DEPLETED MONOLITHIC ACTIVE PIXEL SENSORS IN 180 NM TOWERJAZZ AND - - PowerPoint PPT Presentation

depleted monolithic active pixel sensors in 180 nm
SMART_READER_LITE
LIVE PREVIEW

DEPLETED MONOLITHIC ACTIVE PIXEL SENSORS IN 180 NM TOWERJAZZ AND - - PowerPoint PPT Presentation

DEPLETED MONOLITHIC ACTIVE PIXEL SENSORS IN 180 NM TOWERJAZZ AND 150 NM LFOUNDRY TECHNOLOGY M. BARBERO, P. BARRILLON, I. BERDALOVIC, C. BESPIN , S. BHAT, P. BREUGNON, I. CAICEDO, R. CARDELLA, Z. CHEN, Y. DEGERLI, L. FLORES, J. DINGFELDER, S.


slide-1
SLIDE 1
  • M. BARBERO, P. BARRILLON, I. BERDALOVIC, C. BESPIN, S. BHAT, P. BREUGNON, I. CAICEDO,
  • R. CARDELLA, Z. CHEN, Y. DEGERLI, L. FLORES, J. DINGFELDER, S. GODIOT, F. GUILLOUX, T. HIRONO,
  • T. HEMPEREK, F. HÜGGING, H. KRÜGER, T. KUGATHASAN, C. MARIN TOBON, K. MOUSTAKAS,
  • P. PANGAUD, H. PERNEGGER, F. PIRO, P. RIEDLER, A. ROZANOV, P. RYMASZEWSKI, P. SCHWEMLING,
  • W. SNOEYS, M. VANDENBROUCKE, T. WANG, N. WERMES, S. ZHANG

A COLLABORATION EFFORT OF UNIVERSITY OF BONN, CERN (GENEVA), CPPM (MARSEILLE) & IRFU (SACLAY)

DEPLETED MONOLITHIC ACTIVE PIXEL SENSORS IN 180 NM TOWERJAZZ AND 150 NM LFOUNDRY TECHNOLOGY

slide-2
SLIDE 2

STANDARD MONOLITHIC PIXEL SENSORS

− Combine sensor and readout on same wafer using commercial CMOS technologies − Charge collection mainly by diffusion in epi-layer (typically low-resistivity)

− Too slow and not-radiation hard for high radiation and high rate experiments like ATLAS ITk

− Need depleted sensor volume for fast charge collection and large signal à DMAPS

bespin@physik.uni-bonn.de - HSTD12 2019 2

Kolanoski , Wermes 2015

16.12.2019

slide-3
SLIDE 3

DEPLETED MONOLITHIC PIXELS

− Depletion depth 𝑒 ∝ 𝜍 𝑊

%&'(

− High resistive substrate material (𝜍 = 100 Ωcm − kΩcm) − High voltage add-ons (50 − 200 Vbias ) − Multiple nested wells for full CMOS − Backside processing (thinning) − Fully depleted high-resistive bulk or epitaxial layer as charge sensitive volume

bespin@physik.uni-bonn.de - HSTD12 2019 3 16.12.2019

slide-4
SLIDE 4

CMOS TECHNOLOGY

− Commercial processes with high resistive wafers available

− Large production capabilities − Low cost per wafer − Fast turn around time

− Monolithic designs achievable − Low module cost − Thin modules − Small pixel sizes (50 x 50 µm2 or smaller) − Crucial question: radiation hardness and fast charge collection possible?

bespin@physik.uni-bonn.de - HSTD12 2019 4

and many more… For example:

16.12.2019

slide-5
SLIDE 5

DEPLETED MONOLITHIC ACTIVE PIXELS

Large collection electrode

− Electronics inside charge collection well − Uniform field across pixel volume − Short(er) drift distances

à radiation hard

− Large(r) sensor capacitance

à higher noise at given power

Small collection electrode

− Charge collection well separated from electronics − Longer drift distances & low field regions

à radiation hard?

− Small sensor capacitance

à low noise and power (at given SNR)

bespin@physik.uni-bonn.de - HSTD12 2019 5 16.12.2019

slide-6
SLIDE 6

DEVELOPMENT LINES

16.12.2019 bespin@physik.uni-bonn.de - HSTD12 2019 6

5mm 5mm 10mm 9.5mm

CCPD_LF

  • Subm. Sep. 2014
  • Fast R/O coupled to FE-I4

LF-CPIX (DEMO)

  • Subm. Mar. 2016
  • Fast R/O coupled to FE-I4

LF-Monopix1

  • Subm. Aug. 2016
  • Fast column drain R/O

ALICE ALPIDE

5.7 mm 5mm

INVESTIGATOR

  • Subm. 2016
  • 8 x 8 pixel

submatrices

miniMALTA

  • Subm. 2018
  • measures for
  • rad. hardness

10mm 9.5m m

50 x 250 µm2 50 x 250 µm2 Monolithic

MALTA

TJ-Monopix

36.5 x 36.5 µm2

36 x 40 µm2

MALTA (asynchronous) & TJ-Monopix (column drain)

  • Subm. 2018, large matrices
  • Fast asynchr & col. drain R/O

20 mm 10 mm 20 mm

TJ-MonoPix2

  • Subm. Spring 2020
  • Full size 2x2 cm2
  • improved sensor

and front end

LF-MonoPix2

  • Subm. Spring 2020
  • Full height chip

50 x 150 µm2

33 x 33 µm2

slide-7
SLIDE 7

LF-MONOPIX AND TJ-MONOPIX

LF-Monopix TJ-Monopix

  • Fully monolithic DMAPS prototype chips
  • Complete digital logic inside the pixels
  • FE-I3 like column-drain readout architecture that can cope with ATLAS ITk outer layer hit rate

bespin@physik.uni-bonn.de - HSTD12 2019 7 16.12.2019

slide-8
SLIDE 8

LARGE COLLECTION ELECTRODE DESIGN: LF-MONOPIX LARGE COLLECTION ELECTRODE DESIGN: LF-MONOPIX

slide-9
SLIDE 9

LF-MONOPIX: DESIGN

− Large collection electrode design in LFoundry 150 nm CMOS technology − High-resistive substrate (> 2 kΩcm) − 250 x 50 µm2 pixel size (129 rows / 36 columns), nine flavors

bespin@physik.uni-bonn.de - HSTD12 2019 9 Design: IRFU / CPPM / Bonn 16.12.2019

slide-10
SLIDE 10

LF-MONOPIX: SIGNAL AND NOISE

− Gain (unirradiated) between 10 and 12 µV / e- − Noise (ENC) 180 – 240 e-, dispersion 30 – 70 e- − Typical signal up to 25 ke- and tuned threshold ca. 1400 e- (dispersion 400 e-) − No loss of gain and up to 150 e- noise increase after irradiation to 1015 neq cm-2

bespin@physik.uni-bonn.de - HSTD12 2019 10 16.12.2019

slide-11
SLIDE 11

LF-MONOPIX: EFFICIENCY

16.12.2019 bespin@physik.uni-bonn.de - HSTD12 2019 11

Neutron irradiated 1015 neq cm-2: 98.9 % (130 V)

− 1700 e- threshold (not minimum achievable for unirradiated chip)

Unirradiated: 99.6 % (200 V)

slide-12
SLIDE 12

LF-MONOPIX: IN-PIXEL EFFICIENCY

− High and homogeneous efficiency at 200 V bias voltage before irradiation − Small loss of efficiency (1.8 %) between pixels after irradiation due to a reduced signal shared between adjacent pixels

bespin@physik.uni-bonn.de - HSTD12 2019 12 16.12.2019

Unirradiated Neutron irradiated Horizontal Position

slide-13
SLIDE 13

LF-MONOPIX: TID IRRADIATION

− Gain variation < 3 % for all tested flavours − Noise increase of 15 % for CMOS flavors, 25 % for NMOS flavors due to leakage and changing of CSA bias condition, nominally ENCNMOS < ENCCMOS − Irradiation without annealing before measurements

bespin@physik.uni-bonn.de - HSTD12 2019 13 16.12.2019

slide-14
SLIDE 14

LF-MONOPIX: TID IRRADIATION

− Gain variation < 3 % for all tested flavours − Noise increase of 15 % for CMOS flavors, 25 % for NMOS flavors due to leakage and changing of CSA bias condition, nominally ENCNMOS < ENCCMOS − Irradiation without annealing before measurements

bespin@physik.uni-bonn.de - HSTD12 2019 14 16.12.2019

slide-15
SLIDE 15

SMALL COLLECTION ELECTRODE DESIGN: TJ-MONOPIX

slide-16
SLIDE 16

nwell Deep pwell pwell pwell nwell NMOS pwell PMOS Spacing Low dose N implant

P- Epitaxial layer

P++ Substrate

Depleted boundry

TJ-MONOPIX: DESIGN

− Small collection electrode design in 180 nm TowerJazz technology − Low power consumption (≈ 120 mW/cm2) − Modified process with additional n-layer for full depletion − 36 µm x 40 µm pixel size arranged as 448 x 224 pixels in four flavors

bespin@physik.uni-bonn.de - HSTD12 2019 16

ρ ~1 kΩcm

16.12.2019

slide-17
SLIDE 17

TJ-MONOPIX: THRESHOLD & NOISE

Unirradiated Sample Irradiated sample (1015 neq cm-2)

Threshold 350 e- 570 e- Threshold dispersion 34 e- 66 e- ENC 17 e- 23 e-

bespin@physik.uni-bonn.de - HSTD12 2019 17 16.12.2019

slide-18
SLIDE 18

TJ-MONOPIX: MEAN EFFICIENCY

Unirradiated: 97% Irradiated (1015 neq cm-2): 69%

bespin@physik.uni-bonn.de - HSTD12 2019 18

− Testbeam measurement in 2.5 GeV electron beam in Bonn (ELSA)

16.12.2019

slide-19
SLIDE 19

TJ-MONOPIX: IN-PIXEL EFFICIENCY

− High resolution in-pixel efficiency from MALTA chip − Modification of sensor geometry (gap & additional p-well) show improved charge collection in pixel corners (homogenous)

16.12.2019 bespin@physik.uni-bonn.de - HSTD12 2019 20

More on MALTA in previous talk (# 196) and poster session (# 343)

  • M. Dyndal et al., arXiv:1909.11987

0.5 0.55 0.6 0.65 0.7 0.75 0.8 0.85 0.9 0.95 1

miniMalta in pixel efficiency, sector 1

10 20 30 40 50 60 70 track x pos [um] 10 20 30 40 50 60 70 track y pos [um]

miniMalta in pixel efficiency, sector 1

  • Irradiated to 1015 neq cm-2
  • Homogenous efficiency of 97.9 %

Modifications on sensor and front-end

slide-20
SLIDE 20

TJ-MONOPIX: TID IRRADIATION

PMOS RESET FLAVOR

− Irradiation without annealing before measurements − Due to technical limitation HV diode reset flavor only up to 1 MRad

HV DIODE RESET FLAVOR

bespin@physik.uni-bonn.de - HSTD12 2019 21 16.12.2019

slide-21
SLIDE 21

FUTURE MONOPIX DESIGNS

slide-22
SLIDE 22

THE MONOPIX2 DESIGNS

LF-MONOPIX2

− Chip size increased to 2 × 1 cm2 − Smaller pixels: 50 × 150 μm2 − Larger matrix: 340 × 56 px − Analog FE improvement − Pixel layout improvement

TJ-MONOPIX2

− Full size chip of 2 × 2 cm2 − Pixel size: 33.04 × 33.04 μm2 − Larger matrix: 512 × 512 px − Analog FE improvement and threshold tuning − Downstream data processing,

e.g. data buffering and triggering

− Diode reset for good TID performance − Sensor improvements for better NIEL and TID performance à see talk of H. Pernegger

bespin@physik.uni-bonn.de - HSTD12 2019 23

Submission for both chips in spring 2020

16.12.2019

slide-23
SLIDE 23

CONCLUSION

− Promising results for monolithic active CMOS sensors in both large and small electrode design − Fully functional fast read-out architecture in both chips − Large collection electrode design radiation hard up to 1015 neq cm-2 NIEL and 100 MRad TID damage − Issues for low efficiency after neutron irradiation in small electrode design identified and solved (talk by H. Pernegger and poster from L. Flores) − Modifications on front-end in small electrode design for better TID performance (talk by

  • H. Pernegger)

− Full size prototypes in both technologies currently under development

bespin@physik.uni-bonn.de - HSTD12 2019 24 16.12.2019