IV.

Supercapacitors
Charge storage:
electrical double-layer (EDLC)
fast and reversible faradaic redox
reaction (pseudocapacitance)
Energy storage depends on the ability of electrode to adsorb electrolyte ions under
the applied potential
1 10 100 1000 10000
1
10
100
1000
E
n
e
r
g
y

D
e
n
s
i
t
y

(
W
h
/
k
g
)

Power Density (W/kg)
Li - ion
Fuel
cells
Super-
capacitors
Pb-acid
batteries
NiMH
NiCd
conventional
Higher power
Faster charging (subseconds/seconds vs. minutes/hours)
Less harmful (no heavy metals)
Offer long cycle life (>1,000,000 cycles vs. 500)
Easy to detect the state of charge
Lower fundamental limit for the device voltage (voltage only restricted by the
decomposition of the electrolyte)
Advantages over batteries (for conventional device geometry):
What are the fundamental limits of energy storage ?
1. Empirical estimation:
Energy in supercapacitor device
16
1
2
2
CV
E
Capacitance: (a) 6-30 uF/cm
2
in carbon; (b) up to 200 uF/cm
2
in functionalized carbon;
(c) up to 200 uF/cm
2
in transition metal oxides
If surface area = 2000 m
2
/g capacitance up to 4000 F/g could be reached in an electrode
If max Voltage = 1V the Energy density E = 35 Wh/kg; if 3V, E=315 Wh/kg
due to packaging, two C in a series etc.
Assume formation of a close-packed monolayer of solvated ions (1e; d = 1.5 nm): 0.4 C/m
2
If could be achieved @ 1V: Capacitance = 4,000 uF/cm
2
and E = 1,400 Wh/kg
If could be achieved @ 10V: Capacitance = 400 uF/cm
2
and E = 14,000 Wh/kg
2. Semi-empirical estimation:
3. Fundamental limit:
If 1e/atom in electrode: eNV/2 per electrode (1/2 of batteries but V
supercap
could be >2x higher)
Will micro-fabrication affect supercapacitor performance ?
Power achievable is related to the Equivalent Series Resistance (ESR)
ESR = R
cc/electrode
+ R
electrode
+ R
electrolyte
+ R
separator
Point contacts
> 50 m 25 m
< 0.5 m
< 0.2 m
traditional
Integrated (on-chip)
By moving from traditional to
integrated supracapacitor we:
decrease R
cc/electrode
by >100
decrease R
electrode
by > 100
decrease R
electrolyte
by > 100
decrease R
separator
by > 100
Integrated supercap:
> 100x Power
(gravimetric)
> 200x Power
(volumetric)
< 1% Charging time (t
diff
x
2
)
polymer separator
IV. Supercapacitors
Double Layer Capacitors
+
-
+ + + + +
separator
Gouy-Chapman-Stern Model
Inner
layer
Solvated
ions
Diffuse
layer
walls pore and ions e between th distance separation - d
area; - A charge; - Q e; capacitanc - C
(Farads)
time discharge - e; capacitanc - C power; - P energy; - E
;
2
2
d
A
V
Q
C
E
P
CV
E
r o
Positively
charged
electrode
(+V)
Surface charge
Adsorbed
solution ions
Helmholtz Model
Double Layer Capacitors: Effect of Pore Size on Capacity (Energy)
Energy ~ C
Power ~
R
C
1
C
a
r
b
o
n
Ideal pore size
(~ 3x solvated ion size)
Carbon
2
Pore
Surface
3
Carbon
electrolyte ions in their
solvation shells
Too large pore size Too small pore size
Traditional view
Supercapacitors R&D Status
High Energy Density / High Frequency (100-1000 Hz)
operation has never been demonstrated
Intergated Supercaps never demonstrated
Solid Electrolyte operating at high frequency a challenge
Experimental data on the effects of various parameters on
ion transport in thin films are very limited
Experimental data on the electrical resistance of the
electrodes and interfaces are very limited
Planar and Integrated Supercapacitors
Objectives:
Thin film or planar components to enhance the energy
and power densities for efficient power supply,
conversion and regulation.
Optimization of geometry / architecture for controlled
Energy/Rate performance
Planar supercapacitor components for easier integration
in the packaging
Proposed R&D
Patterned deposition of either CNT (or metal nanowires) on Si wafer
surface via CVD
Surface coating with metal oxide layer (via ALD or CVD)
Developing successful process flows for the formation of integrated
devices based on the proposed architecture
Studying effects of film thickness, porosity, architecture and metal oxide
microstructure on the frequency response, capacitance and power
characteristics
Developing new routes for infiltration of electrodes with polymer
electrolyte and investigate the effects of electrolyte (polymer, solvent,
salt) on frequency response
Double Layer Capacitors: Effect of Pore Size on Capacity
J. Chmiola, G. Yushin, Y. Gogotsi, et al., Science, 313,1760-1763 (2006)
Distortion of solvation
shells in sub-nm pores
results in enhanced ion
storage capacities
C~1/d
Double Layer Capacitors: Effect of Pore Alignment on the Rate of
Ion Transport (Power)
Employed zeolites as sacrificial templates (as originally proposed by Kyotani)
C-coating
Zeolite etching
in HF
1E-3 0.01 0.1 1 10
0
50
100
150
200
250
300
4h, annealed
8h
6h
4h
C
a
p
a
c
i
t
a
n
c
e
,

F
/
g
Frequency, Hz
Frequency response of EDLC based on 0.3 mm
thick electrodes as a function of pore alignment:
aligning the pores allowed nearly 3 orders of
magnitude faster ion transport rate in 1 nm pores
(and thus 3 orders of magnitude higher power)
Pseudo-Capacitors: Polymer-Based (Solution Processing Possible)
-0.6 -0.3 0.0 0.3 0.6
-600
-300
0
300
600
C
a
p
a
c
i
t
y
,

F
/
g
Voltage, V
Regular PANI
After 10,000 cycles
I. Kovalenko, et. al, Advanced Functional Materials, 2010
-0.6 -0.3 0.0 0.3 0.6
-800
-400
0
400
800
after re-mixing
(re-making) electrodes
before re-mixing
C
a
p
a
c
i
t
y
,

F
/
g
Voltage, V
pure PANI after 10,000 cycles
-0.6 -0.3 0.0 0.3 0.6
-800
-400
0
400
800
C
a
p
a
c
i
t
y
,

F
/
g
Voltage, V
3 wt.% purified ND
after 10,000 cycles
0
100
200
300
400
500
600
700
- initial
3

w
t
.

%

N
D

s
o
o
t
3

w
t
.

%

p
u
r
i
f
i
e
d

N
D
2
8

w
t
.

%

a
n
n
e
a
l
e
d

N
D
1
4

w
t
.

%

a
n
n
e
a
l
e
d

N
D
3

w
t
.

%

a
n
n
e
a
l
e
d

N
D
S
p
e
c
i
f
i
c

C
a
p
a
c
i
t
a
n
c
e
,

F
/
g
p
u
r
e

P
A
N
I
- after 10,000 cycles
By incorporating
nanopowder (e.g.,
nanodiamond
powder) into PANI
and by controlling
NP-pani interactions,
greatly enhanced
stability could be
achieved
Pseudo-Capacitors: Transition Metal Oxide Coatings on Carbon
Nanotubes or Metal Nanowires via Vapor Deposition (CVD or ALD)
Al nanowire core
Cu foil
VO
x
coating
1.0 1.5 2.0 2.5 3.0
-3000
-2000
-1000
0
1000
2000
3000
porous carbon
used in commercial
supercapacitors C
a
p
a
c
i
t
a
n
c
e

(
F

g
-
1
)
Potential vs. Li/Li+ (V)
VO
x
coating on
Al nanowires
Transition Metal oxide
provides high capacitance
but suffers from low
electrical conductivity
Metal Nanowires
provide highly conductive
high surface area support
Cu foil
Forest of Al nanowires
2 cm
Gas flow
Vo
x
- coated
Al nanowires
200 nm
Al nanowire growth by CVD Metal Oxide ALD
Pseudo-Capacitors: Transition Metal Oxide Coatings on Carbon
Nanotubes or Metal Nanowires via Vapor Deposition
Aligned CNT on a current collector
Misaligned V
x
O-CNT electrode
0.01 0.1 1 10 100 1000
35 Hz
S
p
e
c
i
f
i
c

c
a
p
a
c
i
t
a
n
c
e

(
a
.
u
.
)
Frequency (Hz)
CNT growth is faster than that of metal
nanowires
However, higher temperatures are
needed (without plasma enhancement)
and the interface between CNT and metal
oxide could be inferior
IV. Supercapacitors
Objectives:
Thin film or planar components to
enhance the energy and power
densities for efficient power supply,
conversion and regulation.
Planar supercapacitor components
for easier integration in the
packaging
Proposed R&D:
Patterned deposition of either CNT (or metal nanowires) on Si wafer surface via CVD
Surface coating with metal oxide layer (via ALD or CVD)
Developing successful process flows for the formation of integrated devices based on the
proposed architecture
Studying effects of film thickness, porosity, architecture and metal oxide microstructure on the
frequency response, capacitance and power characteristics
Developing new routes for infiltration of electrodes with polymer electrolyte and investigate the
effects of electrolyte (polymer, solvent, salt) on frequency response
Challenges:
Optimization of geometry / architecture for
controlled Energy/Rate performance
Experimental data on the effects of various
parameters on ion transport in thin films are
very limited
Experimental data on the electrical resistance
of the electrodes and interfaces are very
limited
High Energy Density / High Frequency (100-
1000 Hz) operation has never been
demonstrated
Intergated Supercaps never demonstrated
Solid Electrolyte operating at high frequency
a challenge