Transcript Electrostatic Discharge in Solids - DigitalCommons@USU

2013 USU Physics Colloquium
Utah State University, Logan, UT
Electrostatic Discharge in
Solids
Allen Andersen and JR Dennison
Physics Department
Utah State University, Logan, Utah
Thank you!
Supported through funding from the
USU Howard L. Blood Fellowship and
NASA Goddard Space Flight Center.
The entire USU Materials Physics Group
Motivation
Experiment
Analysis
Conclusion
Outline
Electrostatic Discharge
What is it and why do we care?
Instrumentation and Procedures
Determining material properties.
Analysis and Modeling
Understanding and comparing material behavior.
Future Work
Quantitative descriptions of microscopic physics.
Motivation
Experiment
Analysis
Motivation
Conclusion
Motivation
Experiment
Analysis
Conclusion
What is Electrostatic Discharge or ESD?
Insulating materials restrict
the flow of charge in an
electric field – up to a point.
If E > EESD → Electrostatic
Discharge (ESD)
Following an ESD:
• Insulator permanently
damaged
• Large currents can flow
• In general THIS IS VERY
BAD!
Motivation
Experiment
Analysis
Conclusion
Examples of ESD
~1 MV/m
in air
10-100 MV/m
in solids
Motivation
Experiment
Analysis
Conclusion
What to expect for solids (via some hand waving).
In Gas
Ionization energy ~10eV
qe * Average distance between particles in air ~10-5 m
=
~106 V/m
=
~109 V/m
In Solids
Ionization energy ~10eV
qe *Average distance between defects in solids
~10-8
m
The electron must gain enough energy traveling through the E-field to
ionize at the next collision to sustain an avalanche discharge.
Motivation
Experiment
Analysis
Conclusion
Spacecraft Charging
The sun gives off
high energy charged
particles.
These particles
interact with the
Earth’s atmosphere
and magnetic field
in interesting ways.
High energy particles imbed charge into spacecraft surfaces.
Motivation
Experiment
Analysis
Conclusion
Spacecraft Charging
• As charge builds up
in insulators the
internal electric field
can exceed the
dielectric strength of
the material.
• ESD accounts for
majority of
“anomalies” induced
by spacecraft
environment
Motivation
Experiment
Analysis
Conclusion
USU MPG
•Space Plasma
Environment
Simulations
More than two decades
of Spacecraft Charging
Research
Charging of power grid
components.
•Physics
Models
Motivation
Experiment
Analysis
Conclusion
Problems with Existing Grid
1. >10 % Power loss in transmission due to radiation and Joule heating.
2. The nation uses several separate out of phase regional power grids.
3. Existing grids are ageing and already being pushed past their limits.
Motivation
Experiment
Possible Solutions
Operate at Higher Voltages ~MV
Currently HV wires operate at ~500kV in US.
In Europe >1MV, in Japan and China>2 MV.
Advantages
• Much higher efficiency for MV transmission.
P=IV=V2/R
Disadvantages
• Higher stress on insulating components
resulting in leaks or failures.
• Coronal discharge for AC lines limits voltage.
Analysis
Conclusion
Motivation
Experiment
Coronal discharge
Analysis
Conclusion
Motivation
Experiment
Analysis
Possible Solutions
Conclusion
DC
Use DC Transmission
Advantages
• Reduces radiation and thermal resistance.
• More cost effective than AC over distances of hundreds
of miles.
• HV transmission without coronal discharge up to 4 MV.
• Could connect the nation’s out of phase power grids.
Disadvantages
•
•
•
•
Requires expensive DC/AC converter stations.
DC components in general are more expensive than AC.
Tradition – War of the Currents.
Higher electric field stress on materials.
AC
Motivation
Experiment
Analysis
Conclusion
Benefits of Improving the Power Grid
HVDC transmission improves efficiency.
Transmission loss effectively halved.
500kV/1000kV ≈ ½
The energy saved if transmission losses were halved nationwide ~30
large coal burning power plants. This is quite possible!
Motivation
Experiment
Analysis
Conclusion
Electronics in General
ESD is an important design
consideration for all electronics.
The problem does not scale
linearly due to quantum
tunneling. In Si/SiO2 transistors
the insulating layer is only a few
atoms thick.
dcircuit≈10-3m→ VESD ≈ 104V
dMOFSET≈10-8m→ VESD ≈ 1V
Motivation
Experiment
Analysis
Experiment
Conclusion
Motivation
Experiment
Analysis
Conclusion
ESD SYSTEM
USU MPG ESD
chamber ~10-6 torr.
>105 times below
Paschen minimum
→ No arcs through
gas around samples.
Motivation
Experiment
Analysis
ESD SYSTEM
Conclusion
• Simple parallel plate
capacitor
• Fully automated
system
• Applies up to 30 kV
• ~150 K<T<300K
with ℓ-N2 reservoir
• 1.98 cm2 sample
electrodes
• 6 electrode carousel
Motivation
Experiment
Analysis
Conclusion
ESD SYSTEM
G
F
E
D
J
H
I
B
A
C
A) Copper electrodes
B) Thermocouple
electrodes
C) Polycarbonate base
D) Conductive sample
plate
E) Thermally conductive,
electrically isolating
layer
F) Liquid Nitrogen
reservoir
G) Adjustable pressure
springs
H) Glassy sample
I) Conductive padding
J) Polymer sample
Motivation
Experiment
Analysis
Conclusion
‘Typical’ Results
Once the critical voltage is reached
the sample breaks down and
current flows freely through the
material.
As voltage is first applied to
The slope of the graph is just the
the sample no current flows
inverse of the current limiting
through it. The insulator is
resistance in the circuit.
doing its job.
V=IRL
I/V = RL-1
The discontinuity
marks our
breakdown voltage.
Motivation
Experiment
Analysis
Conclusion
What ESD data tells us
• The breakdown voltage corresponds to a critical
electric field.
Eesd=Vesd/d
• Breakdown sites vary significantly depending on
material*.
20 µm
Polyimide
20 µm
LDPE
10 mm
ePTFE (Gortex)
Motivation
Experiment
Analysis
Conclusion
LDPE
Low Density Polyethylene – highly electrically insulating material.
Inexpensive and comes in many forms.
• Common spacecraft insulator
• Common power line insulator
• Common everyday material.
Motivation
Experiment
Analysis
Conclusion
Polyimide
Polyimide or KaptonTM– highly electrically insulating material.
Relatively inexpensive and very robust chemically, thermally, and
electrically.
• Common spacecraft insulator
• Common electronics insulator
Motivation
Experiment
Analysis
Conclusion
Glass
Disordered Fused Silica SiOx–electrically insulating material.
Expensive in thin sheets but otherwise inexpensive–very useful.
• Common semiconductor insulator
• Excellent optical coating
• Spacecraft solar array
cover glass
• Window glass
Motivation
Experiment
Analysis
Analysis
Conclusion
Motivation
Experiment
Analysis
Conclusion
LDPE Data
Ohmic slope
I = V/RL
x-intercept at
origin
V10 ≈0V
Prebreakdown
arcing
Breakdown
voltage
Motivation
Experiment
Analysis
Conclusion
Polyimide Data
Ohmic slope
I = V/RL
V10 ≈0V
Prebreakdown
arcing
Breakdown
voltage
Motivation
Experiment
Analysis
Conclusion
Pre-Breakdown Arcs–Oscilloscope
Prebreakdown
arcing
LDPE
Motivation
Experiment
Analysis
Conclusion
Pre-Breakdown Arcs–Oscilloscope
Prebreakdown
arc
Polyimide
Motivation
Experiment
Analysis
Conclusion
Pre-Breakdown Arcs
a) Defect sites can form
in kinks of polymer
chains. These low-energy
defects can be thermally
repaired since
εkinks ≥ 27 meV = kBTRM
b) At higher voltages,
electrons have enough
energy to break bonds,
creating permanent defect
sites.
εbond >> 27 meV = kBTRM
Motivation
Experiment
Analysis
Conclusion
Pre-Breakdown Arcs
Energy
ΔG
F
ΔG
ΔG-qeaoF
ΔG+qeao
F
qeaoF
ao
Position
Under an applied
electric field,
charge can ‘hop’
between defect
sites. The
probability of a
transition in a given
time step depends
upon temperature,
well depth, and
applied field.
Motivation
Experiment
Analysis
Conclusion
Conductivity Model
Trap-to-trap
Tunneling
frequency
Well depth
Density of
Defects
Tests lasting only days can predict decades of behavior!
Motivation
Experiment
Analysis
Conclusion
Time Endurance - LDPE
Pre-breakdown
arcs
Breakdown time
~325 MV/m
Motivation
Experiment
Analysis
Conclusion
Time Endurance - Polyimide
Long wait
time and
higher current
arcs indicate
differences in
defect
populations.
~320 MV/m
Motivation
Experiment
Analysis
Conclusion
Time Endurance
1 week
1 day
1 hr.
1 min.
Motivation
Experiment
Analysis
Conclusion
Fused Silica Breakdowns
Conductive Polyimide pads under fused silica ESD
Motivation
Experiment
Analysis
Fused Silica Data
80µm
Non-ohmic
slope
V10 ≈1000V
Breakdown
Voltage
R1>RL
Conclusion
Motivation
Experiment
Analysis
Conclusion
Fused Silica Data
80µm
2nd non-ohmic
slope
Non-ohmic
slope
V2
0
R1>R2>RL
≈200V
What is going on?!
V10 ≈1000V
Breakdown
Voltage
Motivation
Experiment
Analysis
Conclusion
Fused Silica Data
Actually, we’d already
seen this before.
65µm coating
V20 ≈80V
V10 ≈200V
R1>R2>RL
Motivation
Experiment
Analysis
Conclusion
Fused Silica Data
• Relatively low critical field – we can deal with that.
• Tunneling current seems to be a bigger factor for thinner coatings
• No noticeable pre-breakdown arcs – for SiO2 we don’t have polymer
chains to “kink.”
• Non-ohmic post breakdown slope – possibly only broken down part
way through the sample.
• Transitions to secondary slopes – marks increasing partial breakdowns
Motivation
Experiment
Analysis
Conclusion
We shouldn’t expect the same behavior for
different structures!
Motivation
Experiment
Analysis
Conclusion
Density of States
a) Delta function
b) Constant
c) Linear
d) Power law
e) Exponential
f) Gaussian
Each of these has
different transport
properties.
LDPE – Linear
Polyimide – Exponential
SiO2 – Exponential+Gaussian
Motivation
Experiment
Analysis
Transport Equations
Conclusion
Motivation
Experiment
Analysis
Conclusion
Transport Equations
Complete set of dynamic transport equations
𝐽 = 𝑞𝑒 𝑛𝑒 (𝑧, 𝑡)𝜇𝑒 𝐹(𝑧, 𝑡) + 𝑞𝑒 𝐷
∂
∂z
𝑑𝑛 𝑡𝑜𝑡 (𝑧,𝑡)
𝑑𝑧
{Sum of electron drift and diffusion current densities Ji}
{1D Gauss’s Law}
𝐹 𝑧, 𝑡 = 𝑞𝑒 𝑛𝑡𝑜𝑡 /𝜖0 𝜖𝑟
𝜕𝑛 𝑡𝑜𝑡 (𝑧,𝑡)
𝜕𝑡
− 𝜇𝑒
𝜕
𝜕𝑧
𝑛𝑒 𝑧, 𝑡 𝐹 𝑧, 𝑡
− 𝑞𝑒 𝐷
𝜕 2 𝑛 𝑒 𝑧,𝑡
𝜕𝑧 2
= 𝑁𝑒𝑥 − 𝛼𝑒𝑟 𝑛𝑒 𝑧, 𝑡 𝑛𝑡𝑜𝑡 𝑧, 𝑡 + 𝛼𝑒𝑡 𝑛𝑒 𝑡 𝑁𝑡 𝑧 − 𝑛𝑡 𝑧, 𝑡
{1D Continuity equation with drift, diffusion and source terms}
𝑑𝑛 ℎ 𝑧.𝑡
𝑑𝑡
= 𝑁𝑒𝑥 − 𝛼𝑒𝑟 𝑛𝑒 𝑧, 𝑡 𝑛ℎ (𝑧, 𝑡)
{1D hole continuity equation with Generation and recombination terms}
𝑑𝑛 𝑡 (z,𝜀,𝑡)
𝑑𝑡
= 𝛼𝑒𝑡 𝑛𝑒 𝑧, 𝑡 𝑁𝑡 (z, 𝜀) − 𝑛𝑡 (z, 𝜀, 𝑡) − 𝛼𝑡𝑒 𝑁𝑒 𝑒𝑥𝑝 −
𝜀
𝑘𝑇
𝑛𝑡 (z, 𝜀, 𝑡)
{1D trapping continuity equation for electrons}
Complementary Responses to Radiation
E
Modified Joblonski diagram
--24 meV
--41 meV
• VB electrons excited into CB by the
high energy incident electron
radiation.
• They relax into shallow trap (ST)
states, then thermalize into lower
available long-lived ST.
• Three paths are possible:
1.92 eV
2.48 eV
2.73 eV
4.51 eV
(i) relaxation to deep traps (DT),
with
concomitant
photon
emission;
(ii) radiation induced conductivity
(RIC), with thermal re-excitation
into the CB; or
(iii) non-radiative transitions or e--h+
recombination into VB holes.
--8.9 eV
10/7/13
46
ISU Colloquium
Eeff
F
Complementary Responses to Radiation and Electric Field Stress
E
Modified Joblonski diagram
• VB electrons excited into CB by the
high energy incident electron radiation.
• They relax into shallow trap (ST)
states, then thermalize into lower
available long-lived ST.
• Four paths are possible:
E
--24 meV
--41 meV
1.92 eV
(i) relaxation to deep traps (DT), with
concomitant photon emission;
(ii) radiation induced conductivity (RIC),
with thermal re-excitation into the
CB;
(iii) non-radiative transitions or e--h+
recombination into VB holes; or
(iv) avalanche effect as CB electrons
excite more VB electrons into the
CB, causing ESD.
2.48 eV
2.73 eV
4.51 eV
--8.9 eV
10/7/13
47
ISU Colloquium
Eeff
F
Motivation
Experiment
Analysis
Conclusion
Conclusion
Motivation
Experiment
Analysis
Conclusion
Conclusions
• Polymer and glass structural differences are manifest in ESD
measurements of pre-breakdown arcing and post-breakdown
slopes.
• Pre-breakdown arcing can be understood in terms of thermally
recoverable and irrecoverable defect generation.
• The onset, magnitude, and frequency of pre-breakdown arcing
depends on the density of states for a given material.
• The performance of insulating materials under electric field stress
over time also depends on the density of defects.
Motivation
Experiment
Analysis
Conclusion
Conclusions
Am I really talking about ESD?
The lines between different charge
transport phenomena blur. This is charge
transport on an extreme scale!
Motivation
Experiment
Analysis
Conclusion
Future Work
• Acquire and analyze more ESD data for LDPE and Polyimide.
• Breakdown image analysis – look for patterns.
• Pursue fused silica test options and test borosilicate glass in the
meantime.
• Cast ESD phenomena in terms of the USU MPG charge
transport formalism. Merge ESD, CVC, SVP, SEY, RIC,
cathodoluminescence.
• Derive tbreakdown for the various DOS distributions and fit to data.
• Pursue applications for power grid technologies.
Questions?