1. Diamond detector tests HIE-ISOLDE review meeting 15 June 2009 Fredrik Wenander

2. Diamond goals verify suitability (performance and radiation hardness) replace non-linear(?) and sensitive MCP detectors Wish list - 1 st phase current amplification beam profiler / beam position <1 pA beam intensity <0.5% energy measurement - 2 nd phase TOF cavity phase measurements Beam diagnostics for HIE-ISOLDE between cryostat modules + in experimental beam lines See presentation ‘ Survey of available and innovative beam diagnostics alternatives for HIE-REX’ F. Wenander, Feb 2008

3. Al or Ti contacts, very thin, evaporated SC or PC 10 um thickness -> 3.5 mg/cm2 Radiation hard 1 nA for hours, no increase in leakage current Radiation hard to ~1E15 part/cm2 Rapid response Pulse width 1 ns, up to 1E6 part/s ok Low dead-time also for 1E9 part/s/cm2 Good energy resolution For energy resolution, use single crystal diamonds Single crystal diamond detector <1 cm2 Diamond detector basics Some references (more in the presentation) * RD42 project at CERN * Diamond detectors, presentation by Tom Markiewicz * W. Adam et al., New Developments in CVD Diamond Detector Applications, Eur. Phys. J. (2003) Fluxes up to 1E15 hadrons/cm 2 recorded without degradation for PC-DD * P. Moritz et al., Diamond and Related Materials 10 (2001) 1765-1769 Broadband electronics for CVD-diamond detectors 1. High charge carrier mobility -> sub-ns rise-time 2. Large band gap -> small leakage current 3. High lattice energy and very high thermal conductivity -> radiation hardness * single crystal CVD * poly crystalline CVD From T. Markiewicz

4. + Suitable for beam optimisation + Gain of ~10000 with respect to FC - Response dependent on beam energy and type calibrate with FC and grid? - 1 s response time in current mode * Microwave plasma enchanced chemcial vapor deposition * 26 MeV protons * I-V converter Current amplification Ref: M. Marinelli et al., Diamond and Related Materials, 10 (2001) 706-709 Beam profiling Ref. http://www-w2k.gsi.de/detlab/cvd/CVD-Applications.htm Diamond strip detectors 32 strips possible, 60*40 mm 2, 200 um thick, 1.8 mm pitch difficult / expensive with ~64 ADC channels per unit Refs. M. Marinelli et al., Diamond and Related Materials, 10 (2001) 706-709 http://www-w2k.gsi.de/detlab/cvd/CVD-Applications.htm 100 um pitch microstrips reported Ref. H. Fenker et al., presented at IEEE Nucl Sc Symp Nov 1995, San Franscisco Single particle counting Ref. M. Rebisz et al., Diamond and Related Materials, 16 (2007) 1070-1073 Diamond detectors operated in single particle counting mode similar to anode read-out of MCP, but radiation hard <100 um thickness for fast response Ref. M. Rebisz et al., Diamond and Related Materials, 16 (2007) 1070-1073

5. Energy resolution only DeltaE/E=18%, heteroepitaxial diamond crystal integrate freestanding diamond crystal into a custom 50 ohm stripline mount (ref. A. Stolz, Diamond and Related Materials 15 (2006) 807-810) Worse results with heteroepitaxial diamond crystal reported in ref: A. A. Altuhov et al., Diamond and Related Materials 13 (2004) 718-721 DeltaE/E<1% for p, alfa, 7Li Single crystal Diamond detector detector grade Single Crystal CVD detector from company Element Six 300 to 500 um thickness 17 keV resolution for 5.5 MeV  -particles, 1% for heavy ions Ref: M. Pomorski et al., Phys. Stat. Sol (a) 203 No12 (2006) 3152-3160 Need to reduce beam intensity to single particle level – how? scatter beam on central wire or grids and slits Energy calibration with the bending magnet energy, charge and mass dependent Direct energy measurement Si detector Ref: M. Pomorski et al., Phys. Stat. Sol (a) 203 No12 (2006) 3152-3160 Ref. A. Stolz et al., Diamond and related materials 15 (2996) 807-810 a. Chopper RF – not feasible Need a beam slice of 50 ps -> (Im)Possible? Drift distance of 2 m Use diamond detector as (start &) stop detector resolution <50 ps b. Thin double diamond detectors Ideal with the 20 ps time resolution Only ~70 um range for 4 He 3 MeV/u Only ~30 um range for 207 Pb 3 MeV/u Would need a thickness of ~10 nm Showstopper! No degradation of the time resolution for intensities up to 1E7 part/s/mm 2 Pulsed beam at REX -> attenuation required For a TOF after the linac I believe one would need an RF chopper to select a single RF bucket. Then a drift distance of ~3 m and a fast detector. The detector has a time resolution of <100 ps, and the TOF varies between 60 and 210 ns (from 1.2 MeV/u to 10 MeV/u). For a 1% energy accuracy a 0.5% velocity accuracy is needed, translating to between 0.3 and 1 ns time resolution. * make use of rapid rise time + radiation robust Should only have 1 particle / bucket -> Beam attenuation (<1E6 times) Rutherford scattering on a wire or a foil -Energy loss? - Angular resolution Collimator + grids before 2 m beam tube at an angle assume <50 ps response, 2 m drift distance, beta<20% Time resolution = 0.15% => energy resolution 0.3% Is the bunch structure still maintained? TOF Ref. V. Verzilov et al., DIPAC 2007

6. pCVD, 10x10 mm 2, 500 um thick plated with square 8x8 mm 2 Al electrodes thickness of 25 nm sCVD, 5x5 mm 2, 500 um thick plated with 3 mm diameter Au electrodes thickness of 500 nm Test ‘outsourced’ to: E. Griesmayer, ATLAS/CERN and Bergoz Instrumentation, St Genis, France At REX 12 C 4+ with 1.9 MeV/u => E total =22.8 MeV Collimators + multiple micro-grid layers (single layer transmission 18±2%) < 10 6 ions/s, counted with a scaler > 10 6 ions/s, leakage current measured with a picoammeter Test setup at REX Si substrate removed 4 pCVD, 10x10 mm2, 500 um thick. plated with square 8x8 mm2 Al electrodes with a thickness of 25 nm. one sCVD, 5x5 mm2, 500 um thick. plated with 3 mm diameter Au electrodes, are 500 nm thick. preamplifiers with 20 dB and 40 dB gain Is this really correct? Depending on the polarity of the bias voltage, either the ionised electrons or holes are collected. The drift of these charges produce a signal

7. + Very low noise level (< 1mV) -> Noise discrimination easy + Particle counting up to 1E4 part/s (duty factor => ~1E7 part/s) Single pulse example, +500 V bias Pulse height 109 mV Pulse width 7.7 ns + ~1% energy resolution 12 C 4+ 1.9 MeV/u sCVD with 1000 V bias - Cases with worse resolution Solved with polarity change space charge? charge trapping? T scale =10 ns/div V scale =50 mV/div Signal height, not charge integration Results sCVD

8. pCVD (anti)results 1. fluctuating leakage current (tens pA to nA) -> current amplification mode not viable 2. polarity and time dependent -> counting problems 3. signal size decreases with beam loading / time -> position tuning difficult; always better at fresh pixel -> counting problems Single pCVD pulse T scale =5 ns/div V scale =5 mV/div What happens when the hole is trapped? Hole trapping must be present also for traversing particles? Estimate success rate of future tests – await development * poor resolution due to structural defects, columnar nature of pCVD * formation of additional trapping center by implanted ions * time varying effect - due to polarisation caused by carrier pile-up a contacts or space charge formation in the bulk of the crystal polarization - previously reported by TU Munich / GSI and C. Tuve et al (Italy) Use different pad configuration - pads on same side? C. Tuve et al., Diamond and Related Materials 15 (2006) 1986-1989 Reasons? *charge trapping * polarization * structural defects * contact layer *…

9. * sCVD can work out – but expensive 3 kCHF for 5x5 mm 2 * pCVD still some way to go long term leakage current stability not proven particle counting difficult * Theoretical understanding lacking * Most previous results with: traversing particles / MIPs low flux rates * Production process under control? Future tests considered 1. Current amplification tests using p/s CVD detector with DLC contact? 2. Full-fledged investigation of energy resolution 3. Verify timing properties for TOF and phase setup of cavities Conclusion and Outlook