Radiation Hard Sensor Materials for the CMS Tracker Upgrade The CMS HPK Campaign on behalf of the CMS Tracker Collaboration Institut für Experimentelle Kernphysik, KIT Germany KIT University of the State of Baden-Wuerttemberg and National Research Center of the Helmholtz Association www.kit.edu
Goals of the HPK Campaign! Identify the material baseline for the Tracker Upgrade Phase II! float-zone (FZ) material: 320µm and 200µm! magnetic Czochralski (MCz): 200µm! p-in-n (N) vs. n-in-p (P,Y) technology! Identify the main layout parameter:! Strip pitch! width/pitch-ratio! Isolation techniques of P-type sensors! p-spray (Y)! p-stop (P)! Identify possible new sensor design with intergrated PA or higher granularity! Labeling example for strip sensor: float-zone, 320µm, p-stop = FZ320P 2 September 2nd, 2013 The CMS HPK Campaign
HPK Campaign terms:! Conditions:! Ordered 144 wafers of 6-inch size! Comparabiltiy given -> produced by one vendor: Hamamatsu Photonics K.K.! Measurements done at several Institutes across Europe (17 participating Institutes)! Cross-calibration of measurements! Investigation of:! General dependence on fluences and particle type: Overview of fluences (in10 14 N eq cm -2 ) for different radial positions Radius/cm Protons Neutrons total 40 3 4 7 20 10 5 15! Volume currents! Strip parameters! Charge collection! Annealing behaviour of parameters 50% higher than nominal fluences! T-CAD Simulation studies of sensor material and strip parmeters 3 September 2nd, 2013 The CMS HPK Campaign
Total leakage current! Alpha parameter: current related damage rate! Independent of material, irradiation fluence and particle! No correlation observed with sensor type and material! Measured current for strip sensors higher compared to expectation current of diodes current of strip sensors 4 September 2nd, 2013 The CMS HPK Campaign
Strip parameters I! Coupling capacitance C c:! Interstrip capacitance C int:! Polyresistance R poly! No significant change with irradiation V bias =600V 5 September 2nd, 2013 The CMS HPK Campaign
Strip parameters II! Interstrip Resistance R int! R int drops with fluence from several GOhmcm to some 100MOhmcm! Nevertheless R int still much higher compared to R poly (~2MOhm) -> strip isolation is sufficient! No dependence on annealing, P and N, p-stop and p-spray V bias =600V 6 September 2nd, 2013 The CMS HPK Campaign
Signal FTH200 45.0k 40.0k 35.0k 30.0k p(gev) ALiBaVa; T=-20 C p(gev)+n Pedestal run p(gev)+n Seed Signal (e-) 12.0k 10.0k 8.0k 6.0k 4.0k p( GeV ) p( MeV ) n p(gev)+n p(gev)+n p( MeV ) p(gev)+n p( GeV ) p+n p(gev)+n 0.0 5.0x10 14 1.0x10 15 1.5x10 15 n p(gev) p(m ) FTH200N 600V FTH200P 600V FTH200N 900V FTH200P 900V ( initial, ~2w@RT, ~20w@RT ) Fluence (n eq /cm²) n n p(gev)+n p(gev)+n! non-gaussian noise contribution! just N-type affected! increases with fluence! appears at V bias 600V! See more later...! F 7e14n eq /cm² (R~60cm): N-type at 600V shows higher seed signal than P-type, due to cluster size! After F 7e14n eq /cm 2 (R~40cm): N-type has smaller signal if not noisy! After F ~ 1.5e15n eq /cm 2 (R~20cm): N-type small signal/ high noise! P-type signal ~ 8 ke - at 600V / ~10 ke - at 900V! proposed binary chip threshold: 6 ke - 7 September 2nd, 2013 The CMS HPK Campaign
Signal FZ320 ALiBaVa; T=-20 C! again: red points are noisy measurements! P-type not affected! F < 7e14n eq /cm²: P-type signal lower than N-type! F 7e14n eq /cm 2 : N-type generates much smaller signal! F ~ 1.5e15n eq /cm 2 : N-type below 6 ke -, or noisy! P-type ~8 ke - at 600V/ 11 ke - at 900V; comparable to FTH200P 8 September 2nd, 2013 The CMS HPK Campaign
Signal MCz200 ALiBaVa; T=-20 C! F 7e14n eq /cm 2 : N-type generates smaller signal at 600V for n irradiation! N-type shows RGHs for p irradiation! F ~ 1.5e15n eq /cm 2 : N-type noisy; P-type > 8 ke- for 600V and 900V 9 September 2nd, 2013 The CMS HPK Campaign
Annealing of signal for 200µm F = 7e14 n eq /cm 2 ALiBaVa; T=-20 C ~20weeks! F = 7e14n eq /cm 2 : 200µm sensors well in a band above ~10 to 12 ke -! signal constant for all annealing times 10 September 2nd, 2013 The CMS HPK Campaign
Annealing of signal for 200µm F = 1.5e15 n eq /cm 2 P-type ALiBaVa; T=-20 C N-type 17,5k 17,5k Seed Signal (e-) 15,0k 12,5k 10,0k 7,5k FTH200P 600V FTH200P 900V FZ200P 600V FZ200P 900V MCZ200P 600V MCZ200P 900V Seed Signal (e-) 15,0k 12,5k 10,0k 7,5k FTH200N 600V FTH200N 900V FZ200N 600V FZ200N 900V MCZ200N 600V 5,0k 1.5e15 n eq /cm² 5,0k 1.5e15 n eq /cm² 10 2 10 3 10 4 10 2 10 3 10 4 Annealing (h@rt) Annealing (h@rt)! F = 1.5e15n eq /cm 2 : P-type well above 7.5 ke - until ~20w@RT! N-type drops sharply with annealing time! Most of the time N-type sensors show huge number of RGH (red points >1%) 11 September 2nd, 2013 The CMS HPK Campaign
Annealing of signal for 320µm and 200µm ALiBaVa; T=-20 C! signal of 320µm higher compared to 200µm for most of annealing time! N-type drops strongly after ~40 days of annealing @RT; P-type little less! 320µm annealing not constant over time -> more control during operation necessary 12 September 2nd, 2013 The CMS HPK Campaign
Signal T-CAD Simulation studies! Simulation studies of charge collection efficiency CCE:! radiation damage models developed for neutrons and protons! Reproduction of measurements succesful! Predictions for new sensors and runs possible 13 September 2nd, 2013 The CMS HPK Campaign
MSSD multi-geometry sensor! 12 regions, each 32 strips with different pitches and w/p-ratios P= 70µm, 80µm, 120µm, 240µm w/p= ~0.15, ~0.25, ~0.33 FZ, FTH, MCz; N- and P-type! MSSD samples irradiated:!!!!! 14 20cm: p and n 40cm: p or n September 2nd, 2013 The CMS HPK Campaign
MSSD BeamTest results! cluster signal ~10 ke - in accordance to seed signal of ~8 ke -! noise of N-type much higher compared to P-type 15 September 2nd, 2013 The CMS HPK Campaign
Noise contribution! Symmetric and asymmetric noise seen dependend on material! Pedestal runs in AliBaVa station with BeetleChip! Non-gaussian noise/ Random Ghost Hits RGH mainly on n-type sensors irradiated with charged hadrons! Random ghost hits (RGH): number of hits per strip and event above 5σ of gaussian fit divieded by #strips and #events in a pedestal run without source! A number above 1% was defined as bad, since this would generate 1% occupancy with fakes! fake hits equally distributed over all strips Fig.: Noise histogram from pedestal run 16 September 2nd, 2013 The CMS HPK Campaign
Random Ghost Hits Standard sensor F=1e15n eq /cm² (GeV p) + 5e14n eq /cm² (n) 1000 FTH200N MSSD! Non-gaussian noise seen for N-type FZ and MCz material after short annealing! large RGH phase space for N-type Voltage (V) 800 600 400 200 100 1000 0,00% 0,250% 0,500% 0,750% 1,00% 1,25% 1,50% 1,75% 2,00% 2,25% 2,50% FTH200N! P-type almost not affected Annealing (h@rt) F=1e15n eq /cm² (MeV p) + 5e14n eq /cm² (n) MCZ200N! MSSD:! Increase of RGH occupancy with bias voltage for all regions in N-type! Regions with larger pitch affected first! Regions with small strip width affected first! P-type almost no RGH Voltage (V) Voltage (V) 1000 800 600 400 200 1000 800 600 400 100 1000 Annealing (h@rt) F=1e15n eq /cm² (GeV p) + 5e14n eq /cm² (n) FTH200P 0.00% 0.250% 0.500% 0.750% 1.00% 1.25% 1.50% 1.75% 2.00% 2.25% 2.50% 0,00% 0,250 % 0,500 % 0,750 % 1,00% 1,25% 1,50% 1,75% 2,00% 2,25% 2,50% RGH / Strip / Event RGH / Strip / Event 0.2 0.18 0.16 0.14 0.12 0.1 0.08 0.06 0.04 0.02 0 0.2 0.18 0.16 0.14 0.12 0.1 0.08 0.06 0.04 Region 1-120 Region 2-240 Region 3-80 Region 4-70 Region 5-120 Region 6-240 Region 7-80 Region 8-70 Region 9-120 Region 10-240 Region 11-80 Region 12-70 100 200 300 400 500 600 700 800 900 1000 Region 1-120 Region 2-240 Region 3-80 Region 4-70 Region 5-120 Region 6-240 Region 7-80 Region 8-70 Region 9-120 Region 10-240 Region 11-80 Region 12-70 Bias voltage / V! w/p: 0.13 0.23 0.33! Pitch: 70 um 80 um 120 um 240 um MCz200N FTH200P 0.02 200 100 1000 Annealing (h@rt) 0 300 400 500 600 700 800 900 1000 Bias voltage / V 17 September 2nd, 2013 The CMS HPK Campaign
Neutrons only 7e14n eq /cm² mixed 10e14n eq /cm² neutrons 1000 F=3e14n eq /cm² (GeV p) + 4e14n eq /cm² (n) FTH200N 1000 F=1e15n eq /cm² (n) FTH200N Voltage (V) 800 600 400 0,00% 0,250% 0,500% 0,750% 1,00% 1,25% 1,50% 1,75% 2,00% 2,25% 2,50% Voltage (V) 800 600 400 0,00% 0,250% 0,500% 0,750% 1,00% 1,25% 1,50% 1,75% 2,00% 2,25% 2,50% 200 100 1000 200 100 1000 Annealing (h@rt) Annealing (h@rt)! Effect is much reduced for neutron only irradiation! Dependence on ionizing radiation hints towards a combined effect of bulk damage and surface charge Q f 18 September 2nd, 2013 The CMS HPK Campaign
Simulation of electric fields in N- and P-type! Peak electric field is higher for N-type compared to P-type sensor! P-type: peak E-field decreases with increasing surface damage Q f! N-type: peak E-field increaes with Q f! Micro-discharge possibility higher for N-type sensors 19 September 2nd, 2013 The CMS HPK Campaign
Sum up! Strip parameters slightly affected by radiation for N- and P-type but R int which is still high enough to ensure strip isolation! N-type material shows higher seed signal than P-type when:! irradiated with pure protons up to ~3e14n eq /cm² for 320µm thickness! up to 7e14n eq /cm² for 200µm! Above 7e14n eq /cm² signals in N-type material are smaller than in P-type, since the required higher operation voltage generates RGHs! Long annealing leads to strong decrease of CCE in thick N-type sensors! Signal of thin sensors constant over annealing time; ensures correct operation! All N-type sensors show non-gaussian noise after hadron irradiation, P-type sensors don t! T-CAD Simulation studies:! Very good agreement with measurements even for the irradiated sensors! N-type show higher electric field strenghts compared to P-type 20 September 2nd, 2013 The CMS HPK Campaign
Thank you and especially to the sensor group and the people who provided results and the plots! 21 September 2nd, 2013 The CMS HPK Campaign
Backup 2S-module 22 September 2nd, 2013 The CMS HPK Campaign
23 September 2nd, 2013 The CMS HPK Campaign
24 September 2nd, 2013 The CMS HPK Campaign
Temperature Dependence of RGHs 0,005 fake hits (hits/strip/event) 0,004 0,003 0,002 0,001 MCZ200N 1.5e15 400V 263h MCZ200N 1.5e15 600V 8400h FTH200P 7e14 600V 240h MCZ200N 1.5e15 400V 263h MCZ200N 1.5e15 600V 8400h FTH200P 7e14 600V 240h 60,0µ 40,0µ 20,0µ current (A) 0,000-24 -22-20 -18-16 0,0 temperature ( C)! Fake hit rate increases with temperature following current!! The increase of fake hit rate is typically linked to an increase in current. But an increase in current does not necessarily lead to fake hits. 25 September 2nd, 2013 The CMS HPK Campaign
Symmetric Noise of MSSD sensors 26 September 2nd, 2013 The CMS HPK Campaign
Five trap model 27 September 2nd, 2013 The CMS HPK Campaign