Diamond Manufacturers for ATLAS Upgrades Brief Overview Next

Diamond Manufacturers for ATLAS Upgrades Brief Overview: Next Upgrade (IBL): Diamond Beam Monitor (DBM) News from two diamond manufacturers E 6/DDL II-VI News from diamond cutting & thinning companies Summary March 26, 2012 R. Kass 1

Diamond Beam Monitor Part of IBL upgrade – Bunch-by-bunch luminosity monitor (aim < 1 % per BC per LB) • Finer segmentation & larger acceptance than BCM • Never saturates • Internal stability monitoring – Bunch-by-bunch beam spot monitor • Need triple-module telescopes for (limited) tracking • Can distinguish hits from beam halo tracks • Unbiased sample, acceptance extends far along beam axis BCM – Baseline: 4 telescopes of 3 IBL modules per side → 24 total diamonds – Avoid IBL insertion volume and ID acceptance (η>2. 5) – Place in pixel support structure close to detector and beam pipe DBM: 3. 2<η<3. 5 March 26, 2012 R. Kass 2

DBM Diamond Sensor Plan Diamond Sensors for DBM: type: polycrystalline CVD diamond size: 21 x 18 mm 2, 525 ± 25 mm thickness number: 40 -45 need for DBM modules 24 + spares 5 for Irradiation studies Two diamond suppliers involved: DDL/E 6 (UK based) 21 x 18 mm p. CVD diamond II-VI (US based) 2 Some parts already in hand that need cutting and/or thinning March 26, 2012 R. Kass 3

Sensors from DDL Ten Detectors ordered from DDL/E 6 (thick E 6 wafer – Wafer 9) – Plan was for wafer to be tested at OSU → wafer characterization → device selection – Wafer 9 received from E 6 11 -Jan-2012 • Rind still attached • Defect level looks ok – Wafer 9 returned to E 6 - rind removal – Wafer arrived at OSU, test grid applied, being testing March 26, 2012 R. Kass 4

Wafer 9 from DDL Growth side Substrate side 5 inches March 26, 2012 R. Kass 5

Thickness of wafer 9 from DDL As grown thickness varies from ~1. 24 to 1. 48 mm March 26, 2012 R. Kass 6

Collection Distance & Current Characterisation of DDL’s wafer 9 CCD (mm) current (n. A) Good regions have I <5 n. A at 1000 V in air All regions of wafer 9 look good We are almost finished measuring the CCD & I in all regions of the wafer Expect to finish measuring the CCD & ship back to DDL/E 6 mid-week March 26, 2012 R. Kass 7

Electric Field Characterisation of DDL’s wafer 9 Need to take into account the varying thickness of the wafer Scale previous CCD plot to E=0. 66 V/mm March 26, 2012 This information allows us to make a “cut map” R. Kass 8

Cut Map Example Based on the CCD and thickness info we divide the wafer into “sensors” wafer 8 Wafer 8 was cut into eleven 2 x 2 cm 2 sensors March 26, 2012 R. Kass 9

E 6/DDL Production Capabilities Get 10 -15 FE-I 4 sensors per wafer Ordered 10 DBM Sensors detectors from DDL’s wafer 9 21 x 18 mm 2 with CCD>200 mm at 1000 V Each piece will be thinned to 525 mm Expect the pieces to arrive in June Processing takes 6 -10 weeks after return of wafer Expect to have access to 10 -20 wafers/year determined by the orders we place March 26, 2012 R. Kass 10

Work with II-VI is the “ 2 nd Company” 5 inches www. ii-vi. com II-VI makes “optical grade” cvd diamond laser windows. . Wafer Results • Can grow thick wafers - 2 mm thick – grown for another application • Very good CCD results – 300 µm @ 0. 5 V/µm • Problems with N 2 and growth rate • problems showed up at the edges March 26, 2012 R. Kass 11

Sensors from II-VI Proceeding to develop additional supplier of detector grade material based on their samples • Good CCD results – 300 µm @ 0. 5 V/µm even though grown for another application and problems with N 2 – Modified growth process • ATLAS committed to produce one detector grade wafer by June with option for second wafer • Quote received 9 -Feb: specified ccd >250 µm @500µm thickness • ATLAS placed order for 10 parts with option for 10 more March 26, 2012 R. Kass 12

Cutting & Thinning Parts in Hand Have tested part thinning (750μm→ 525μm) – 1 cm x 1 cm part used, came back fine Sent first 2 cm x 2 cm parts for thinning returned with edge problems we are looking into a laser trimming repair looks do-able Sent: one 2 x 6 for cutting & thinning four 2 x 2’s for thinning Expect three weeks to get 2 x 2 parts back If ok → send remainder of parts March 26, 2012 R. Kass 13

Summary Two manufacturers are in place: DDL, II-VI Three orders of sensors from two manufacturers: DDL: 11 (wafer 8) + 10 (wafer 9) wafer 9 being tested II-VI 10 (with an option of another 10) looking forward to receiving their pieces in May CCD measurements on DDL’s wafer 9 just about finished will ship back to DDL/E 6 shortly Progress on wafer thinning working with 2 companies in the US Can now get 100’s of sensors/yr March 26, 2012 R. Kass 14

Extra Slides March 26, 2012 R. Kass 15

Introduction: Diamond as sensor material Property Diamond Silicon Band gap [e. V] Low leakage 5. 5 1. 12 Breakdown field [V/cm] 107 3 x 105 Intrinsic resistivity @ R. T. [Ω cm] > 1011 2. 3 x 105 Intrinsic carrier density [cm-3] < 103 1. 5 x 1010 Electron mobility [cm 2/Vs] 1900 1350 Hole mobility [cm 2/Vs] 2300 Saturation velocity [cm/s] Density [g/cm 3] 0. 9(e)-1. 4(h)x 480 107 0. 82 x 107 3. 52 2. 33 6 14 5. 7 11. 9 43 13 -20 ~2000 150 13 3. 61 Radiation length [cm] 12. 2 9. 36 Interaction length [cm] 24. 5 45. 5 Spec. Ionization Loss [Me. V/cm] Aver. Signal Created / 100 μm [e 0] 6. 07 3. 21 3602 8892 4401 8323 Atomic number - Z Dielectric constant – ε Low cap Displacement energy [e. V/atom] Rad hard Thermal conductivity [W/m. K] Heat spreader Energy to create e-h pair [e. V] Low Noise, Low signal Aver. Signal Created / 0. 1 X 0 [e 0] Radiation Studies Single-crystal CVD & poly CVD fall along the same damage curve Proton damage well understood At all energies diamond is >3 x more radiation tolerant than silicon 16

Radiation Damage - Basics Radiation induced effect Leakage current Space charge Charge trapping § Diamond small & decreases ~ none Yes Operational consequence none Silicon Operational consequence I/V = αΦ Heating α ~ 4 x 10 -17 A/cm ΔNeff ≈ -βΦ Thermal runaway β ~ 0. 015 cm-1 Increase of full depletion voltage Charge loss 1/τeff = βΦ Charge loss Polarization β ~ 5 -7 x 10 -16 cm 2/ns Polarization none Charge trapping the only relevant radiation damage effect û NIEL scaling questionable a priori G Egap in diamond 5 times larger than in Si Ø Many processes freeze out Ø Typical emission times order of months § Like Si at 300/5 = 60 K – Boltzmann factor ü A rich source of effects and (experimental) surprises ! OSU, Nov 9, 2011 R. Kass: DOE Review 17