields(dortmund) Material characterization: IR: ITME Warsaw, B. Surma SIMS: SIMS Lab Warsaw, A. Barcz 4-point probe and management: ITME, E. Nossarzewska DLTS, C/V + I/V: Uni-HH Radiation damage studies (Uni-HH): Irradiations with p (PS-CERN), n (TRIGA-Ljubljana), Co-6 Gamma (BNL, collab. with Z. Li, E. Verbitskaja) Tools: DLTS + Laplace-DLTS, TSC, TCT, C/V + I/Vannealing, Simulation-studies Project started: November 2 First irradiations: p (July + Sept. 1), n (Oct. 1), γ (Sept. 1) Gunnar Lindström, CiS report, CERN 29-Nov-1

2 Resistivity profiles on 4" CiS wafers Wacker <111> and Topsil <1> measured from depletion voltage on diodes Wacker <111> Standard process <ρ> = 3.71 kωcm ± 4% Topsil <1> Standard process <ρ> = 1.12 kωcm ± 16% Wacker <111> exhibits less radial dispersion than Topsil <1> Gunnar Lindström, CiS report, CERN 29-Nov-1

3 Resistivity profiles on 4" SINTEF wafers Wacker <111> and Topsil <1> measured from depletion voltage on diodes Wacker <111> DOFZ 72h/115C <ρ> = 3.2 kωcm ± 3% Topsil <1> DOFZ 72h/115C <ρ> = 3.85 kωcm ± 9% Also for SINTEF process and after DOFZ: Wacker <111> exhibits less radial dispersion than Topsil <1> Gunnar Lindström, CiS report, CERN 29-Nov-1

4 Depth profiles of Oxygen concentration DOFZ: 8 72 h at 115 C measured with improved SIMS on bevelled samples 1. A first survey: theoretical and SIMS profiles of O-concentration for diffusion at 115 C (theor.): C S = 2.e17 cm -3 ; diffusion constant: D = 2.e-1 cm 2 /sec 5 O-concentration [cm -3 ] h diffusion in N 2 12 h diffusion in N 2 16h diffusion in N 2 24h diffusion in N 2 48h diffusion in N 2 72h diffusion in N depth [µm] SIMS-COMPARISON-CIS REPORT 31-JAN The final choice: all experimental profiles normalised to theoretical curves wrt integrals between 5-1 µm F theor/exp = 1.3 +/- 1% SIMS profiles of O-concentration for DOFZ at 115 C and for as grown Cz-wafers - all wafers: thickness = 28 µm O-concentration [cm -3 ] Cz as grown, <O> = 8.14e17 O/cm 3, <C> = 1.7e15 C/cm 3 FZ 72h/115 C, <O> = 1.14e17 O/cm 3, <C> = 2.9e15 C/cm 3 FZ 48h/115 C, <O> = 9.86e16 O/cm 3, <C> = 5.8e15 C/cm 3 FZ 24h/115 C, <O> = 5.68e16 O/cm 3, <C> = 3.3e15 C/cm depth [µm] SIMS-PROFILES FZ AND CZ FOR RAINER SIMS profiles reliable, symmetric wrt 14 µm Theory predictions excellent [O] in Cz extremely homogeneous as expected Gunnar Lindström, CiS report, CERN 29-Nov-1

5 Depth profiles of Oxygen concentration DOFZ: 8 72 h at 115 C Integrated SIMS compared to IR absorption results Examples of IR absorption spectra IR and SIMS results in comparison: Sample# t diffusion [h] [O] SIMS [1/cm³] [O] IR [1/cm³] IR/SIMS 1 None, (Cz) 8.13e e e e e e e e e e e e e16 1.4e e17 1.3e e e SIMS results: integrated average between d= and 14 µm IR results: averages of 4 measured samples in each case <[O] IR /[O] SIMS > = 1.1 ± 1% Agreement between SIMS and IR excellent Gunnar Lindström, CiS report, CERN 29-Nov-1

6 Resistivity profiles on 4" wafers Change due to oxidation? measured with 4-point probe FZ <111> wafer #12 FZ <111> wafer # Wacker <111> ρ = 2 5 kωcm, before, after oxidation FZ <1> #1 FZ <1> # Wacker <1> ρ = 1 6 kωcm, before, after oxidation Little change in resistivity after oxidation slight improvement of radial dispersion Gunnar Lindström, CiS report, CERN 29-Nov-1

7 Resistivity profiles on 4" Cz-wafers Influence of thermal donors ρ measured with 4-point probe Cz Sumitomo <1>, ρ 6 Ωcm CZ <1> wafer #23 before and after oxidation after ox. before ox. Before oxidation: <ρ> = 59 Ωcm ± 11% 4 After oxidation: <ρ> = 355 Ωcm 2 kωcm kωcm ± 11% CZ <1> wafer #23 after thermal donor anneal to flat _ to flat Comparison: After oxidation and TD kill: <ρ> = 82 Ωcm ± 12% Thermal donor kill essential for final process 1h/8 C + 1min ramp to RT sufficient (?) final ρ larger than on as grown wafer Gunnar Lindström, CiS report, CERN 29-Nov-1

8 Resistivity of Cz-wafers Influence of thermal donors Results from DLTS Cz Sumitomo <1>, ρ 6 Ωcm DLTS and N eff by V dep on test diodes: Diode #12: N eff = 6.88e12 cm - ³ 625 Ωcm N thd = 1.32e12 cm - ³ (double charged), measd. by DLTS effective concentration: 2.64e12 cm - ³ Net doping conc.: ( )e12 cm - ³ = N eff,corr. =4.24e12 cm - ³ 114 Ωcm Diode #35: N eff = 9.1e12 cm - ³ 472 Ωcm N thd = 2.16e12 cm - ³ (double charged), measd. by DLTS effective concentration: 4.32e12 cm - ³ Net doping conc.: ( )e12 cm - ³ = N eff,corr. =4.78e12 cm - ³ 9 Ωcm After thermal donor correction: 96 Ωcm ± 6% 4-point probe measurements: (on different wafer) ρ after oxidation: 36 Ωcm ± 11% ρ after oxidation and TD kill: 82 Ωcm ± 11% TD-concentration measured by DLTS can fully explain the change of resistivity proper TD killing needed in process Gunnar Lindström, CiS report, CERN 29-Nov-1

9 Resistivity profiles on 4" CiS-Cz-wafers ρ measured from depletion voltage on processed diodes Cz Sumitomo <1>, ρ 6 Ωcm CZ <1> resistivity from test diodes after processing After processing including TD kill: <ρ> = 1.34 kωcm ± 6% CZ <1> wafer #23 after thermal donor anneal to flat _ to flat Comparison: Before processing But after oxidation + TD kill (4-point probe) <ρ> =.82 kωcm ± 12% Final process including TD kill is working homogeneity of ρ excellent difference in V dep and 4-point probe not understood Gunnar Lindström, CiS report, CERN 29-Nov-1

10 Reliability of annealing experiments at elevated temperatures the "bistable defect" and 1 st order vs. 2 nd order CiS Wacker <111> 24h/115 o C, TRIGA, Φ n = 1.e14 n/cm 2 N eff [1/cm 3 ] prompt measm. after 8C-step 2nd order fit: N Y = 6.8e12, N C = 3.e12 24h/RT after 8C-step 1st order fit: N Y = 5.1e12, N C = 2.8e annealing time at 8 o C [min] C1212-DNEFF-COMPARISON CiS ATLAS wedge, standard, PS protons, Φ 1MeVn-eq = 9.1e13/cm 2 N eff [1/cm 3 ] prompt measm. after 8C-step 2nd order fit, N Y = 7.2e12, N C = 2.6e h/RT after 8C-step 1st order fit: N Y = 5.6e12, N C = 2.6e annealing time at 8 degrc [min] CW9A3-DNEFF-P Bistable effect both appreciable for n- and p-irradiation 24 h/rt relaxation after each 8C step necessary (τ relax = RT) Resulting annealing curves can only be fitted with 1 st order Large differences in N Y between prompt/2 nd order and relaxed/1 st order Gunnar Lindström, CiS report, CERN 29-Nov-1

11 Proton Irradiation of Test Diodes from CiS ATLAS Wedge Wafers, E p = 24GeV/c -Annealing Results at 8 C CiS ATLAS wedge test diodes Wacker <111> ρ = 3.4 kωcm standard process, PS-proton irradiation 9.94e14 p/cm e14 p/cm N eff [1/cm 3 ] e14 p/cm 2 4.9e13 p/cm e13 p/cm annealing time at 8 degrc [min] CW-STANDARD-DNEFF-P CiS ATLAS wedge, test diodes Wacker <111> ρ = 3.4 kωcm DOFZ process: 24h/115C, PS-proton irradiation N eff [1/cm 3 ] all fits: 1 st order with 2 exponentials 9.94e14 p/cm e14 p/cm e14 p/cm 2 4.9e13 p/cm e13 p/cm annealing time at 8 degrc [min] CW-DOFZ-DNEFF-P Test Diodes with Identical Mask Design and Process as ATLAS Wedge Strip Detectors, Manufactured on Same Wafer For Reliable Analysis Relaxation of Bistable Effect Important (24h RT after Each HT Annealing Step) Annealing Functions only to be fitted by 1 st order fits Gunnar Lindström, CiS report, CERN 29-Nov-1

12 Reverse annealing amplitude for CiS ATLAS wedge test diodes in comparison to ROSE results after 24GeV/c p-irradiation with Φ p = p/cm² CiS ATLAS wedge, PS/7/1, Φ p => 1e15 p/cm 2 Reverse annealing amplitude as function of Fluence N Y [1/cm 3 ] CiS ROSE (standard) CiS ATLAS (standard) g Y = 5.8e-2 cm CiS ATLAS (24h/115C) N sat = 1.72e13/cm -3 CiS-ROSE (16h/115C) V dep [V] for d = 28 µm MeV neutron equivalent Φ [1/cm 2 ] N_Y-COMPARISON ROSE results are reproduced in ATLAS wedge for similar material (Wacker <111>) & oxygenation DOFZ process => saturation of N Y for Φ p > 1 15 p/cm² change in V dep < 1V independent of Φ p and annealing Gunnar Lindström, CiS report, CERN 29-Nov-1

13 CiS SRD test diodes Wacker <111> and <1> standard process CERN scenario for Φ p = p/cm² CiS - SRD, Wacker <111> and <1> standard process PS proton irradiation, low Φ range N eff (4min/8C) [1/cm 3 ] Wacker <1> β = 2.6e-2 Wacker <111>, β = 1.4e V dep [V] for d = 28 µm Φ 1MeVn-eq [1/cm 2 ] NEFF-STANDARD-LOW PHI CiS - SRD, Wacker <111> and <1> standard process PS proton irradiation, total Φ range N eff (4min/8C) [1/cm 3 ] <1>, low Φ: β = 2.1e-2 Wacker <1> Wacker <111> <111>, low Φ: β = 1.4e V dep [V] for d = 28 µm Φ 1MeVn-eq [1/cm 2 ] NEFF-STANDARD-ALL PHI Model description only possible for low Φ range Total donor removal, β compatible with ROSE results Change of N eff for <1> larger than for <111> Gunnar Lindström, CiS report, CERN 29-Nov-1

14 CiS SRD test diodes Wacker <111> DOFZ process CERN scenario for Φ p = p/cm² Comparison with ATLAS wedge stable damage CiS - SRD, Wacker <111> DOFZ process PS proton irradiation, Φ p = p/cm N eff (4min/8C) [1/cm 3 ] h/115 o C, β = 9.1e-3 48h/115 o C, β = 9.22e-3 72h/115 o C, β = 9.33e V dep [V] for d = 28 µm Φ 1MeVn-eq [1/cm 2 ] NEFF-OXY-ALL PHI CiS ATLAS wedge, Wacker <111>, standard and DOFZ process stable damage as function of fluence N min [1/cm 3 ] β = 1.47e-2 β = 1.11e-2 standard process DOFZ: 24h/115 o C V dep [V] for d = 28 µm Φ 1MeVn-eq [1/cm 2 ] N-MIN-COMPARISON CERN scenario: model description possible for total Φ range β compatible with ROSE results, independent of [O] for 24 to 72 h/115 C Comparison with ATLAS wedge reasonable,? for high Φ Measurements: N min = <N eff > for t anneal = 4-8 minutes Gunnar Lindström, CiS report, CERN 29-Nov-1

15 Summary This talk: SRD project is getting under way Wacker FZ shows less ρ dispersion than Topsil Resistivity not affected by DOFZ process New SIMS technique reliable, [O]-results IR Reliable annealing only after relax of bistable defect Analysis reveals 1 st instead of previous 2 nd order Extracted parameters compatible with ROSE results Cz can be used for processing taking care of TD Many thanks to: M.Moll, M.Glaser (CERN), V. Cindro (JSI); E. Nossarzewska, B. Surma (ITME), A. Barcz (ITE) and M. Zielinski (SIMS Lab, Warsaw) Further Results: SIMS, see talk given by A. Barcz Characterization of unirradiated diodes: J. Stahl Co-6 gamma irradiation macroscopic: E. Fretwurst Co-6 gamma irradiation microscopic: I. Pintilie Gunnar Lindström, CiS report, CERN 29-Nov-1