Preliminary results on LGAD - CERN
Transcript of Preliminary results on LGAD - CERN
E. G. Villani STFC Rutherford Appleton Laboratory
Preliminary results on LGAD
E.G.Villani, 37th RD50 Workshop, Zagreb (online)
Overview
• LGAD project description and goals
• design simulations
• preliminary results
• next steps
1E.G.Villani, 37th RD50 Workshop, Zagreb (online)
3
A RAL, University of Oxford, University of Birmingham and Open University project in collaboration
with Teledyne e2v foundry for LGAD production
• Three types of cathode size of single cells (4,2 & 1mm) and one of 2 x 2 array of 1mm cells. Up to four
different cell layout flavors (A,B,C,D) are implemented with different distances of guard ring to the
cathode
• LGAD and PiN diodes share the same layouts , only difference being the presence or not of the gain
layer
LGAD Project description and goals
1E.G.Villani, 37th RD50 Workshop, Zagreb (online)
PiNLGAD
4mm
2mm A 2mm B
1mm - A 1mm - B 1mm - C 1mm - D 1mm - D 1mm - C 1mm - B 1mm - A
2mm B 2mm A
4mm
2x2mm A 2x2mm B 2x2mm B 2x2mm A
Field 1
4
• Additional devices are available on each wafer, which has 3 fields, to investigate various effects,
including separate biasing of P-stop, retraction of gain layer from JTE contact, further extension or
reduction of metal cathode over JTE
• Capacitors, diodes and resistors available too
LGAD Project description and goals
2E.G.Villani, 37th RD50 Workshop, Zagreb (online)
Field 2
Field 3
5
• Eight 6’’ wafers of 50m thick HR P- epi layer
• Different gain layer implant doses and energy to sample gain region according to
TCAD simulations
LGAD Project description and goals
3E.G.Villani, 37th RD50 Workshop, Zagreb (online)
Wafer # GL dose GL energy GL type
1 1 1 Boron
2 1.07 1 Boron
3 0.92 1.05 Boron
4 1 1.05 Boron
5 1.07 1.05 Boron
6 1.15 1.05 Boron
7 1 1.11 Boron
8 1.07 1.11 Boron
Table 1: devices variants TCAD summary – Gain vs. Dose (E)
LGAD only
JTE P31 IMPLANT
(7,27)
EPITAX growth
50 um
SUBS
6
DRIVE IN [GL implant]
SPROCESS
2D process flowchart
Oxidation Resist depo
JTE etch
Strip resist
JTE drive in Resist depo
Pstop implant
Strip resist
Resist depo
ACTIVE etch
Strip resist
Oxidation Resist depo
GL implant
Strip resist
Resist depo
SHN implant (A&JTE)
Strip resist
LTO deposit ACTIVATION
LTO DENSIFICATION
Resist depo
CONTACTS ETCHING
Strip resist
METAL DEPOSITION
Electrodes with Al WF
in SDEVICE
REMESHING FOR
DEVICE
SIMULATION
IV plots Charge collection SDEVICE
SPROCESS
• Full process simulation implemented for LGAD and PiN
TCAD simulations
4E.G.Villani, 37th RD50 Workshop, Zagreb (online)
7
GR P-stop CathodeFL#1A
• Extension of GL changes by around 0.5 um in going from
LOW DOSE to MAX DOSE
*
*
*
P-stop
Electrical simulation setup, common to PiN and LGAD, with RC
network
• IV plots
• Bulk radiation damage not included in this iteration, but
SiO2-Si interface traps implemented
• CCE for vertical MIP hit (80 e/h /um) through centre
* with no SRH Qcoll Qinj 0.1%
GRMIP
* Qcollected [LGAD]/Qcollected [PiN]
where Qcollected = Irin [0…150 ns]
Irin [A]
GR
LGAD – 50 um HR epi
TCAD simulations
5E.G.Villani, 37th RD50 Workshop, Zagreb (online)
Vbias [V]
LGAD
PiN
Gl Dose: HIGH
Gain
Gl Dose:
NOMINAL
1.7fC
8.8fC
8
4.7fC
• Gain is defined as ratio of LGAD/PIN collected charge for a 50 ns transient and normalized to 0.3,0.6,0.9 BV
• Each LGAD gain curve is normalized w.r.t. PIN that uses the same Impact Ionization model ( 3 models used: Unibo,
Okuto, Lackner)
• A gain of > 10 is predicted for 2/3 I2 models used for the MIDDLE dose
• Design based on Impact ionization models predicting lowest gain (UniBo)
Gl Dose: LOW
TCAD simulations
6
T = 300K
GR floating
E.G.Villani, 37th RD50 Workshop, Zagreb (online)
9
• Fabrication completed November 2020
• First wafer currently being tested at OPMD, Oxford and Birmingham
LGAD Project description and goals
7E.G.Villani, 37th RD50 Workshop, Zagreb (online)
10
• IV plots of 5 LGAD-4A and 5 PiN-4A devices
on WF 21 (E=1,D=1)
• GR floating, T = 21C
LGAD/PiN IV tests - 4mm
8
LGAD-4A and PiN-4A
E.G.Villani, 37th RD50 Workshop, Zagreb (online)
5-LGAD
5-PiN
GR
Wafer 21
I [A]
Vbias [V]
11
• CV plots of 5 LGAD-4A and 5 PiN-4A devices
on WF 21 (E=1,D=1)
• F = 100kHz, AC = 35 mV
• GR floating, T=21C
• GL depletion voltage 20V, Vfd 32V
LGAD/PiN CV test - 4mm
9
LGAD-4A and PiN-4A
E.G.Villani, 37th RD50 Workshop, Zagreb (online)
5-LGAD
5-PiN
Vbias [V]
1/C2 [F-2]
Wafer 21
GR
VGL
VFD
10
Summary and next steps
• A RAL, University of Oxford, University of Birmingham and Open University project in collaboration with Teledyne e2v foundry has produced the first batch of LGAD devices
• First fabricated samples available November 2020• Eight 6’’ 50um thick p-epi wafers processed with different
levels of dose and energy of implanted GL• First wafer currently being IV-CV tested at Oxford and
Birmingham. Laser dicing on it to be performed next
THANK YOU
E.G.Villani, 37th RD50 Workshop, Zagreb (online)
11
Acknowledgments
• Two years project supported by STFC, grant ref.nr ST/T002298/1
• STFC, Teledyne e2v, University of Birmingham University of Oxford, Open University
E.G.Villani, 37th RD50 Workshop, Zagreb (online)
I
backup
E.G.Villani, 37th RD50 Workshop, Zagreb (online)
Lights ON
Lights OFF
LGAD 1A
MC gain min
MC gain MAX
MC gain fluctuations -5 runs