Fine-Pixel Detector FPIX Realizing Sub-micron Spatial Resolution - PowerPoint PPT Presentation

TIPP2017, Beijing Fine-Pixel Detector FPIX Realizing Sub-micron Spatial Resolution Developed Based on FD-SOI Technology D. Sekigawa, S. Endo, W. Aoyagi, K. Hara, S. Honda (U. Tsukuba) T . Tsuboyama, M. Yamada, S. Ono, Y . Ikegami, Y . Arai, I. Kurachi, T . Miyoshi, J. Haba, K. Hanagaki. M. Togawa (KEK) Project supported by JSPS Grant-in-Aid for Scientific Research on Innovative Areas Japan/US Cooperation Program in the Field of High Energy Physics KEK Detector Technology Project

SOI PIXEL DEVICES SOI: SILICON-ON-INSULATOR CMOS circuitry fabricated on buried oxide(BOX) LAPIS 0.2um FD-SOI Features: monolithic (no metal bumps) • SOI-CMOS (FETs fully isolated) • Can choose* substrate of optimum • resistivity (fully depleted CMOS sensors possible) *SOITEC SmartCut TM many excellent features Material budget single event effects speed • • • S/N latch up • cost • • power dissipation Operation temp. (0.3K ～ 570K) • Pixel size • • TID tolerance improved to 1MGy by introducing double SOI wafer 2 K. Hara, TIPP2017 May 22-26, 2017 BEIJING

FINE-PIXEL DETECTOR: FPIX2 Pixel size: 8 μ m □ • On-pixel circuit #Pixels: 128 × 128 • Handle wafers: • >single SOI 25k Ω ・ cm p, 500 um t >double SOI 1k Ω ・ cm p, 300 um t Rolling shutter RO • 8 parallel outputs In Development:  to demonstrate excellent spatial resolution achievable with SOI technology (=>tracker for SOFIST TB) chip layout (3mm-sq)  as demonstrator of TID tolerance (FPIX2 equipped with three middle-SOI regions) TID: hole accumulation in BOX/GOX Middle-SOI: compensate TID effects by applied negative voltages PIXEL DECODER 3 I/O 3 K. Hara, TIPP2017 May 22-26, 2017 BEIJING Courtesy of Lapis semiconductor

FNAL TEST BEAM FTBF: 120GeV protons from Main Injector scint. 4.2s beam spill every 1 minute (view from downstream) Trigger generated by a SEABAS2 board using Scint.(5mm-sq) and ATLAS • FE-I4 (2mmx1.75 ｍｍ ROI) Data of 4 FPIX2 and 2 SOFIST sensors acquired per TLU request. • All R/O boards (SEABAS2) implemented with same TimeStamp firmware Last FPIX2 made accessible for exchanging to irradiated DSOI • SEABAS2 （ Soi EvAluation BoArd with Sitcp): 16ch 12bit 40MHz ADCs, Giga-bit Ethernet 4 K. Hara, TIPP2017 May 22-26, 2017 BEIJING

SINGLE/CLUSTER CHARGE Qsum(7x7) vs Q(ADC) of single pixel ClusterSize(Qi>Qsum/49) w/ max. in an event Max intensity 300 kWChits/spill FPIX2 @-9V Qseed>80ADC FPIX2 @-70V Define Qcluster=Qsum(5x5) FPIX2 @-70V Qseed>200ADC intensity max Intensity min 300 kWChits/spill 60 kWChits/spill Beam intensity effect is negligible 5 K. Hara, TIPP2017 May 22-26, 2017 BEIJING

BIAS DEPENDENCE 5x5 cluster charge about the maximum charge Pedestal spread measured w/ beam pixel in an event 2.9 ± 1.0 ADC (very uniform!) VDET=4V VDET=140V Landau Gauss Fit 1600 5x5 cluster charge [ADC] 1400 ~400um/500 Cluster Charge [ADU] depleted 1200 1000 800 V DET = 70V 600 Signal=1134 ± 3ADC 400 200 36 100 140 =14.5 0 0 5 10 15 S/N~78 6 K. Hara, TIPP2017 May 22-26, 2017 BEIJING

FPIX2:DSOI 100KGY 5x5 cluster charge about the maximum charge pixel in an event 5x5 cluster charge [ADC] Non-irrad (200V) 600 ~150um/300 VSOI2=0 irrad depleted 500 non irrad 400 300 100kGy (200V) 200 VSOI2 : I/0=-4V , Pixel=-8V , Dec=-12V 100 100 200 0 0 5 10 15 Innovative double-SOI allows operation of SOI devices to 100kGy Recent study extended to 1MGy 7 K. Hara, TIPP2017 May 22-26, 2017 BEIJING

TRACKING Hit position X2 [ μ m] V DET =70V 300kWC/spill T roll shutter =1ms #clusters/event in 2 nd FPIX Hit position X1 [ μ m] Alignment&Cut Hit position X2 [ μ m] 5 [ μ m 2 ] ｃ In X In Y ｃ Σ (hit residual) 2 [ μ m 2 ] Hit position X1 [ μ m] 8 K. Hara, TIPP2017 May 22-26, 2017 BEIJING

RESIDUALS in X Residuals of DUT hit wrt the track reconstructed using other three FPIXs DUT= DUT= 1 st FPIX2 2 nd FPIX2 σ = 1.075 σ = 0.871 ± 0.009 μ m ± 0.008 μ m 8 8 4 4 [ μ m] [ μ m] -8 -8 -4 0 -4 0 DUT= 4 th FPIX2 DUT= 3 rd FPIX2 σ = 0.854 σ = 3.86 ± 0.008 μ m ± 0.09 μ m 4 8 [ μ m] -8 -4 0 8 4 [ μ m] -8 -4 0 9 K. Hara, TIPP2017 May 22-26, 2017 BEIJING

RESIDUALS in Y Residuals of DUT hit wrt the track reconstructed using other three FPIXs DUT= DUT= 2 nd FPIX2 1 st FPIX2 σ = 0.875 σ = 0.720 ± 0.009 μ m ± 0.006 μ m 4 8 -8 [ μ m] -4 0 8 4 [ μ m] -8 -4 0 DUT= DUT= 4 th FPIX2 3 rd FPIX2 σ = 0.815 σ = 3.84 ± 0.010 μ m ± 0.09 μ m 8 4 [ μ m] -8 0 -4 8 4 [ μ m] -8 -4 0 10 K. Hara, TIPP2017 May 22-26, 2017 BEIJING

INTRINSIC SPATIAL RESOLUTION Observed residual spread: � ��� � ��� in Y in X � ���� � � ������ 1 st : 1.075 ± 0.009 μ m 0.875 ± 0.009 μ m � ��� � 2 nd : 0.871 ± 0.008 μ m 0.720 ± 0.006 μ m under an assumption all four FPIXs 3 rd : 0.854 ± 0.008 μ m 0.815 ± 0.010 μ m have the same intrinsic resolution and 4 th : 3.86 ± 0.09 μ m 3.84 ± 0.09 μ m contributions of misalignment/ multiple scattering are negligible (=conservative σ track value ), we can calculate σ int analytically � ��� in Y in X 1 st : 0.80 ± 0.01 μ m 0.65 ± 0.01 μ m 2 nd : 0.71 ± 0.01 μ m 0.59 ± 0.01 μ m 3 rd : 0.73 ± 0.01 μ m 0.70 ± 0.01 μ m 4 th : 0.76 ± 0.02 μ m 0.76 ± 0.02 μ m Intrinsic resolution: 0.71~0.80 μ m in X 0.59~0.76 μ m in Y (effect of imperfect alignment included) 11 K. Hara, TIPP2017 May 22-26, 2017 BEIJING

CHARGE SPREAD ABOUT PIX(Q MAX ) 1 st FPIX 2 nd FPIX 1 st FPIX Bias=700V Resolution evaluated better in the order 2 nd ~3 rd >>1 st in X 2 nd >>1 st ~3 rd in Y Better resolution of 2 nd Y could be explained by the beam injected at normal… Suspect X-talk in X for degraded resolution… under investigation 3 rd FPIX 12 K. Hara, TIPP2017 May 22-26, 2017 BEIJING

COMPARISON TO PREVIOUS STUDIES 2.0 “Physical limitations to the spatial resolution of solid-state detectors”, M. Boront et al., spatial resolution [um] DEPFET 24um IEEE TNS 62-1, p381 (2015) 1.5 1.0 0.5 FPIX 8um 0.0 10 100 1000 d=450um S/N （ 1.4/24 ） (0.71~0.80)/8 From X (0.59~0.76)/8 From Y FPIX2: S~1000(70V)/N~3 … well in the saturated region (1.07/13.75) Other device with ~1 μ m resolution Fig.4 13.75um pixel (SOI) ⇒ 1.07 ± 0.05 um “a simple formula σ ~p/(S/N) saturates due to M.Battaglia et al., NIMA676,50(2012) δ electrons at higher S/N” FPIX2 achieved a sub-micron spatial resolution 13 K. Hara, TIPP2017 May 22-26, 2017 BEIJING

CONCLUSION  Performance of FPIXs of 8 μ m pixel size has been evaluated successfully in 120 GeV test beam.  Intrinsic spatial resolution of 0.71~0.80 μ m in X 0.59~0.76 μ m in Y obtained (imperfect alignment may remain) by simple charge-weighted mean hit position calculation method.  Silicon detector achieving sub-micron spatial resolution was possible with a 0.20 μ m SOI technology.  SOI detector irradiated to 100 kGy showed no degradation in charge collection. 14 K. Hara, TIPP2017 May 22-26, 2017 BEIJING

SOFIST SOFIST residual to FPIX track Bias=130V (~500um depletion) Readout: on-chip 8-b ADCs 15 K. Hara, TIPP2017 May 22-26, 2017 BEIJING

FTBF BEAM Beamline Beam spot measured with FE-I4 ScanTime=280ns/pix wirechamber 1mm bin 5mm 5mm Beam divergence dy/dz dx/dz mrad mrad 16 K. Hara, TIPP2017 May 22-26, 2017 BEIJING