Development of Radiation Hard Sensors for Very High Luminosity Colliders - CERN-RD50 project -

Transcription

1 11 th International Workshop on Vertex Detectors, VERTEX 22, Hawaii, November 3-8, 22 Development of Radiation Hard Sensors for Very High Luminosity Colliders - CERN-RD5 project - Michael Moll CERN - Geneva - Switzerland On behalf of CERN RD5 Collaboration cern.ch/rd5

2 Outline Motivation and Introduction to RD5 Radiation Damage Microscopic defects (changes in bulk material) Macroscopic damage (changes in detector properties) RD5 - Approaches to obtain radiation hard sensors Material Engineering Device Engineering RD5 future work plan Summary RD5 A very young collaboration! 11/21 Workshop on rad-hard devices 2/22 - R&D Proposal 6/22 - Approved as CERN-RD5 1/22-1 st RD5 Workshop now: startup of specialized working groups

3 RD5 - Development of Radiation Hard Semiconductor Devices for High Luminosity Colliders 274 Members from 52 Institutes 45 European and Asian institutes (34 west, 11 east) Belgium (Louvain), Czech Republic (Prague (2x)), Finland (Helsinki (2x), Oulu), Germany (Berlin, Dortmund, Erfurt, Halle, Hamburg, Karlsruhe), Greece (Athens), Italy (Bari, Florence, Milano, Modena, Padova, Perugia, Pisa, Trento, Triest), Lithuania (Vilnius), Norway (Oslo (2x)), Poland (Warsaw), Romania (Bucharest (2x)), Russia (Moscow (2x), St.Petersburg), Slovenia (Ljubljana), Spain (Barcelona, Valencia), Sweden (Lund) Switzerland (CERN, PSI), Ukraine (Kiev), United Kingdom (Exeter, Glasgow, Lancaster, Liverpool, London, Sheffield, University of Surrey) 6 North-American institutes Canada (Montreal), USA (Fermilab, Purdue University, Rutgers University, Syracuse University, BNL) 1 Middle East institute Israel (Tel Aviv) Detailed member list:

4 Motivation to form a new R&D Collaboration LHC upgrade ( Super-LHC later than 21) 1 years LHC: L = 1 34 cm -2 s -1 f( R=4cm) ~ cm -2 f( R=75cm)~ cm -2 Technology available However, serious radiation damage! 5 5 years S-LHC: L = 1 35 cm -2 s -1 f( R=4cm) ~ cm -2 Focused and coordinated R&D mandatory develop radiation hard and cost-effective detectors LHC experiments ( starting 27) Radiation hardness studies also beneficial before a luminosity upgrade. Radiation hard technologies now adopted have not been completely characterized: Oxygen-enriched Si in ATLAS/CMS pixels Linear collider experiments Deep understanding of radiation damage will be fruitful for linear collider experiments where high doses of e, γ will play a significant role.

5 Main Objective: RD5 - Scientific objectives and strategies To develop radiation hard semiconductor detectors that can operate beyond the limits of present devices. These devices should withstand fast hadron fluences of the order of 1 16 cm -2, as expected for example for a recently discussed luminosity upgrade of the LHC to 1 35 cm -2 s -1. Three R&D strategies: Material engineering - Defect engineering of silicon (oxygenation, dimers, ) - New detector materials (SiC, ) Device engineering - Improvement of present planar detectors (3D detectors, thin detectors, cost effective detectors, ) Variation of detector operational conditions - Low temperature operation - Forward bias operation Further key tasks: Basic studies Defect modeling and device simulation } RD5 Center of gravity

6 Scientific Organization of RD5 RD5 - Development of Radiation Hard Semiconductor Devices for High Luminosity Colliders Spokesperson Mara Bruzzi INFN and University of Florence Material Engineering Deputy-Spokesperson Claude Leroy University of Montreal Device Engineering Defect/Material Characterization Bengt Svensson (Oslo University) Defect Engineering Eckhart Fretwurst (Hamburg University) New Materials Juozas V.Vaitkus (Vilnius University) Pad Detector Characterization Stanislav Pospisil (Prague University) New Structures Mahfuzur Rahman (Glasgow University) Full Detector Systems Gianluigi Casse (Liverpool University) Characterization (pre/post irradiation) SIMS, IR, PL, Hall, TSC, DLTS, EPR, - V 2 O, O-dimers, - clusters,.. Theory Defect formation, defect annealing, energy states,.. Oxygenation DOFZ High resistivity CZ Epitaxial Silicon Dimerization Thermal donors Other impurities H, N, Ge, Pre-irradiation treatments SiliconCarbide CdTe other materials Irradiation tests Test structure characterization IV, CV, CCE Irradiation tests NIEL Device modeling Operational conditions 3D detectors Thin detectors Cost effective solutions Simulations Irradiation tests LHC-like tests Links to HEP Links to R&D of electronics Device simulations Comparison: pad-mini-full detectors CERN contact: Michael Moll

7 Radiation Damage Microscopic Effects V Vacancy particle Si + S E K >25 ev I Interstitial point defects (V-O, C-O,.. ) 6 Co-gammas Compton Electrons with max. E g»1 MeV (no cluster production) E K > 5 kev Electrons More point defects point defects and clusters of defects E e > 255 kev for displacement E e > 8 MeV for cluster Neutrons (elastic scattering) E n > 185 ev for displacement E n > 35 kev for cluster More clusters 1 MeV protons 24 GeV/c protons 1 MeV neutrons Initial distribution of vacancies in (1mm) 3 after 1 14 particles/cm 2 [Mika Huhtinen ROSE TN/21-2]

8 Radiation Damage 3 Macroscopic Effects Annealing Irradiation (e.g. at 6ºC) U dep [V] (d = 3µm) Change of V dep (N eff ) 2. Increase of leakage current n - type type inversion "p - type" 1 14 cm Φ eq [ 1 12 cm -2 ] 6 V N eff [ 1 11 cm -3 ] [ Data from R. Wunstorf 92] I / V [A/cm 3 ] n-type FZ - 7 to 25 KΩcm n-type FZ - 7 KΩcm n-type FZ - 4 KΩcm n-type FZ - 3 KΩcm p-type EPI - 2 and 4 KΩcm Φ eq [cm -2 ] n-type FZ - 78 Ωcm n-type FZ - 41 Ωcm n-type FZ - 13 Ωcm n-type FZ - 11 Ωcm n-type CZ - 14 Ωcm p-type EPI - 38 Ωcm [M.Moll Ph D Thesis] 6 N eff [1 11 cm -3 ] N a = g a Φ eq g C Φ eq annealing time at 6 o C [min] N Y, = g Y Φ eq N C N C α(t) [1-17 A/cm] oxygen enriched silicon [O] = cm -3 parameterisation for standard silicon [M.Moll PhD T hesis] annealing time at 6 o C [minutes] Gamma irradiation: type inversion and increase of leakage current, but: no annealing!

9 3. Deterioration of the charge collection efficiency Two mechanisms reduce collectable charge: Trapping (electrons and holes) Underdepletion (detector design and geometry) ATLAS microstrip + RO electronics Q/Q [%] Max collected charge (overdepletion) / Q standard oxygenated Data: [G.Casse, Liverpool] Φ p [1 14 cm -2 ] Oxygenation has no influence on trapping. After p/cm 2 (24GeV/c) - 8% of charge collected (25ns) - overdepletion needed! More details in later talk: Charge collection in silicon Gianluigi Casse, Liverpool Data: Gianluigi Casse; 1 st Workshop on Radiation Hard Semiconductor Devices for High Luminosity Colliders; CERN; 28-3 November 22

10 Defect Engineering of Silicon Influence the defect kinetics by incorporation of impurities Best example: Oxygen Idea: Incorporate Oxygen to getter radiation-induced vacancies prevent formation of Di-vacancy (V 2 ) related deep acceptor levels Observation: Higher oxygen content less negative space charge (less charged acceptors) One possible mechanism: V 2 O is a deep acceptor V 2 in clusters V O VO VO (not harmful at room temperature) V 2 O (negative space charge) E c VO V 2 O(?) Experimental evidence for V 2 O? Yes! E V - Acceptor at E c ev [I.Pintilie: 1 st RD5 Workshop, APL 81(22)165] - Double-acceptor at E c -.43eV(-/) and E c -.23eV (--/-) [E.Monakhov 1 st RD5 Workshop, PRB 65(22)23327]

11 Oxygen enriched silicon DOFZ (Diffusion Oxygenated Float Zone Silicon) 1982 First oxygen diffusion tests on FZ [Brotherton et al. J.Appl.Phys.,Vol.53, No.8.,572] 1995 First tests on detector grade silicon [Z.Li et al. IEEE TNS Vol.42,No.4,219] 1999 Introduced to the HEP community by RD48 (ROSE) ROSE RD48 N eff [1 12 cm -3 ] First tests show clear advantage of oxygenation Carbon-enriched (P53) Standard (P51) O-diffusion 24 hours (P52) O-diffusion 48 hours (P54) O-diffusion 72 hours (P56) Carbonated Standard Φ 24 GeV/c proton [1 14 cm -2 ] [RD48-NIMA 465(21) 6] Oxygenated Vdep [V] (3 µm) N eff [1 12 cm -3 ] Later systematic tests reveal strong variations with no clear dependence on oxygen content 15 K Ωcm <1 11> - stan dar d 15 K Ωcm <111 > - ox ygenated [M.M oll] Φ 24 G ev / c p ro to n [1 1 4 cm -2 ] V dep [V] (3 µm)

12 Characterization of Oxygen concentration in DOFZ O-concentration [cm -3 ] SIMS [O] depth profile DOFZ 72h/115 o C DOFZ 48h/115 o C DOFZ 24h/115 o C Cz as grown depth [µm] [G.Lindstroem et al.] Abs. coeff. (cm -1 ) IR-absorption: average [O] T=3K Si :O L1-1 L6a-1 L8a Wavenumber (cm -1 ) Sample CZ DOFZ 24h/115C DOFZ 48h/115C DOFZ 72h/115C SIMS [O/cm 3 ] 8.1e17 5.7e16 9.9e16 1.2e17 IR [O/cm3] 8.5e17 6.8e16 1.e17 1.4e17 Average: [O] IR /[O] SIMS = 1.1 1% Oxygen in DOFZ is mono-atomic (no clustering) Details: [G.Lindstroem et al. 1 st RD5-Workshop]

13 Annealing after 24 GeV/c proton irradiation N C [1 13 cm -3 ] N eff [1 11 cm -3 ] DN eff (F eq,t) = N a (F eq,t) N a = g a Φ eq g C Φ eq annealing time at 6 o C [min] 24 GeV/c proton irradiation standard FZ g c = cm -1 oxygenated FZ g c = cm Φ eq [1 14 cm -2 ] + N C (F eq,t) + N Y (F eq,t) = beneficial annealing + stable damage + reverse annealing N Y, = g Y Φ eq N C N C Stable damage reduced by a factor of 3 delayed reverse annealing Saturation of reverse annealing NY/NY N Y [1 13 cm -3 ] standard FZ standard FZ 24 GeV/c proton irradiation normalised reverse annealing τ Y = 5 min ROSE RD48 oxygenated FZ 24 GeV/c proton irradiation Φ eq [1 14 cm -2 ] oxygenated FZ τ Y = 22 min annealing time at 8 o C [min]

14 How much [O] is enough? Reverse annealing amplitude N Y after proton and neutron irradiation N Y [1 13 /cm 3 ] 2 GeV/c proton irradiation standard DOFZ 24h/115 o C 8 standard DOFZ 72h/115 o C DOFZ 24h/115 o C Φ p [p/cm 2 ] [G.Lindstroem et al.] Φ n [n/cm 2 ] [G.Lindstroem et al.] V dep [V] (28 µm) Saturation of reverse annealing for protons and neutrons Higher beneficial effect for protons No big difference between 24h and 72h oxidation at 115ºC N Y [1 13 /cm 3 ] N eff [1 11 cm -3 ] N A g C Φ eq annealing time at 6 o C [min] neutron-irradiation [M.Moll] V dep [V] (28 µm) Details: [G.Lindstroem et al. 1 st RD5-Workshop] N C N Y N C

15 Updated Damage Projection - ATLAS Pixel B-layer N eff (1 12 ) [cm -3 ] Radiation level: F eq LHC-scenario: 1 year = RD48 standard RD48 oxygenated (DOFZ) CiS standard CiS oxygenated (DOFZ) eq (year) = cm -2 (full luminosity) > 85% charged hadrons 1 year = 1 days beam (-7 C) 3 days maintenance (2 C) 235 days no beam (-7 C) RD48 standard CiS standard RD48 oxygenated CiS oxygenated V dep (2µm) [V] New: Std. Silicon: rad.harder than predicted by RD48 DOFZ: reverse annealing delayed and saturating with high fluences time [years] Aug simulation : M.Moll, CERN - parameters : G.Lindstroem, Hamburg

16 DOFZ : Spectacular Improvement of g-irradiation tolerance Vdep [V] CE: <1> STFZ CF: <1> DOFZ 24h CG: <1> DOFZ 48h CH: <1> DOFZ 72h [E.Fretwurst, Hamburg] dose [Mrad] No type inversion for oxygen enriched silicon! Slight increase of positive space charge (due to Thermal Donor generation?) N eff [1 11 cm -3 ] [E.Fretwurst et al. 1 st RD5 Workshop] See also: - Z.Li et al. [NIMA461(21)126] - Z.Li et al. [1 st RD5 Workshop]

17 Material Engineering: Czochralski silicon (CZ) Very high Oxygen content cm -3 (Grown in quartz (SiO 2 )crucible) High resistivity (>1KWcm) available only recently (Magnetic CZ technology) CZ wafers cheaper than FZ (RF-IC industry got interested) Vdep [V] (3 µm) Cz-TD generated Cz-TD killed 19 MeV pions "CERN scenario" standard FZ (CiS) DOFZ 24 h/115 o C (CiS) Φ π [cm -2 ] [G.Lindstroem et al.] N eff [1 12 cm -3 ] Irradiation of test-structures: Only small change in V dep up to p/cm 2 No type inversion Leakage current and charge trapping as for FZ silicon Very high oxygen content: Beware of thermal donors! Details: [G. Lindstroem, 1 st RD5 Workshop] [Z.Li, 1 st RD5 Workshop] First full size detectors produced: MCZ, 9 Ωcm, 32.5 cm 2, strip 1µm, pitch 5µm Electrical performance: 3µA@9V Details: [E.Tuominen, 1 st RD5 Workshop] Test beam: Signal/Noise 1 ; Efficiency 95% ; Resolution of 1µm Radiation tests: under way

18 Defect Engineering: Oxygen Dimers in Silicon Idea: Transform Oxygen into Oxygen dimers (O 2 ) Standard Si : V+O VO; V+VO V 2 O Dimered Si : V+O 2 VO 2 ; V+VO 2 V 2 O 2 How to produce silicon containing dimers? Co 6 -γ or electron irradiation at 35ºC V+O VO; VO+O VO 2 ; I+VO 2 O 2 Does it work? deep acceptor (neg. charged) Simulations: deep acceptor (but shallower than V 2 O) neutral in SCR? 21 L.Lindstroem et al.: IR measurements show dimer line after 35ºC electron irrad. of CZ silicon. [ Lindstr.om et al. PhysicaB 38 (21) 284] 22 S.Watts et al.: First test on FZ detectors performed. No clear evidence for dimers. [Watts et al NIMA485 (22) 153] 23 - Systematic tests under way (RD5 - Dimer Task Force)

19 Epitaxial SiC A material between Silicon and Diamond Property Diamond 4H SiC Si E g [ev] E breakdown [V/cm] µ e [cm 2 /Vs] µ h [cm 2 /Vs] v sat [cm/s] Z 6 14/6 14 ε r e-h energy [ev] τ h [s] Wigner En.[eV] Wide bandgap (3.3eV) lower leakage current than silicon Signal: Diamond 36 e/mm SiC 51 e/mm Si 89 e/mm more charge than diamond About Diamond: See later talk Diamond by Mara Bruzzi Higher displacement threshold than silicon radiation harder than silicon (?)

20 Semi-insulating 4H-SiC with Ohmic Contacts 9 Sr-source ρ ~1 11 Ωcm polarization effects due to deep traps/micropipes fi CCE deterioration Nucl.Phys.B Proc. Suppl.78 (1999), 516 Schottky Barrier Epitaxial SiC Schottky contact Au n + 4H-SiC, 3µm, substrate Ohmic contact Ti/Au 2mm n ~ cm -3, 3-7µm, epitaxial low defect density in epi layer fi negligible polarization effects Present status: e.g. CREE (USA) 2-5 µm epitaxial 4H-SiC 9.$ - 2 wafer

21 Device Engineering: 3D detectors Electrodes: narrow columns along detector thickness- 3D diameter: 1µm distance: 5-1µm Lateral depletion: lower depletion voltage needed thicker detectors possible fast signal n Present size up to ~1cm 2 n p n Hole processing : Dry etching Laser drilling Photo Electro Chemical Electrode material n Doped Polysilicon (Si) Schottky (GaAs) Irradiation tests p/cm 2 (55 MeV, 23GeV ) p/cm 2 (19 MeV) p n n [P. Roy, Glasgow, 1 st RD5 Workshop] More details on Friday: Molecular Biology Sherwood Parker, Hawaii

22 Device Engineering: thin detectors Benefits: - better tracking precision and momentum resolution - low operating voltage - more precise timing - improved radiation tolerance: 5µm thick, 5Ωcm Si detector (V dep = 2V): type inversion after 1 15 cm -2 Drawbacks: - mip signal ~ 35e-h pairs - relatively broad Landau distribution at higher values Technical Approach: - Epitaxial Si device - Thinning with chemical attacks and micro-machining thinned Si Thinning Si by chemical attack (IRST Trento)

23 RD5 Work plan What is happening right now? All RD5 institutes: Inter-calibration of microscopic/macroscopic measurement equipment/procedures Exchange of already produced samples and on-stock material Design of common masks for test structures Evaluation of irradiation facilities, common irradiation runs RD5 subgroups already working on: DOFZ (influence of process, particle dependence of damage) Epitaxial silicon, Hydrogen and Thermal donors in silicon High resisitivity CZ detector characterization/production V x O x defect identification and simulation, Oxygen dimer production Irradiation of samples up to very high fluences > cm -2, CV CCE comparison Evaluation of semiconductors other than silicon, production of 4H-SiC wafers 3D and thin detectors optimal processing technique? Cross check of detector simulation tools within RD5 much more

24 Summary Future detectors for - LHC upgrade will face fluences up to cm -2 - Linear Collider will face high flux of e-m-radiation Newly formed CERN-RD5 collaboration following three strategies: Material Engineering Silicon: Oxygenation, CZ, dimers, other impurities, New Materials: SiC and others Device Engineering Geometries with short carrier drift length: 3D and thin detectors Device modeling, cost effective solutions, n-in-p, n-in-n, Modification of Operational Conditions Forward Bias operation, charge injection, Different particles cause different displacement damage Increase of leakage current Change of V dep : Increase of negative space charge (not for CZ silicon) Charge loss: Trapping and loss of active volume To obtain ultra radiation hard sensors a combination of the above mentioned approaches depending on radiation environment, application and available readout electronics will be best solution. Further information: e.g.: Proposal and all transparencies of 1 st RD5 Workshop (CERN, 2-4 Oct 22)