Download 11th-RD50-Ioana_Pintilie

Cluster related hole traps with
enhanced-field-emission
- the source for long term annealing in
hadron irradiated Si diodes -*
I. Pintilie1,2, E. Fretwurst2 and G. Lindstroem2
1
National Institute of Materials Physics, Bucharest, P.O.
Box MG-7, 76900, Romania
2Institute for Experimental Physics, Hamburg University,
D-22761, Germany
* Article submitted to APL
Ioana Pintilie, 11 th RD50 Workshop,
12-14 November 2007, Geneve
1
Materials
MCz-Silicon wafers: <100>, n/P, 100 μm, 1090 cm, Nd = 3.96x1012 cm-3, CiS
process (samples 8556-02-25 and 8556-05-35)
EPI-Silicon wafers: <111>, n/P, 72 μm on 300 μm Cz-substrate, 169 cm, CiS
process
-standard Oxidation (EPI-ST) (sample 8364-02-35), Nd = 2.66x1013 cm-3,
- diffusion oxygenated for 24 h/1100°C (EPI-DO) (sample 8364-05-36) Nd =
2.48x1013 cm-3
Irradiation: - 1MeV neutrons at Ljubljana; fluences of 5x1013 cm-2
- Co60- gamma irradiation at BNL (300 Mrad dose)
Annealing : Isothermal treatments at 80 0C
Investigation methods: - I-V, C-V
- TSC
Ioana Pintilie, 11 th RD50 Workshop,
12-14 November 2007, Geneve
2
TSC results-for fully depleted diodes
EPI-DO
12
TSC signal (pA)
20
15
VO
V2+?
H(152K)
10
E(35K)
H(140K)
E(28K)
5
H(40K)
BDA
BDB
H(116K)
0
Defect concentrations cm
Forward injection at 5K, RB=150V
as irradiated
20min@80C
80min@80C
1370min@80C
34270min@80C
Electron injection at 5K
1370min@80C
-3
7x10
25
12
6x10
12
5x10
EPI-DO
H116K
H140K
H152K
all
12
4x10
12
3x10
12
2x10
12
1x10
0
1
0
20
40
60
80 100 120 140 160 180 200
Temperature (K)
10
100
1000
10000
0
Annealing time at 80 C
• H(116K) was detected previously*
• H(152K) ~ was atributed so far to CiOi
*
M. Moll, H. Feick, E. Fretwurst, G. Lindstroem, Nucl. Instrum. Meth. Phys. Res. A 409, 194 (1998)
Ioana Pintilie, 11 th RD50 Workshop,
12-14 November 2007, Geneve
3
TSC results-for fully depleted diodes
TSC signal (pA)
25
20
VO
15
V2+ ?
H(152K)
10
E(35K)
H(140K)
5
E(28K)
++/+
BD
BD
0/++
H(116K)
7x10
12
6x10
12
5x10
12
4x10
12
3x10
12
2x10
12
1x10
12
-3
F - 150 V 20min80C
F - 150 V - as irradiated
F - 150 V- 80 min@80C
F - 150V - 160min@80C
F - 150V - 320min@80C
F - 150V - 640min@80C
F - 150V - 1370min@80C
F - 150V - 2810min@80C
F - 150V - 5700min@80C
F - 11460min@80C
F - 16980 min@80C
Defect concentrations (cm )
EPI-ST
30
EPI-ST
H116K
H140K
H152K
all
0
1
0
25
50
75
100
125
150
175
200
10
100
1000
10000
0
Annealing time at 80 C
Temperature (K)
Ioana Pintilie, 11 th RD50 Workshop,
12-14 November 2007, Geneve
4
TSC results-for fully depleted diodes
MCz
TSC signal (pA)
80
H(40K)
70
60
50
H(152K)
E(35K)
40
V2+?
H(140K)
30
VO
E(28K)
H(116K)
20
+/++
BD
-/0
IO2
10
0/++
12
7x10
MCz
-3
Forward injection at 5K, RB=300V
as irradiated
20min@80C
80 min@ 80C
320min@80C
1370min@80C
34270min@80C
Defect concentrations (cm )
90
H116K
H140K
H152K
all
12
6x10
12
5x10
12
4x10
12
3x10
12
2x10
12
1x10
0
BD
1
10
100
1000
10000
0
0
0
20
40
60
80 100 120 140 160 180 200
Annealing time at 80 C
Temperature (K)
Ioana Pintilie, 11 th RD50 Workshop,
12-14 November 2007, Geneve
5
H(116K), H(140K) and H(152K)- hole
traps with enhanced field emission
13
EPI-DO irradiated with 1 MeV neutrons,  = 5x10 cm
-2
13
EPI-DO irradiated with 1 MeV neutrons,  = 5x10 cm
30
5
Forward injection at 5K
250V
200V
150V
100V
70V
0/++
BD
T = 2.5K
H(116K)
T = 5.3K
25
H(152K)
0/-
H&E traps
T = 6K
0/-
H(140K)
0/-
T = 3.5K
TSC signal (pA)
TSC signal (pA)
15
10
-2
20
Forward injection at 50K
250V
200V
150V
100V
70V
H&E traps
0/+
+/0
H(152K) +CiOi
15
T = 3.5K
10
5
0
0/++
BD
T = 3K
H(116K)
T = 5K
H(140K)
0/-
T = 3.5K
0
90 100 110 120 130 140 150 160 170 180 190
100
120
140
160
180
Temperature (K)
Temperature (K)
Filling of traps at 5 K
Filling of traps at 50 K
Ioana Pintilie, 11 th RD50 Workshop,
12-14 November 2007, Geneve
6
Point defects and cluster related defects
60
EPI-DO after 300 Mrad Co gamma irradiation
13
EPI-DO irradiated with 1 MeV neutrons,  = 5x10 cm
Forward injection at 50K
250V
200V
150V
100V
25
20
CiOi
+/0
25
15
10
BD
5
++/0
20
Forward injection at 50K
250V
200V
150V
100V
70V
0/+
+/0
T = 3.5K
10
0/++
BD
T = 3K
0
H&E traps
H(152K) +CiOi
15
5
Tpeak = 2.5 K
-2
30
Tpeak = 0K
TSC signal (pA)
TSC signal (pA)
30
H(116K)
T = 5K
H(140K)
0/-
T = 3.5K
0
80
90 100 110 120 130 140 150 160 170 180
100
Temperature (K)
120
140
160
180
Temperature (K)
All H(116K), H(140K) and H(152K) are cluster related defects
Ioana Pintilie, 11 th RD50 Workshop,
12-14 November 2007, Geneve
7
How much CiOi can overlap to H(152K)
peak ?
60
EPI-DO irradiated with Co -, 300 Mrad dose
For Tfilling< 30K no
CiOi can be filled
TSC signal (pA)
60
Forward injection at different T
Fw - 20K
Fw - 30K
Fw - 40K
Fw - 50K
Fw - 60K
Fw - 80K
50
40
CiOi
+/0

30
20
BD
10
0/++
V2
-/0
0
80
100
120
140
Temperature (K)
160
180
As long as the forward
injection is performed at
Tfill< 30K the magnitude of
H(152K) peak is not
affected by the CiOi defect
Ioana Pintilie, 11 th RD50 Workshop,
12-14 November 2007, Geneve
8
TSC spectra analysis for traps with
enhanced field emission
In general, the TSC signal resulting from a hole trap in the diode is given by:
TSC (T ) 
q  A  w  Pt  e p (T )
2
T


 1

 exp   e p (T )  dT 
T


0


where Pt is the concentration of the filled traps, ep is the emission rate of holes, w the depleted
width of the diode, A is the active area of the diode.
• This formalism is not suited for traps with enhanced electric-fieldemission, it may lead to huge errors in evaluating the defect level
position in the band gap
• An identification of the mechanism behind the enhanced-fieldemission depends on how the experimental data fits when different
potentials are accounted for the defect
Ioana Pintilie, 11 th RD50 Workshop,
12-14 November 2007, Geneve
9
Mechanisms leading to an enhancedelectric-field-emission
• Poole Frenkel effect is characteristics to coulombic wells
– where the carrier climbs over a barrier lowered by the electric field
– seen for electric fields 103 - 106 V/cm
– gives a clear indication about the charge state of the defect
• pure tunneling
– the carriers are tunneling the barrier at high energy
– possible only for electric fields > 107 V/cm
– gives no indication about the defect charge state
• phonon assisted tunneling
– the carriers absorb thermal energy from the latice and then tunnel through the
barrier at a higher energy
– seen for electric fields 104 - 106 V/cm
– gives no indication about the defect charge state
Ioana Pintilie, 11 th RD50 Workshop,
12-14 November 2007, Geneve
10
Poole Frenkel effect
V(r) = -q2/40r –qFrcos,
Ei= V(rmax)=-q(qFcos/ 0r )1/2
• For one dimensional PF effect =0 and epPF=ep0exp(-Ei/kT)
• For three dimensional PF effect 0<  < /2 and
ep
where,
PF


 e p (1 /  2 ) e  (  1)  1  1 / 2
0
  (qF / 0 r )1/ 2 q / kBT
• for F has to be considered the local electric field in the p+n diode:
F(x) =
F(x) =
qN eff (d  x) /  0 r  (V  Vd ) / d
qN eff ( w  x) /  0 r
Ioana Pintilie, 11 th RD50 Workshop,
12-14 November 2007, Geneve
if V>Vd
if V<Vd
11
TSC spectra analysis for traps with
enhanced field emission
For Coulombic wells the emission rate depends on the local electric field and
the corresponding TSC peak should be calculated according to:
q  A  Pt
TSC PF (T ) 
2

w PF

e p ( x, T )  exp 
0

1
T



T0



e PF
(
x
,
T
)

dT
dx
p
Ioana Pintilie, 11 th RD50 Workshop,
12-14 November 2007, Geneve


12
Results for the fit of TSC peaks with 3D-PF
13
EPI-DO irradiated with 1 MeV neutrons,  = 5x10 cm
-2
TSC signal (pA)
15
250V
200V
150V
100V
70V
3D-PF, 250 V
3D-PF, 200V
3D-PF, 150V
3D-PF, 100V
3D-PF, 70V
10
5
0/++
BD
T = 2.5K
H(116K)
The 3D-PF effect formalism
describes the experiments

Forward injection at 5K
H(152K)
0/-
H&E traps
T = 6K
The H(116K), H(140K) and
H(152K) centers have acceptor
levels in the lower part of the gap

0/-
T = 5.3K
H(140K)
0/-
T = 3.5K
0
90 100 110 120 130 140 150 160 170 180 190
Temperature (K)
The H(116K), H(140K) and
H(152K) centers contribute
with their full concentration
as negative space charge to
Neff0 in the initially n-type
silicon diodes
Ioana Pintilie, 11 th RD50 Workshop,
12-14 November 2007, Geneve
13
The levels position in the bandgap
The parameters for the zero field emission rates
describing the experimental results are:
Ei116K = Ev + 0.33eV (0.285eV*)
p116K =410-14 (410-15*)cm2
Ei140K
= Ev + 0.36eV and p
Ec
and
140K =2.510-15
P0/+
VO-/0
V2-/0
Eg/2
cm2
C O +/0
i
/H152K0
Ei152K = Ev + 0.42eV (0.36eV*) and
p152K =2.310-14 (210-15*) cm2
H140K0
i
/H116K0
Ev
*
- aparent ionization energy and capture cross sections when no field dependence of
emission rates is accounted
Ioana Pintilie, 11 th RD50 Workshop,
12-14 November 2007, Geneve
/-
14
-3
Concentration (cm )
The impact of H(116K), H(140K) and H(152K)
defects on Neff
3.5x10
13
3.0x10
13
2.5x10
13
2.0x10
13
1.5x10
13
1.0x10
13
5.0x10
12
EPI-DO
Neff at RT
Nd - [H116K] - [H140K] - [H152K]
EPI-ST
Neff at RT
Nd-[H116K]-[H140K]-H[152K]
These defects are
responsible for the
long term
annealing of Neff
MCz
Neff at RT
Nd - [H116K] - [H140K] - [H152K]
0.0
-5.0x10
12
1
10
100
1000
10000
0
Annealing time at 80 C (minutes)
Ioana Pintilie, 11 th RD50 Workshop,
12-14 November 2007, Geneve
15
Summary
•Three cluster related acceptors in the lower part of the
gap have been detected by TSC experiments.
•Their zero field activation enthalpies resulted from
modelling the TSC peaks with the 3D-Poole Frenkel
model and taking into account the spatial distribution of
the diode electric field.
•It is shown that these acceptors are responsible for the
long term annealing of Neff in hadron irradiated silicon
diodes.
Ioana Pintilie, 11 th RD50 Workshop,
12-14 November 2007, Geneve
16