Thin-Film Semiconductor Technology Applied to
Large Area Radiation Detectors
CIRMS Conference 2012
October 23, 2012
Bruce Gnade
UT Dallas Flexible Electronics Research Group at UTD
•
POST‐DOCS/STAFF SCIENTISTS
– Dr. I. Mejia, Dr. Kurtis Cantley, Dr. M. Jia, Dr. I. Trachtenberg, Dr. A. Carrillo, Dr. N. Hernandez, Dr. J. Conde, Dr. Dick Chapman •
GRADUATE STUDENTS/PROJECTS
–
–
–
–
–
–
–
–
•
Ana Salas– PhD. – Alternate Contacts (AFOSR)
Duo Mao – Ph.D. – Organic Memory (ARL)
Mike Perez, Ph.D. – n‐type flexible TFTs (NSF)
John Murphy Ph.D – Large Area Neutron Detectors (ARL, FUSION)
Martha Rivas – Pulsed Laser Deposition of Chalcogenides
Dewan Lutful Kabir – Reliability and Electrical Characterization
Lindsey Smith – Neutron converter layers (DNDO)
Kevin Larosa – Backplane electronics (DNDO)
VISITING SCHOLARS
– J. Ramos (UACJ), V. Martinez, (CIMAV), G. Gutierrez (CIMAV), Alfredo Luque (CNyN)
• Collaborators
– Dr. David Allee (ASU), Dr. Eric Forsythe (ARL), George Kunnen (ASU)
•
FUNDING
– U.S. Army Research Labs,Military Tech, DOE, UT Dallas, Texas MicroPower, NSF, Texas instruments, FUSION, DNDO, AFOSR, DARPA
Agenda
•
•
Large Area Radiation Detectors
– Why thin‐film devices
– Large area neutron detector project
Current state of thin‐film semiconductor devices
– Transistors
– Memory
– CMOS
– Circuits
Why large area detectors?
Large area neutron detector
2” 3He tube
Neutron Source
8 ft
0
8 ft
Probability of neutron hitting 2” tube ~ 1 part in 36000
Probability of neutron hitting large area detector ~ 1 part in 3
Future Flexible Electronics
Throughput vs cost/cm2
Throughput (m2/sec)
100
1
Displays
Source: Organic and Printed Electronics, Organic Electronics Association (OE‐A) Broucher,
3rd Edition, 2009 Digital X‐ray imagers
10‐4
10
1
100
500
Minimum feature Size (m)
# of TFTs
per min
Area per
min
~108/min
4m2/min
Yield zero
redundancies
99.99997%
Display
~cost/cm2
$0.05/cm2
Si-CMOS
($5k/wafer)
$6.86/cm2
Thanks to Dr. Eric Forsythe – Army Research Laboratory
Digital X-ray
1-up/subs
Digital x-ray
4-up/subs
$29/cm2
$7/cm2
Large Area Sensor Arrays: The Concept with ASU
ROW
RESET
• Neutron Detection
•VIS / IR Imager
• Millimeter Wave
• MEMs
•
•
Blast
Acoustic
• Electronic Textiles
AMPLIFIER
PIXEL
READ
Energy to charge conversion
Very large area, flexible, low power neutron detectors
Three main pieces
‐ energy conversion layer
‐ sensing diode
‐ backplane electronics
Neutron conversion layer strategies
Isotope
Reaction
3He
3He(n,p)3H
6Li
6Li(n,α)3H
10B
10B(n,α)7Li
natGd
natGd(n,γ)
157Gd
157Gd(n,γ)
235U
235U(n,f)
Converter‐on‐diode
6Li 10B nanoparticle composite
Thermal σ
(barns)
5333
940
3835
49700
259000
681
Charged particles and energies (keV)
p: 573, 3H: 191
3H: 2727, α: 2055
α: 1472, 7Li: 480
Conversion e‐: 29‐191
Conversion e‐: 29‐182
Various fission products
Converter‐in‐diode
6Li
nanoparticles 10B
Au (100nm)
Au (100 nm)
Sensing diode
Sensing diode
Al (100 nm)
Al (100 nm)
Flexible substrate (~1 mm)
Flexible substrate (~1 mm)
Neutron Detector Modeling in MCNPX and MATLAB
Detector Layers simulated in MCNPX:
Weapons Grade Plutonium
 Example neutron conversion layer
optimizations result in 2.6um 10B thickness
•
•
•
•
Front end captures Neutron and emits charged alpha particle,
Alpha detected with PIN diode and active pixel sensor circuit
MCNPX used to model and optimize detector front‐end layers
Detector layers capable of being fabricated with FDC process
MATLAB model of arrayed pixel sensors with simulated alpha particle exposure (Red Boxes):
With Correlated Double
Sampling applied:
• Each cell represents an active pixel sensor
• Sensors modeled to reflect FDC process variations.
• CDS dramatically improves ability for detector to resolve alpha particle strikes
Thin‐film diode optimization ‐ MCNPX
TABLE II. Intrinsic gamma efficiencies for selected thicknesses and gamma
energies. (LLD = 300keV and a 2.8 µm 10B conversion layer)
Thickness
(μm)
γ energy
(keV)
Si
CdTe
GaAs
ZnO
3
511
1460
0
0
0
9×10-14
0
0
0
8×10-14
10
511
1460
0
1×10-13
6×10-13
9×10-12
0
2×10-12
1×10-12
8×10-12
30
511
1460
6×10-11
1×10-10
6×10-7
2×10-7
1×10-7
6×10-8
3×10-7
2×10-7
100
511
1460
3×10-6
2×10-6
4×10-4
1×10-4
2×10-4
6×10-5
2×10-4
9×10-5
300
511
1460
3×10-4
2×10-4
4×10-3
3×10-3
2×10-3
2×10-3
2×10-3
2×10-3
Sensing Diode Optimization: Schottky vs. MIS
Integrating over channels for peak
Detector bias (V)
Particles detected
0V
5V
20V
562 595 651 Measured source strength
1,938 1,954 2,245 Efficiency
60%
60%
69%
Integrating over channels 100‐450
Detector bias (V)
0V
5V
20V
100V
Particles detected
785 772 752 869 Measured source strength
2,486 2,445 2,382 2,752 Efficiency
77%
76%
74%
85%
Radiation measurement setup
Preamp
Oscilloscope
Amp and det bias
MCA
Sample chamber
TFT‐based Active Pixel Sensors The detection of two alpha particle strikes on a commercial diode and amplified with the two stage low noise thin film transistor amplifier.
Flexible CMOS
NMOS
PMOS
Why CMOS
– Dramatically reduced power consumption
– Analog circuitry possible
–Sensor applications
12
VGS = -20V
W/L = 400m/7m
10
IDS (
8
VGS = -15V
6
4
c
VGS = -10V
c
2
VGS = -5V
How to do F‐CMOS
0
0
-5
-10
-15
-20
-25
-30
VDS (Volts)
102
–Combine Processes
5x10-3
W/L = 400m/7m
VDS = -25V
4x10-3
100
3x10-3
10-1
2x10-3
10-2
1x10-3
10-3
10-4
0
10
0
-10
VGS (Volts)
-20
-30
Sqrt (IDS) (A)1/2
log(| IDS |) (A)
101
• N‐Type a‐Si:H TFTs
• N‐Type inorganic TFTs
• P‐Type Organic TFTs
–Standard Cell Approach
a‐Si / pentacene CMOS Devices – with ASU
Inverters
NAND Gates
12
25
B
20
15
Input (V)
Input (V)
20
15
10
5
8
pmos
Vin
Vout
4
5
nmos
GND
0
0
0
5
10
VIN (V)
15
20
B
15
10
5
A
0
0
25
25
NOR Gate Output (V)
VDD
10
NAND Gate Output (V)
A
Gain
VOUT(V)
25
16
20
NOR Gates
20
15
10
5
0
20
15
10
5
0
-0.02
-0.01
0.00
0.01
Time (seconds)
0.02
-0.02
-0.01
0.00
0.01
Time (seconds)
0.02
Technology Comparison
Chalcogenides
250
650
ZnSe
600
SnSe2
27
InS
50
SnS2
18
10 21
N
10 20
N
P
N
N
10 19
N
P
P
P
P
10 18
10 17
P
P
P
N
N
P
N
10 16
10 15
P-TYPE
N
P
P
P
N
N
N
10 14
CuGaS2
20
CuInS2
15
ZnTe
100
PbS
5
Cu2S
10
Zn
O
C
dS
C
dT
e
Zn
S
Zn
Se
C
dS
e
Zn
C T
uI e
nS
C e2
uI
C nS
uI 2
n
C Te
uA 2
C lSe
uG 2
aS
e2
CdSe
Concentration (cm-3)
CdS
N-TYPE
Electron or Hole
Semiconductor
Mobility (cm2/V‐s)
e‐
h+
10 22
Organic semiconductors ~ 6x10−2 to 1 cm2V-1 s-1
Amorphous silicon: ~ 0.01 to 1 cm2 V-1 s-1
Inorganic semiconductors: 100 cm2 V-1 s-1
CBD CdS/ Evaporated Pentacene Inverters
VI
VOUT
N
100
20
V O U T
G A IN
• Optimized pentacene and CBD CdS deposition
80
15
10
40
5
20
0
0
0
5
10
V IN [V ]
15
20
G A IN
V O U T [V ]
60
• Inverters yield gains of about 80
Ferroelectric‐based organic Memory (FeRAM)
PVDF ‐ TrFe
P
MFM Capacitor – 1T‐1C memory cell
‐ relatively easy to fabricate
‐ destructive read
‐ complicated read / write circuitry
HH
HH
C
C
C
C
F F
F F
Au
Ferro‐
electric
Gate – Au
Fully integrated TFT used for 1T1C
1.
2.
3.
4.
CBD CdS as the semiconductor
ALD HfO2 as the gate dielectric
Parylene as ILD
Low temperature process, maximum temperature is 100 oC
1E-3
120.0µ
0.016
VG = 0 V
VG = 2.5 V
VG = 5 V
VG = 7.5 V
VG = 10 V
VG = 12.5 V
VG = 15 V
100.0µ
SQRT ID
ABS ID
0.014
1E-5
0.012
1/2
60.0µ
W = 160 um
L = 5 um
40.0µ
1E-6
W = 160 um
L = 5 um
0.010
1E-7
0.008
FET = 1.5 cm /V-s
1E-8
0.006
VT = 5.5 V
1E-9
2
0.004
20.0µ
0.0
0
2
4
6
8
VD [V]
10
12
14
ID [A]
SQRT ID [A ]
80.0µ
ID [A]
1E-4
1E-10
0.002
Saturation
0.000
VD = 15 V
-5
0
5
VG [V]
10
1E-11
1E-12
15
2T2C FRAM Characterization BAL
DWL
WE
DWL
SL
GND
SR
CL0
CL1
PL
WL0
CL2
CL3
WL1
WL2
CL4
CL5
WL3
WL4
CL6
WL5
CL7 GND WL7 WL6
64‐bit FRAM layout
20
In‐pixel charge amplifier – Ramirez design
Prototype neutron detector
• The output signal from the sensor element is a low voltage pulse with 10’s ns width, with very high impedance due to the capacitive nature of the device.
• Charge amplifiers provide very high input impedance ‐ they integrate weak charge pulses to convert them into voltage pulses with low impedance.
[1] F. J. Ramirez‐Jimenez et al., Nuclear Instruments and Methods in Physics Research A, 497, (2003), 577‐583.
[2] Bernd J. Pichler et al., IEEE Transactions on Nuclear Science, 48(6), (2001), 2370‐2374.
[3] HAMAMATSU, model H4083.
Potential Advantages
•
•
•
•
•
Very large aperture at moderate cost based on AMLCD In‐pixel electronics provide low capacitance for pixelated array
Potential for high γ‐ray rejection rate due to TFT array electronics
Potential for fission neutron multiplicity measurements Potential for determination of directionality of the neutron source
Thank You!