1. Impact of Calendering and Silver
Addition to Carbon Nanotube-based
Electrodes Used in Printed Multi-
layer Capacitors
Ramea Al-Mubarak
May 19, 2015
1
2. Outline
• Motivation
• Introduction
• Hypothesis
• Research Goals
• Experimental Methods
• Results and Conclusion
• Recommendations for future studies.
2
3. Motivation
• Energy harvesting
• Produced electrical energy
storage
• Energy storage device:
minimization of energy loss
creating advanced energy storage
devices with large storage
capacity and high efficiency.
• Energy storage technologies:
Batteries
Super-capacitors
3
Demand for global
energy
Conventional
energy resources
Sustainable & clean
energy
development
4. Introduction: Capacitors
Devices that physically store electric charge
Energy stored through charge
separation
(static electricity between
two oppositely charged plates)
C α
𝐴
𝑑
Capacitance(C): the ability of a
body to store an electrical charge
(Measured by Farad)
Traditional Capacitor
4
5. Some groups printed carbon nanotube sheets*
• The lowest achieved sheet resistance was 78
Ω with 200 printed layers. (0.2 microns).
• Taking the print layers into account, the
lowest recorded sheet resistance was 760 Ω
with 12 prints.
* Ref: Ryan et al., Inkjet printing of carbon nanotubes . Nanomaterials, 3,
453-468.
5
6. • Some of the carbon nanotubes might be completely
isolated and have no contact with any other carbon
nanotubes.
• Electrons are constrained to isolated CNTs, which do
not contribute to the conductivity of the printed sheet.
• Some of the CNT may be in contact with other CNT;
this contact creates an electron pathway, ultimately
resulting in electrical current, which is the reason for
conductivity of printed sheets
Electron pathway
Isolated CNT
CNTs Network
6
7. Hypothesis
* Addition of silver flake ink and silver
nanowires into a CNT ink, used for printing
the electrode layers of capacitor, will improve
the device capacitance:
• Lacing and eliminating the isolated tubes of
CNT
• Conductivity of silver is very high which will
improve the conductivity of the CNT sheet
conductivity
7
8. Cont. Hypothesis
*Calendering the electrode layers using
different pressure values would smooth the
film surface, eliminate and flatten the
peaks resulting from the protruding CNTs.
8
9. Before calendering
Electrode
Dielectric
After calendering
Dielectric layer needs to completely cover the electrode to
prevent shorting
The smoother the electrode layer is, the thinner the dielectric
layer needed the higher the capacitance of the device
9
10. Research objectives
• optimize silver percentage and structure (nano-
wires or silver flake) added to CNT ink on the sheet
conductivity.
• Study the influence of calendering the multi-layer
capacitor electrodes.
• Fabricate multi-layer capacitors.
• Formulate carbon ink from a synthetic graphite
powder and conductive carbon black powder.
10
11. Experimental methods
1- Carbon ink formulation
5 wt% ammonia in DI
water (45wt.%)
25.6 wt.% Graphite
powder (172.08 𝑚2/𝑔)
Mix 5 min
Adjust PH ≈ 9
Solution resin 23.5%
4.3% Eth
Mix 45 min
1.6% film forming
emulsion resin
Slowly mix 15 min
Printed onto PET with a
Byrd bar#7
Good film adhesion ,but the sheet
resistivity was high, 770 Ω
Ink A
11
12. Cont. carbon ink formulation
5 wt% ammonia in DI
water (13.6 wt.%)
23 wt.% Conductive
carbon black powder
(S.A=254 𝑚2
/𝑔)
Mix medium speed 2 hours
Adjust PH 8-9
Solution resin 37.6%
8.3% Eth
Slowly mix 1 hr
17.6% film forming
emulsion resin
Printed onto PET with a
Byrd bar#7
High viscose, and it showed very poor adhesion.
Films readily peeled off the PET substrate after
curing.
Ink B
12
13. Ink A + Ink B Ink C
Obtain Ink C with better adhesion and conductivity
properties.
0
20
40
60
80
100
120
140
160
180
47.8 wt.% Ink A 67.5 wt.% Ink A 74 wt.% Ink A
SheetResistivityΩ
52.2 wt.% Ink B 32.5 wt.% Ink B 26 wt.% Ink B
13
14. Inks specifications
Ink A Ink B Ink C
Carbon wt.% 25.6% 23.0% Theoretical 25.10%
Exp. 23.62%
- Graphite 20.80%
- Conductive carbon black 4.30%
Solution resin
wt.%
23.5% 37.6% 26.10%
Emulsion resin
wt.%
1.59% 17.58% 4.58%
Sheet resistivity Ω 770 - 104.2
32.50% of Ink B to Ink A sheet resistivity 86.4%
Surface area of conductive carbon black≈ 32.3% higher
than the surface area of graphite powder
14
16. Ink blends preparation: (Manually mixed)
Graphite ink + Silver flake ink
Silver wt.%: 5, 10, 15, 20, 25, 30, and 35 wt.%
CNT ink + silver nanowires
Silver wt.%: 2, 5, 8, 10, 20, and 30 wt%
CNT ink + Silver flake
Silver wt.%: 5, 10, 20, 35, 50, 75, and 90 wt.%
For instance , a blend of 10 wt.% of silver nanowire in
CNT was prepared by adding 7.56 ml silver nanowire to
8.5 g CNT ink
16
17. DOE for films drown down printed by blends of silver with CNT or graphite ink.
Draw-down printed onto PET by Byrd bar.
Cured in an oven at 120 ºC for 10 min.
Silver flake ink + carbon samples sintered using a NovaCentrix
PulseForge 1200 unit at 370 V, 1400 μs, and 1.5 OLF at 20
FPM (Sheet resistivity 28% less than samples cured in the oven
alone)
17
18. Sheet resistivity measured
(Four-Point Collinear Probe)
Average thickness and roughness
measured
(Bruker GT-K interferometer
microscope)
Bulk resistivity and conductivity
calculated.
Bruker ContourGT-K -
Optical Microscope
bruker.com
18
19. 3- Multi-Layer Capacitor Fabrication
Phase I: Printing the electrode layers and screen selecting
DOE for screen printing the electrodes layer
Screen 1:Mesh count 325SS
and 0.0011" wire diameter
Screen 2: mesh count 105SS and
0.003" wire diameter
19
20. Phase II: Printing Complete Multi-Layer Capacitors
Stack of 3 layers was printed.
A dielectric layer (ELECTRODAG PF-
455B) was printed on top of the electrode
layer.
Cured using a Fusion UV drier equipped
(passed through the UV drier 2 times)
Third layer was printed on top of the
dielectric.
The capacitance of was measured using
an E4980A Precision LCR Meter (Agilent
Technology).
Complete multilayer printed
capacitor (Three layer
structure)
20
21. 4- Calendering
Nip pressures (10, 25, 30
psi). 1 pass through the
calender.
Roughness, thickness, and sheet
resistivity .
Compared to the values without
calendering .
Study the influence of
calendering on sheet conductivity
and its surface topography.
21
22. Results and Discussion
1- Effect of Silver Addition to Commercial CNT Ink
Silver Nanowires
Blends Incompatible and clearly unprintable and unstable
at wt.% beyond 10 wt.% silver nanowires
a- CNT blended with 5% silver
nanowires ink
b- CNT blended with 20% silver
nanowires ink
Compatibility of CNT with silver nanowires.
Silver nanowires
will spontaneously
aggregate when
added to aqueous
solutions with a
low amount of
ions.
Commercial CNT
ink contained 0.5-
15 wt.% water,
22
23. % by which Silver Nanowires addition decreases sheet and
bulk resistivity in comparasion to the control samples, CNT
80
82
84
86
88
90
92
94
96
98
100
2 5 8 10
%bywhichsheetandbulk
resistivitydecreases
Silver Nanowires wt.%
% of Sheet
Resistivity
change
% of Bulk
Resistivity
change
2 wt.% silver nanowires : Decreases sheet resistivity by ≈ 88%, and the bulk
resistivity by ≈ 98%.
23
24. 0.00
20.00
40.00
60.00
80.00
100.00
120.00
140.00
160.00
0 2 4 6 8 10 12
SheetResistivity(Ω/□)
Silver nanowires wt.% in ink
The effect of silver nanowire addition to CNT ink on sheet resistivity
Taking into consideration the cost and conductivity
improvement optimum % silver nanowires 3 wt.%.
- High electrical conductivity of silver
- High aspect ratio (= 277.78)
connecting the isolated CNT and forming
long-range connectivity in the random
systems, according to percolation theory
24
25. Silver flake
0.00
20.00
40.00
60.00
80.00
100.00
120.00
140.00
160.00
0 10 20 30 40 50 60 70 80 90 100
SheetResistivity(Ω/□)
%Novacentrix silver flake in ink
In over the range of 0-10 wt.%.
Max. value of 154.1 Ω at 20% silver (0.328% higher)
% by which the resistivity is improved is almost the
same beyond 50 wt.%.
Optimum 35 wt.% silver flake (87.29% lower)
-5
15
35
55
75
95
5 10 20 35 50 75 90
%bywhichsheet
resistivitydecreases
Silver flake wt.%
25
27. • Bulk resistivity (98% lower):
75 wt.% silver flake ink = achieved by adding 2 wt.% silver
nanowires.
• Cost and bulk resistivity results :
adding silver flake is not beneficial.
silver flake is micro-sized which might separate and insulate
the CNTs instead of connecting them.
• At 35% resistivity starts to decrease:
might have nothing to do with percolation or connecting the
CNTs
Silver flake might begin to dominate the resistivity response
and the CNT began to act like an impurity
The improvement might be due high conductivity of silver.
27
28. 2- Effect of Silver Flake Addition on Fabricated
Graphite Ink
0
50
100
150
200
250
300
350
400
0 5 10 15 20 25 30 35
Sheetresistivity,Ω/□
Silver flak wt% in ink
-5
5
15
25
35
45
55
65
75
5 10 15 20 25 30 35
%bywhichsheetresistivitydecreases
Silver flake wt.%
Compatible at all %
No aggregation was
observed.
5 wt.% silver flake
reduced the sheet
resistivity to 50.69%,
compared to the
100% graphite ink.
28
32. 2- Influence of Calendering on Sheet Thickness,
Resistivity, and surface topography
No influence on electrodes thickness, roughness, and sheet resistivity.
No effect on high CNT peaks.
Eliminated the short peaks.
Before calendering After calendering
32
33. 3- Printed capacitors capacitance
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
CNT CNT CNT+Silver flake
capacitance,nF
CNTS14, t=0.03
microns
CNTS22, t=0.16
micron
CNTNS35S22, t=0.8
microns
Capacitors fabricated using a 35 wt. % silver flake and
65 wt. % CNT blend for printing the electrode layers
enabled an 8.6 % increase in capacitance over CNT
electrode-based capacitors
Optimum point 3
wt.% provides
sheet resistivity ≈
5 Ω. 72.4% higher
than the sheet
resistivity of 35%
silver blind
33
35. Conclusion
• Ink was formulated from graphite powder (327.8 Ω, 0.432
Ω.cm).
• Commercial CNT ink (153.6 Ω,0.002 Ω.cm)
• To compare the two inks:
CNT has a sheet resistivity about 53% higher than
formulated ink.
bulk resistivity 99.5% higher than formulated ink.
CNT is more conductive, but it is more expensive.
• By adding 5 wt.% silver flake to Ink C, sheet resistivity was
reduced 50.7% and its bulk resistivity was decreased 12%.
35
36. • By adding 5 wt.% silver flake to CNT ink, sheet
resistivity was reduced 7.5% and bulk resistivity
increased by 632.6%.
• Adding silver flake to synthetic graphite ink is
efficient. But adding it to CNT ink not effeciant.
• The optimum silver % in the CNT ink was found to
be 3 wt.% silver nanowires.
• Results showed that calendering did not improve the
conductivity of the CNT printed sheets.
36
37. Recommendations for future studies
• Mixing nano-sized silver particles with the CNT ink
instead of using larger micro-sized silver flake particles.
• Fabricating an ink from scratch using CNT and silver
nanowire powders to avoid any aggregation and
incompatibility issues.
• Finally, print a stacked capacitor with more than three
layers to increase overall capacitance.
37
38. Acknowledgements
• Dr. Margaret Joyce, professor, supervisor.
• Dr. Paul D. Fleming III, professor, committee member.
• Dr. Thomas Joyce, professor, committee member.
• Colleagues: Bhushan Hiralal Pati, Veronika Husovska,
James Atkinson, Michael Joyce, Ali Eshkeiti.
38
