Download TCAD device simulation of different pixel geometry

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts
no text concepts found
Transcript
Characteristics of fabrication
processes, pixel designs and
simulations
A. Dorokhov
09/05/2017
1
9/05/2017
A. Dorokhov, IPHC, Strasbourg, France
Monolithic active pixel sensors
substrate/EPI
substrate or/and
epitaxial layer (p-type)
charge collecting
electrode (s) (nwell)
particle
chip is fabricated in
CMOS technology:
readout electronics
fabricated next to charge
collecting electrode at
the same substrate ->
9/05/2017
choice of proper
fabrication process is
important
A. Dorokhov, IPHC, Strasbourg, France
2
Fabrication of MAPS in CMOS process
Low resistive
substrate/epitaxial
layer (<10 Ohm cm): not
easy (or not possible) to
deplete detector volume
MAPS development :
AMSxx, XFABxx technology
High resistive
substrate/epitaxial
layer (~>1 kOhm cm),
easier to deplete
detector volume
Twin/Double well process:
one can use simple
circuitry in pixel-> only
NMOS transistors are
allowed
first MAPS detector for particle
experiment (STAR at RICH), EUDET :
AMS technology
Triple/quadruple well
process -> can use
advantage of NMOS
and PMOS in the pixel
volume
ALICE tracker upgrade,
CBM tracker: TOWER
technology
In order to help understanding of operation, get new ideas,
verify the performances of the MAPS detector simulation of
the sensing elements and readout circuit is needed:
device simulation with TCAD and Spectre simulation
(Cadence)
9/05/2017
A. Dorokhov, IPHC, Strasbourg, France
3
TCAD device simulation of pixel
Pwell
Nwell
layout geometry of
one pixel
MIMOSIS0 chip,
transistors are
inside Pwell
18um EPI of 10 Ohm cm (low resistivity)
Bias=1V
Bias=40V
well exclusion
area
due to symmetry
simulate only one
quarter of
rectangular pixel
Depletion zone comparable, at 1 V and
40V of bias for different substrates
Bias=1V
Bias=20V
Bias=40V
TOWER 18um EPI
depletion
9/05/2017
A. Dorokhov, IPHC, Strasbourg, France
4
TCAD device simulation different EPI
Bias=1V
Bias=20V
Bias=40V
TOWER 18um EPI
TOWER 25um EPI
TOWER 30um EPI
5
9/05/2017
A. Dorokhov, IPHC, Strasbourg, France
Results of TCAD device simulation
6
9/05/2017
A. Dorokhov, IPHC, Strasbourg, France
Results of TCAD device simulation
7
9/05/2017
A. Dorokhov, IPHC, Strasbourg, France
Results of TCAD device simulation
8
9/05/2017
A. Dorokhov, IPHC, Strasbourg, France
Results of TCAD device simulation
9
9/05/2017
A. Dorokhov, IPHC, Strasbourg, France
Conclusions for presented example of comparison
high resistive substrate and low resistive
Low resistive substrate:
1. Almost no depletion
2. Leakage current is smaller
High resistive substrate:
1. Helps a lot to deplete pixel
volume
2. Leakage current is larger
3 .capacitance almost not changed
4 . going from pixel pitch 26.88x30.24 to 22x22 does not
change much leakage, capacitance and depletion volume,
because we do not have full depletion, only the depletion
fraction is much larger (x ~2)
10
9/05/2017
A. Dorokhov, IPHC, Strasbourg, France
TCAD device simulation of different pixel geometry
40um of 600 Ohm cm Si, 22x22um^2 pitch
surface of Nwells and Pwell exclusions are independent on shape (circular, octagonal or square)
circular Nwell, circular PWell exclusion
octagonal Nwell, circular PWell exclusion
circular Nwell, square PWell exclusion
octagonal Nwell, square PWell exclusion
11
9/05/2017
A. Dorokhov, IPHC, Strasbourg, France
Results of TCAD device simulation
12
9/05/2017
A. Dorokhov, IPHC, Strasbourg, France
Results of TCAD device simulation
13
9/05/2017
A. Dorokhov, IPHC, Strasbourg, France
Results of TCAD device simulation
14
9/05/2017
A. Dorokhov, IPHC, Strasbourg, France
Results of TCAD device simulation
15
9/05/2017
A. Dorokhov, IPHC, Strasbourg, France
Conclusions for presented example of comparison
different Nwell and Pwell exclusion shapes
1. The shape of charge collecting diode has marginal influence<few %
2. The shape of Pwell exclusion has small influence (<10%)
16
9/05/2017
A. Dorokhov, IPHC, Strasbourg, France