Lawrence Livermore National Laboratory
Rebecca Nikolic
A.Conway1, C. Britton3, C. Cheung2, M. Dar2, M.N. Ericson3, L. Fabris3, R. Radev1, Q. Shao1, L. Voss1, T.F. Wang1
LLNL1, Univ. of Neb.2,ORNL3
This work performed under the auspices of the U.S. Department of Energy by
Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344, LLNL-PRES-474261
Si Based Pillar Structured Thermal Neutron Detectors
Funded by
DHS/DNDONebraskaUniversity of
Lincoln
[email protected] IAEA March 23, 2011
2Lawrence Livermore National Laboratory
0 500 1000 1500 200010
-1
100
101
102
103
104
105
Co
un
ts
Energy (keV)
Neutron 3V
Gamma 3V
No source 3V
We are developing high-performance neutron
detectors
252Cf
DUT
252Cf
DUT
252Cf
DUT
137Cs
20 % recent result (26 µm pillar)
105 neutron/gamma discrimination
Applied Physics Letters, Vol. 93, Issue 13, Sept. 2008, p. 133502
50µm Pillar, > 90% fill
Detector noise
Neutron events
200 keV discriminator
3
Lawrence Livermore National Laboratory
Thermal neutron detectors are needed for the detection of SNM: 3He tube is current standard
10Boron is interspersed between silicon pillars to detect thermal neutrons
10Boron chosen for high thermal neutron cross-section 3837 barns and process compatibility
4Lawrence Livermore National Laboratory
Comparison: Thermal Neutron Converter Materials
10B+n → 7Li +
10 100 1000
0.0
0.2
0.4
0.6
0.8
1.0 10B
10B
5C
B
B5C
6LiF
Pro
ba
bili
ty o
f N
eu
tron
In
tera
ctio
n
Thickness ( m)
ηint = 1-exp(-t/mfp)
1st Excited State Ground State
Energy (MeV) 0.84 Li 1.47 α 1.02 Li 1.78 α
Distance in 10B (µm) 1.85 3.6 2.0 4.4
Distance in Si (µm) 2.4 5.2 2.8 6.4
Pillar height determined by
probability of n interaction
Pillar spacing determined by
& Li travel ranges
Mean free path = 1/N [ m]
N = number density, nuclei/volume
= microscopic cross-section
5
Lawrence Livermore National Laboratory
Pillar Detector enables
high efficiency neutron detection
Standard 2D Design LLNL 3D Design
Thermal
neutrons
Si
10B
Electron-hole pairs
Charged
particles
pitch
+ -
+ -
+ -
( , 7Li)
4 m
50 m
10B efficiently produces particles
Most /7Li do not reach the detector
Limited efficiency: 2-5 % (@ thermal)
Simulations show 3-D structure will
increase efficiency towards 50+%
Device geometry:
etch depth (50 m) no capture
pitch (4 m) alpha particle range
+ - Electron-hole pairs
De
tec
tor
C
on
ve
rte
r
Thermal neutrons
10Boron
Charged particles( , 7Li)
( , 7Li)
3 m
6
Lawrence Livermore National Laboratory
Pillar Detector Fabrication
1. Pattern pr mask 2. Deep RIE etching 3. Conformal Boron
Deposition
boronMulti-step Process
Si
photoresist
4. Plasma etching
of boron5. Deposit contacts
electrodes
7
Lawrence Livermore National Laboratory
OUO
OUO
Inductive-Coupled Plasma (ICP)
Control ion density and RF power independently
1. Less damages to etched materials; 2. Flexibility in mask materials;
3. High etch rate; 4. Low substrate temperature
RF biased wafer
chuck
RF input
1000 W
N S
Magnet Magnet
RF waveguide
Plasma
H2, Ar2, CH4, Cl2,SF 6 Gas
Electron Cyclotron Resonance (ECR)
SF6 & C4F8 Gas
High-density plasma processing for
high aspect ratio structures
8
Lawrence Livermore National Laboratory
OUO
OUO
“Pillar” Chips for 10B R&D and Detector Fabrication
ICP Etch
Pillar Litho well-developed
Large area pillar litho developed
Developing negative pr process for large area
50 µm pillar process developed
Compatible with 2’’ and 4’’ wafers
50µm Pillar Chips, 2µm Width
nLOF 2020
Vertical
sidewalls
Vertical CVD system provides improved film reproducibility and safety.
Vertical vs. Horizontal CVD System for 10B Deposition
Horizontal CVD System
Features
Vertical CVD System
Pump
Substrate
heater
• Better control of precursor and sampleheating
• Quick sample loading• Large conductance → Effective pump
rate & lower operation pressure• Tailor for scale up with larger wafers• Low chemical exposure
• Inexpensive set up• Easy cleaning of quartz tube chamber• Limited control of precursor and
heating profiles for samples• Long sample loading time• Potential for chemical exposure
Pump
Mass flow
Controller:
B10H14
B10H14
source
CVD
Furnace
Features
Heat shielding improves heat profile and boron filling.
Activating New Vertical CVD System
Vertical CVD system50- m silicon pillar samples
before boron coatingSubstrate
heater
Side wall
shielding
Top side shielding
Sample location
Shower head
High fill factor achieved
First device runs had low
efficiency
Optimizing vertical process in
parallel with 1’’ CVD tube
11
Lawrence Livermore National Laboratory
We have developed specialized processing and characterization
tools for 10B
Raman Spectroscopy
Stress induced on silicon pillars
calculated from Raman shift
1E+14
1E+15
1E+16
1E+17
1E+18
1E+19
1E+20
1E+21
1E+22
1E+23
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Depth (microns)
B C
on
ce
ntr
atio
n
1E+02
1E+03
1E+04
1E+05
1E+06
1E+07
1E+08
1E+09
Co
un
ts P
er
Se
co
nd
Si (sec. ion counts)->
B
C06M1250B05.SWF
LLNL, Sample 228
9/12/2006
Cs Evans Analytical Group
Boron CVD 600 C – 900 C
SIMS:No B diffusion Si/B interface
Boron CVD and Etch Back
A) Pillar Platform 1:25 aspect ratio
B) Conformal boron filling
C) Rapid removal of boron with
high B:Si selectivity
Stress in 10B Coating
4 µm
12
Lawrence Livermore National Laboratory
0%
10%
20%
30%
40%
50%
60%
70%
80%
0 20 40 60 80
Pillar Height ( m)
Ne
utr
on
Eff
icie
ncy (
%)
1.0E+04
1.0E+05
1.0E+06
Ne
utr
on
to
Gam
ma
Dis
cri
min
ati
on
Neutron Effeciency
200 keV
662 keV
1 MeV
104
105
106
Simulation Tool-box
• MCNP Neutron history
•TRIM alpha, lithium trajectory
• Silvaco semiconductor device physics
Efficiency by simulations and measurement
Simulations Measurements
Applied Physics Letters, vol. 93, issue 13, Sept. 2008, p. 133502
0 500 1000 1500 200010
-1
100
101
102
103
104
105
Co
un
ts
Energy (keV)
Neutron 3V
Gamma 3V
No source 3V
20 % recent result (26 µm pillar)
105 neutron/gamma discrimination
13
Lawrence Livermore National Laboratory
We are increasing detector area and channel number
Interposer board 4 mm x 4 mm detectors
Scaling to Larger Area
Excellent uniformity
across 9, 4 mm x 4 mm detectors
Jleak: 10-3 A/cm2
1.E-08
1.E-07
1.E-06
1.E-05
1.E-04
1.E-03
1.E-02
1.E-01
1.E+00
-10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3
J (A
/cm
2 )
V (V)
4mm x 4mm
5mm x 5mm
10mm x 10mm
14
Thermal neutron detection efficiencies measured at LLNL
Part # Efficiency
EPI 26-1-1* 22
EPI 26-1-4 23
EPI 26-7-1 ~20
EPI 26-6-3 16.4
EPI 26-7-2 16.1
EPI 26-7-3 19.2
EPI 26-7-4 17.8
EPI 26-1-2 ~20
EPI 26-2-3 ~20
EPI 26-2-4 ~20
Process for measuring thermal
neutron efficiency:
MCNP of source and lab for # of incident
thermal neutrons
Eff = Measured counts / simulated incident
counts
Measure “calibration” detector using poly
moderated 252Cf source
Other detectors:
• Compare to calibration detector to
determine relative efficiency
All detectors from “EPI-26” family
1’’ CVD system
*Calibration detector
Packaged LLNL Pillar Detector
TO can
2 mm x 2 mm detectors
15
Lawrence Livermore National Laboratory
Characterization of 9 Element Detector Array
Detector size 2 mm x 2 mm
Moderated 252Cf sources
1.5 m distance
Detector unbiased
Measurement time 3000 sec.
Total counts Channels 1-9 = 16k
14 counts/sec/cm2
1 2 3
3
54 6
7 8 9
Element 1
Element 4
Element 7
Measurements with
rack-mount electronics
16
9 channel custom circuit with discrete components
Custom design and layout with
discrete COTS components
Integrates 9 detectors with individual
readout chain
• Packaged individually
• Or, on interposer board
Readout chain includes:
• Charge sensitive preamp
• Shaping amplifier, 250 ns
shaping time
• Low level discriminator
• Driver circuit for TTL output
• LED display and driver circuits
• Can readout either electron or
hole signal
CSP Shaper LLD
Output to
LED display
Individually
packaged
detectors
18
Lawrence Livermore National Laboratory
Example for single channel, 8 channel was implemented
Read Out Design: ASIC
May be a full detector
or a sub-area depending
on the size of the pixel
capacitance and leakage
To Counter/Multiplicity
Logic
Pillar
DetectorShaper
Leading Edge
Discriminator
Charge Integrating
Front-End
t0 t0n
LVDS
Logic Stage
Single ASIC Channel
ASIC
2.8 mm x 4.0 mm
8 Channels per ASIC
Mosis Foundry
20 ASICs received Dec. 2010
Differential signaling
19
Silicon Detector for HLC: High radiation tolerance possible
NIEL - nonionizing energy loss
IEL – ionizing energy loss
1013 5 1014 5 1015 5 1016
eq [cm-2]
5000
10000
15000
20000
25000
sig
nal
[el
ectr
on
s]
n-in-n (FZ), 285 m, 600V, 23 GeV p
p-in-n (FZ), 300 m, 500V, 23GeV p
p-in-n (FZ), 300 m, 500V, neutrons
p-in-n-FZ (500V)n-in-n FZ (600V)
M.Moll - 08/2008
References:
[1] p/n-FZ, 300 m, (-30oC, 25ns), strip [Casse 2008][2] n/n-FZ, 285 m, (-10oC, 40ns), pixel [Rohe et al. 2005]
FZ Silicon Strip and Pixel Sensors
strip sensorspixel sensors
*Silicon FZ Strip Detectors for High Luminosity Colliders
Radiation Damage to Sensors:
Bulk damage due to NIEL
Change of effective doping concentration
Increase of leakage current
Increase of charge carrier trapping
Surface damage due to IEL
(accumulation of positive charge in oxide
& interface charges)
Determine if Pillar Detector has elevated damage due to:
large surface area
composite material structure
20
High fluence measurements: gamma
15 cm
Pillar Detector
137Cs
Test setup Electrical measurements done before and after
irradiation
Time, 60ksec (16 hours)
Gamma flux: 2.64×109 photons/(cm2·sec)
Total gamma fluence delivered: 2.0×1014 photons/cm2
Electrical properties of I-V, C-V and transient voltage
decay remain same:
I-V remains the same
No trapping centers introduced
No damage in the depletion
region
Transient voltage decay remains the same
No change in effect lifetime: ~ 1.0 µs
No damage in bulk or pillar sidewalls
C-V remains the same
No change in voltage pulse for a
certain amount of induced charges
Neutron detection performance not
degraded
1.E-07
1.E-05
1.E-03
1.E-01
1.E+01
-10 -8 -6 -4 -2 0 2
J (A
/cm
2)
V (V)
before 60k sec irradiation
after 60k sec irradiation
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
0.E+00 5.E-05 1.E-04
V (V
)
Time (sec)
before irradiation
60k sec irradiation
1.E-11
1.E-10
1.E-09
-10 -5 0
Cap
acit
ance
(F)
V (V)
before 60k sec irradiation
after 60k sec irradiation
21
High fluence measurements: fast neutron
Electrical measurements done before and after
irradiation
Time, 60ksec (16 hours)
Neutron flux: 3.2×103 n/(cm2·sec)
Total neutron fluence delivered: 1.9×108 n/cm2
Electrical properties of I-V, C-V and transient voltage
decay remain same
1.5 m
Pillar Detector
252Cf
Test setup
I-V remains the same
No trapping centers introduced
No damage in the depletion
region
Transient voltage decay remains the same
No change in effect lifetime: ~ 1.0 µs
No damage in bulk or pillar sidewalls
C-V remains the same
No change in voltage pulse for a
certain amount of induced charges
Neutron detection performance not
degraded
1.E-06
1.E-05
1.E-04
1.E-03
1.E-02
1.E-01
1.E+00
1.E+01
1.E+02
-10 -8 -6 -4 -2 0 2 4
J (A
/cm
2)
V (V)
before irradiation
after 60k sec irradiation
0
0.1
0.2
0.3
0.4
0.5
0.6
0.E+00 5.E-05 1.E-04
V (V
)
Time (sec)
before irradiation
after 60k sec irradiation
1.E-11
1.E-10
1.E-09
-10 -8 -6 -4 -2 0
Cap
acit
ance
(F)
V (V)
before irradiation
after 60k sec irradiation
22
Lawrence Livermore National Laboratory
High fluence measurements summary
Thermal neutron counts monitored before & after irradiation testing
Thermal Neutron Counts Set-up
• Moderated 252Cf source 4x108 n/sec
• 30 cm diameter D2O covered with 1 mm Cd
•Time 60k sec (16 hours)
•0.5 micro-sec shaping time
• Detector bias = 0V
Gamma (#/cm2s)
109
Gamma and Neutron Radiation
Neutron (#/cm2s)
103
No reduction in n counts
23
Lawrence Livermore National Laboratory
3He detector can be used for efficiency
calibration
He-3 detector used:
• Cylindrical 3He from LND, Inc.
• Active volume
1.55 cm diameter, 2.39 cm long
• 4 Atm 3He, 850V
• Data sheet efficiency ~30% thermal
neutrons
Initial measurements performed
Efficiency calculations underway
Measurement of He-3 detector for comparison with Pillars
Conditions Counts
No Source 1393
Bare 252Cf 622413
Moderated* 252Cf 887504
2.0 m
3He Tube252Cf
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
0 50 100 150 200 250 300 350
LOG
(nu
mb
er o
f co
un
ts)
Channel number
*15cm D2O sphere
Background
Bare 252Cf
Moderated 252Cf
24
Lawrence Livermore National Laboratory
3He vs Pillar Detector
Detector 3He tube*
(LND-25177)
Pillar detector
(EPI26-1-1)
Active dimensions
(cm)
Diameter:1.55
Length:2.390.2 x 0.2
Operation voltage (V) 850 0
Total counts (1 hour) 886,111 3067
CPS/area
(#/cm2·s)66.44 21.30
*Gas pressure: 3040 torr
3He
Pillar detector
2m
2m 252Cf
D2O moderator
0
20
40
60
80
100
120
140
0 20 40 60 80 100 120 140
Nu
mb
er o
f co
un
ts
Channel Number
0 DEG
45 DEG
90 DEG
180 DEG
Isotropic source
No angular dependence (as shown in measurements)
Non-Isotropic source
Little streaming 4.6 ° solid angle to deflect out Si
Isotropic source
Pillar Detector
25
Lawrence Livermore National Laboratory
OUO
OUO
Mapping Pillar Detector Over 3He Tube Form:
Efficiency Comparison
Configurations Pillar EfficiencyNumber of
detectors
Total counts
per second
Sensitivity
normalized to 3He
1Single side 22 % 7 x 11 (77) 65.45 0.27
Single side 44 % 7 x 11 (77) 130.9 0.53
2
Single side 22 % 12 x 11 x 2 (264) 224.4 0.91
Single side 44 %12 x 11 x 2 (264)
449 1.82
Comparison of 3He and pillar detector(246.14 cps for 3He)
L
D
Config. 1 Config. 2
D
LD=1.55 cm
L=2.39 cm
D=1.55 cm
L=2.39 cm
Pillar efficiency can increase by lining multiple layers within the form factor
26
Lawrence Livermore National Laboratory
Summary
Pillar based neutron detectors can be a 3He replacement
(small form factor applications)
Ideal material system is Si (electron/hole collection) and 10B
(neutron conversion)
High aspect ratio structures of 2 m x 2 m pillars, with a 4 m
pitch and scaled to 50 m can achieve 50 % thermal neutron
detection efficiency, currently at 20 %
8-channel ASIC read-out integration underway
Scaling for efficiency and area
Output from
ASIC
ASICDetector
Interface
Discriminator
Threshold Control
ASIC Power