Slide 1 - MOS-AK

Download Report

Transcript Slide 1 - MOS-AK

Towards Predictable Compact Model Descriptions
for Organic Thin-Film Transistors
S. Mijalković, D. Green, A. Nejim
Silvaco Technology Centre, St Ives, Cambridgeshire, UK
A. Rankov, E. Smith, T. Kugler, C. Newsome, J. Halls
Cambridge Display Technologies, Godmanchester, UK
MOS-AK / GSA Workshop, 3 April 2009, IHP, Frankfurt / Oder
Organic Electronics:
“Harry Potter’’ and “Star Trek” Technologies are Coming your Way!
o
o
Fast development of organic electronics is supported by applications that require low cost electronic
circuits covering mechanically flexible large areas.
These include e-skin, e-paper, e-nose, smart-fabrics, flexible displays, printed electronics or radio
frequency identification tags (RFID).
Flexible display
Organic Solar Cells
e-skin
e-nose
Smart fabrics
Organic Solar Cells
o
o
Organic electronics can be fabricated using faster and cheaper low temperature processes.
Basis of the future organic electronic circuits are organic TFTs (OTFTs).
MOS-AK / GSA Workshop, 3 April 2009, IHP, Frankfurt / Oder
Organic Electronics:
Forecast and Opportunities
o The organic electronics will be a $30 billion
business in 2015 mainly due to logic, displays
and lighting.
o It will be a $250 billion business in 2025, with
at least ten billion dollars sales from
logic/memory, OLED displays for electronic
products, OLED billboard, signage etc, nonemissive organic displays, OLED lighting,
batteries and photo-voltaics, with sensors
almost at that level.
o Organic lighting will severely dent sales of both
incandescent and fluorescent lighting in the
second decade from now.
o Organic electronics in the form of electronic
billboards, posters, signage and electronic
books will revolutionize the conventional
printing and publishing industry.
o The future organic market will be newly created
without replacing much from the inorganic
semiconductors in existing electronics
products.
MOS-AK / GSA Workshop, 3 April 2009, IHP, Frankfurt / Oder
Organic Electronics:
Challenge for Electronic Design Automation (EDA)
o
Inorganic semiconductor industry relies
extensively on EDA software to support the
iterative cycles of process, device and
circuit technology improvements.
o
To further develop organic electronics
industry, equivalent design tools are needed.
o
EDA tools essentially depend on numerical and
analytical device models which are, in case of
OSCs, not yet matured and quite sparsely
implemented in commercial EDA tools.
o
Cambridge Display Technology (CDT) and
Silvaco Europe have joined forces together in
a TSB funded project entitled PMOS to
enhance EDA tools for use in the organic
electronics and to help move organic
transistor technology from the lab to the
shop floor.
MOS-AK / GSA Workshop, 3 April 2009, IHP, Frankfurt / Oder
UK Technology Strategy Board (TSB) Project:
Physical Modelling of Organic Semiconductors (PMOS)
Project partners
•
Cambridge Display Technology (CDT)
– Expert in polymer light emitting diode (PLED) technologies.
– Leader in development of solution processable (printable) organic.
semiconductors for display fabrication.
– Expertise in development of PLED materials and deposition processes.
Silvaco
– Leading provider of TCAD and EDA software for IC design
– Provides established products for TCAD process and device simulation, spice
parameter extraction, circuit simulation, custom IC design and verification.
•
Project activities
•
•
•
•
Design of OTFT devices using physical TCAD modelling.
OTFT spice modelling and parameter extraction.
Measurements and modelling of device reliability and aging effects.
The focus is on display device (OLED) drivers as these will be the first
large scale organic semiconductor products.
MOS-AK / GSA Workshop, 3 April 2009, IHP, Frankfurt / Oder
OFET Architectures and Peculiar Features
o OSC morphology varies from amorphous to (poly)crystalline.
o OFETs are commonly realized using an OSC layer without
a deliberate doping.
o The carriers that contribute to the charge distribution and
transport in OFETs must be injected from the metallic
contacts.
o Without particular semiconductor type of the OSC layer,
OFET can operate in the electron or hole carrier
accumulation mode depending on polarity of the gate
voltage and capabilities of the contacts to inject particular
carrier type.
o The source and drain have no junction isolation. A
drain/source leakage current is limited by intrinsic OSC
conductivity and contact resistance rather then reverse
junction current.
o Contact resistances often dominate the OFET
performance and represent a bottleneck to achieve full
potential of the intrinsic transistor effect.
o OFETs are typically characterized with much lower carrier
intrinsic carrier mobility having different bias physical
origin then their inorganic counterparts.
MOS-AK / GSA Workshop, 3 April 2009, IHP, Frankfurt / Oder
Density of States and Carrier Concentration
E
E
LUMO
CB
TRAPS
VB
HOMO

n   g A ( E) f A ( E) dE

1. Intrinsic DOS from molecular LUMO (HOMO) energy
levels ranging from delocalized conduction and valence
bands in molecular crystals to localized trap-like energy
distributions amorphous OSCs.
2. DOS from localized in-gap trap energy levels.
MOS-AK / GSA Workshop, 3 April 2009, IHP, Frankfurt / Oder
Silvaco Atlas:
OFET Example
MOS-AK / GSA Workshop, 3 April 2009, IHP, Frankfurt / Oder
Variable Range Hopping Conductivity Model
Percolation Theory
Gij  G0 exp sij 
sij  
2 Rij
a

Ei  EF  E j  EF  Ei  E j
2kT
Gc  G0 exp sc 
Gij  Gc
N b  g ( E ) sc  sij dRij dEi dE j
Bc sc  

 2.8
N s  g ( E ) sc kT  E  EF dE
n 
0
qn
MOS-AK / GSA Workshop, 3 April 2009, IHP, Frankfurt / Oder
exp( sc )
Sheet Channel Conductivity
Exponential DOS distribution

Carrier concentration
 q  VC 
n  ni (T ) exp 

kT


0
T / T0
ni (T )  ni 0
sin( T / T0 )
 qE  EC 
NA
g A (E) 
exp 

kT0
kT


0
Local channel conductivity
T /T
 (T0 / T )3
 q  VC   0
  0
ni 0 exp 

3
kT


0
 2  Bc
Vissenberg and Matters, Phys. Rev. B, 1998.
MOS-AK / GSA Workshop, 3 April 2009, IHP, Frankfurt / Oder
Sheet Channel Conductivity (and Mobility)
GC  C  QC 

 
0   QC


C  C 
 CiVC 
VC 
2kT0 ni 0 S
Ci
Exact
 T0 
 1
T

  2
 0  ni 0 (T0 / T ) 4 sin(T / T0 ) 
0
C 


qni 0 
2 3 Bc

Model
T0 / T
MOS-AK / GSA Workshop, 3 April 2009, IHP, Frankfurt / Oder
Channel Sheet Charge:
Surface Potential (SP) Model
Surface Potential (GAMMA=1e-4)
5
Exponential DOS carrier concentration gives
0
 q  VC 
n  ni (T ) exp 

kT


0
(Psi-Vc) [V]
-5
-10
-15
SP equation in accumulation operation mode:
-20
-25
-25
-20
-15
-10
-5
0
5
10
15
20
(Vg-Vfb-Vc) [V]
Surface Potential (GAMMA=1)
5
0
(Psi-Vc) [V]
-5
  s  VC 

VG  VFB   s   Vt 0G  h
 Vt 0 
2q s ni (T )
G
Ci
2
2
h( x)  exp( x)  x  1
QC  CiVG  VFB   s 
-10
-15
-20
-25
-25
-20
-15
-10
-5
0
(Vg-Vfb-Vc) [V]
5
10
15
20
There is an approximate analytical solution
(Gildenblat, et al., IEEE J. SSC, 2004)
MOS-AK / GSA Workshop, 3 April 2009, IHP, Frankfurt / Oder
Channel Sheet Charge:
Unified Charge Control Model (UCCM)
 s
CC

VC V
CC  Ci
VG
Ci
s
G
QC
 Ci
 s V
G
1
1 
  dVC

 Ci CC 
CC dQC 
n
CC  
QC
2Vt 0
 Q 
  QC  CiVt 0 ln  C   CiVG  VT  VC 
QCP
 
 QCP
 V  V  VC   ln( 2) 
2V C 
  2CiVt 0
QCP

QC   t 0 i ln 1  exp  G T
ln( 2) 
2Vt 0


MOS-AK / GSA Workshop, 3 April 2009, IHP, Frankfurt / Oder
Intrinsic Drain-Source Current
z
W
I DS
L
d
S
D
y
W

L
VD
 G  d
C
n
VS
x
Surface-Potential Based Model
W V
dn

d s
 GC ( s )
L V
d s
Charge Based Model
D
I DS
S
dn
2V C 
 1 t0 i
d s
QC ( s )
I DS
W

L

QCD

GC (QC )

QCS
dn
1 V
  t0
dQC Ci QC
drift
diff
I DS  I DS
 I DS
  QC S  2   QC D 2 


  2Ci


 1
 1
W   QC S   QC D 
 2Vt 0 

  1
L 

drift
I DS
diff
I DS
dn
dQC
dQC
W

L
MOS-AK / GSA Workshop, 3 April 2009, IHP, Frankfurt / Oder
Parameter Extraction in UTMOST IV
MOS-AK / GSA Workshop, 3 April 2009, IHP, Frankfurt / Oder
Model Verification:
Surface Potential Based
-4
0
10
-0.01
-5
-0.02
Ids [mA]
-Ids [A]
10
-6
10
-0.03
-0.04
-0.05
-7
10
-0.06
-8
10 -40
-35
-30
-25
-20
-15
-10
-5
0
-0.07
-40
-35
-30
-25
-15
-10
Vds [V]
Vg [V]
Comparison between simulated (lines) and
measured (circles and squares) transfer
characteristics of the OTFT
-20
Comparison between simulated (lines) and
measured (circles and squares) output
characteristics of the OTFT
MOS-AK / GSA Workshop, 3 April 2009, IHP, Frankfurt / Oder
-5
0
Model Verification:
Charge Based
Comparison between simulated (lines) and
measured (circles) transfer characteristics of the
OTFT in the linear operation region with Vds=-3V
(blue line and circles) and saturation operation
region with Vds=-30V (red line and circles)
Comparison between simulated (lines) and
measured (circles) output characteristics of the
OTFT for Vg=-10V, -20V, -30V and -40V.
MOS-AK / GSA Workshop, 3 April 2009, IHP, Frankfurt / Oder
Model Verification:
Temperature Variations
Comparison between simulated (lines) and measured (circles)
transfer characteristics of the OTFT in the saturation operation
region at different temperatures: T=270K (dark blue) T=280K (light
blue) T=300K (green) T=310K (pink) T=330K (red)
MOS-AK / GSA Workshop, 3 April 2009, IHP, Frankfurt / Oder
Origin of Contact Resistance
Surface Contamination
Geometric
Top gate, bottom contact has greater area for
injection, thus our preferred architecture
G
Contamination layer
S
S
Oxide layer will grow on many metals
Residues from resist processes etc
Oxide layer
G
Bottom Gate
Top Gate
Morphology
Electronic
Electronic barriers
to injection into
air stable, deep
HOMO OSC
LUMO
WF
S
HOMO
Voids in OSC to source/drain lead
to poor contact
Crystal orientation important
MOS-AK / GSA Workshop, 3 April 2009, IHP, Frankfurt / Oder
Strong Influence of the Series Resistances
Output Characteristics (no series resistances)
2.0e-04
0.0e+00
-2.0e-04
Id [A]
-4.0e-04
-6.0e-04
-8.0e-04
-1.0e-03
-1.2e-03
-1.4e-03
-40
-35
-30
-25
-20
-15
-10
-5
-0
Vd [V]
MOS-AK / GSA Workshop, 3 April 2009, IHP, Frankfurt / Oder
Contact Resistance Extraction
Contact resistance at L=0
VGS  10V
Total Channel
Resistance
Vg = -20V
500um
Vg = -30V
Vg = -40V
VGS  20V
VGS  30V
VGS  40V
Channel Length, L
RON  RSD (VGS ) 
L  L
1
 RSD (VGS )  L  L S (VGS )
 1
Weff K0 VGS  VT 
1  LGB  1 LGB 1
 1

0  Leff  G Leff  B
 F 

 kT 
0 ( F )  0 exp 
F  2 106
V
cm
MOS-AK / GSA Workshop, 3 April 2009, IHP, Frankfurt / Oder
Non-Quasi-Static Effects
Rt 
Leff
Weff  GT 0
Ct  LeffWeff  CT 0
1
1
I

 K ss  DS
RSX RDX
VGS V
MOS-AK / GSA Workshop, 3 April 2009, IHP, Frankfurt / Oder
DS
0
Some Other Open Issues
•
•
•
•
•
•
Influence of the bias dependent bulk and interface traps:
- extended surface potential equation including forward and reverse,
(accumulation and depletion) operation mode,
- a bias modulated back channel conductivity.
Leakage currents
- bulk and contact limited leakage effect,
- temperature dependence.
Short channel effects:
- physics based channel length modulation,
- effects of the depletion and strong lateral electric field on the drain side,
- space-charge limited transport.
Gate leakage current model.
The influence of the OSC film thickness on the interface electric field
- layered distribution of the accumulated charge due to the elongated molecular
geometry and regular assembly,
- floating body potential.
Aging and hysteresis of the OTFT characteristics within the model and the
corresponding circuit design.
MOS-AK / GSA Workshop, 3 April 2009, IHP, Frankfurt / Oder
Acknowledgement
We want to thank Prof. Benjamin Iñiguez and his group
for valuable recommendations regarding compact
organic TFT modelling.
This work is supported by the UK Technology Strategy Board
through the PMOS project TP/J2519J.